t.

I .
I '

.
Id:
I
---:..--
'
~ -'----"'
THE PANOPTIC SORT

•
 l
'

Critical Studies in Communication

and in the Cultural Industries
Herbert I. Schiller, Series Editor

The Panoptic Sort: A Political Economy of Personal Information,

Oscar H. Gandy, Jr.
Triumph ofthe Image: The Media's War in the Persian Gulf- A Global Perspective,
edited by Hamid Mowlana, George Gerbner, and Herbert I. Schiller
The Persian Gulf TV War, Douglas Kellner
Mass Communications and American Empire, Second Edition, Updated,
Herbert I. Schiller

FORTHCOMING
The Social Uses ofPhotography: Images in the Age ofReproduction, Hanno Hardt
Introduction to Media Studies, edited by Stuart Ewen, Elizabeth Ewen, Serafina
Bathrick, and Andrew Mattson
Music Television, Jack Banks
Media Transformations in the Age of Persuasion, Robin K. Anderson
A Different Road Taken: Profiles ofFive Critical Communication Scholars,
John A. Lent
Hot Shots: An Alternative Video Production Handbook, Tami Gold and
Kelly Anderson
The Dallas Smythe Reader, edited by Thomas Guback
Public Television Versus the Market, William Hoynes
The Communications Industry in the American Economy, Thomas Guback
THE PANOPTIC SORT
A Political Economy
of Personal Information

Oscar H. Gandy, Jr.

Westview Press
BOULDER • SAN FRANCISCO • OXFORD
Critical Studies in Communication and in the Cultural Industries

AU rights reserved. No part of this publication may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopy, recording, or any information storage and re·
trieval system, without permission in writing from the publisher.

Cop}'Tight © 1993 by Westview Press, Inc.

Published in 1993 in the Uni ted States of America by Westview Press, Inc., 5500 Centra l Avenue, Boul-
der, Colorado 80301-2877, and in the United Kingdom by Westview Press, 36 Lonsdale Road,
Summertown, Oxford OX2 7EW

Library of Congress Cataloging-in-Publication Data

Gandy, Oscar H.
The panqpticsort: a political economy of personal information I
Oscar H. Gandy, Jr.
p. cm.-(Critical studies in communication and in the
cultural industries)
!ocludes bibliographical references and index.
ISBN 0-8133- 1656-1 (hc.)-TSBN 0-8133-1657-X (pbk.)
l. Privacy, Right of-United States. 2. Records-United States-
Access control. l. Title. II. Series.
JC596.2.USG36 1993
323.44' 8' 0973-dc20 92-37434
C!P

Printed and bound in the United States of America

e T he paper used in this publication meets the requirements

of the American National Standard for Permanence of Paper
for Printed Library Materials Z39.48-1984.

10 9 8 7 6 5 4 3 2
CONTENTS

List ofTables and Figures IX

Acknowledgments xi

1 Prologue
Introduction, l
The Diverse Theoretical Origins of Panoptic Thinking, 3

2 Information and Power 15

Introduction, 15
A Question of Power, 18
Panopticisrn, 21
Structuration, 25
Technology, Markets, and Culture, 35

3 Operating the Panoptic Sort 53

Data and the Panoptic Sort, 53
Government Information Gathering, 55
The Corporate Data Machine, 60
Data Processing, 71
A Technology of Power, 80
List Vendors, 90

4 Corporate Perspectives on the Pan optic Sort 95

In Pursuit of the Corporate View, 95
The Telemarketing View, l 02
Direct Marketing Association, 104
Robert Posch and Corporate Opinion Leadership, 106 .
The Mail Survey, 108
Corporate Policy and Practice, 111
Business Perspectives on the Panoptic Sort, 122

5 Relationships and Expectations 123

[n Search of Public Perceptions, 123
The Group Interviews, 126
Thinking About Privacy, 126
Defining Privacy Invasions, 128

vii
r
Vlll CONTENTS

Profiles, 129
Technology of Surveillance and Limits on Data Gathering, 13 l
Data Sharing Limits, 134
The Capsule View, 135

6 The Social Origins of Views on Privacy 137

Introduction, 137
The Penumbra of Fear, 137
A Question ofTrust, 140
How Did You Get My Name? 141
. Can They Really Do That? 143
There Ought to Be a Law! 144
Oppositional Tendencies, 147
Structuration, or the Social Origins of Social Opinion, 149
In Search of Social Explanations, 150
Summary, I 75

7 A Data Protection Regime 177

Need We Involve the Law at All? 177
Personal Privacy and Data Protection: ATangled Web, 178
The Heritage ofWarren and Brandeis, 182
A Question of Dignity and Other Intangible Injuries, 185
All ofThis and Democracy Too? 189
A Complex of Tnformational Rights, 189
Information Privacy and Corporate Speech, 193
Who's the Enemy Here? 194
Sensitive Information, 198
Reasonable Expectations and the Destructive Gales of
Technology, 201
Fighting Back, 203
The Marketplace Solution, 205
Public Policy and the Panoptic Sort, 209
Fairlnformation Practices, 221

8 Conclusion 227
And So to Conclude, 227
What Is to Be Done? 230

Notes 233
About the Book and Author 269
Index 271
TABLES AND FIGURES

Tables
3.1 Personal contributions to machine-readable, network-
linked data files 63

4.1 Corporatetendencies 112

4.2 Marketing techniques l 1'3
4.3 Actors seen as responsible for the escalation of the customer
information issue L14
4.4 Correlates ofbusiness practices 115
4.5 The association between organizational characteristics and
marketing techniques 117
4.6 Rotated factor matrix-business information orientations 118
4. 7 Multivariate analysis of variance factor scores with three
independent variables 120

6.1 Organizational rankings by trustworthiness 141

6.2 Orientation toward a legislative response 146
6.3 Social origins of orientations toward the panoptic sort 151
6.4 Correlations between explanatory factors 167
6.5 Multivariate influences on privacy orientations
(hierarchical regression) 167

Figures
6.1 Conservatism and age 158
6.2 Political cynicism 158
6.3 Trust in government by age 158
6.4 Trust in business by age 159
6.5 Extent of concern by age 159
6 .6 Failed to apply by age 159
6.7 Fearoftechnology byage 161
6.8 Trust index by age 162
6.9 Trust ofinsurers by age 162

ix
x TABLES AND FIGURES

6.10 Trust of data handlers by age 162

6.11 Trust, class, and exposure 173
6.12 Trust, class, and history 173
6.13 Trust, life cycle, and exposure 174
6.14 Trust, life cycle, and history 174
ACKNOWLEDGMENTS

Tftis project has been a long time in development. It began to pick up steam once I
decided to move from Howard University to the University of Pennsylvania.
George Gerbner is largely responsible for that move and for many of the good
things that have happened as a result. Among those things was an introduction to
Ken Laker, the head of Penn's computer science department and guiding force for
the Penn/AT&T Telecommunications Technology Program that provided three
years of support for my research. Ben Peterson's confidence in my potential led an
understandably cautious organization to eventually pair me with Emmanuel
Gardner, whose generous assistance led to my national telephone survey. With the
AT&T grant, and with support from the Annenberg School, I was privileged to
work with a continually ch~nging but always wonderful team of graduate research
assistants including Jerry Baber, Todd Kristel, Eleanor Novek, Catherine Preston,
Michael Schunck, Nikhil Sinha, Beth Van Horn, and Csilla Voros.
As I began to share my thoughts about privacy and the panoptic sort, I was
pleased to be included in the stream of comment and conversations about tech-
nology, rights, and responsibilities. Although we did not always agree, I value the
advice and counsel of Ed Baker, Mary Culnan, Janlori Goldman, Jim Katz, Gary
Marx, Robert Posch, Priscilla Regan, James Rule, Rohan Samarajiva, and Alan
Westin. As the final hours approached and it seemed as though I would never take
the next critical step, Kathleen Jamieson made an offer that I could not refuse.
Time off and an implied deadline made it both possible and necessary to trans-
form boxes into pages. Her support and encouragement are much appreciated.
I would also like to thank Herb Schiller and Gordon Massman whose persever-
ance put me in the capable hands of Shena Redmond, Sabina Vanish, and the
other good folks at Westv:iew who have helped to finish this project in good form
and in record time.
Finally, l would like to acknowledge the continued support and understanding
of my wife, Judy, whose gentle counsel and editorial insights were always on call.
For our daughter, Imani, who always wanted to know "how many pages is it?" I
can now say "ask your mom!"

Oscar H. Gandy, Jr.

xi
1
PROLOGUE

INTRODUCTION
In i934, the Spiegel corporation was an industry leader in the development of a
pointing system, which it used to evaluate applications for credit. Spiegel devel-
oped what it called the "vital question system,'' which gathered data in four critical
areas that were then used as the primary factors in the decision to grant credit. The
four questions were (1) amount of the order, (2) ocwpation of the applicant, (3)
marital status, and (4) race of the applicant. Other data gathered in the rating pro-
cess included an assessment of the importance of the geographic territory to the
overall marketing plan. 1 Although race and marital status are no longer legally per-
missible components ofthe credit authorization process, the evidence is dear that
a similar discriminatory process that sorts individuals on the basis of their esti-
mated value or worth has become even more important today and reaches into ev-
ery aspect of individuals lives in their roles as citizens, employ~es, and consumers.
I refer to this process as the "panoptic sort:' th:e all-seeing eye of the difference
machine that guides the global capitalist system. Kevin Robins and Frank Webster
have coined the phrase "cybernetic capitalism" tO underscore the nature of the to-
talizing system of social control that depends on the ability of state and corporate
bureaucracies to collect, process, and share massive amounts of personal informa-
tion to track, command, coordinate, and control each and every one of us to an
extent we would not have considered possible.2 Other descriptive terms appeared
over the years as I have gathered examples and insights about the nature of this
process. One that still holds an attraction is the notion of triage. The popular un-
derstanding of the term is that associated with medical decision making: "the
sorting and allocation of treatment to pati en-ts, especially battle and disaster vic-
tims, according to a system of priorities designed to maximize the number of sur-
vivors."3 The original meaning of the teri11,4 however, is detived from the French
trier, meaning to pick or to wll, but the word emerged into the English language
as having to do with the "grading of marketable produce," and more specifically,
referring to "the lowest grade of coffee benies, consisting ofbroken materia1:' 5 Al-
though some metaphors speak for themselves, l~t me be clear. I see t]1e panoptic
sort as a kind of high-tech, cybernetic triage th.rough which individuals and
groups of people are being sorted according to their presumed economic or polit-
 2 PROLOGUE

ical value. The poor, especially poor people of color, are increasingly being treated
as broken material or damaged goods to be discarded or sold at bargain prices to
scavengers in the marketplace.
This book is a project in critical theory. The purpose of critical theory is gener-
ally to understand the world in order-to change it, to make possible the realization
of some social ideal. Critical theory should then, ultimately, contain both an anal-
ysis and a prescription. This book can claim to meet only the first requirement. It
provides a critical assessment of the character and consequences of the develop-
ment and implementation of the panoptic sort. In my view, this sorting mecha-
nism cannot help but exacerbate the massive and destructive inequalities that
characterize the U.S. political economy as it moves forward into the information
age. It is a process that feeds on itself. Although there are already some signs of re-
sistance that have emerged in some quarters, the response of the panoptic system
is very much like that of a child's straw finger puzzle: Once you have placed your
fingers in either end of the tube, the more you stuggle to escape, the more it tight-
ens its grip. I would like to suggest that these inequalities are emerging in an area
that is critical to the maintenance of a democratic polity and to the operation of
an efficient market. These inequalities have to do with differential access to infor-
mation that is necessary for informed decision making.
The operation of the panoptic sort increases the ability of organized interests,
whether they are selling shoes, toothpaste, or political platforms, to identify, iso-
late, and communicate differentially with individuals in order to increase their in-
fluence over how consumers make selections among these options. At the same
time that the panoptic sort operates to increase the precision with which individu-
als are classified according to their perceived value in the marketplace and their
susceptibility to particular appeals, the commoditization of information increases
the dependence of these interests on subsidized information. 6 To the extent that
the panoptic sort, as an extension of technical rationalization into the social realm
of consumer and political behavior, depends on a ~eduction of the skills of indi-
viduals in the same way that automation reduces the skills oflaborers in the fac-
tory or the modern office, the market and the political or public sphere as we un-
- derstand them are transformed and are placed at risk. 7
As the panoptic sort matures and increases in scale and scope, a number of
contradictory developments seem likely. First, because the sorting mechanism
utilizes data about past behaviors, it tends.to limit the options that are presented
for individuals to choose. When the options concern choices about information,
this tendency has the potential to increase the knowledge and information gap be-
tween the haves and the have nots. Also, to the extent that these conservative mod-
els aim at the lowest common denominator, the panoptic sort will contribute to a
generalized lowering of the average level of public understanding. Second, be-
cause the sorting mechanism is based on· theoretical models that reflect quite
transitory fads or trends in social, economic, and political thought, increasing in-
stability in markets and political action wilJ become the rule rather than the ex-
PROLOGUE 3

ception. Third, as people's awareness of the panoptic machine grows, some will
find ways to resist, and others will attempt to withdraw. Both responses will invite
further attempts at inclusion and containment within the panoptic sphere. The
same technology that threatens the autonomy of the individual seems destined to
frustrate attempts to reestabHsh community and shared responsibility because it
destroys the essential components of trust and accountability.
Whereas Jurgen Habermas holds out hope for the liberatory potential of the
rationalist program, I argue that it is this very rationalism that is at the core of the
panoptic sort.8 The freedom that Habermas seeks and necessarily assumes for
communication to produce understanding, is consistently threatened by the stra-
tegic rationalism of this technology, which is designed to identify, classify, evalu-
ate, and assign individuals on the basis of a remote, invisible, automatic, and
comprehensive sensing of pe_rsonhood. What is left of the public sphere will be de-
scrioed as little more than a phantasm, a ghostly afterimage that will appear in
different forms to different individuals according to their profiles. The communi-
cations competence- that Habermas suggests is required before the ideal speech
situation can emerge- will be systematically denied to individuals and groups.9
Although communication may take place within more narrowly defined sociocul-
tural communities, the ability of people to engage in communication across these
lines will be lost as segmentation moves apace.

THE DIVERSE THEORETICAL ORIGINS

OF PANOPTIC THINKING
This book represents an effort to integrate the perspectives from several social sci-
ence disciplines. A political economy of personal information is at once set apart
from the mainstream focus and approach of economics. The narrow vision and
unrealistic assumptions of the neoclassical canon would make it impossible to
take note of the real influence of people's everyday lives, the influence of powerful
institutions, and, indeed, the presence of power itself. Randall Bartlett suggests
that "there is, as yet, no well-structured theory of power in economic relationships
that can be used to look for power," and although institutionalists and Marxists
have made power a central part of their analysis, he finds it either merely declared
or assumed, rather than defined or explored. 10 Multiple sources in sociology, eco-
nomics, and political science have contributed to the sense of disciplinary power
that is inherently a part of the operation of the panoptic sort.
These sources have helped to construct an understanding of power as a com-
plex relationship rather than as something to be weighed and measured. This
power is seen as a dynamic relationship, one that changes as it changes those who
are part of the relationships that define it. Although there are any number of rela-
tionships that might be described in terms of the relative inequalities that obtain,
my primary focus is on the inequality that obtains between the individual and the
bureaucratic organizations ofbusiness and government interests. And within this
4 PROWGUE

nexus, the focus will be sharper still on the asymmetry between individuals and
busjness establishments.

From Marx, a Dialectical View

The influence of Karl Marx on my thinking about social change is difficult to
measure, but there is little doubt that it is pervasive. In part, this difficulty is a re-
flection of the e.''tent to which other writers who have influenced my thinking
have not recognized or have failed to acknowledge Marx's influence on their own
sociology, philosophy, history, economics, and yes, even psychology.H Thus,
much of what these writers have contributed to my understanding of technology
and social change has been enhanced by Marx's social theory, but the credit lines
have become blurred. Ideology i-s thought to have been a victim of its own success.
The choosing ofsides within Marxist social theory has transformed this core con-
struct into a contested battle site where each contending group seeks to stdp off
the distasteful or vulgar aspects of its derived meaning and keep only those that
justify the study of a particular social practice. 12 Because argume11ts about relative
autonomy and determination have challenged the traditional force with which the
terms economic base and superstructure might be offered as an anchor for a mate-
rialist theory of social change, the fonnal concepts have also fallen into relative
disuse. 13 It is clear, however, that the same conceptual elements in less obvious
Marxian garb are at the heart of much contemporary theorizing about an Infor-
mation Age.
We are regularly chaUenged to examine the relationships between changes in
technology, which occasion, or are necessitated by, changes in the stTucture of the
market, the regulatory infrastructure, or the perceptions o f consumers. New de-
vices that facilitate production, distribution, display, or reproductjon of mass me-
dia products nearly always tlueaten the equilibrium of the marketplace, and those
actors most at risk from competitive alternatives are first in lioe to demand regu-
latory protection-in the public interest, of course. 14 Anne Br anscomb under-
scored the impact of technological change on the ability of individuals and corpo-
rations to exercise what they have come to see as their rights to information. 15
Changes in technology make it easier for relatively unsophisticated users to make
high-quality copies of information, thereby transforming private resources into
public goods. The call by media firms for copyright enforcement falls on deaf ears
when it is recognized that the general public believes that making copies of televi-
sion programs for later viewing or personal tapes offavorite songs for use in por-
table or automobile tape players is a perfectly legitimate use. Changes in technol-
ogy heighten our awareness of the strictures that old laws, old contracts, and old
customs represent for tlte exploration of me newly possible. The real possibility of
genetic engineering of humans, in addition to rodents and plants, raises an ethical
storm. 16
Jn the modern discourse of the science policy debate, the language of Marx ap-
pears stodgy and old-fashioned. Why speak of " fetters" when the linguisti~ cur-
PROLOGUE 5

rency demands talk about hyperghettoization am! the incontestability of markets?

However, the dynamic tension between the technological means of production
and the traditional or contractual relationships through which goods are finally
exchanged against capital, revenue, and wages continues to be played out in the
superstructural realm of the "legal, political, religious, aesthetic or philosophi-
cal -in short. ideological, forms in which men become conscious of this conflict
and fight it out." 17
The debates within Marxism over the nature of nonproductive labor bring me
closest to the root of my search for a political economy of personal information. 18
It is Marx's labor theory of value and through that his explication of the core con-
cept of surplus value and capitalist exploitation that serve as the critical distinc-
tiQn between a Marxian political economy and the mainstream neoclassical or-
thodoxy.19 Although Joseph Schumpeter was generous in his praise of Marx's
social theory, including much of his economics, he suggested that Marx need not
have placed all his bets on a concept as weak as that of surplus value. 20 We even
find a curious revisionism in the writing of Arun Bose, whose assessment of mod-
ern Marxian political economy includes the claim that

Marx himselfhad doubts and misgivings about this labour value approach, from the
beginning to the end of his career as a political economist. A few years before his
death in 1883 (in i875 to be precise), he finally called for its rejection for mystically
and "falsely ascribing supernatural creative power to labour." Jn other words, Marx
renounced the labour value approach because it was based on a mystique about Ja-
bour which was as unacceptable to him as the mystique about capital inherent ia the
trinity formula.2 1

Dan Schiller suggests that this association between productive labor, which is
defined as labor that produces surplus value for the capitalist, and nonproductive
labor, much of which may be expended in marketing and other activities neces-
sary for the realization of surplus value, stands in the way of the development of a
Marxian analysis of the information economy.22 To the extent that those who are
involved in the design and administration of the panoptic sort are nonproductive
by traditional definition, their growing importance for capitalist economic devel-
opment mµst be realized and their status as "peripheral phenomena" must be
changed.

From Ellul, a Perspective

on the Technological Imperative
The panoptic sort is a technological system. Its rationale is techriicist. To some de-
gree, concern about the panoptic sort reflects a reticence to embrace technology.
It is a posture that is well founded. Jacques Ellul stands as the most visible exem-
plar of those who see technology as guaranteeing a dystopic future. Ellul's is an
untempered pessimistic view. His view of technology is a totalizing model that
6 PROLOGUE

evokes a sense of resignation and hopelessness, although that is not his purpose.
Those who do not share his view of technology are seen as simply misinformed: ·

It has not yet been appreciated that this entry of technology means control over aU
the persons involved, all the powers, aU the decisions and changes, and that technol-
ogy imposes its own law on the different social organizations, disturbing fundamen-
tally what is thought to be permanent ... , and making politics totally futile .... No
decisions can be made that run contrary to technological growth. All decisions are
dictated by the necessity of technological development. Nothing else matters. 23

In Ellul's view, it is not only the physical world and the social institutions that
must adapt to the demands of the technological imperative, even human nature
itself must ultimately be transformed. He suggests, for example, that

genetic manipulation is designed to produce exactly the type of people that we need .
. . . From birth, individuals are to be adapted specially to perform various services in
society. They are to be so perfectly adapted physiologically that there will be no mal-
adjustment, no revolt, no looking elsewhere. The combination of genetic makeup
and educational specialization will make people adequate to fulfill their technologi-
cal functions. 24

This chilling vision of the future does not promise an egalitarian transformation
by which perfection is the lot of all,25 but reveals a landscape of shameful inequal-
ity of the sort that William Julius Wilson26 has described with.many contemporary
examples:

On the one hand there will be a kind of aristocracy marked by its total and infallible
adaption to technical gadgets and the technological system, and on the other hand
there will be a vast number of people who are outdated, who cannot use·th~ technol-
ogy, who are powerless, who are still at the social stage but who live in a technologi-
cal environment for which they are totally unprepared. 27

Ellul is frequently criticized for a tendency to represent technology as an au-

tonomous force, one that is guided by the goal of efficiency. 28 Robert K. Merton ,
in his Foreword to the Knopf edition of Ellul's The Technological Society, makes
this autonomous trajectory clear: "The essential point, according to Ellul, is that
technique produces all this without plan; no one wills it or arranges that it be so.
Our technical civilization does not result from a Machiavellian scheme. It is a re-
sponse to the 'laws of development' of .technique." 29 Technologies seem to de-
mand, because they require for their own efficient operation, the devdopment of
ancillary or support technologies. The technological imperative then generates
something akin to an inflationary spiral pr a positive feedback, deviatio'n amplify-
ing loop, with each response cycle increasing the level of technological sophistica-
tion. Although the process may be dialectical, Ellul's system seems to be devoid of
 PROWGUE 7

any conceptual role for human agency, beyond those who serve rather than direct
the technological wave. 30 Joseph Schwnpeter shares Ellul's sense of the impor-
tance of technological development and offers a similarly dialectical vision of dy-
namic change. However, Schumpeter's "gales of creative destruction" are not au-
tonomous but are the result of entreprenurial investment in product and process
innovations valued for their potential contribution to the capitalist's "bottom
. ,,31 .
1me.

From Weber, an Understanding

ofRationalization
Although Max Weber's contribution to modern social science is extensive and is
enjoying a sort of rebirth, not unlike that which had been enjoyed by Marx, two
aspects of his theoretical project have emerged as central to my analysis of the
panoptic sort. The first aspect is Weber's development of the theme of rationality
as a characteristic of Western industrialization, and the second is his exemplifica-
tion o( that theme in the rationalization of human activity through the mecha-
nism of bureaucratization. The dominant theme of the contemporary discourse
on modern society is one of efficiency, but efficiency is only one aspect of Weber's
construct. And more important, perhaps, the current elevation of efficiency to the
level of a moral imperative ignores Weber's warnings about the negative conse-
quences of rationalization. "Jn no sphere of life according to Weber, has rational-
iza'tion unambiguously adv~nced human well-being. The rationalization of eco-
nomic production, for example, has created the 'iron cage' of capitalism, a
'tremendous cosmos' that constrains individuals from without, determining their
lives 'with irresistible force'. "3 2 This rationalization, especially the formal instru-
mental rationality, assumes that everything is subject to measmement and calcu-
lation. Substantively rational goals and purposes are pursued in markets, where
the transactions are guided by reliable calculations of demand. Bureaucratization
ensures calculability by organizing the factors of production in ways that guaran-
tee their centralized control.
Perhaps we might see Weber's emphasis on the influence of bureaucracy as re-
. fleeting a similar degree of comfort with the notion of an autonomous, self-di-
rected technology. For Weber as for Ellul, those associated with bureaucratization
act to ensure its continued growth. Weber expressed the sense that "bureaucracy
had an inherent tendency to exceed its instrumental function, and to become a
separate force within society, capable of influencing the goals and character of that
society. It constituted a separate power group within the state, a separate status
stratum within society at large." 33 Bureaucracy is an organizational technology,
and the spread of this technology is seen as inevitable. According to Reinhard
Bendix, we must find in Weber's work a view that "a system of bureaucratic rule is
inescapable.... Universal bureaucratization was for Weber the symbol of a cul-
tural transformation that would affect all phases of modern society. If this devel-
opment ran its full course it would result in a new despotism more rigid even than
8 PROLOGUE

the ancient Egyptian dynasties because it would have a technically efficient ad-
ministration at its disposal." 3 ~
The somewhat tempered pessimism of Weber may be based in part on his ap-
preciation of the potential for meaningful action on the part of individuals. Like
Marx, Weber believed that history is the product of individual action, even if the
designs and circumstances are not of any actor's choosing. David Beetham's ex-
ploration of Weber's political activism reflected the similarity in the analyses of
Marx and Weber but aJso noted the differences. "Both Marx and Weber, therefore,
recognized the same power relationships, the same structure of power, in modern
society; where they differed was the point at which they sought to apply the level
of political action to this structure."35 Weber recognized, however, that the politi-
cal opposition would not really be able to resist the siren call of bureaucratization
but would be required, if it hoped to succeed, to develop its own "coun terbureau-
cracy." Thus for Weber, unlike Marx, socialism offered no promise of escape, but
rather an even more extensive bureaucracy. We find signs of this Weberian criti-
cism in Frank Webster and Kevin Robins's exploration of various responses by the
Left to information technology and its promise for socialism. With few differences
among writers on the Left they noted. "a remarkable consensus as to how it should
be approached: neutral, inevitably to be accepted, and, more or less, determin-
ing."36 For them, this view sees the scientific and technological revolution as
something to which people must learn to adapt and perhaps to enjoy under the
guidance of a beneficent state. I find dear parallels in the disclaimers offered by
representatives of progressive political action groups who invest in and use the
same manipulative and exclusionary technologies of segmentation and targeting
as those used by reactionary c.onservative groups because they believe that with-
out these technologies they must surely lose.37
A critical aspect of Weber's view of bureaucracy was his association of power
with knowledge-the knowledge of official statistics, which, when linked ro the
science of mathematics, provided an additional margin of legitimacy. Christopher
Dandeker emphasized the surveillance aspect of bureaucratic administrative
power and underscored the importance of controlled access to the files developed
by the bureaucratic organization in ways that ensure "that administrative reason-
ing can in principle be understood and replicated by anyone with access to the
same information." 38
Weber's economics is valued for its recognition of the existence and power of
institutions and for his criticism of marginal utility analyses, which depend on
unrealistic assumpti.o ns about the distribution of that power. For Weber, con-
sumer sovereignty is denied by the .existence of power inequities, including the
power of producers to influence consumers' preference functions. It is suggested
that Weber "rejected any view that market relations within a capitalist society
were or could be 'free: in the sense that they dispensed with uneqLial power rela-
tions."39 Weber's assessment of the rationality of the legal system also took note of
PROWGUE 9

the persistence of inequality. From Rogers Brubaker's reading, we see that Weber
understood that

formal freedom of contract does not guarantee that everyone will be equally able to
stipulate the terms of contractual agreements, for legally protected inequalities in
the distribution of property generate inequalities in bargaining power. Freedom of
contrae,"t enables an economically powerful employer, for example, to "impose his
terms" on the worker who is constrained to accept them by his "more pressing eco-
nomic ~eed." 40

This is what Is meant by the claim that individuals are largely "contract term
takers" in the bulk of their economic relations with organizations, especially with
regard to those in which personal information is required before the transaction
can move forward.41 Weber's distinction between formal and substantive rational-
ity finds expression in this area as well. T he formally rational pursuit of capitalist
goals is seen to generate greater substantively irrational inequality and scorn. In-
deed as Brubaker suggests, to the extent that it is "characterized by a high degree
of formal rationality, the modern capitalist economic order maximizes the value
of calculability, efficie~cy and impersonality b ut it is deeply inhospitable to egali-
tarian, fraternal and caritative values:' because formal rationality favors economi-
cally powerful groups.42 Bartlett's generously detailed exemplifications of the
broad range of inequalities in the relations between consumers and employers,
producers, and government officials largely finishes the critique that Weber's
analyses had begun. 43

From Foucault, the Underlying Concept ofPanopticism

The influence of Michel Foucault o~ my work is so substantial that it threatens to
dominate the construction of my arguments about power and social control. It is
from Foucault that I derive the underlying concept of panopticism as a technol-
ogy of power realized through the practice of d isciplinary classification and sur-
veillance referred to as the panoptic sort.44 The Panopticon is the name given by
Jeremy Bentham to the design for a prison that would facilitate the efficient obser-
vation or surveillance of prisoners by guards or supervisors who might periodi-
cally occupy a central tower. The major effect of the design was to create a "state of
conscious and permanent visibility that assures the automatic functioning of
power." 45 In Foucault's view, the panoptic design need not be limited to prisons
but would.apply equally well to other institutions that share a discipUnary, educa-
tional, or rehabilitative purpose. The panoptic technology was not limited to sur-
veillance alone but included the classification and isolation of subjects by category
or type. Once divided, a panoptic structure could be used as a "laboratory; it
could be used as a machine to carry out experiments, to alter behavior, to train or
correct individuals. To experiment :with medicines and monitor their effects. To
try out different p1:Jnishments on prisoners, according to their crimes and charac-
10 PROLOGUE

ter, and to seek the most efficient ones." 46 Thus, panopticism serves as a powerful
metaphorical resource for representing the contemporary technology of segmen-
tation and targetillg, which involves surveillance of consumers, their isolation
into classes and categories, and their use in market tests that have the character of
experiments. 47 Kevin Robins and Frank Webster are expUcit in their suggestion
that the new communications and information technologies proyjde "the same
dissemination of power and control, but freed from the architectural constraints
of Bentham's stone and brick prototype. On the basis of the 'information revolu-
tion,' not just the prison or factory, but the social totaHty, comes to function as the
hierarchical and disciplinary Panoptic machine." 48
More generally, Foucault is to be credited with emphasizing the linkage, in-
deed, the inseparability of power and knowledge. An important focus of his his-
torical work was to de~cribe those forms of power/knowledge that are bound up
in the classificatory activities Foucault refers to as "dividing practices," which
identify and then separate the deviant, the diseased, or the dissenters. 49 "Essen-
tially, 'dividing practices' are modes of manipulation that combine the mediation
of a science (or pseudo-science) and the practice of exclusion-usual1y in a spatial
sense, but always in a social one."50 The process of the "objectification of the sub-
ject:' which is implied by the dividing practices, finds an important match in the
more subjective forms through which the individllal actively participates in trans-
forming himself or herself into a disciplinary subject. From my perspective on
marketing, I see this subjectification in consumers' identification of themselves
with products and brands, as well as in standards of consumption that marketers
have defined for "yuppies" or the "thirty-something'' generation. The distin-
guishable character of popular dress is the prominent display of the nanie of a par-
ticular brand of garment, car, or beverage preferred by the "in" group. Mary
Douglas suggests in this regard that "as fast as new medical categories (hitherto
unimagined) were invented, or new criminal or sexual or moral categories, new
kinds of people spontaneously came forward in bordes to accept the lab~ls and to
live accordingly. The responsiveness to new labels snggests extraordinary readi-
ness to fall into new slots and to let selfhood be redefined." 5 1 This process of "ac-
tive self-formation" is one that Foucault described as being mediated by some ex-
ternal authority figure. ln the contemporary stage of this process, such authority
figures may be selected and defined instrumentally but reinforced by an autono-
mous cadre of cultural e:>..i>erts whom we rely on to telJ us "what's in, and what's
not!"
This process of objectification involves yet another aspect of the power/ knowl-
edge relationship that informs my critical project. Foucault js credited with
adding the concept ofnormalization to that ofefficiency as the goaJ of rationaliza-
tion. "By 'normalization,' Foucault means a system of finely graded and measur-
able intervals in which individuals can be distributed around a norm-a. norm
which both organizes and is the result of this controlled distribution.'' 52 T heim-
portance of statistical measures in the definition of what is normal serves to pull
PROLOGUE 11

the logic of the law away from ~trict definitions of right and wrong. We find in the
process of normalization the emergence of the empirically derived standard of the
"public's interests" replacing a more ethically derived definition of the public in-
terest.53 This process of normalization may also be seen in the increasing central-
ity ofeconomic efficiency considerations in the ruling ofjudges and courts, as well
as the emergence of risk analysis as a statistical basis for the offer of contractual
services.54
This process o( normalization contains an inherent contradiction that is evi-
dent in the increasing fractionalization of markets. Just as specialists emerge to
discipline and correct new classes of deviants, another branch of the same techno-
structure identifies new forms of perversion and new classes of deviants. As the
ability to measure and classify expands, so, almost automatically, do the emergent
classes come to be filled with cases demanding attention. According to one ob-
server, "this is what Foucault meant when reminding us that power is not just a
force which excludes and says 'No,' but a form of creation."55 Having heard and
understood Foucault, Stanley Cohen then turns away to issue his own plea: "Some
ways must be found to halt the seemingly inexorable process by which society
keeps classifying, controlling, excluding more and more groups according to age,
sex, race, behavior, moral status, ability or psychic state."56 As will be suggested in
the pages that follow, success appea_rs nowhere on the horizon.

From Giddens, a Perspective on Complexity

Contemporary theoretical debates within the social sciences reflect the ebb and
flow of fashion as well as insight. In opposition to the heavy-handed theories of
social domination that characterize much of the critical theory associated with the
Frankfurt School before attempts at its rehabilitation by Jiirgen Habermas, we
find a growing number of neo-Althusserians who suggest that existence and
change in social systems are the product of a complex of influences that almost de-
fies description.57 This claim of complexity is referred to as "overdetermination,"
in the sense that entities are mutually constitutive. In the most extreme or radical
view, this mutuality means that it is impossible to identify any cause or influence
as being more important than any other. This "antiessentialist" posture leaves
theorists free to choose an "entry point" from which to begin the ongoing process
of describing the complex relations between entities.58
Anthony Giddens's theory of structuration provides a useful analytical frame-
work through which one might begin to examine the relations between entities in
a way that does not deny the possibility of assessing the nature of influence or
power within those relations. 59 Giddens's contribution to this project is quite sub-
stantial, as I wiU attempt to make dear in this brief introduction.
The concept of structuration implies a dynamic rather than a static vision of
influence. Although there is well-deserved criticism of the haziness of Giddens's
explication of the concept of structutation, it is more than suggestive of a way to
understand the production and reproduction of a given social order as a recursive
r

12 PROLOGUE

process through which knowledgeable actors, in pursuit of goals, interact with

other actors and institutions in ways tnat reproduce or reconstitute both the ac-
tors and the relations. 60 Giddens's emphasis on the knowledgeable actor's under-
standing of her circumstances and her interests represents an important distinc-
tion betw'een structuration and the dominant versions of Antonio Gramsci's
notion ofbegemony. 61 Hegemony describes a circumstance or status in which a
dominant, or hegemonic, view of reality comes to be diffused throughout a soci-
ety. This hegemonic ideology is that which George Gerbner et al. finds in the
"mainstream;• toward which people are moved through a process of"cuJtivation:'
primarily attributed to television as the dominant cultural form.62 This eq_uilib-
rium status is arrived at through a process ofleadership and negotiated consensus,
rather than through more coercive forms of social control. However, the privi-
leged position that Giddens assigns to the agency of individuals stops well in ad-
vance of the extreme positions taken by some cultural theorists who deny influ-
ence all together. Indeed, Giddens's definitions ofknowledgeability invites efforts
to describe the sorts of limiting constraints that other actors' resources may repre-
sent.
Part of Giddens's contribution to our understanding of the influence of struc-
ture is his identification of the "rules and resources" that are implemented
through the routine interactions that make up the daily Lives of his agents. Indi-
viduals come to understand the rules they have to follow when they interact with
other actors and institutions, and they also come to understa1:id and appre.ciate ·
the differences in the resources that they can bring to bear in these interactio~s.
Thus, we see the exercise of power and influence in the context of what people ac-
cept as legitimate or appropriate demands and sanctions. Actors are Limited by the
contemporary rules and resources that govern a variety of interactions between
them. Some of these rules and resources develop such a stability and reach that
they rise to the level of institutions or social systems.
Giddens's elaboration of the theory of structuration involves his expansion of
relations into the dimensions of.time and space. There are important differences
between interactions that take place between individuals face-to-face in real time
and those that take place across vast distances through the means of advanced
communications technology. The absence of authority in space and time is no
longer a limit on the ability of agents in power to coordinate the system. Modem
communication technology provides a means of remote control.
With Giddens's notion of distanciation, it becomes possible to th~orize the in-
teraction between individuals and figures of authority who are not only not pres-
ent but who are conceivably "interacting" with hundreds, even millions, of others
at the same time. The testimonial advertisement in a prime time television pro-
gram places youngsters in an interaction with an established cultural hero who
develops or is assigned authority in a related area as he or she declares how much
he or she likes or depends on a particular brand of shoes. A somewhat different
form of "distanced" authority can be seen in the "authorization code;' which is
PROLOGUE l3

received from a transaction processing organization handling a vendor's point of

purchase request for a charge against a customer's credit line. This distanced in-
teraction differs from, but has much in common with, the remote monitoring of a
telemarketer's processing of an inquiry to an 800 number. Giddens is rather ex-
plicit in his analysis of the power inherent in this remote surveillance capacity of
communications systems. 63

Administrative power now increasingly enters into the minutiae of daily life and the
most intimate of personal actions and relationships. In an age more and more in-
vaded by electronic modes of storage, collation and dissemination of information,
the possibility of accumulating information relevant to the practice of government is
almost endless. Control of information can be directly integrated with the supervi-
sion of conduct in such a way as to produce a high concentration of state power. 64

Of course, it is not necessary to limit one's examination of structured relation-

ships and interactions to those between citizens and the state. Indeed, the primary
emphasis in this book will be on relations between individuals and the corporate
private sector.
With the knowledgeable actor as a focal point for an analysis of the emerging
institutionalization of the panoptic sort, Giddens's critical project suggests a
number of key empirical targets: (1) practical consciousness and the relations be-
tween personality and social perceptions, attitudes, and opinions; (2) routiniza-
tion and the relation between repetitive interactions with bureaucratic organiza-
tions that collect and claim rights in personal information; (3) social positioning,
through which different social roles are adopted, assigned, or denied on the basis
of computer-aided assessments; and (4) the ever-ex-panding settings and locales in
which interactions and relationships become defined and structured. Each of
these will be examined in some detail in the pages that follow.
 2
INFORMATION AND POWER

INTRODUCTION
The panoptic sort is the name I have assigned to the complex technology rhat in-
volves the collection, processing, and sharing of information about individuals
and groups that is generated through their daily lives as citizens, employees, and
consumers and is used to coordinate and control their access to tl1e goods and ser-
vices that define life in the modern capitalist economy.
The panoptic sort is a system ofdisciplinary surveillance that is widespread but
continues to expand its reach. The operation of the panoptic system is guided by a
generalized concern with rationalization of social, economic, and political sys-
tems. The panoptic sort is a difference machine that sorts individuals into catego-
ries and dasses on the basis of routine measurements. ft is a discriminatory tech-
nology that allocates options and opportunitjes on the basis of those measures
and the administrative models that they inform. The panoptic sort has been insti-
tutionalized. It is standard operating procedure. It is expected. It has its place. Its
operation is even required by law. 1 And where it is not, people call out for its in-
stallation. Its work is never done. Ead1 use generates new uses. Each application
justifies another. lt is efficient, having largely been automated. Like a voice-
activated recorder, it moves into action solely in response to an action by the ob-
ject of its control. The panoptic sort is a system of actions that governs other ac~
tions. 2 The pan optic sort is a system of power.

Identification
The panoptic sort can be understood to involve three integrated functions or pro-
cesses: identification, classification, and assessment. Although its operation is by
no means limited to identifiable individuals, it depends to a large part on the abil-
ity ofits users to reliably identify the objects to be controlled. The identification
will never move to the level of personhood as we may understand the person as the
subject of religion, philosophy, and ideaJized systems of justice. The attention of
the panoptic Sort moves only to levels of identification that have administrative
and instrumental relevance. Here we refer to the identification of persons with
histories, records, and resources when those persons or agents of those persons
present a card, fonn, signature, claim, or response, or when they present them-
16 INFORMATION AND POWER

selves at a particular place or time. Identification is associated with authorization

and authentication of claims. Identification is associated with the assumption of
responsibility for actions, transactions, interactions, and reactions, which may be
recorded by the panoptic system.
The level of identification required by the panoptic system is indicated by the
importance of the transaction that is about to take place. As the level of risk in-
creases, more sophisticated technologies are called into play. The signature gives
way to the physicaJ description, which gives way to the photograph, which gives
way to the fingerprint, or the voice prin t, or the retinal scan. But that is not
enough.
The pan optic sort frequently requires third-party validation. You may be who
you say you are, but we need verification that you are what you say you are. Are
you old enough? Are you, perhaps, too old? Are you trained and certified; is your
license currently in force? Are, you creditworthy, reliable, stable, honest, entitled?
Are you one of us? This form of identification, more often than not, involves some
form of classification.

Classification
Classification involves the assignment of individuals to conceptual groups on the
basis of identifying information. Class membership is based on measurement of
one or more attributes of an individual's identifying array of attributes. As we
have suggested with regard to identification, the data matrix may be infinitely
complex, depending on the requirements and resources of the panoptic system
brought into play at any given point of interaction. The identification of class
membership will always be made on the basis of less information than is at hand
or is readily available. As we have suggested with regard to preprocessing, infor-
mation is thrown away so that more efficient means of control may be put in
place.
In discussing the process of classification in the emergence of natural history as
a scientific discipline, Foucault suggested that "it reduces the whole area of the
visible to a system of variables all of whose values can be designated, if not by a
quantity, at least by a perfectly clear and always finite description. It is therefore
possible to establish the system of identities and the order of differences between
natural entities." 3 Of course, as Foucault later reminded us, not all classification
and differentiation were limited to the visible. Important transformations in the
classificatory enterprise involved making connections between the seen and the
unseen, drawing inferences about deep structure from surface appearance. Esti-
mates of honesty, based on the movement of pens on moving paper or pencils
checking boxes on lengthy questionnaires, have become commonplace, and un-
easiness about the accuracy, reliability, and validity of these measures does little to
restrict their use.
The panoptic sort institutionalizes bias because the blind spots in its visual
field are compensated for by a common tendency to fill in the missing witli the fa-
INFORMATION AND POWER 17

miliar or with that which is expected. When the paradigmatic vision of the
pan optic machine is linked with the futures ofbureaucratic organizations and the
individuals who stand at their helms, the incentive to find precisely what has been
predicted is often too powerful to resist. A disciplinary profession that depends on
treating a particular kind of problem has every incentive to calibrate its instru-
ments to find ever more cases of the dysfunction that are in need of expert atten-
tion. The discovery of epidemics is very difficult to separate from the interests of
the agencies whose responsibility it is to keep them under control. Bureaucratic
records reflect the local custom. The definition of what is a crime depends as
much on the social status of the perpetrator and the victim as it does on the ac-
tions that have allegedly taken place. Troy Duster notes that "if one looks at the rec-
ord in 250 years of U.S. history, no white man ever commi't ted the crime of rape on
a black woman in twelve southern states."4 Is this a statement about the violent
sexual behavior ofsouthern white males, or is it a statement about a racially biased
system of classification? There are no objective standards; classification always in-
cludes an assessment, whether expressed or not.

Assessment
Assessment represents a particular form ofcomparative classification. Individuals
are compared with others. Individuals are compared with hundreds and thou-
sands of others whose measured attributes help to establish norms and the
bounds of reasonableness and acceptability. Assessment involves the use of stan-
dards and assumptions about the normality and the independence of distribu-
tions. Social distributions are often highly skewed rather than normal. Social dis-
tributions are often highly correlated, because they share a common cause, rather
than being independent.
Once classification has occurred, assessment frequently involves the examina-
tion of probabilities- that is, the likelihood that a person will act, react, or inter-
act in a particular way to a situation or circumstance. Individuals may be classed
and evaluated on the basis of the responses they give, as well as on estimates of
how likely similar responses are to occur. As with the requirement for precision in
identification, the demand for precision in. assessment is based on an assessment
of the consequentiality of error. Given the tendency of humans to be risk averse,
the privileging of avoidance over gain is not without a basis in fact. 5
The panoptic sort is more concerned with the avoidance of loss than with the
realization of gain. Although on the face of it this statement may seem to be in
conflict with assumptions about profit maximization, if we recognize that cost re-
duction and the avoidance ofloss are what make the realization of profit possible,
the emphasis is not so far afield. Yet the claim made here is meant to be as provoc-
ative as it sounds. The panoptic sort is primarily a defensive technology. It oper-
ates through victimization, through avoidance. Although marketing targets are
eventually identified and selected, these targets are the individuals who remain on
the list after the high risks and the sure losers have been eliminated from the pool.
18 INFORMATION AND POWER

The panoptic sort is a screen that excludes, a filter that blocks, a magnet that ig-
nores fine wood in preference for base metals. The sorting process works primar-
ily by eliminating those who are too much, too little, too late ... too bad! By estab-
lishing a criterion, such as those in the 95tb percentile and above, the panoptic
sort requires that you discard 95 pecent of the population. Th<1t is unequivocally
what it means to sort and then choose.
The panoptic sort victimizes because it decontextualizes. Status is divorced
from circumstance. The circumstance cannot be recaptured; an assessment will
always be incomplete. However, the ways in which context is misrepresented are
not randomly distributed but reflect an institutionalized bias; a bias established
by race, gender, age, class, culture, and consciousness.
Just as capitalism as a form of social organization has neither fully matured nor
been extended to the same degree in all areas of social existence in even the most
advanced industrial societies, the spread of panoptic technology is uneven and in-
complete. This chapter is organized so as to provide a vantage point from which to
observe the development and spread of the panoptic sort as an institutionalized
system of power.

A QUESTION OF POWER
Randall Bartlett offers a definition of power that may serve us well as we venture
into battle with those who would ignore the role that information plays in its use.
He defines power as "the ability of one actor to alter the decisions made and/or
welfare experienced by aJ1other actor relative to the choices that would have been
made and/or welfare that would have been experienced had the first actor not ex-
isted or acted."6 Defined in this way, power is a relative measure. All actors may be
seen to have some power. The importance of the question is based in the desire to
determine, or to demonstrate, which is perhaps more clearly the intention of this
work, that the power of individuals is frequently overwhelmed by the power of bu-
reaucratic organizations.
Bartlett identifies several forms of power that are useful io characterizing our
response to the various manifestations and influences of the panoptic sort.
Ungranted event power describes a circumstance in which the other creates· a situ-
ation, or a situation may be created by accident or by an unrelated superior force,
in which an iJwjdidual's options are limited. What economists refer ·10 as external-
ities are seen to be exemplary forms of ungranted eve11t power. The telemarketing
call that disturbs your train of thought is an e.xternality. It is an objective reduc-
tion in your welfare for which you are unable to claim compensation. The person
who is able to make such a call has ungranted event power. An organization with a
computerized telemarketing operation is able to exercise this power simultane-
ously on hundreds of individuals. This power has a particular irony jf tbe call is
from the local Bell operating company offering you caller identification service,
which would protect you against just this sort of unwanted telephone harassment.
fNFORMATION AND POWER 19

Bartlett also includes in this category of ungranted event power the influence
on welfare that might flow from a lobbyist's modification of the regime of rights
that affects the welfare of thousands of individuals. Few if any of the persons af-
fected by this modification of their rights willingly consented to this change.
Agenda power is related to this, in that some actors may influence how other ac-
tors understand the options that are available to them. As this understanding in-
fluences tl1e choices of the latter, it may represent substantial power. Finally, in the
same vein, Bartlett identifies what he calls value power-the ability to influence
the subjective evaluation of events. For example, in combination with ungranted
event power, "If I can actually 'make you like' an item that you previously did not,
I have exercised power of an extraordinary kind:' 7
Amitai Etzioni's discussion of value power adds the dimension of social or
moral value as an additional force beyond the economic evaluations of worth
common to the neoclassical economic paradigm.8 And, as the moral dimension is
influenced by the dominant values of the society, oppositional stances are less
powerful in that it would take a greater investment to change the average value po-
sition. This is an issue to which we will return because of this critical claim that in-
dividual preferences are not given. Instead, preferences aJe learned, indeed taught,
and more often than not, reflect structures of domination. 9
As we explore the political economy of personal information, the relative
power of individuals in comparison with that of institutions and organizations
becomes highly relevant. As Bartlett suggests, "Whenever knowledge is a scarce
good, it confers power on its possessors." 10 But the power that the individual is
able to exercise over the organization when she withholds personal information is
almost always insignificant in comparison with the power brought to bear when
the organization chooses to withhold goods or services unless the information is
provided. This inequality is inherent in the concept of a market in which there is a
large number of sellers facing a large number of buyers. Theoretically, none has
power over the other because there is always another place to buy or there is al-
ways another person to sell to. In reality, however, economic transactions always
take place in the context of substantial inequality. 'o\'hen sellers are organizations,
they are few, rather than many, and they have acted through marketing to create
the impression that the options are fewer still.
Samuel Bowles and Herbert Gintis join Bartlett in identifying similar
inequalities in power that characterize the relation of labor to capital. 11 Bowles
and Gintis suggest that the "effectiveness of capitalist command over workers de-
pends ultimately on the capitalist's power to terminate the labor contract:' 12 They
suggest that capital is different from labor in that it is able to move more freely.
Unlike labor, which is embodied in people, capital can move across the globe in-
stantaneously at the push of a button. 13 To the extent that information is capital,
its ability to move is unprecedented. This heightened mobility increases the power
of capital relative to that oflabor. The laborer, on his own, generally does not per-
ceive the option of voluntary unemployment or the breaking of an unsatisfactory
20 INFORi'vlATION AND POWER

contract as a viable option. The theoretical labor market, in which there are no
costs associated with the negotiation of new labor contracts, does not ex.ist. There
are substantial costs that flow from laborers who quit, as well as from those who
are fued. More important, in terms of the panoptic sort,

to the extent that being fired is a signal that G must carry with her into the market,
they may reduce the quality of jobs made available. If she is able to hide the partim-
lar job experience, she is also hiding the positive experience aspects of that position,
and pe.tbap~ raising questions about the hole in her employment record. C's ability
to fire is an ability to return G to the labor market with a published statement of her
incompetence. That is a potential ability to alter the agenda of job opportunities
available to Gin the future. 14

Thus, employers have substantial power over labor, which limits the worker's op-
tions regarding the labor contract. On the one hand, the actions of an employer,
without the consent of the employee, can change the present and future welfare of
any worker by influencing the record that follows ru1 employee throughout work-
ing life. A chain reaction may also set in. A period of unemployment may cause an
arrears in the payment of bills, which will adversely affect a worker's credit rati11g.
As access to a credit report is considered to be a legitimate business need for an
employment decision, a poor credit report may thus mean that future employ-
ment possibilities are also limited. This is substantial ungranted event power.
Awareness of these consequences makes the average employee highly cautious
about testing the limits of her power. On the other hand, unless they are barred by
the presence of a strong union or by the threat of government action, capital
strikes or reductions in force are con;unonplace and are initiated without apparent
risk to the owners of the firm.
Although it will not be developed here to the extent it deserves, there is an ap-
proach to power that focuses not on individuals, fu-ms, or the state, but on the
networks that the elites in each of these spheres develop and maintain in the pur-
suit of their common interests and objectives. 15 H ere I refer to the tradition of
power structure research associated with G. William Domhoff, resurrected in part
in a treatise on the policy process. 16 Domhoff finds himself in an argument be-
tween the instrumentalists, with whom he is most closely aligned, and the struc-
turalists, who discuss the limitations on the ability of the state to act indepen-
dently. The articulated structuralist view, represented by the work of Nicos
Poulantzas, suggests that "ultimately, even a president cannot press beyond the
constraints of government structure and must succumb to the internal dynamics
which push government to maintain stability, provide a favorable investment cli-
mate for privately held businesses, and suppress any discontent with the outcomes
of these policies!' 17 Against this view, the instrumentalists explore the power aud
influence of organizations and institutions outside of the formal structure of the
government apparatus, such as the foundations, policy centers, or think tanks,
 INFORMATION AND POWER 21

and the business organizations and political action committees (PACs). A more
comprehensive analysis of the organizational influence on policy is to be found in
the work of Edward Laumann and David Knoke. 18 Particularly useful about their
analysis is their focus on the ability of organizations to mobilize and process infor-
mation as a resource in the production of influence.
Michel Foucault's primary project was concerned with the exploration of the
techniques-the microphysics of power- that were involved in the formation of
the individual as the object of a disciplinary technology. 19 His project was radical
and intentionally destabilizing. In a conversation in which he answered criticisms
from those involved in the penal system, such as social workers, Foucault re-
sponded forthrightly to the challenge that he provided little guidance:

It's true that certain people, such as those who work in the institutional setting of the
prison-which is not the same as being in prison-are not likely to find advice or in-
structions in my books that tell them "what is to be done." But my project is pre-
cisely to bring it about that they "no longer know what to do:' so that the acts, ges-
tures, discourses which up until then had seemed to go without saying become
problematic, difficult, dangerous. This effect is intentional20

Foucault's conception of the relation between power and knowledge is com-

plex. Rather than to claim, as was (and is) popular, that power equals knowledge,
Foucault discussed the two as being inextricably linked and mutually constitutive.
Power cannot operate without a constitutive field of knowledge, and the opera-
tion of power reveals both its possibilities and its limits; knowledge is necessary
for power to continue to operate. The social science disciplines tl1at were the focus
of much of Foucault's writing about power are defined in terms of their special-
ized and evolving knowledge of persons as objects of power. This knowledge is
gained through a form of disciplinary surveillance that he defined as panopticism.

PANOPTICISM
Paoopticism is based on the belief that control over individuals is made possible
through a system that facilitates the continuous, automatic, disciplinary surveil-
lance of persons who have been determined to be in need of correction or normal-
ization. As a technology of power it involves the organization of individuals into
space through their partition into categories that are at their base dichotomous,
either inside or outside. 21 Subsequent examination of people through observation
and surveillance serves the goals of correction and control
~ Foucault began his discussion of panopticism with a description of late
" seventeenth-century efforts to control the plague. The panoptic procedures of
lf .segmentation and surveillance are readily identifiable in the routines described
k>r locating, observing, and taking note of the status of the residents of the town.
Locked into their individual homes, the residents were required to appear at a
22 fNFORMATlON AND POWER

window and report the status of their health, under pain of death. The surveil-
lance depended on the creation of a record of every resident that formed the basis
of a permanent registration and that could be added to after each inspection. In
this description Foucault found the underlying principles of panopticism
through which we will see clear reflections of the much more sophisticated sys-
tems in place today:

This enclosed, segmented space, observed at every point, in which the individuals
are inserted in a fixed place, in which the slightest movements are supervised, in
which all events are recorded, in which an uninterrupted work of writing Links the
centre and periphery, in which power is exercised without division, according to a
continuous hierarchical figure, in which each individual is constantly located, exam-
ined and distributed among the living beings, the sick and the dead-all this consti-
tutes a compact model of the disciplinary mechanism. 22

And the power of the mechanism lay in its analysis.

Foucault identified the Panopticon ofJeremy Bentham as an architectural real-
ization of the disciplinary principle.23 The prison that Bentham wanted to have
built was a model of efficiency. Bentham believed that this design, orginally devel-
oped by his brother, a naval architect, would serve a utilitarian purpose of the
highest order. His presentation of the design to the National Assembly of
postrevolutionary France won him the distinction of honorary citizenship.24
Bentham's vision was not lim.ited to the use of the panoptic form as a prison, per
se; he also thought that it might serve a dual function as a self-sufficient, even
profitable workhouse for the indigent and unemployed who were w1willing or
unable to join the emerging proletariat.
The Panopticon, as designed, would be a many-sided construction with a cen-
tral observation tower. Each cell would be supplied with two windows, one to pro-
vide backlighting for the occupant, the other to allow the observer in the tower
unimpeded visual access. Unlike the dungeon, which locked prisoners away in
darkness, the panoptic design provided control through complete visibility. Be-
cause each occupant of a cell would be isolated from others, it would reduce the
possibility of the collective, cooperative, or infectious spread of problematic con-
ditions or behavior. In Foucault's assessment, the design could be readily adapted
for any number of uses, to house patients, students, soldiers, or workers. All could
be managed more efficiently if isolation and surveillance could be facilitated. In
Bentham's design, power would operate because it was both visible and unverifi-
able. The central tower, visible at all times, would be a constant reminder that an
unseen observer might be watching at any time. The Panopticon would be effi-
cient because it would require fewer guards or supervisory personnel. In one
sense, the system was self-monitoring in that residents would soon come to expect
continuous surveillance or, at the very least, would always be uncertain about I
when the next glance would fall on their cell. The likelihood of periodic observa- j
DIFORMATION AND POWER 23

tion would operate like random reinforcement in conditioning each person to his
status.
The efficiency of the pan optic design was also ensured by the fact that it did not
matter who exercised the power of observation. Almost anyone could perform the
function. Indeed, as Foucault suggested, there need not ·even be uniformity of
purpose or motivation behind the operation of the Panopticon by observers with
access to the tower- any would serve ~s well. "The more numerous these anony-
mous and temporary observers, the greater the risk for the inmate of being sur~
prised and the greater his anxious awareness of being observed."25 Anyone could
be watching, and nearly all would tell someone should a prisoner step out ofline.
The fact that the Panopticon was never built says more about Bentham's fail-
ings as a politician than it says about the utility of the Panopticon as a metaphor
for a system of control. Foucault was dear on this point. He suggested that the
Panopticon must "be understood as a generalizable model of functioning; a way
of defining power relations in terms of the everyday life of men."26 The metaphor
is powerful. Although some may find its totalizing character harsh and unrealis-
tic, it is not one from which I shrink. Instead, I wish to echo Foucault's claim, and
to offer extensions of my own:

It is polyvalent in its applications; it serves to reform prisoners, but also to treat pa-
tients, to instruct schoolchildren, to confine the insane, to supervise workers, to put
beggars and idJers to work. Lt is a type of location ofboclies in space, of distribution
of individuals in relation to one another, of hierarchical organization, of disposition
of centres and channels of power, of definition of the instruments and modes of in-
tervention of power, which can be implemented in hospitals, workshops, .schools,
prisons. Whenever one is dealing with a multiplicity of individuals on whom a task
or a particular form of behavior must be imposed, the panoptic schema may be
used. 27

It is only the locational constraints, the notion of separation by space, occasioned

by the. initial conceptualization of the panoptic system as a building and by the
surveillance as visual that limits Foucault's construct. But in an age of electronic
networks, virtual memory, and remote access to distributed intelligence and data,
disciplinary surveillance is no longer limited to single buildings, and observation
is no longer Umjted to Une of sight. This fact has been stated forcefully by Frank
Webster and Kevin Robins, whose work has been influential in guiding my pur-
suit of the mechanics of the panoptic sort. 28

Disciplinary Surveillance
Foucault's emphasis on the social sciences, the disciplines of sociology, psychol-
ogy, criminology, and psychiatry, reflected his interest in particularly salient
forms of social control that have historically been legitimated behind the cloak of
science. The professionals involved in market research and risk analysis are not
24 INFORMATION AND POWER

substantially less scientific in their methods and certainly no less consequential in

their conditioning of behavior than any who fall within Foucault's category of the
objectifying social sciences. They are all characterized by the same "vast compila-
tion of data, the proliferation of dossiers, and the continuous expansion of new
areas of research developed concurrently with a refinement and flourishing of disc
ciplinary techniques for observing and analyzing the body, so as to make it more
available for manipulation and control." 29
It is critical to note that the purpose of the modern prison and of other
panoptic systems that imitate its technology is not punishment, but transforma-
tion, rehabilitation, and correct training. The same may be said of the panoptic
sort: It is not limited to identification, classification, and assessment, but includes
the goal of normalizing behavior within categories. This process of correct train-
ing involves three simple components, one of which we already met-surveil-
lance, or hierarchical observ~tion. The second component, normalizing judg-
ment, which seeks to punish nonconformity, or deviation from the norm, may
teach conformity more efficiently through judicious supply of rewards. The dis-
tribution of rewards, or of freedoms on the basis.of ranks, grades, or scores, de-
scribes and defines at the same time that it instructs. Like the colors of the epau-
lettes on the boys at the Ecole Militarie,30 the colors of American Express (Amex)
cards, green, gold, and platinum, are only glimpses of the much finer gradations
that operate to lead cardmembers through their paces.31 The possibilities for dis-
cipline with the Amex membership are quite similar to the possibilities within the
military academy, and the process of normalization is the same because it depends
on a system that "compares, differentiates, hierarchizes, homogenizes, [and] ex-
cludes."32
The third component of correct training is the examination, which facilitates
the process of normalization. It is through examination that the evaluation on
which normalization is based may be produced. Amex is reported to examine
each of its cardmembers on as many as 450 categories on a daily basis, because
such an examination facilitates both the authorization of charges and the market-
ing of cardmember, auxiliary, and affiliate services. For Foucault, the examination
is important because it not only places individuals within the field of surveillance,
but it also "situates them in a network of writing; it engages them in a whole mass
of documents that capture and fix them. The procedures of examination were ac-
companied at the same "time by a system of intense registr~tion and documentary
accumulation:m In addition to the creation of an individual record, which de-
fines the individual as a case, this documentation eases the disciplinary accumula-
tion of records, which eases further classification, and the establishment of empir-
ical norms. The computerization and linkage of these written documents via
telecommunications move the panoptic sort to a different level of significance.
Part of the challenge we face in the integration of Foucault's vision of a system
of dispersed but continually evolving sytems of power is understanding how these
JNFORMATION AND POWER 25

systems come to be institutionalized and legitimated at different levels of social

systems. We are especially interested in the ways in which these sytems of power
come to be integrated with other systems of domination that may be mutually re-
inforcing. The evolution and elaboration of systems of power may be seen as a
process ofstructuration.

STRUCTURATION
Structuration is one of the more acceptable neologisms generated by Anthony
Giddens as he has labored to produce a theoretical alternative to functionalism.
Giddens is unremitting in his criticism of all forms and varieties of functionalism,
even to the finding of damaging traces of its presence in Marx and Marxists of a
variety of stripes. 34 His major criticism of functionalism has been that it provides
no true explanation in a claim that a system has needs for some thing or some rela-
tionships that may come to characterize it. Saying that a capitalist system needs a
reserve army of the unemployed does nothing to explain how it is that a reserve
army comes to be. Giddens suggests that functionalist theories ought to pu.rsue an
explanation through consideration of the counterfactual- what would happen to
the system in the event that there were no such army? What would it do. to ensure
that such an army developed and was maintained? It is clear that Giddens's cri-
tique in part attacks the tendency of functionalism to assign an intelligence to a
system as though it were an entity, a life form, or, at best, an organization with a
leadership. To say that capitalism as a system needs something is quite different
from saying that a group of capitalists, an elite core that recognizes a common
class interest, may act in ways its leaders believe will serve those needs. This is also
different from arguing that a capitalist logic demands certain choices, because that
logic, so defined, is one that is either implemented or not by capitalists them-
selves. Similarly, leaders of the state as an organization may perceive what the
needs of groups and organizations may be and may act to meet some of those
needs as a part of their mission and presumed responsibility. But in each of these
examples, knowledgeable human beings are acting on the basis of an analysis of
circumstances and options. This is a continu¢g problem in the pursuit ofanalysis
at the level of institutions.
Organizations and institutions do not think American Telephone and Tele-
graph (AT&T) has no mind, no consciousness, and no social conscience other
than that which may be written down, reproduced, and transmitted to its leader-
ship through a process of socialization. Social systems have no needs, because sys-
tems are not entities but are theoretical constructs. Social systems cannot act, but
they may be seen to operate in a particular fashion. It may even be argued that so-
cial systems have requirements, but this claim cannot be demonstrated conclu-
sively unless the definition of the system is tautological in that its requirement is
26 INFORMATION AND POWER

part of its definition-a capitalist system is one that requires the exploitation of
labor.
Giddens's emphasis on knowledgeable human beings, acting in pursuit of their
interests as they understand them, provides an entry point through which to be-
gin thinking about structuration as a process, rather than as a sta'tic framework.
For Giddens, "social systems are composed of patterns of relationships between
actors or collectivities reproduced across time and space.'' 35 Giddens's knowledge-
able agents are privileged in terms of their individual and collective agency.
Knowledgeable actors not only have an understanding, but also act on the basis of
that understanding. This action, taken primarily through interaction with other
knowledgeable agents, produces and reproduces the social system. Giddens's con-
cept of structuration is not unlike the process of socialization, except that his em-
phasis on agency avoids defining a process through which persons are acted on.
Rather, structuration describes a process that unfolds by working through. Fur-
thermore, structuration differs from socialization in that relationships, as well as
beliefs and values, are reproduced through the process.
This notion of reproduction is to be distinguished from the familiar biological
sense of reproduction in which similar, perhaps functionally identical, beings are
produced through genetic rules and biological resources contributed by parents,
but it is more recursive in the sense of being constitutive of itself. That is, in the
process of engaging in interactive and contextually situated practices, in which the
actors are constrained by the rules as well as by the resources that they bring to the
relationship, each actor is produced again, or reproduced, in an identifiable rela-
tionship that we may understand to be structured.
Interactions between parents and children take place in a variety of sites and
circumstances, each characterized by a particular set of expectations and gov-
erned by a set of rules and understandings about the powers and freedoms of each
knowledgeable agent. To the extent that each individual acts consistently with his-
torically and idiosyncratically generated sets of expectations, the parent-child re-
lationship is reproduced. Of course, parent-child relationships usually change.
Sometimes the changes are conflictual and unpleasant. The nature of the change
is not mysterious. Most parents are able to understand and accommodate these
changes, which they accept as part of a normal, if ritualized, process by which
their children pass into adulthood and independence. These are the sorts of
changes that Giddens's emphasis on time and space might suggest we explore
through a focus on the middle range ofhistoricaJ time, which is captured by the
concept ofthe life cycle. Beyond the minute-by-minute, day-by-day, even year-by-
year spans of time, which characterize the dimension of temporality, we find a
much longer period that Giddens associates with institutions. The primary em-
phasis in this book will be within the shorter historical moments of people's time,
although I do not wish to lose sight of the structures of time in which institutions
are .seen to change.
INFORMATION AND POWER 27

Roles, Resources, and Routines

Giddens has been subjected to considerable criticism for the often frustrating lack
of clarity with which he defines some of the core concepts in his theory of
structuration. Among the most problematic areas are those having to do with the
use of power in various interactions. Knowaedgeable agents understandably differ
in terms of the resources that they can br~ng to bear to increase the probability
that the outcomes of those interactions are ones that they favor. The use of an
agent's resources, however, is constrained by the existence of rules that govern the
use of rewards as well as sanctions. Rules are moral, professional, social, and in-
herited and may even be negotiated on the spot and apply only in limited special
circumstances. These rules apply to communications as well as more overt forms
of behavior that define any interaction. For Giddens, "all human interaction in-
volves the communication of meaning, the operation of power, and modes of nor-
mative sanctioning." 36
Depending on the site and the nature of the interaction, referred to as a struc-
tural set, Giddens suggests that we may focus on different structural properties,
which he labels signification, domination, and legitimation. Structures of domi-
nation are further distinguished in terms of capabilities relating to influence over
persons (authorization) and command over material resources (allocation).
Giddens suggests that through rules the structuring may acquire a certain degree
of stability as the rules permeate social systems, and in this way they may be seen
as having become institutionalized. That is, some rules become institutionalized
in bureaucratic forms and have tbe force oflaw, and the instrumentations of prec-
edent and expertise help to guide their interpretation.
Different institutions may therefore be characterized in terms of the modalities
of domination that are most characteristic of their process. Although all three
structural properties are always present and interacting, legitimation may be seen
to be in the forefront of the legal realm, and allocative considerations migbt char-
acterize the economic realm. The authoritative modality is in the forefront of in-
teractions within political institutions.37
John Thompson suggests that there are numerous problems with the concept
of rules in Giddens's theoretical system.38 Part of the difficulty has to do with the
fact that "the rules which comprise structure are embroiled in struggles, are sub-
ject to rival interpretations and are continually transformed in their very applica-
tion. Rules, il1 other words, cannot be conceptualized in isolation from the re-
sources which facilitate the exercise of power:'39 Rules, therefore, cannot be
specified objectively because, like systems and relationships, they are conceptual
and vary in the degree to which they are believed to be operative in any interac-
tion. An agent's awareness of the rules that apply in any interaction may be influ-
enced strategically by another actor, as well as by a habitually structured process of
"checking" as an individual takes stock reflexively of the circumstances in which
he or she finds himself or herself at any point in time.
28 lNFORMATION AND POWER

Consciousness
This checking may take place at different levels of consciousness or with different
degrees of self-awareness. Giddens makes an important contribution in his dis-
cussion of the differences between nvo levels of consciousness. As I have sug-
gested, individuals acting in the pursuit of their interests are guided by their un-
derstandings of the environment, the rules of the game, and their own capacities
and resources. Much of this knowledge is seen to operate below the level of one's
conscious awareness, a kind of autopilot that draws on the accumulated knowl-
edge and experience that individuals have stored, perhaps even at the neurological
level. People may know how to drive a car, ride a bicycle, throw a ball, return a
serve, dodge a wild pitch, and perhaps even how to get their computer to print out
a design onto a page. They also know how to engage in a conversation, an argu-
ment, or a negotiation. Yet they may be far from articulate about the mechanics
involved in any of these activities. This is a distinction that Giddens makes be-
tween practical consciousness and discursive consciousness, in which language is
required to give knowledge and reason expression. Discursive consciousness is
not limited to communication with others, but includes communication with
oneself- the ability to reflexively examine the reasons, the arguments, and the
meanings behind the myriad of choices that might be involved in the completion
of any routi1~e process, such as preparing an evening meal or voting for a referen-
dum issue. Giddens helps us to understand the methodological difficulties in-
volved in trying to understand how what people know influences what they do.
"Where what agents know about what they do is restricted to what they can say
about it, in whatever discursive style, a very wide area of knowledgeability is sim-
ply occluded from view."40
Giddens introduces the influence of history and culture into his theoretical
mix through his suggestion that consciousness, both practical and discursive, has
multiple sources, personal, direct, and mediated. The "stocks of knowledge," on
which actors draw as they engage in the interactions that produce and reproduce
the systems in which they are situated, are neither fixed nor fully given. It is im-
portant to understand the extent to which the power within the various relation-
ships that define a system may be bound up in the structuring of access to knowl-
edge.
Although a distinction is made between practical and discursive conscious-
ness, I take from Giddens an emphasis on an indi,,idual's reflexive monitoring of
her own actions, in the context of reassessments of progress and purpose: "All
(competent) actors in a society are expected to 'keep in touch' with why they act as
they do, as a routine element of action, such that they can 'account' for what they
do when asked to do so by others:'41 It is important at this stage to ·n ote that
Giddens recognizes that the reflexive monitoring of individuals is almost by defi-
nition incomplete and inadequate. When Giddens characterizes his agents as
knowledgeable, he admits to the limitations on that knowledge. It is clear, for ex-
ample, that an actor will not fully apprehend all of the conditions that constrain
 ..<ft
INFORMATION AND POWER 29

(help or hind.er) the actions she may pursue. Similarly, this actor will not know,
indeed cannot possibly know, of all of the unintended consequences of whatever
actions she may take. Indeed, as Bartlett suggests, many of the actions that an in-
dividual will take will be constrained by actions taken by other actors, and not all
of those constraints will have been planned by those actors; they may be uncon-
trolled externalities. 42
An individual's motivation for action works through her reasons for the ac-
tion, which include both her sense of her present status, assessed reflexively, and
her sense of the conditions in which ac6ons are taken in pursuit of purposeful
goals. We may add, although Giddens is unclear on this point, that the unknown
may also include the unconscious motivations, reasons, and awarenesses that may
interact with and influence understanding and action. The monitoring of actions
will take note of those that. were intended and those that were unintended. Be-
cause vision is physiologically as well as psychologicaJJy limited, however, even
many of the unintended consequences will not be noticed. I might also suggest,
consistent with the notions I share with Bartlett and others, that much that is un-
acknowledged in the conditions, including the unintended consequences, is not
available for monitoring because other interested actors have purposively hidden
it from view. lt is through this vision of agency that Giddens is able to make an as-
sociation with Marx's dictum about age11cy. People make (and are made by) the
world, but it is not a world constructed according to their designs.
As has been suggested in the discussion of rules and resources, a critical aspect
of knowledgeability has to do with self-identification and estimates of the pre-
sumed appropriateness or legitimacy of one's claims or of the claims that others
make on the basis of that identification. Relations among individuals that are de-
fined or are definable in terms of class involve considerations of differential re-
sources appropriate to one's position or status. Awareness of the rules governing
interactions, which is also likely to involve an evaluation of the appropriateness or
fairness of the rules, may be seen as an index of class consciousness.

Class Consciousness
Critical consciousness frequently involves an analysis of the power of others as
having unjustly limited the life chances of entire groups or classes of individuals.
Group consciousness, defined as a recognition of commonality ofstatus, options,
and purpose, has been seen as a necesssary precondition of resistance and social
transformation. Class consciousness has been identified as the primary condition
necessary for a revolutionary transformation of capitalism. The failure of class
consciousness to develop into a revolutionary form has been a core problem in so-
ciaJ theory. 43
The definition of class and the origins of its consciousness have been funda-
mental concerns of Marxists and reflect the centrality of the concept of class con-
sciousness in Marx's writings. Contemporary Marxist scholars struggle to explain
the considerable variety in class orientations that can be found within the working
30 INFORMATION AND POWER

class. Erik Olin Wright has pursued his concept of contradictory class positions as
a way to understand the complexity of class relations and to reclaim class analysis
as a theoretical force. 44 Wright recognizes that one's existel'lce in a class location
means that one is subject to a set of mechanisms and forces that influence one's
class consciousness. For Wright, the "task is to understand the ways in which
macro-structural contexts constrain micro-level processes, and the ways in which
the micro-level choices and strategies of individuals can affect macro-stuctural
arrangements:' 45 Class consciousness, as a common w1derstanding, is derived
from shared material interests, which may or may not be derived from their com-
mon experience as labor. Wright's empirical work has been troubled, as has been
the work of many who sought the link between experience and consciousness, by
the fact that people within the same objective class will have quite different life ex-
periences and, as a result, will understand the world and their place in it quite dif-
ferently. Wright finds in Giddens a failure to specify an alternative to the determi-
nations of class consciousness that are based on property relations. Indeed, he
suggests that many of the examples that Giddens provides are in fact very close to
Marxist understandings of class. 46 He finds in Giddens no reason to doubt that
the influence of class is primary: "This does not imply that individuals are neces-
sarily 'class conscious' in the sense of being aware of their class position and class
interests, but simply that their forms of social consciousness are more systemati-
cally shaped by class relations than by any other relations and should therefore be
accorded primacy." 47 Wright has also expressed a reluctance to join the Weberian
bandwagon and to set aside theory in preference for an empiricist glance toward
whatever relationships may appear.
Max Weber distinguished between class and status groups in ways that set him
apart from Marx. The distinctions between Marx and Weber on the q uestion of
class were not limited to their differential emphases on class conflict as a primary
force in the transformation of society. The distinction for Weber between classes
and status groups reflected a diffe,rence in his conception of the nature of power
within a society. With slightly different terms, Weber's conception of the relations
between the economic base and the superstructure can be read from his discus-
sion of power, wherein he finds a determination of class situation in a person's
market situation. 48 An individual's class situation is determined by economic con-
ditions, based on that person's position in the commodity or labor markets. Such
a position is defined in terms of an individual's possession of goods and opportu-
nities for income. Weber suggests that we can speak of a class when a given num-
ber of people share, in common, these determinations of life chances. These
classes are not communities of common interest, but definitional sets.
Weber steps back from the Marxian position that class position determines

ll
class consciousness. Although he recognizes that a common class situation may
generate "similar" reactions, such a result is far from certain. Instead, Weber ar- •
gues that social consciousness, which may be reflected in associations and social
action, is dev<loped in the mltural •eolm, in intelle<tu•I dis<eum. Clos• •itu•tion
INFORMATION AND POWER 31

has to be understood before it can support class action, and this understanding is
not inextricably linked to one's position in the market. The kind of social con-
sciousness that can lead to social action is, in Weber's view, more likely to take
place in status groups, which are real, rather than co_nceptual, social forms.
Status groups, differentiated in terms of the status honor its members may
claim, are expressed by a "style oflife" that their members and those who aspire to
membership must adopt. 49 The difference between classes and status groups is
seen ultimately to turn on their relation to the production or the consumption
spheres. Class situation is defined and determined by the relations of production
and the acquisition of wealth, whereas "status groups are stratified according to
the principles of their consumption of goods as represented by special styles of
life." 50 This· is not to suggest that status position is independent of class position,
because the ability to maintain a life-style is largely dependent on economics, but
only to indicate that the correlation is far from perfect.
Wright's desire to find a more clearly theoretical basis for understanding the
variance in class consciousness may also be seen in the empirically grounded the-
orizing of Arthur Stinchcombe. Stinchcombe defines class consciousness as the
"tendency of people to think of their position i.q the larger society in terms of their
position in an employing organization. Workers are class .conscious when they
think of their grievances at work and their interests in politics as both derived
from their employment relation to particular organizations."51 Similarities among
workers. in industrial societies reflect the similarities in the contnicts that deter-
mine whlch "slots" workers are assigned to within a particular production system.
The political content of class consciousness is based on the interaction between
the processes that assign workers to slots and on the rights and freedoms that are
common to workers in particular slots. Stinchcombe utilizes the contributions of
E. P. Thompson and David Lockwood to aid our understanding of the diffuse and
nonrevolutionary class consciousness that has come to characterize workers in the
service or information sectors of modern economies.
In contrast to ~he neoclassical notions of free labor negotiating contracts with
employers, industrialization involved a "bureaucratization of the wage contract:'
by means of which workers found themselves in a "take it or leave it" situation
with regard to a standard contract. The uniformity in the labor contract is seen to
have been.matched by a kind of uniformity in the political contract defined in
terms of democratic suffrage. The uniformity in the labor contract makes it possi-
ble for labor to identify its common interest in seeking a general improvement in
the wage payment in general, rather than in terms of a personal contract. Collec-
tive barg.aining through unions is an understandably bureaucratic response to
such an analysis of common interests deriving from common circumstance.
After noting substantial differences in the nature of the labor movements in
the United Kingdom, the United States, and Japan, Stinchcombe focuses on the
cultural aspects of the process through which class consciousness develops. The
developmen t of the left-wing Enlightenment project of Thomas Paine is seen to
32 INFORMA'.110N AND POWER

have departed from the rightist commitment to tradition and the wisdom of one's
betters, which restricted the development" of class consciousness among some.
Stinchcombe finds in Thompson a thread that is central to Giddens's theory of
structuration. The development of class consciousness was the result of the ongo-
ing "intellectual struggles of a bunch of workers to figure out what was going on,
in terms of the cultural heritage that they had."52 Class consciousness was the
product of a process involving a goal oriented search for understanding, condi-
tioned but not fully determined by class position. A common conclusion that
emerged from this process was that the system was not trustworthy. It was not"a
system in which the rulers cared about their fates, respected their action to defend
those fates, and regaJded them as persons with inherent value to be respected and
protected."53 As we shall see quite clearly as we examine public understanding of
the pan optic sort, the question of trust is a critical component in the formation of
consciousness and the demand for state action.
Problems with trust emerge when organizations create the expectation that the
standard contract will guarantee standard treatment, but when the concrete; ob-
jective experience of workers is anything but equal. Ensuring equal treatment has
been traditionally a prime concern of unions, but the decline in unionization in
the U.S. work force may be seen as both cause and consequence of ill-formed class
consciousness.
Stinchcome identifies two factors to explain the absence of class consciousness
among workers. The first is the creation of a dual labor market in which the sub-
ordinate market contains low-skilled, low-commitment workers, generally fe-
male, young, and minority group members. These workers tend to be excluded
from positions that have strong union control, and they fail to perceive a common
purpose in standard contracts because such contracts do not exist where turnover,
or the flow oflabor between jobs, is so rapid.54 The second and more interesting
factor in terms of OUJ efforts to understand the cultural component of conscious-
ness is the claim that one's consciousness as a worker comes from interaction and
contact with workers.
It is the nature of work in the service sector that "many service jobs isolate the
person from social contact with other workers and require intense contact with
clients."55 This influence does not work in a unified direction but rather works
similarly to the mainstreaming process identified by George Gerbner and his col-
leagues.56 Contacts with clients in upper-class professions are seen to lead service
workers leftward, whereas contacts with clients in lower-class professions lead
those workers rightward. Neither movement leads to the development of a pro-
gressive class consciousness.
The strength of this influence is determined by the extent of a worker's contact
with clients. "The more the demography of the workplace indudes clients, and the
more intense the interaction ofservice workers with dien tS is, the more those who
might otherwise be disposed to join unions or to vote left will instead be non-
union and vote to the right."57 This confusion in class conscious1wss is heightened
lNFORMATION AND POWER 33

by the fact that many service workers are themselves engaged in selling the sym-
bol$ of class status. To sell effectively, service workers must place themselves "in-
side the heads" of their clients. This process cannot help but infl uence their own
consciousness and values. You become what you sell.
Stinchcombe suggests that when the interaction with clients involves a neces-
sary deference to the client, a similar form of identification takes place. He sug-
gests further that the nature of the organization of the service sector means that
many workers will be in smaller organizations where they will be in close direct
contact with the profit-oriented owner of the enterprise. The influence on con-
sciousness from .this contact is far greater than anyone would imagine might re-
sult from such contact between workers and owners in a massive organization
such as AT&T. The scale of the organization and the individualization of the labor
contract also support the development ofa conservatizing ambition, which is less
likely to develop on the factory floor or at a telemarketing console.
It is through this focus on the nature of the work process that Stinchcombe
helps us to understand some of the co~flicting.and contradictory aspects of class
consciousness that are derived from Wright's notion. of contradictory class loca-
tions. It is his emphasis on the day-to-day, routine interactions that help .to condi-
tion, structure, and reproduce class orientations that are among the most useful,
especially as they are closely linked to Giddens's construct of structuration. How-
ever, in the modern (or postmodern) era, a great many routine interactions no
longer take place face-to-face.

Distanciation
Giddens's concept of distanciation helps one understand the significance of tele-
communications and telecommunications nehvorks. Interactions, which involve
situated practices generally conceived of as taking place in a restricted time-space
nexus, increase in variety through the technology of communication. The tech-
nology of communication, no doubt including communication across distances
by means of the drum, but perhaps more significantly, the communication that
spans temporal limits as well th rough writing, changes the structures of domina-
tion. Allocative and authoritative resources may be used to coordinate social sys-
tems across time and space by means of communication. This process of
distanciation, which transforms the presence/absence character of the power that
is reproduced through interaction, is qLrnlitatively different in the context of mod-
ern communications neh'l'orks. Communications neh'l'orks change tl1e nature of
the resources that are required to maintain a level of domination within a social
system. GeoffMulgan suggests that the"'social organization of control is substan -
tially dependent on available communications technologies. Without direct and
rapid communication, control must depend on the use ofagents." 58 Coordination
is a fundamental aspect of structuration and involves the application of authorita-
tive resources. The ability of some actors to exercise this authority remotely and to
effectively multiply the presence of any given authority by allowing its appearance
34 lNFORMATION AND POWER

in multiple sites at a fraction of the cost that would have been required for a physi-
cal presence is a res~urce of great importance and historical moment. Giddens
suggests that "in general (although certainly not universally) it is true that the
greater tlie time-space distancia ti on ofsocial systems-the more their instHutions
bite into time and space--the more resistant they are to manipulation or change
by any individual agent."59 ·
Giddens also suggests that we give due consideration to the storage and access
qualities of media that help to determine the nature of tinie-space distanciation:

The storage of authoritative and allocative resources may be understood as involving

the retention and control of information or knowledge whereby social relations are
perpetuated across time-space. Storage presumes media ofinformatiom1l represen·
tation, .modes of information retrieval or recall and, as with all power resources,
modes of its dissemination. Notches on wood, written lists, books, files, films,
tapes-all these are media of information stora.ge of widely varying capacity and de-
tail. All depend for their retrieval on the recall capacities of the human memory but
:Jso on skills of interpretation that may be possessed by only a minority within any
given population. 60

As Thave suggested elsewhere, the inequallty in the distribution of individual ca ~

pacities to access and retrieve information is compounded when similar limita-
tions also constrain the ability of individuals to express their knowledge discur-
sively in ways that have force. 61
Derek Gregory's criticism ofGiddens's theoretical project focuses on the chal-
lenges of empirical research that might be involved in the coordination that is fa-
cilitated through interactions with others who an: absent rather than p1esent in
time-space.6i Gregory notes that Giddens places capita)jst societies in a structure
of domination that is characterized by a high degree of time-space distanciation,
which supports the realization of power through the control of allocative (eco-
nomic) resources through survei1lance and commoditizatio He suggests, however,
that Giddens draws back too soon from hisn. responsibifay to specify what some
of the locales and settings might be in which this modern form of structuration
takes place. Clearly, he suggests that "actors routinely encounter locales in the
conduct of everyday life, no doubt, but any genuinely critical theory must surely
go beyond these mundanities to show how particular places and spaces are pro-
duced. " 63 Gregory's demand for geographical specificity misapprehends the as-
pects of a virtual reality in a networked environment that defies locational preci-
sion. It is not a geographical space but a transactional space that begins to identify
and define the structurational reality of tlie panoptk sort.
As we have suggested wit11 regard to Foucault's reconstruction of the
Panopticon, the operation of the pan optic machine has escaped the constraints of
.location or space just as effectively as it has overcome limitations on time. Gregory
is correct, however, in his charge that Giddens's contribution remainsi for tlx
most part, at a very high level of abstraction, and perhaps it falls to us to push
INFORMATION AND POW ER 35

those insights into the empirical realm to discover how technology, markets, and
culture enable as well as constrain the freedom of knowledgeable agents.

TECHNOLOGY, MARKETS, AND CULTURE

It is my goal in this book to explore the changes in technology, especially the inte-
grated technical system that I refer to as the panoptic sort, in the context of the
changes that are taking place in the markets of late capitalism. In the face of the
collapse of the communist, or state capitalist, forms of governance that have oc-
curred so dramatically as the decade of the i99os began to reveal its character, I
would only invite ridicule by speaking as though capitalism was nearing its end.
Yet the transformations in capitalism that have been predicted by Marx, Weber,
and Schumpeter have surely produced a kind of capitalism that Adam Smith
would quickly disown. When some social theorist succeeds in popularizing a
name for this new form of socioeconomic organization, then capitalism will have
come to an end. Thus, to the extent that the old nomenclature no longer fits the
data very well, lateness is a fitting characterization.
The cultural changes of interest are those that reflect and are reflected in the
laws, practices, and customs regarding the treatment of personal information, as
the coordination of capitalism requires ready access to increasingly detailed infor-
mation about the history and status of individuals. The panoptic sort is an inte-
grated system that is involved in the identification, classification, assessment, and
distribution of individuals to their places in the arrayoflife chances, which are de-
termined in the play of tensions within late capitalism. The operation of the sort
reflects the extent to whjch individuals and institutions share understandings
about the necessity as well as the fairness, appropriateness, or legitimacy of the
system that is evolving at the same time as it is being put into place. It may operate
smoothly, or it may operate in fits and starts, like an automobile that is out of ad-
justment or under the control of a new driver unsure and unskilled in the opera-
tion of the clutch.

The Role ofTechnology

The automobile is a useful metaphor for thinking about the panoptic sort in the
context oflate capitalism. The automobile is certainly a technology. It has become
quite sophisticated in terms of the number of its systems that are under the con-
trol of microprocessors. Although it is not difficult to imagine a time when the
amount of active control by drivers will have been reduced to a bare minimum
and the level of any operator's knowledge of its systems will be negligible at best,
we are not well served by thinking of automotive technology as autonomous. The
automobile serves a variety of instrumental goals, not limited to getting us from
here to there. The use of automotive technology for transportation, as well as for
status enhancement and the expression of personality, has had an immeasurable
influence on the environment and on the social, economic, and legal systems that
36 LNFORMAT!ON AND POWER

have developed to structure its use. Yet refusing to accept a perspective that views
technology as an autonomous force does not at the same time mean that one must
deny the fact of technological influenc~. .
Technology and technological systems represent resources that knowledgeable
actors may use in their interact ions with other actors that may systematically con-
strain their options. An orientation toward technology may also come to charac-
terize the ways in which these knowledgeable actors understand their goals and
options. I believe that there is a technological imperative. It resides not in the ma-
chines but in the people who use them. By locating the imperative in individuals,
rather than in machines, l do not mean to associate myself with those who suggest
that technologies are neutral. 64 This is simply not so. Technologies are designed by
humans for specific uses. Although creative (or destwctive) individuals may dis-
cover or invent novel uses or applications for particular technologies, the range of
those applications is limited by the character of the technology. Some technolo-
gies are more flexible in this regard than others. By design they have a broad range
of applications. The disclaimer "guns don't kill, people do" is true as far as it goes.
But it is also true that guns can be used far more efficiently for lcirnng than they
can for hammering nails into plasterboard. Automobiles also kill, but they have
been designed. for other ends, which they serve more efficiently. Information tech-
nologies maybe seen to have quite a broad range ofapplications and may even be
argued to have a great potential for increasing the democratic potential of
societies. Part of this apparent flexibility is inherent in the fact that the technology
of the computer-based information system is not limited to the hardware but in-
cludes the applications software, which more completely defines its use.
The operational effectiveness of any inherent technological constraint depends
on the awareness, understanding, and acceptance of this apparent limitation by
different individuals. Jacques Ellul helps us to see that this understanding is also
subject to the influence of technical systems in use. Ellu) is generally identified as
belonging to the class of technophobics. Yet i1 is his mission to teach understand-
ing and to engender resistance rather than fear. His master work on the nature of
technology has stood the test of time as an articulate·expression of insights into a
way ofthjnking that permeates all spheres of social existence.65 It is important to
understand Ellul's thinking with regard to propaganda as being integrally related
to his views on technique in general. 66 ln his Preface, Ellul suggests that "not only
is propaganda itself a technique, it is also an indispensable condition for the de·
velopment of technical progress and the establishment of a technological civiliza·
tion . And, as with all techniques, propaganda is subject to the law of efficiency."6'
Later he continues, "propaganda is called on to solve problems created by tech-
nology, to play on maladjustments, and to integrate the individual into a techno-
logical world." 68
Clifford Cbristians69 finds Ellul's vision of the technological society and its to-
talizing systems of control to be superior to the theory of ideological hegemony
associated with Antonio Gramsci,7° in that Ellul understands that the sophisti-
INFORMATION AND POWER 37

cated use of mass media to eviscerate public opinion is one that is itself guided by
the rule of efficiency. This same technicism that distorts the public sphere is also
seen to have transformed political participation into an abstract illusion. 7 1 This
conclusion is directly opposed to those who believe that communications tech-
nology can never be a democratizing force. ·
John Wilkinson, translator of the Knopf version ofEUul's The Technological So-
ciety, warns the readers that EUul's writing betrays an "insistence on rendering a
purely phenomenological account offact, without causal explanation of the inter-
relation of the subordinate facts. . .. The important questions concerning the
technological society rarely turn for EUul on how or why things came to be so, but
rather on whether bis description of them is a true one."72 Yet it is difficult to avoid
the bald functionalism that Ellul uses to personify technique. Although there are
alternative ways to read Ellul to bring his vision more in line with the theoretical
construction we have in mind, the most damaging criticism launched against
Ellul is that he sees technology as autonomous. This, in common with his ten-
dency to give technique an anthropomorphic stance, leads one to wince when
reading a passage such as this:

Thus technique theoretically and systematically assumes to itself that liberty which it
has been able to win practically. Since it has put itself beyond good and evil, it need
fear ·no limitation whatever. It was long claimed that technique was neutral. Today
this is no longer a useful distinction. The power and autonomy of technique are so
well secured that it, in its turn, bas become the judge of what is moral, the creator of
a new reality:73 ·

Yet Ellul's descriptions of what has taken place and his predictions of what lies
ahead are so clear, sharp, and ripe with illustrative potential that they cannot be
abandoned on the basis of this partially stylistic flaw. There is, as C. George
Benello suggests, a need to find a way to introduce the concept of instrumental ra-
tionality as a behavioral orientation into the history that Ellul writes for us. 74 In
this way we can produce another reading of the hundreds of examples of a deci-
sion pertaining to technology, made in the context of technicist vision, which
leads to yet another set of decisions made along the path of technical efficiency.
This is part of our search to find insights into the patterns that emerge as we in-
clude social systems oflarger and larger scale within our vision. For us the prob-
·lem of theory is to find an explanation for the empirical fact that local actions lead
to global patterns. The problem of elusive theoretical closure is reflected in the
isolated struggles of analysts who focus on the macroscopic level of systems and
institutions and refuse to engage those who toil at the microscopic level of indi-
viduals, cognitive processes, or even the molecular level of communication be-
tween cells.
In Ellul's description of the attempt to mechanize the production of bread, we
find that.an attribute of the wheat made it difficult for the machines as designed
 INFORMATION AND POWER

to reproduce the characteristic flour of a premechanical age; the result was a dif-
·ferent kind of bread. "In the last resort, the ultimate success of mechanization
turned on the transformation ofhum~n taste. Whenever technique collides with a
natural obstacle, it tends to get around it either by replacing the living organism
by a machine, or by modifying the organism so that it no longer presents any spe-
cifically organic reaction.""; The "iif' in the example is not truly any kind of auton-
omous technology, but an entrepreneur seeking to find a way to make use of a
technology that facilitated an efficient (read: profitable) processing of grain. All
that was necessary for this entrepreneur to realize his goals was the creation of a
demand for a new kind of bread. David Lovekin offers a contemporary example to
demonstrate the lengths to which this process has come:

Thus, a simple food like potatoes becomes Tater-Tots, something that is not clearly
food at all, and that contains elements of no clearly known nutritional value. What is
clear is that each piei:e is made to fook like the other piece, identities which are also
different, new. McDonald's markets and produces sameness.... To understand fast
food, a purely technological phenomenon, one must look to the walls and notice the
pictures of the food. One buys the picture, which will never nourisli, but which will
always keep the customer corning back for more, the ever-perfect, indeed, the same
hamburger, designed in the laboratory and cooked by computers. 76

This does not suggest that humans are not somehow convinced that it makes
good sense to accommodate themselves to the limitations of technological design,
nor does it suggest that they are not also somehow transformed in the process. Ali
that we resist is the assignment of the power, design, and reason to a machine or a
tool. Technique is an orientation, not an entity.

Rationalization and Technique

Reading Ellul with the sideways glance that he adopts in his book on propaganda
assists us in finding an instrumentalist position from which to ·understand tech-
nique as it emerged at the time of the industrial revolution. In Propaganda: The
Formation ofMen's Attitudes, Ellul identifies propagandis~s as goal-oriented, tech-
nologically sophisticated individuals, organized through institutions, who act
with purpose. 77 Although he also suggests that the technique that is propaganda
constrains those who would use it, the sense of instrumentalism can be easily
iden tilled and then found throughout his work on technique: "From this point of
view, it might be said that tech.nique is the translation into action of man's con-
cern to master things by means of reason, to account for what is subconscious,
make quantitative what is qualitative, make clear and precise the outlines of na-
ture, take hold of chaos and put o rder into it." 78 He finds a transformation in the
rationale behind science that in the eighteenth century began to admit more and
more of a concern with utility, rather than knowledge, for the sake of curiosity and
discovery. The latter parts of the century were characterized by a kind of
 L\l'ORMATION AND POWER 39

optimism, and for Ellul this "state of mind created .. . a kind of good conscience
on the part of scientists who devoted their research to practical objectives. They
believed that happiness and justice would reswt from their investigations; and it is
here that the myth of progress had its beginning."79 A form of technical con-
sciousness was seen to be derived from the influence of special interests, not only
those of capital, but also those of a revolutionary state, as in France. Ellul even
saddles Marx with the responsibility for having "rehabilitated technique in the
eyes of the workers. He preached that technique can be liberating. Those who ex-
ploited it enslaved the workers, but that was the fault of the masters and not of
technique itself." 80 Technique expands in response to the expression of interest,
and interest, with the aid of propaganda, bas been expanded over the years to ac-
commodate the reach of technique.
Ellul's examination of the role of technology in the economy and the role of
economists as social technologists is an insightful extension of Marx's contribu-
tions to our w1derstanding of the role of a technological change in the economic
base by reminding us that, despite their protestations to the contrary, economists
have taken Marx's challenge to heart: The goal is not to understand the world, but
to change it! Ellul argues that "at the same time that the economist has created a
technique for knowing, he bas created a technique for acting:• and this involves
the introduction of plans and norms that, over time, assUllle a coercive force. 81
Positive in method, normative in purpose, economics and economists have
helped to spread a technicist orientation into all corners of the society, not merely
those clearly defined as economic.82 Ellul's discussion of the human technologies
bears much in common with Foucault's emphasis on the disciplines.
For Ellul, the technologically oriented human sciences are being aimed toward
the task of "rounding up those elements of the human personality that are still
free and forcing ('reintegrating') them into the expanding technical order of
things. What yet remains of private life must be forced into line by invisible tech-
niques, which are also implacable because they are derived from personal convic-
tions."83 This all-encompassing grasp of technique seeks to eliminate all forms of
"social maladjustment or neurosis. Man is to be smoothed out, like a pair of pants
under a steam iron."84
Ellul asks rhetorically, why did history produce such a dramatic growth in
technique at the time at which it occurred. James Beniger asks the same question
but limits it to information technology. 85 Beniger ignores Ellul's response in favor
of his own version ofWeberian rationalization-an instrumental need to reestab-
lish control through the efficient management of information.
Progressive rationalization has been identified as the unifying theme of
Weber's assessment of Western civilization. Rationalization is clearly akin to
Ellul's notions of technique. By rationalization "Weber meant the process by
which explicit, abstract, intellectually calculable rules and procedures are increas-
ingly subsiruted for sentiment, tradition, and rule of thumb in all spheres ofactiv-
ity."86 Rogers Brubaker suggests that there are a great variety of meanings of ratio-
40 CNFORMATION AND POWER

nal in Weber's work, yet we can see most of them as being inherently compatible
with an underlying theme. Brubaker includes "deliberate, systematic, calculable,
impersonal, instrumental, exact, quantitative, rule-governed, predictable, me-
thodical, purposeful, sober, scrupulous, efficacious, intelligible and consistent."87
Although Weber identified rationalization as a characteristic of Western soci-
ety, he was less"than sanguine about the results that flowed from its adoption and
spread. The rationalism that undergirds capitalism creates a constraining "iron
cage." With regard to bureaucratic organization, rationalism has dehumanization
as its consequence. Scientific rationalism in Weber's view also heightens the possi-
bility for political, social, and economic manipulation. 88
Rationalization involves the application ofmethod to tbe pursuit of identifi-
able goals in administration in the same ways that method supports the discovery
of knowledge in the science of sociology. Yet there are problems and contradic-
tions. Robert Holton and Bryan Turner join Christopher Dandeker in noting
Weber's ambivalence about the demand for equality as a matter of social rationali-
ty. 89 "To guarantee equality of condition or equality of outcome requires extensive
political intervention, which may involve the increasing surveillance and subordi-
nation of population to government regulation." 90 Thus, equality and freedom are
placed in opposition by the structural and procedural requirements of rational-
ization,
Although rationalization was a goal, Weber recognized the existence of prob-
lematic, perhaps insurmountable, methodological difficulties involved in the ra-
tionalization of social policy. Indeed, David Beetbam's comments suggest that al-
though Weber's emphasis on rationalization and the technical superiority of tbe
bureaucratic form runs throughout his economic writing, Weber's "political writ-
ings were concentrated explicitly on its negative, on what it could not achieve."9 1
Holton and Turner suggest that "Weber was particularly adamant on the impossi-
bility of aggregated measures of welfare, due to the incommensurable subjective
values given by individuals to their utilities." 92 This problem continues to vex ef-
forts to develop measures of well-being that might serve in the contemporary ad-
ministration of social justice.93
Weber held that there was a critical distinction between substantive and formal
rationality. The formal calculative aspects of rationality were generally not at is-
sue. Individuals will depart in their specifications and assessments at the level of
goals and values and, as a result, their subjective estimates of rationality will differ.
As Holton and Turner note, "what counts as substantively rational within a group
or a culture cannot be reduced to a consideration of the formal rationality of the
process of want-satisfaction. Hence the conflict between optimal economic ratio-
nality and the substantive rationality of protesting classes or disprivileged status
groups striving for justice is endemic."94 ·
Weber's analysis of the development of a rationalized system of law took note
of the preference that patriarchal and theocratic powers had for a system of laws
that reflected a substantive rationality. The law that worked best was one that
INFORMATION AND POWER 41

served their interests best, and this often meant that formality or even logical con-
sistency represented a hindrance. 95 However, when the interests of an emerging
bourgeoisie coincided with the willingness of the monarch to exchange legal flexi-
bility for economic support, the movement of the law toward formal rationality
accelerated. Ofsome interest is the emergence of the disciplinary specialization of
trained jurists who would come to replace the "lay notables of the old folk justice:'
This transition was brought about in part by "the growing need for their special
skills in the administrati~n of justice. These skills consisted above all in the 'capac-
ity to state dearly and unambiguously the legal issues involved in a complicated
situation,' and that capacity resulted from special professional training in a uni-
versity."96 Here, as elsewhere, rationalization involved professionalization and the
emergence of yet another discipline.

Efficiency
Weber's economic sociology was not concerned directly with microeconomic
questions of efficiency. Marxist scholars have made the most i.nfluential contribu-
tions to our understanding of the social consequences that flow from the rational-
1

ization of th.e process of exploitation under capitalism. Harry Braverman has been
identified as having made one of the most significant contributions to our under-
standing of rationalization in the manufacturing process. 97 Braverman's discus-
sion of the progressive division of labor emphasizes the instrumental pursuit of
savings in labor time. The less paid labor time that might be required for each unit
of manufacture produced would mean an increase in the amount of surplus value
that became available for capitalist investment or consumption. Frank Webster
and Kevin Robins explore the pursuit of efficiency and control over labor as a joint
product and goal of the technological system. 98 This process of rationalism is seen
to have moved from the time-and-motion studies of Frederick Wmslow Taylor,
through the intensification of managerial control, to the assembly line referred to
as FordisQJ..
Characteristic of this pursuit ofefficient control is the "separation of head from
hand:' which is involved in capturing and installing the knowledge and intelli-
gence of skilled labor into the hardware and software of automated p rocess and
control. 99 The '{ieskiUing" of the work force may be seen as an externality rather
than as an explicit goal of management. Webster and Robins describe the exten-
sion ofFordist principles to the realm of governance and political rationalism w1-
der the rubric of Social Taylorism, which reaches its extreme in the complete mo-
bilizati?n of ~ociety within the panoptic system. 100
Webster and Robins's analysis shares much in common with Beniger's analysis
of the "control revolution." 101 Crisis and a technological response to crisis are at
the heart of Beniger's dialectical chain of events. Beniger explains the growth in
the use of information technologies as a response to crises that emerged in pro-
duction as information controlled technology and managerial authority in-
creased output beyond the ability of the transportation system to move goods to
42 INFORMATION AND POWER

market. The need to coordinate the schedules of railroads was satisfied, at least
temporarily, by the invention of timetables and was facilitated by the spread of
control through the telegraph. Crises of consumption, which emerged as the dis-
tributional bottlenecks were overcome, were themselves regulated by the develop-
ment of controls over mass consumption through marketing, retailing, and mar-
keting research. This application of scientific management to the consumption
function is described by Robins and Webster as "Sloanism," after Alfred P. Sloan
of General Motors who presaged "just in time manufacturing" by not producing
an automobile until a buyer for it had been identified. 102 Similar control of gover-
nance is described in the spread of the bureaucratic system.
A key aspect of Beniger's definition of rationalization is the efficiency gains as-
sociated with what he calls preprocessing. Control is enhanced through the elimi-
nation of "unnecessary" information. That is, rationalization, as preprocessing,
"might be defined as the destruction or ignoring of information in order to facili-
tate its processing." 103 In asking us to think·about the amount of paperwork that
might be involved in the processing of the hundreds of forms that aFe generated as
increasing complexity in systems of production, distribution, or governance re-
quires more information for the management of the entire process, Beniger su·g-
gests we ought to imagine "how much more processing would be required, how-
ever, if each new case were recorded in an unstructured way, including every
nuance and in full detail, rather than by checking boxes, filling blanks, or in some
other way reducing the burdens of the bureaucratic system to only the limited
range of formal, objective, and impersonal information required by the standard-
ized forms." 104 This preprocessing accelerates the dehumanizing standardization
and the disciplinary normalization that depersonalizes and restricts the freedom
of individuals who are, as Ellul suggests, made to fit the technology, which has be-
come part and parcel of the panoptic sort.
Preprocessing is a standardizing technology. As Beniger suggests, standardiza-
tion increases the efficiency of most technical systems. It makes the processing of
information efficient in the same ways that efficiencies are realized in material
production. Standardization facilitates routinization, reduces the skill needed,
and allows for the replacement of labor with machines. The efficiencies realized
through standardization are not costless. Success in marketirig involves, at least to
some degree, the creation of the impression of difference. The control of workers
and consumers alike, at some level, requires the appearance of individuation.
Standardization is relative, is conceptual, and may take place at the level of ap-
pearance. Thus, as Spiros Simitis suggests, the appearance of great variety and
choice in the marketplace is fundamentally illusionary, as consumers must choose
not from an infinite range but from within the predetermined and highly stan-
dardized categories presented by the market. With reference to "high-tech;' inter-
active media offerings, Simitis suggests that "the media supplier dictates the con-
ditions under which communication takes place, fixes the possible subjects of the
dialogue, and, due to the personal data collected, is in an increasingly better posi-
INFORMATION AND POWER 43

tion to influence the subscriber's behavior. Interactive systems, therefore, suggest

individual activity where in fact no more _than stereotyped reactions occur." 105
What is important for us to emphasize at this stage is that the creation of stan-
dards, categories, classes, and options, although they may be contested from time
to time, is the product of a process dominated by ageI).tS in organizations, not in-
dividuals who are the objects of a system of power.

Measurement and Calculation

Preprocessing and standardization, wherever they occur, involve measurement,
calculation, and comparison with a priori or empirical standards or distributions.
The measurement, almost by definition! is imprecise. 106 Think for a moment
about the simple measurements that might be required to facilitate the identifica-
tion of persons: gender, race, age, height, and weight are all standard measures
with varying degrees of imprecision. Although sexuality and sexual identity have
become issues of_ some importance in recent times, we need not be informed
about the core of these debates to agree that we could differentiate between indi-
viduals along some continuum that we might identify as maleness/femaleness.
Preprocessing allows only a dichotomous response: You are one or the other.
Questions of individual choice in self-definition arise in this area almost as readily
as they do in terms of classification by race. Where is the line one crosses in order
to become black rather than white? ls there a blood test? When racial identifica-
tion becomes a question of ethnicity, the question of"who is Jewish" or "who is a
Native American" becomes a question of considerable political and economic
concern. The rules tend to be determined administratively, for administrative
purposes, and reflect powers inherent in traditions that have become institution-
alized in law.
Not even measures of height and weight are immune from a challenge to their
accuracy. Many of the measures are based on self-report and tend toward what
people believe to be socially preferred. The tendency to ignore that ~hich cannot
be easily measlired is a criticism that is frequently directed at empiricist social sci-
ence. It is an especially important aspect of the radical critique of political econo-
my.10' Benefits and costs that are not easily measured in terms of dollars and cents
are excluded from models seeking to evaluate the welfare effects of a given social
policy. Emphasizing the easily measured cannot help but distort the decisions of
actors whose reflexive monitoring is influenced by the limited availability of envi-
ronmental cues. The same distortions play through and reproduce systemic con-
straints when institutional decision making comes to depend on the accessible.
It might be argued that each of the measurements that become part of a per-
son's history ought to include relatively detailed information about the circum-
stances in which an action, such as being fired, laid off, or even changing jobs vol-
untarily, has taken place. The same actions are bound to have had quite different
structuring circumstances, and this information is lost to a system that forces that
complexity into a form that demands that you "list the jobs you have held in the
44 INFORMATION AND POWEH

past ten years.'' The same is also the case with regard to the record of late or missed
payments that might come to seriously mar a person's credit history. Yet, from the
perspective of the record keeper, who wants to either include or exclude an event
from a specific class, the details are unnecessary, and the costs of organizing and
storing them are excessive. Mario Bunge adds: "The elements of the colle.c tion will
be the same in certain respects, never in all particulars.'' 108
It is in the context of measurement as interpretation that Giddens's notion of
the double hermaneutic demands consideration. l 09 In the social sciences, proba-
bly more than in the case of the physical sciences, 110 the concern with reactivity is
well founded. 111 Yet Giddens describes a more circuitous route through which
measurement and analysis transform an object. The theories and concepts that
are developed by social scientists are routine) y inco1pora ted into the language and
understandings that help to structure the societies that they have studied. Giddens
suggests:

Because i.n some respects it is only possible for the social scientist to keep "one jump
ahead" of those whose behavior he or sl1e is investigating, much soda! science ap-
pears relatively banal to lay members of society. Yet this seeming banality disguises
the tremendous practical impact which social science has had and which is in sub-
stantial part constitutive of modernity.. .. My point is not just that social science
concepts and finctin gs "influence" "the ways in which we think;' but that they be-
come in large part constitutive ofthe practices whicl.1 form institutions of modeniity. 112

Social interest in 01easurement is not limited entirely or even primarily to a

concern with the explanation of causation. Bunge identifies a complex ofdetermi-
nations, none of which are necessarily causal. 11 ~ As he suggests, "determination is
a vector in a space of a large, as yet unknown number of dimensions, causal deter-
mination being just one of its components or projections." 114 Beyond causal de--
termination, Bunge identifies quantitative self-determination, such as we see
when physical systems in motion continue in motion (to the extent that motion is
not hindered by gravity and/or friction) through the force of inertia. Social sys-
tems may be argued to continue in a similar manner until changed by some force.
In biological systems, there is also a form of interactiona.l~reciprocaJ causation, a
functional interdependence of the sort that describes some glandular function. In
addition to mechanical and statistical determination, Buuge identifies structural
determination, which describes the influence of an organism on its constituent
parts. Social organisms, such as groups and organizations, may be seen to be de-
terminative, although not causal, with regard to the behavior of their membe~
The members of an organization may be influenced by a form of teleological de-
terminism, wherein the ends determine the means. A biological example can be
seen in the production of saliva in the presence of, or even in response to the men-
tion of, a lemon. CJearly the lemon did not cause the salivation, but the body
preparation for processing and consumption (goal) helped to determine
lNFORMATION AND POWER 45

means (saliva). In the social realm, we might consider the influence of goals on the
behavior of group members who are committed to those goals as determinative.
Bunge also identifies a form of qualified self-determination, which he refers to
as dialectical determination. Here we find the contrasting interests of contending
groups acting to change the very social structure of these groups in a qualitative
rather than a quantitative manner. In recognition of the complexity of determina-
tion, it is frequently sufficient to demonstrate a relationship. Jn the context of the
exercise of power, the value of a demonstrated relationship is in the interpretation
of a consistent pattern as an instrumentalism we know as prediction.

Prediction and Coordination

Prediction is of central importance to the panoptic sort. It is not only the insur-
ance industry that is interested in predictions of the outcomes likely for classes of
individuals. Predictability, the reduction of uncertainty about individual
behavior, is a valued aspect of social systems and the social relations within them.
Power may be seen as the ability to act in such as a way as to induce a desired and
predictable reaction. It is in the pursuit of improved predictability that the
panoptic sort is engaged, and it is calibrated on the basis of a limited amount of
information about a great number of individuals.
In the context of predictability, we see the interplay between generality and
specificity. Whereas disciplinary power operates on individuals through their
membership in groups, the systemic orientation to groups is based on informa-
tion gathered about the experience of individuals. "Very few facts in the concrete
world are predictable with near certainty, and none in all detail, because scientific
forecast is based on the knowledge of laws and of specific information regarding
singular facts, neither of which is ever complete and exact:' 115 In addition, with re-
gard to Giddens's insights about the influence of social theory, the identification
of the criterion as well as the independent variables in any predictive model are re-
flective of the thinking of the tin1e, rather than determined by any ontologically
determined necessity. Interdependence was certainly a fact long before social the-
orists recognized it and demanded that it be incorporated into analytical mod-
els. 116 According to Thomas Haskell,

Social scientists in the i89o's and in the Progressive era located the effective causa-
tion of human affairs sometimes in heredity, sometimes in environment; sometimes
in the play of economic interests, sometimes in particular historical conditions, such
as the frontier experience; sometimes in stimulus-response associations, sometimes
in instinct, sometimes in culture, sometimes in that alien area within the person but
remote from his conscious mind, the subconscious. They seldom located causation
close to the surface of events or in the conscious, willing minds of individuals. 117

Giddens's emphasis on human agency may take us closer to this surface, but be-
cause he places agency in the context of a theory of structuration, the heritage of
positivism still weighs heavily on his thinking.
46 lNFORMATION AND POWER

Yet it is not clear that the operation of the pan optic sort depends on a compre-
hensive theory of causation. Indeed, incompatible theories of agency may charac-
terize its application at different sites, and at the intersections between them.

Bureaucratization and Surveillance

Christopher Dandeker's investigation of surveillance since the l700S links both the
collection and the maintenance of information in files as essential characteristics
of the bureaucratic form of organization. As a technology, the bureaucratic orga-
nization depends on information and knowledge being controlled and modified
if necessary by the organization. For Weber, the ability of rhe bureaucratic form to
manage this information efficiently explained the growth of bureaucracy:

The decisive reason for the advance of bureaucratic organisation has always been its
purely technkal superiority over every other fo rm. A fully developed bureaucratic
apparatus stands to these other forms in much the same relation as a machine does
to non-mechanical means of production. Precision, dispatch, clarity, familiarity
witb the documents, contin\llty, discretion, uniformity. rigid subordination, savings
in friction and in mate.rial and personal costs-all these things are raised much more
effectively to the optimal level by a strong bureaucratic. especially a monocratic, ad-
ministration with trained individual officials than by any form of collegiate, honor-
ific or avocational admioistration. 118

Centralized control of a common informationa.l core introduces a reliable

standardization, avoiding the rigidity of traditional authority and, at the same
time, avoiding the apparent irrationality of individual personality and style. In
Dandeker's view, the bureaucracy is very much like a machine "to the extent to
which subjective or irrational elements of will and mood are eliminated." 119
Dandeker suggests that there is considerable variety in the cap acityofbureau-
cracies to engage in surveillance. Organizations vary in terms of the amount of in.-
formation they gather and maintain in centralized files, in terms of the spe·e d with
which information flows into and out of files and between members of the organi-
zatfon, and also in terms of the number of points of contact between the organiza-
tion and the population under surveillance. Advances in computerization and
telecommunications have the potential to increase each of these indices of bu-
reaucratic scope, which, by definition, increases the power and reach of the bu-
reaucracy.
Dandeker intimates that, historically, surveillance of the work force was both
facilitated and necessitated by the application of the principles of scientific man-
agement in ways that resulted in an increased division oflabor. Standardization,.
monitoring, record keeping, and statistical analysis were the cornerstones of bu-.
reaucratic surveillance and control of the labor process. This detailed burea
cratic specification not only increased managerial control, but it also xeduc.ed
level of skill and the training time required to reach it. This deskilling further
l)IFORMATION AND POWER 47

duced the bargaining power of individual laborers because they could easily be re-
placed by others who could learn to perform the job just as well in a relatively
short time. Bureaucratic rationalization through surveillance made it possible for
management to compare individual workers to empirical or managerial norms of
output or method, and performance could be linked with compensation.
With computerization, remote surveillance of labor has come to mean that a
supervisor in another town might be responsible for assigning calls to telephone
operators on the basis of continuous assessment of their performance and the per-
formance of others.120 Employees in the growing telemarketing sector are under
continuous surveillance, and some even receive computer-generated printouts at
the end of the day that locate their average performance plotted against organiza-
tional norms. 121
Bureaucratization of the capitalist firm involved an expansion ·of scope from
the surveillance and control of its employees to include surveillance of its custom-
ers and competitors. Just as surveillance of the work force has moved beyond the
confines of the factory and the office into the personal lives and the psychology of
the workers, surveillance of consumers has moved into life-styles and psycho-
graprucs deemed relevant to the rationalization of the marketing function. As the
modern corporation grows in size and global reach, 122 telecommunications be-
comes even more important:

Taking advantage of electronic means of surveillance and techniques of transferring

information, financial and other resources, such enterprises have sub-divided their
operations and located them in different countries depending on product markets,
labour costs and so on. Modern large corporations can control these dispersed
operational units because electronic techniques of surveillance break through the
space-time limitations of communications systems based on word of mouth, hand-
writing and the printed word. 123

Dandeker shares with Weber the belief (although perhaps not the same level of
concern) that bureaucracy will continue to expand, and he rejects the claims of
the futurists such as Alvin Toffler who project the emergence of an "ad hocracy,"
or some other post-Fordist organization of economic activity. This is not to sug-
gest that Dandeker does not envision changes in the nature of organizations, as
there are dearly changes under way, especially with regard to the specification of
the boundaries between organizations. He suggests that we should not see these
changes in terms of a "decline in the significance of the vertical principle of bu-
reaucratic organization" because a decentralized structure does not mean a weak
center; it suggests only that there is an alternative and perhaps less visible system
of power. 124 Whether the structure of power and authority is centralized or dis-
persed and operated interdependently, there is no question that postmodern sys-
tems will still depend on knowledge of their parts and a shared sense of the legiti-
macy of the actions taken by seemingly autonomous units.
 INFORMATION AND POWER

Legitimation
Weber believed that "every type of domination depends on an administrative ap-
paratus under a chief or ruling body and on a shared belief in the legitimacy of
rules or decisions." 125 The apparent predominance of rules and regulations in the
bureaucracy, with its formal policies and procedures manuals outlining standard
practices, provides one form of legitimation seen as the absence of arbitrariness.
Legal authority differs from traditional or charismatic authority in that it is based
on the presumed rationality of the laws, policies, or procedures that have been de-
veloped. 126 The formal bureaucracy represents the institutionalization of the rules
and resources that are at the heart of Giddens's process of structuration. Not only
are rules in bureaucratic organizations operational, they are written down and
avaHable to all and provide an external source of legitimation of action in that
rules can be referred to as a source of guidance in the case of dispute. The bureau-
cratic organization also specifies which individuals have which resources or
powers. The rules define spheres of competence and responsibility. The resources
within a bureaucratic organization resl not with the individual but with the posi-
tion. This "separation of the office from the incumbent" allows a bureaucratically
structured organization to function more efficiently because it can function con-
tinuously through the routine replacement of individuals. 127 Thus, even though
thepopularityofprofessional specializations may ebb and flow, the importance of
legitimacy applies to expertise. For the power of disciplines to have their force, le-
gitimacy has meant that the professiotts have to be invested with the cloak of au-
thority. Thomas Haskell suggests that the professionalization of social science was
a movement to establish authority, and, further, "with ironic regularity, the move-
ment to establish authority rose .above even ils most selfish intentions to defend
not class interest, but authority as a gen.eral principle." m
With regard to the function of a state bureaucracy,, however, a lack of stal?ility
in the norms and goals of the political center can serve to threaten the efficient op-
eration of the governmental systems and raise questions about the Jegjtimacy of
its decisions. Reaganism and Thatcherism, which represented attempts to redirect
the focus and priorities of the state and iJwolved an attack on the government bu-
reaucracy through privitization, are modern examples of the kinds of distur-
bances that bureaucratic systems have to overcome in the face of contradictions
between structure and goals. 129 The "Reagan revolution" represented a challenge ,
to the Keynesian model of an economically active state and elevated "the market"
to a position of centrality in the ideology of change. In the view of one analyst, we
find that the " market is reconstituted as a major ideological force and crucial dis-
tinctions between the productive and unproductive, private and public, wealth
creating and wealth consuming, come to be lhe yardsticks for judging policy." 130
In one sense, the market stands in opposition to the bmeaucratic ideal. The
market as a conceptual entity is autonomous and unpredictable in all but the
character of its theoretical equilibria. Yet its hypothetical flex.ibility was thought to
make its efficiency superior to that of a government bureaucracy rigidified by the
C-'FORMATION AND POWER 49

interests of bureaucrats and professionals, as well as the pervasive influence of or-

ganized interests that had come to depend on the government budget The pursuit
of"rule by markets" changed the purpose of government surveillance to a more
explicitly disciplinary focus, and it legitimated its expanded scope on the basis ofa
critical need to "eliminate waste, fraud, and abuse" in government.131
According to Andrew Gamble, the "achievement of the New Right was to place
the supporters of public provision and pubUc services on the defensive. The New
Right attacks on the grounds of cost and efficiency were predictable, but what was
less expected was the New Right claim that they had a superior moral vision of
what a free society should be like. Their concept of citizenship saw freedom and
equality being achieved through the daily plebiscite in the market, not through
the infrequent plebiscite in the political system:' 132 The government bureaucracy
faced a legitimation crisis that was heightened by the fact that the executive
branch saw its professionals as enemies and its rank and file as dishonest
slackards. 133
Thus we see that as social systems evolve differences in ideology can modify the
approaches taken by knowledgeable agents to realize their goals. Changes in pol-
icy reflect changes in beliefs about which technologies, including those of a state
bureaucracy, are the most efficient means available for the realization of identifi-
able goals. In the next section, Texamine information technology as a particular
response to the demand for rationalization.

Information Technology
and the Ratfrmalization ofControl
A considerable amount of work has been put forward that attempts to distinguish
between the structures and orientations that characterize the United States and
other advanced industrial economies from those of earlier stages after their entry
into the industrial revolution. A good part of the debate has surrounded the issue
of whether or not this period, which is characterized as an Information Age, rep-
resents a significant, epochal change or is merely a continuation of capitaHsm as
we have known it. Without suggesting that the issue is really a nonquestion, let us
consider the possibility that it really will not matter unless the response of people
and the institutions they control behave in a fundamentally different way because
they do or do not accept that view. We are, for example, somewhat at odds as to
whether the Wormation Age is upon us or whether it is just ahead in the not too
distant future.
Wilson Dizard entitles his book The Coming Information Age. 134 On the one
hand, be discusses changes that are taking place but he suggests that we are only in
the first stages of this new age. On the other hand, James Beniger speaks of his
control revolution as having begun more than a century ago. Peter Hall and Pas-
chal Preston, reflecting the views of long wave theorists, suggest that we are not
approaching an epoch at all but are merely on the verge of the fifth Kondratieff
upswing. 135 Jorge Schement and Leah Lievrouw's chapter in their edited volume
50 lNFORMATION AND POWER

offers a compromise position that suggests that ''the information phenomena re-
flect the continuing evolution of industrial capitalism, which has resulted in an
information-oriented society in the United States." 136 Some of the distinctions
drawn by these analysts revolve around precisely what it is that observers take as
markers of significance in defining the period of interest.
Dizard notes a flood of publications that discuss the current period as being
post-something-or-other (as in postindustrial from Daniel Bell) and takes due
note of the variety in indexes that these authors use as markers of this change. A
chain of researchers from Fritz Machlup through Daniel Bell and, more recently,
Marc Porat, have emphasized the changes in the composition of the labor force,
the members of which can b e identified as working primarily in the production or
transformation of information or knowledge, rather than in agriculture, manu-
facturing, or noninformation services. Porafs work137 stimuJated hundreds of
imitative studies that sought to determine the extent to which other nations were
or were about to become information societies. 158 If, on the one hand, one's pri-
mary concern is the projection of the character of the labor force and the demand
for a particular kind oflabor, such predictions are perhaps meaningful.
If, on the other hand, one's interests are more general and one is concerned
with economic pJanning at a national level, then perhaps with the aid of an input-
output model, questions of investment, trade, and sectoral linkage become more
1
important than labor force character per se. 139 For example, Merheroo Jussawalla
disCllsses several input-output analyses of Pacific rim countries arid notes that
Singapore's investments in information networks appear to be appropriate given
the tight information related linkages of five of the ten most important sectors in
the economy. 110
However, if one's interests lean toward a critique of capital and an understand-
ing of the ways in which the new technologies have been incorporated into (all the
while transforming) capitalist social relations, then other measures become more
important. Traditional Marxists struggle with the conceptualization of productive
and nonproductive labor and wonder aloud if the growth in the number of non-
productive information workers represents a critical crisis. 141 Although the theo-
retical basis for the construct is markedly different from that of neoclassical eco-
nomics, there is an underlying commonality in the concern about the
contribution of information technology and the information work force in the re-
alization of the goals of capitalism. Information, information technology, and tht
informational work force all represent the products of a broadly shared instru-
mental rationalism linked to the realization of profit, or surplus value. Part of the
challenge in making sense of the apparently cyclical "swarms" of innovations 1~
that are differentially linked to a decline in economic expansion is to identify
underlying logic that selects data processing, pattern recognition, or mobile co
munications as the leading edge technologies at one period rather than
other. 143 The nature of the interaction of organizations with their environm
l'rORMATION AND POWER 51

helps to determine the kinds of information and information technology that

their leaders adopt and adapt to meet their objectives. 144
Although Beniger identifies information technologies as basic as the manual
typewriter, most contemporary writing about the Information Age is focused on
the twin technologies of the computer and the telecommunications networks that
facilitate their interoperability. 145 Advances in both technologies are discussed in
terms of the increased speed, capacity, reliability, and ease of operation that char-
acterize these systems. This increase in capacity, or power, has been accompanied
by a phenomenal reduction in the costs of processing or transmission. Miniatur-
ization increases the number and variety of devices that can be made more "intel-
ligent" with the addition of microprocessors. Miniaturization also increases the
ease with which individuals can be tied into the telecommunications network, no
longer bound by wires, cables, or unwieldy interface devices. 146
A good part of the contribution of computer and telecommunications systems
to the character of the Information Age is based on the ability of these systems to
process information in a way that provides intelligence about the environment.
The computer-high speed, high capacity, perhaps designed to operate in parallel
with hundreds of other microprocessors-can scan unimaginable amounts of in-
formation in search of patterns and relationships that help decision makers make
sense of the world. Although a great many observers are quick to challenge the ac-
curacy, and warn of the consequences, that flow from acting on this intelligence,
the computer makes it possible to take far more information into consideration
than has ever been possible. 147 The telecommunications network makes it possi-
ble to access this capacity from virtually any place on the earth (and of course,
places not on the earth, as well). This low-cost capacity to collect, store, process,
and share intelligence increases the pressures on institutionalized actors to gather
and make use of even more information. 148 A question remains as to the effect
that this intelligence has on the productivity of individuals, organizations, or
economies as a whole. 149
The claims of the Futurists, from Daniel Bell through the popularizers such as
Alvin Toffler, are challenged by Marxist and non-Marxist analyses of the produc-
tivity of the information work force. Charles Jonscher's analysis suggests that in
the United States, at least, the rapid growth in the size of the information work
force had already peaked and was beginning to decline as continued rationaliza-
tion required the introduction of labor-saving technology in the office just as
surely as it had been introduced into the factory and on the farm. ' 50 Webster and
Robins take special note of the likely consequences of this rationalization on
women in the work force, who happen to be concentrated in the routine informa-
tion jobs most susceptible to elimination through automation. 15 1 Yet it is also
dear that the pressure to develop intelligent systems through a process of knowl-
edge engineering not unlike the time-and-motion studies of Frederick Taylor will
only increase in the future., and reductions in force will be visited on the ranks of
middle-level managers. 152
52 INFORMATION AND POWER

As recognized by Katz 153 and Arriaga 154 and others, much of the information
work force in Third World nations, especially those on the verge of economic de-
velopment, is actually employed in nonproductive government bureaucracies. In-
ternally and externally applied pressure to accelerate privatization will require
substantial reduction in this part of the information economy.
Futurists, economists, and Marxists alike have identified another aspect of this
period that demands more attention than it has been given to date. Not only is in-
formation a critical input into the process of rationalization and control, but it
has also emerged as a commodity in its own right. Information producers have
been organized into independent organizations that serve a variety of clients. The
commoditization of information has not proceeded smoothly, in part b~cause of
the peculiarities of infonnation as a resource. Critical questions emerged and re-
main about the ownership of information, not the least of which has to do with
the ownership of information about individuals that is produced as a byproduct
of their interaction with organizations in their roles as employees, citizens, and
consumers. The panoptic sort, as the integrated contsol technology of the Infor-
mation Age, depends for its operation on ready access to information about indi·
viduals. Tbe debates about informational privacy are a reflex.ive index of the
stresses and strains that are created as knowledgeable actors, at various sites in in-
dependent and intersecting systems, struggle to identify, define, and institutional-
ize an orientation toward personal information that will allow the process of ra-
tionalization to continue in ways that are consistent with their goals.
Yet the issue is not simply a question of privacy. Ln the report to the Club of
Rome regarding the role of information, Klaus Lenk suggests that there is an error
in focusing our attention on the issue of privacy as it is generally formulated. "Tbr
real issue at stake is not personal privacy, which is an ill-defined concept, greatfr
varying according to the cultural context. [tis power gains of bureaucracies, boda.
private and public, at the expense of individuals and the non-organized sectors
society, by means of gathering of information through d irect observation and
means of intem1ive record keeping." 155 The issue is the panoptic sort.
3
OPERATING THE PANOPTIC SORT

DATA AND THE PANOPTfC SORT

The panoptic sort depends on data that are made from the raw material of human
experience. This chapter will explore the ways in which this information is gath-
ered, data are generated, and intelligence is shared among the bureaucracies of
business and government.
The panoptic sort, as a djscrimjnatory technology, has much in common with
content analysis as a technique for drawing inferences about the source of sym-
bolic materials. Klaus Krippendorff's cliscussion of the kinds of methods used and
inferences drawn by content analysts over the years helps to make the connection
clear. 1 Core figures involved in the development of the method share an interest in
the pragmatic, instrumental uses of a technology that will support the develop-
ment of inferences about the causes, circumstances, or unseen states of the source
of communications. The source has been detioed variously to include individuals,
especially political actors, as well as entire nations. Krippendorff defines content
analysis as "a research technique for making replicable and valid inferences from
d.ata to their context." 2 This context may be understood to refer to the environ-
mental circumstances, which may include tbe psychological state of the commu-
nicative source, which gives r ise to symbolic content with a particular character.

The target is what the analysis will want to know about. Since content analysis pro-
vides vicarious knowledge, informatiou about somel'hing not directly observed, this
target is located in the variable portion of the context of available data.... The task is
to make inferettces from data to certain aspects of the.ir context and to justify these in-
ferences i.n terms of the knowledge about the stable factors in the system of interest. 3

The use of content analysis in the study of propaganda to assess military strat-
egy, to predict troop movements and the readiness of an enemy for surrender, has
ready parallels in the assessment of consumer culture and the prediction of trends,
fads, and tendencies on the basis of content analysis of newspapers and popui<u
media. Both mllitaryand marketing strategists rely heavily on such assessments to
plan their moves against targets.
In the production of intelligence about the relationship between context and
content, analysts frequently rely on independent sources of information about the

53
54 OPERATING THB PANOPTIC SORT

environment. Studies that link the content of presidential speeches to the status of
the economy must depend upon government and corporate statistics to define the
environment. The panoptic sort is similarly dependent on multiple sources for in-
formation about the contex1:s in which its targets ex:ist. Concerns about the reli-
ability and validjty of such data are common, but the panoptic system acts to ad-
dress the problems of data standards and compa tability in ways that clearly exceed
the resources and vision of any single content analyst. 4
The panoptic sort differs from content analysis in other dimensions as well.
Content analysis avoids sources of reactive error in that it is generally an unobtru-
sive technique and does not depend on compliance of the source as long as the
symbolic materials are readily available (as is the case with mass media content).
The panoptic sort, however, depends in part on subjects' awareness of the fact (m;
at least the possibility) that thefr behavior is under surveillance at the numerous
points of contact they have with bureaucratic systems. A, similarity with content
analysis emerges again, however, in the fact that individuals are never aware of the
variety of interests that will have access to personal information, nor can they
imagine all the analytical and strategic uses to which this personal information
may be put. Thus, at many levels of the system, analysts engaged in surveillance
and in the production of strategic information about citizens, employees, and
consumers function behind a cloak ofinvisibility-a one-way mirror.
The panoptic sort, as a technical system, has an orientation to data similar to
that of content analysis. Both analytical systems must be able to handle large vol-
umes of data. Both must develop methodological strategies to ensure the useful-
ness of the intelligence or the inferei1ces that are produced by the analysis.
Krippendorff refers to au analytical process that includes four primary steps: data
making, data reduction, inference, and analysis. The process of data making in-
volves establishing a link between attributes and measures and always contains the
risk of error, both systematic bias and errors associated with the instability in the
behavior ofinterest. Errors are introduced into the pan optic sort in the same ways
that they are introduced into a content analysis. The errors may be inherent in the
specification of the construct, or they may reflect unreliability in the process that
captures the behavior through observation and tt:ansforms it into data suitable for
processing.
Content analysis may involve unmanageable amo1mts of data, such as a cou.nt
of the number of times in which a female character is physically abused by a male
character in a year's productions of theatrical film. Because the behavior is so
common, a method of sampling these films is necessary for the task to be made
manageable. By the same token, for some purposes, the panoptic sort must rely on
data generated from probability or purposive samples. For many purposive sam-
ples, the panoptic technology searches for key words, events, or relationships.
Files that display any of those key markers may be selected for further analys~
This analysis of raw hits is common to the disciplinary surveillance of bureaucra-
cies concerned with truces, credit, and insurance.
OPERATING THE PANOPTIC SORT 55

Data reduction in content analysis finds ready parallels within the panoptic
sort. It is part of the analytical process but is similar to the strategy of preprocess-
ing that seeks to reduce the amount of information to that which is the minimal
efficient scale for the production and sharing of useful intelligence. Data reduc-
tion involves the search for similarity and difference expressed in convenient,
communicable terms. It may be visual--charts, maps, graphs, or models- and it
may be linguistic, as in the labels assigned to the forty geodemographic clusters
like "back-country folks" used by Claritas/PRIZM to help marketers understand
their clientele or their circumstances.5
Finally, the panoptic sort, like content analysis, is a dynamic process. Measures
and methods are constantly being adjusted as the inferences about the targets of
surveillance are evaluated in terms of their contribution to the realization of the
organization's goals.

GOVERNMENT INFORMATION GATHERING

The U.S. government is a major source of information in the economy and about
the economy: that which it produces on its own and that which it brings forward
through a system of contracts and grants. My emphasis here is on the role of the
government at the federal and state levels in the collection or generation of infor-
mation about individuals and groups in society. A distinction is to be made be-
tween government statistics and personal data gathered by agencies of the govern-
ment for the purposes of administering a particular bureaucratic function. The
distinction is a tenuous one because there is considerable evidence that data move
back and forth between administrative files and those of statistical agencies, and
more important, representations of normality and deviance derived from general
statistics are used in bureaucratic decisions about the benefits or services that in-
dividuals are to receive. The distinction between government data and the data
collected and maintained by private bureaucracies is also difficult to maintain.
Data gathered by government agencies find ready use by private firms, and private
surveillance is increasingly relied on by government agencies to enhance their
own investigations.
The numbers of computerized files maintained by the agencies of government
continue to grow at a nearly exponential rate, despite efforts by the Office of Man-
agement and Budget (OMB) to restrict reporting requirements for business and
industry. 6 A study of the U.S. Office ofTechnology Assessment (USOTA) revealed
that 12 government agencies maintained some 539 records systems classified un-
der the Privacy Act, and that these systems contained more than 3.5 billion re-
cords, 60 percent of which had been computerized. Estimates of the number of
such records are likely to be grossly underestimated, because the act of recording
their existence has been complicated by the growth in the use of personal comput-
ers by workers in these government agencies. 7
 OPERATING THE PANOPTIC SORT

Much of the public's concern about government data gatheri11g has been fo-
cused on the system of files maintained by those government agencies responsible
for law enforcement, tax collection, and tl1e protection of national security inter-
ests against domestic and foreign enemies, real and imagined. Herbert Mitgang,
former president of the Authors' Guild, compiled a detailed analysis of the use of
dandestine surveillance to develop files on thousands of writers and artists be-
lieved to represent some threat to the U.S. government.8 A quote from John
Kenneth Galbraith illustrates tJ1e tension between actual and potential harm that
might result from the development of such files: "While the impression of other
people's paranoia is great, my own was diminished by d1e fact that while the docu-
ments are full of deeply damaging intentions, virtually nothing unpleasant ever
happened as a consequence. But om: can see how the only slightly more vulnera-
ble must have suffered." 9
Frank Donner's analysis of government surveillance is much less sanguine.'0
As director of the America11 Civil Liberties Union (ACLU) Project on Political
Surveillance, Donner is of the opinion that government surveillance represents a
substantial threat to political freedom in the United States. He suggests that the
Congress, the Federal Bureau of lnvestigatjon (FBI), the Central lnteUigence
Agency (CIA), the various armed.services, and grand juries have all abused their
power in the search for enemies. This continuing search for subversives by the FBI
included a program that would recruit librarians to identify "suspicious" users of
unclassified material of a "sensitive" nature who might conceivably represent ii
threat to the competitive or security status of the United States. 11 In a recent book,
Gary Marx focuses on the techniques of undercover police surveillance that i~
dude a variety of "stings," which brutally distort the definition of entrapment:. n
Marx, like Donner, emphasizes ·the development of data through surveillance as a
means of prevention that faci]itates an anticipatory controlling response.
Donner, "open-ended political surveillanee is really the collection of e.vidence
potential criminal conduct." 13 ·
It is important to note th.at the files gathered for purposes of surveillance
different from the files that are part of records of the criminal justice syste
such as the Criminal History System, which provides a n<1tional electronic in
to the files of persons with any federal or state criminal history. 14 However, the·
tegration of such files with surveillance records can be triggered in a great vari
of ways, making for a distinction without a difference. A recent proposal to im
menta point-of-sale system of automated record checks for persons attemptin
purchase firearms would access a proposed "national felons file," which w
conceivably include "suspect" flags triggering a notification to tbe FBI or o
agency. 15 Citing1974 data, Donner reports that the FBl had some 6.5 million
containing a great variety of raw data in different forms: photog~aphs, m
randa, and transcripts. Those largely paper files expanded and were tranform
the computer and its digital storage replaced some 58 million index cards. Do
also caUs our attention to the history of the lnternal Revenue Service (IRS)
OPERATING THE PANOPTIC SORT 57

notes its frequent subversion to political ends unrelated to its primary function of
tax collection. Donner claims that "the undercover capacity of the IRS outweighs
that of all the other federal civiLian agencies combined." 16
Political surveillance is not the only purpose for which the government main-
tains files on individuals. The criminal justice system and the records of the courts
have already been noted. The administration of the public welfare system repre-
sents a massive system of files and has been identified as a primary mobilizing
force behind the expansion of file matching programs. 17 For the purposes of veri-
fying the eligibility of individuals for social services or economic support, the fed-
eral government, under the Deficit Reduction Act of J984, required states to match
records in their files against dozens of other computerized files maintained by
public and private o rgani~ations. 1 8 The involvement of the government in the
maintenance of public health has also stimulated the creation of personal files.
Persons who are patients or who have received care in government facilities, who
have participated in government research programs, or even who have been diag-
nosed as possessing a genetic "defect" will become part of a government file, list,
or record. 19 David Flaherty suggests that the Social Security Administration has
data that "constitute one of the richest single aggregations of current demo-
graphic and economic information on individuals in the United States."20
Government data at the state levels include the countless public records noting
births, deaths, weddings, divorces, transfers of real property, voter registration,
and the issuance of licenses to drive or to provide professional services. In addi-
tion to the administrative purposes they may serve, these extensive records are a
valuable resource to countless providers of private goods and services, including
those who provide insurance, credit, and related financial services.

The Census
Paul Starr reminds us that the original purpose of the census was somewhat dif-
ferent from its present administrative purpose and dominant use. 2 1 Rather than
serving as a source of quantitative data, the census was primarily an aid to a state's
efforts in the surveillance of its population, especially with regard to the assess-
ment and collection of taxes. The differences between the current census and
those of the prernodern age are striking. The modern census is a count of the en-
tire population, whereas earlier counts may have been limited to households, and
within households, to adult males. For a considerable period, the census of old
was taken continuously, whereas the modem census process demands the expen-
diture of substantial effort and expense to complete the count within a limited
time. Although Starr sees the most important difference between pre- and post-
modern census forms in the contemporary separation of the activities of the col-
lection of government statistics from the collection of taxes, from the perspective
of the panoptic sort an equally important distinction is the fact "that statistical
data from modern censuses are typically expected to be published; prernodern
censuses were generally state secrets." 22
58 OPERATING THE PANOPtlC SORT

Starr's emphasis on the transfo rmation of the census from an instrument of

surveillance to a source of government social statistics can be understood as an at-
tempt by government to ensure the public trust and ready compliance with its
data gathering requirements. Concerns about privacy have generally lowered in-
dividual responsiveness to a variety of surveys, inclucling the U.S. census. Such
concerns have resulted in the cancellation of the census in the Federal Republic of
Germany because of fears about the use of the data by the state for the purposes of
disciplinary surveillance. David Plaherty's discussion of German surveillance
notes that in i983, some 53 percent of the population had indicated a mistrust of
the census, and at least 25 percent of the households would not cooperate by filling
out the forms. Flaherty echoes the sentiment of Starr in suggesting that the "fun-
damental problem with the planned census was the failure to maintain an abso-
lute functional separation between statistical and admjnistrative uses of personal
data." 23 The 1983 German census was postponed, and the government began a
process oflegislative and bureaucratic adjustment in response to the protest, but
the initial response was not enough. In 1987, political opposition to the census be-
came part of the platforms of the Green party and the Social Democrats, leading
the statistical office to spend some 45 million deutsche marks on public relations
to calm the fears that the political rhetoric had stirred. 24
fn the United States, the Bureau of the Census has been more successful in
maintaining public trust in the confidentiality of the data that it gathers. The pri-
mary means of guaranteeing confidentia)jty is the separation of inclividually iden-
tifiable information from their records if t11ere is to be any transmission of the in-
formation in the records to other parties, including other agencies of the
government. 25 Tt is suggested that in an effort to reduce the possibility of reestab-
lishing the linkage between a name and a particular record, the bureau will not
even return records to other government agencies that initially provided them for
verification or for estimation use if the borrowed records were individually identi-
fiable. 26
The United States has been a leader in the collection of government statistics by
means of the census. The role of the census in apportionment was defined by its
inclusion in the Constitution. A regular population census, established in 1790,
did more than count the population, however; it made distinctions on the basis of
race, gender, and status as slave. The numbers of questions and_the range of their
coverage of personal information increased steadily after i840. 27 In planning for
the 1990 census, interested parties, such as those represented by the American
P lanning Association, petitioned the bureau not to eliminate questions abour
housing, especially estimates of house sale and rental value. Despite the fact tha
census data are not to be used to the detriment of anyone complying with the re-
quest for information, planners suggested that they fea1·ed the loss of valua
data that they used for rental housing administration and enforcement.28 Al-
though the use of the data may not be the proximal cause, it is not difficult to
tablish the linkage between census data and the redlining of neighborhoods
 59

unities in ways that bring Lmquestionable harm. As will be discussed later,

rommercial and political importance and consequentiality of census data far
the formal purpose ofproviding guidance to the federal government.
In addition to the periodic censuses, the Census Bureau conducts hundreds of
timwys each year tracking population trends, trends in employment, and trends
ii consumer e.'{_penditures. Because of its expertise, it conducts many of these sur-
~ for other agencies of the government
Leadership in the development of the census in the United States was not
8i11Ched in other areas of statistical information gathering at the federal level.
. One interpretation of the lag was a distrust of centralized government, which
'-md the state governments collecting and publishing most of the public statis-
lics before the Civil War. 29 It was because of the expansion of federal government
RSp<>nsibilities during and after the Civil War that the collection of statistics be-
pn to expand and involve other departments and bureaucratic units. The cre-
.-ion of federal agencies and departments, such as the Department of Agriculture,
~Bureau of Education, and the Fish Commission, which became the Bureau of
fisheries, and later the Department of Commerce, ensured a steady expansion of
!he statistical reach of the government.
Several government agencies with responsibility for regulatory oversight
nnerged during times of heightened concern about harmful business practices at
the turn of the century. The creation of a federaUncome tax and an IRS after pas-
sage of the Sixteenth Amendment marked the birth of still another magnet for
personal data and statistics. The demands of war and military preparedness saw
the creation of a statistics branch within the War Department in i918. The infla-
tionary impact of war mobilization increased the importance of data about prices,
md the Bureau of Labor Statistics increased its responsibility during the war, in-
cluding its work on the development of the cost-of-living index.
After 1942 government statistics took a dramatic turn as a result of the whole-
sale adoption of the methods of probability sampling. These methods increased
the efficiency with which the government could keep tabs on more and more as-
pects of the economy under its charge. The Census Bureau developed a sampling
method to facilitate more efficient estimation of the rather e.'Xtensive list of per-
sonal attributes that was being sought at the time. Although there remained a set
of questions that the heads of every household would complete, other questions
\ornuld be asked of one out offive, one out offifteen, and one out of twenty house-
holds. 30 In the i97os the Statistical Reporting Service played a significant role in
the development of remote sensing methods to estimate crop production. In one
sense, we might consider all statistics based on probability samples as a form of re-
mote sensing, in that data preswned to represent classes or categories of individu-
als ru:e generated without any direct contact with the overwhelming majority to
whom the statistics apply.
Social scientists31 and others have expressed concern over the fact that more
and more of the data generated by government is available only through elec-
60 OPERATING THE PANOPTlC SORT

tronic means, which may serve to limit professional as well as public access to
those data. 32 Th.is concern is heightened by a concurrent tendency on the part of
the federal govenment to implement a policy of privatization, which makes access
to these data dependent on an ability to pay the third-party vendors who have
been brought into competition with traditional government information
sources. 33 At the same time that the electronic storage and access to government
data and statistics are made more efficient, a technical resource and skill barrier
has been introduced that serves to concentrate power at the centers, or nodes, of
the panoptic network.

THE CORPORATE DATA MACHINE

As with government, modem business enterprise depends on data. Corporate de-
cision makers require accurate and timely iinformation about theix own competi-
tive environments, the efficiency of their production process, and the extent of
success that is currently enjoyed by each of their product Lines. In addition, the
modexn business enterprise needs information about the future to use as a critical
input into strategic planning.J4 Information about the environment includes as-
sessments of the competitive potential of products in development compared with
those of their competitors. Industrial espionage represents the underside of a
search for intelligence that includes more legitimate data gathering methods such
as the continuous scanning of scientilic, ·technical, and trade publications, partici-
pation in conferences and seminars, and the anxious perusal of costly, limited clis-
tribution consultants' reports.
The short-term competitive advantages derived from product and process in-
novations require access to information that is generated by internal research
development, infused into tl1e organization through mutually beneficial linka
with former competitors or through purchase on the open market. Aith
Stinchcombe suggests that the form of the organization will reflect
uncertainties it faces and the means that it develops to deliver necessary info
tion to decision points with.in the system.35 Information about employees as co
ponents in the organization's production function is of particular import
with regard to the panoptic sort.
Organizations face an almost impossible task of selecting employees w•
whom they wiJJ enter into a long-term contractual relationship if that selectio
to include an assessment of that employee's potential contribution to a variety
tasks that have yet to be defined. There is ooJya slight improvement when the
player must make decisions about which tasks to assign to workers already ·
the organization. Stinchcombe suggests that there is tremendous uncertain
the labor contract from the perspective of the employer, which is due to the
difficulty in predicting performance. Still, he suggests, "a,n organization needs
formation about the immecliate past work performance of a worker so as to
cate rewards fairly. and it needs information about future work perform
OPERATING THE PANOPTIC SORT 61

decisions about hiring, promotion and retention." 36 The bureaucratic imperative

demands that this information be "auditaible" or quantifiable so that it can be
used more easily in analytical models of performance and potential. This demand
for auditability leads organizations to increase the collection of those bits of infor-
mation that are more easily gathered, including that which may be gathered more
efficiently through automation. Yet, because the quality of the information is so
suspect, susceptible as it is to errors of measurement, misinterpretation, and stra-
tegic modification by workers who are aware of the monitoring, firms compound
the errors by increasing the reach of the informational net. More flawed informa-
tion does not improve the analysis, it only compounds the errors.
Drug testing of current and prospective employees represents just one of the
more controversia] aspects of the modem corporation's attempt to reduce uncer-
tainty about its work force. The demand for information by business and govern-
ment organizations has stimulated the development of a drug testing industry es-
timated to exceed $1 billion in revenues by 1990.37 The rationales for such testing
are varied and controversial. One concern is the risk of harm to consumers or
other "innocent bystanders" when a person under the influence of drugs causes
an accident. A second concern is associated with the loss to society when popular
athletes and cultural heroes die as the result of drug use. The primary concern for
business is that associated with the assumed relationship between drug use and
productivity. This relationship is not well established, and some critics argue that
the selective emphasis on " illegal" drugs is primarily ideological because most
testing ignores the more serious productivity losses associated with use of alcohol,
tobacco, and prescription drugs, all of which are legal at the present time. 38
The political aspect of the almost faddish demand for information about drug
use was revealed when the Reagan administration made drug abuse a focal point
of its domestic campaign to reduce crime and eliminate waste in the U.S. econ-
omy. The report from the President's Commission on Organized Crime, which
was released coincidentally with a scheduled conference on drug testing by the
National Institute on Drug Abuse (NIDA) in i986, marked a dramatic shift in cor-
porate policy, made possible by an apparent shift in public opinion. No longer
was it radical to suggest the testing of persons who were employed in critical, sen-
sitive, or high-risk positions. The dominant line, echoed by the press, established
the necessity of testing everyone who wanted to work. 39
Information about the actual and potential consumers of goods and services
represents an area of the most explosive growth and ultimate significance for the
reach of the panoptic sort. Information about the quantities and prices of goods
sold represents a critical form of feedback to the decisions by firms inside and
outside a given market. Firms naturally keep track of their own sales, but informa-
tion about the sales made by other firms also provides important strategic insight.
Information about sales and purchases within and beMeen industries provides
the basis for economic planners to assess the status of the economy. Data gathered
by the Department of Commerce and the Bureau of the Census provide critical
62 OPERATING THE PANOPTIC SORT

assessments of the nature and status of U.S. business. Input-output analysis of the
flows of goods and services within and between sectors provides the basis for so-
phisticated assessments of the status of a nation's economy a11d its potential for
competition in the global marketplace.'10
Most of this inforo1ation is at a fairly high level of aggregation. lt is when firms
seek to improve their performance in the market by increasing the precision with
which they are able to assess the responsiveness of consumers to variations in
price, quality, and representation that the demand for personal information
grows more quickly. Cor:porate planners come to believe that it is no longer suffi-
cient to gather information about the fact that a product was sold at a given price,
in response to a particular advertisement, at a particular time of the year. Busi-
nesses increasingly demand to know more about the ldnds of people who bought
and the kinds of people wbo did not. Information about their own customers
takes on a new value as the technology of database marketing makes it possible for
firms to target thefr promotions to those most likely to respond to similar appeals
in the future. lnformation about potential customers who are not already known
to the firm represents the critical aspect of a growing demand for personal infor-
mation about consumers. Firms can gather this information through their own
research and they can purchase it in the growing market of primary and second-
ary sources. Frequently, organizations that advertise will be provided with infor-
mation about potential consumers by media representatives who seek to incre~
the perceived value of the audience produced by their publication or information
service.

The Corporate File:

It Begins with the Application
Although corporate bureaucracies are not the source for all the personal infor
tion that enters the pa.noptic sort through applications and bureaucratic interac::
tion (Table p), all of the data generated have become readily available thro
electronic means.
Applications are not generally involved in the purchase of most cons
goods. They are more likely in the realm of services. The justification for the
ation of a file in the area of services is based in part on the length of the rela ·
ship and the likelihood that more than one person may be involved in deliv ·
that service over the life of the relationship. Occasionally applications are req -
to classify the applicant in terms of eligibility or in relation to the assignm
tl1e applicant to one or more classes of service. Insurance represents a unique
gory of service that begins with the limited amount of information that am_
gathered on a simple form but that may expand to include microscopic lev.
detail.
H. Laurence Ross's assessment of the extent of personal information in -
ance files in the late 1960s was that it was extremely liroited. 41 The appli
forms for individual insw·ance policies gathered information that located
OPERATING THE PANOPTIC SORT

TABLE 3.1 Personal Contributions to Machine-readable, Network-linked Data Files

1. Personal Information for Identification and Qualification.
Includes birth certificate, driver's license, passport, voter's registration, automobile
registration, school records. marriage certification, etc.
2. Financial Information.
Includes bank records, savings passbooks, ATM cards, credit cards, credit reports/
f iles, tax returns, stocklbrokerage accounts. traveler's checks.
3. Insurance Information.
Includes insurance for health, automobile, home, business, general and specific lia-
bility, group and individual policies.
4. Social Se/Vices Information.
Includes social security, health care, employment benefits, unemployment benefits,
disability, pensions, food stamps and other government assistance, veterans' bene-
fits, senior citizens' benefits/subsidies.
5. Utility Services Information.
Includes telephone, electricity, gas. cable television, sanitation, heating, garbage,
security, delivery.
6. Real Estate Information.
Involved with purchase, sale, rental, tease.
7. Entertainment/Leisure Information.
Includes travel itineraries, recreational profiles, automobile and other rentals/leases,
lodging reservations, airplane and ship reservations, entertainment tickets/reserva-
tions, newspaper and periodical subscriptions, television/cable ratings.
8. Consumer Information.
Includes store credit card s, accounts, lay-aways; leases and rentals: purchases;
purchase inquiries; subscriber lists; dress, hat, and shoe sizes.
9. Employment Information.
Includes applications, medical examinations, references, performance assess-
ments, employment histori es. employment agency applications.
10. Educational Information.
Includes school applications, academic records, references, extracurricular activi-
ties/memberships, awards and sanctions, rankings.
11 . Lega/lnformation.
Includes records of the court, attorney's records, newspaper. reports, index and ab-
stract services.

dividual by name, residence, and place of employment. A second class of descrip-

tors approached the construction of a profile of the applicant from the perspective
of risks associated with a particular life-style, but the primary indicator was age,
because of its statistical association with life expectancy. Life-style information
about occupations and hobbies could have been used to adjust life expectancy, as
some activities are revealed to be statistically more dangerous than others. A de-
tailed medical history, including questions about the health of dose relatives, was
frequently expanded or validated by physical examination. Related documents
from the applicant's medical history might have been obtained on the basis of a
broad release form signed as part of the application.
This initial application frequently triggered what Ross euphemistically referred
to as an "inspection;' which seems more akin to an investigation, through which
 OPERATING THE PANOPTlC SORT

specialized firms gathered behavioral information from friends, neighbors, and

associates. These inspections were particularly concerned with gathering infor -
mation about an individual's personal "habits," including the use of alcohol and
other drugs. The information generated by applications and inspections joined
the medical record and moved to a centralized data exchange called the Medical
Information Bureau (MIB). The MJB functions very much like a parallel system
of credit bureaus in that organizations in the insurance industry and the medical
field submit information for their mutual benefit and use in making decisions
about individuals. Unlike the case of credit bureaus, however, consumer access to
the information in their medical files is many times more difficult. At the tin1e of
Ross,s writing, computerization was just becoming a factor in the insurance busi-
ness. Ross,s predictions about their use were very much on target: "It would seem
reasonable that among the uses to which these computers will be put is sharing
available and relevant personal information concerning insureds within and be-
tween departmeuts of individual companies, as well as among companies.:" 42
Ross suggested that a trend in the area of life insurance toward more group in-
sW'ance would reduce the need for personal information. The same would not be
true for automobile insurance, for which information about personal habits, driv-
ing records, and environmental risks would maintain its instrwnental value.
Homeowners' insurance would also reflect a continuing need for personal infor-
mation.

Surveys and Samples

The simple and direct way to gather information about consumers is to ask for it
through a marketing survey or questionnaire. Surveys are mailed to people in
their homes, completed over the telephone, and filled out in response to appeals
on the street or in shopping malls. Consumers respond freely to marketing sur-
veys inserted in their newspapers and magazines. Information about conslllmers is
gathered when customers complete the questions asked o n the warrantee forms
usually included with electronics and h ome appUances. Generally, ·there is no
compensation provided in exrnange for the information or for the time involved
in providing it. Opinion researchers have noted that a substantial proportion
the population believes that such information helps manufacturers better meet
their individual needs, and thus an improvement in the quality of goods is rew
enough. OccasionaJly there are smal l premiums or the possibility of w·
prizes of one sort or another, which are offered as compensation for the incom
nience and value of the personal information provided. Increasingly, data ab
consumers is gathered under false pretenses, such as when saJes represeutatn
pose as members of a public opinion research firm. 43 This practice, as well as
dramatic increase in the number of marketing surveys attempted by ph<:>ne,
apparently brought about a decline in the willingness of people to participate.'"
Refusal to participate in surveys or the exclusion of some categories of res
dents from the surveys threatens the reliability and validity of the infonna ·
OPERATING THE PANOPTIC SORT

that such surveys produce. Increased demand for access to personal information
from consumers directly has begun to produce a variety of responses that may dis-
tort this information in critical ways. As more consumers refuse to participate,
there is an increase in the use of those who do respond as being representative of
those who do not. This occurs in two ways. The first way is through weighting.
The responses of those in the sample are multiplied by a weighting factor to make
their proportion in the sample match that of a presumably reliable assessment of
their presence in the population. As a result, the attitudes, opinions, and
behaviors of willing participants then are overrepresented in tbe ~ubsequent anal-
yses of data. This problem had been pointed out in the criticism by minority
broadcasters of the dominant ratings services. Because African-Americans who
agreed to participate in surveys or who were "qualified" to receive a meter o r were
able to complete a diary were different from those who listened to "black-format-
ted" radio, the audiences for these stations were seriously underestimated. 45
The same overrepresentation occurs when the same people are caUed on to
participate in more and more surveys because they are known to be willing. These
consumers develop a kind of expertise as professional subjects, and their re-
sponses are less valuable as cues about the distribution or the characteristic re-
sponses of the population. This problem has particular implications for market-
ing research:

While public opinion surveys typically sample a cross-section of the general public,
marketers' growing use of market segmentation requires that they look for specific
population groups which are thought to be the best consumers for their goods or
services. This study suggests that some of these groups- the young, better educated,
higher income, females- are of particular interest to many, and that they are being
oversurveyed. If the trend continues, it may become more difficult to interview these
popular segments, and both refusal rates and repeat participation may rise.46

In the face of continuing concerns about the representativeness of the samples

that are drawn to serve as panels for the estimation of the media consumption
behavior of Americans, there is increased pressure to speed the development and
implementation of more precise and reliable means of audience measurement.47
The eventual domination of the market for People Meter technology by the A. C.
Nielsen Company has not led to any noticeable slowdown in corporate willing-
ness to pay for more precise estimates of individual media use. The Nielsen com-
pany was reported to be working in cooperation with t11e David Sarnoff Research
Center on the development of a sca1ming technology that would utilize infrared
sensors to take note of the entry and exit of individuals from the television viewing
space. This technology would use optical pattern recognition to identify individu-
als and whether or not their faces were oriented toward the screen. 48 This automa-
tion of the process of identification is designed to lessen concerns about the ten-
dency of some members of the family, especially youngsters, to forget to indicate
when they are or are not watching the program.
66 OPERATING THE PANOPT!C SORT

Transaction Records
The initiation of an individually identifiable file in a corporate database may be-
gin with the first transaction. It is easy to understand why a file may be initiated
when consumer purchases are made on credit, such as with a store card, with per-
sonal checks, or even with a bank or nonbank credit card. For security and for rec-
ord keeping purposes, stores have traditionally requested that consumers provide
adilitional personal identification, which is either printed on checks or added by
clerks. It is only recently that some firms, such as American Express, have warned
their clients not to allow their card numbers to be written on checks nor to allow
their telephone numbers to be added to charge slips to reduce fraudulent use of
that information. However, even cash transactions in some establishments gener- 1
ate a computerized record. The electronics firm, Radio Shack, uses an inventory
management system that generates a computerized sales slip as well as a customer
file. The customer who buys even a single battery either initiates a file or adds to
an existing file with each purchase. The key to the files is the last four iligits of the
consumer's telephone number. Customer refusal to provide these numbers, the
absence ofa previously keyed 6le, on he provision of a fictional number generatfS'
a receipt with the name J. Doe.
The importance of the development of scanning technology cannot be over
stated. Scanning from point~of-purchase terminals, such as the checkout count
in the supermarket, provides data at high speed and in real time about the sta
of the market as well as the responsiveness ofconsumers to variations in price
representation. This information helps in the coordination of the distributi
system that supplies the market with products in the right size, style, color, and
on to match the apparent tastes of the shoppers who frequent a particular st
Such coordination reduces the losses associated with excess or insufficient inv
tory in a region or in a particular store. But the scanning technology also provi
the organization with the option of gathering this information at the level of
chases by identified individuals. Special mailings or other distributions of pro
tional materials to persons whose identities are scanned at the time they pay
their purchases facil itate the linkage between inventory control and mark -
central to the emerging just-in-time approach to manufacturing, which links
duction to consumplion. Of course, not all ventures into the area of sc
linked market surveillance have been unqualified successes. Indeed, a highly ·
ble failure in an expensive start~up effort by Citicorp to develop its point-of.
(POS) Information Services unit, suggests that much is still to be learned
diseconomies of scale and scope in the data business. 49
Amex, like Sears, Roebuck and Company (Sears), has vast pools of data
its millions of customers. A.mex has more than thirty-fow· million names in it5
ternational database of customers, and it has detailed knowledge of where
travel, where they eat, and, jncreasingly, what they buy. 50 Not only does Sears
a massive database of customers who have Sears cards, but also the company's
OPERATING THE PANOPTIC SORT

cent entry into the competition for nonbank credit cards adds to the number of
sources of transaction data from the corporation's subsidiaries involved in real es-
tate and financial services. AT&T has joined this group with its offer of Visa and
Mastercards and the transaction records it collects through its position as the
dominant carrier oflong distance and 800-number information calls.
Not even welfare recipients are excluded from this card-initiated system of
transaction records. Experimental automatic teller machine (ATM) cards were is-
sued in 1991 to welfare recipients in Baltimore, Maryland, and contracts with Af-
filiated Computer Systems promised the spread of computerized welfare pay-
ments to other states.51 The card could be used at ATMs within the community to
withdraw cash up to the amount of the monthly benefit, and it could also be used
at participating stores in the community and charged against an equivalent in
food stamps. Although the primary rationale for introducing the cards is a reduc-
tion in administrative costs, there is no reason to expect that the surveillance po-
tential inherent in the transaction data wiU be ignored for long. 52
The importance of telecommunications in the generation and routing of trans-
action information has led to the creation of an acronym to characterize this tele-
phone transaction-generated information (TTGI).53 Thomas McManus describes
several classes or types ofTTGI:

1. 'White pages information, which includes the alphabetical or address-

based listings of telephone subscribers by name. Such a list is initiated
through the application for service. In most service areas, subscribers are
charged a fee not to be listed in such a directory.
2. New service order information, which includes the former number and
address of subscribers who have obtained a new number or listing.
1
3. Calling detail records, which can be generated within organizations as
well as by the telephone service organizations, include the date, time,
and duration of calls from particular instruments to identifiable num-
bers. These records provide management with a means of controlling the
use of the organization's telecommunications resources in the pursuit of
efficiency, including restrictions on nonbusiness activities. Calling detail
records also provided a powerful means of surveillance when corporate
phones may have been used by whistle-blowers to contact the press or
government agencies regarding questionable corporate activities. Tele-
phone companies utilize calling detail records to identify customers who
might be interested in special telephone services, such as long distance
rates, or residential versions of 800-number service so that incoming
calls from youngsters, friends, and distant relatives can be made more
easily.
4 Billing and credit records, which add information about the customer's
68 OPERATING THE PANOPTIC SORT

history of payment and the variety of special services that might have
been acquired to enhance the "plain old telephone service" (POTS).
). Calling number identification, which refers to tJ1e controversial practice
of forwarding the telephone number of the calling party through a resi-
dential or small business service called "Caller-ID," or to corporate users
of Boo numbers or other specialized long distance services. By forward-
ing the number of the calling party, a record of the transaction can be
generated at the receiving end, even if the telephone call has not been
completed. Thus, a client calling an agent or his broker would leave a re-
cord that indicates not only that she called but how frequently and at
what times the calls were made. Calls to Boo numbers for information or
to issue complaints generate a transaction record. Calls to 900 numbers
or premium information services generate a similar record, but one that
can be quite precise in linking a calling party to a particular class of in-
formation.

McManus describes a complex and contentious struggle over the determina-

tion of who bas what kinds of rights and responsibilities with regard to the use of
these forms ofTTGI. The issues involve the privacy interests of the consumers as
well as the competitive interests of other telephone and information service pro-
viders who believe themselves to be disadvantaged by a monopoly firm's control
over this information.
TTG1 is not limited to that collected by telephone service providers. Increas-
ingly the telephone or the telecommunications network is the critical link
through which millions of transactions are accomplished each day between indi-
viduals and organizations in different areas of the globe. Many of these transac-
tions are processed by third parties, which have no direct interest in the transac-
tion but have access to the data generated by the transaction. The limjtations OD
the uses to which this information may be put by these different parties have yet to
be fully specified. These data have rapidly become an increasingly valuable ad
tion to the panoptic sort.

Experiments
Each day, thousands of U.S. consumers participate in experiments without
benefit of having signed any infonned consent forms. If we define an experim
broadly as a tightly controlled exercise in which individuals are provided '' ·
stimuli under varying circllJ}1stances to assess tbeir response, then the thous
of market tests administered by the staffs of the nation's advertising and m
research firms would qualify. These experiments, like those administered in
interest of science, vary in the extent of experimental control over the sample,
ex1'osure, and the measurement of behavioral or attitudinal responses. Deve
men ts in communication technology have made important contributions to
OPERATING THE PANOPTIC SORT

reliability and validity of these field experiments, which come from increases in
experimental control.
It may be the case that true experimental control-that which aJlows a re-
searcher to claim the ceteris paribus condition, that all other things are equal-is
not only beyond the reach but is perhaps outside the region of interest of the mar-
keting community. The marketing researcher is interested not in basic human
truths but in the response of particular classes of individuals to offers and appeals
to buy a particular good or service. General insights about human nature, of the
sort generated by scientific investigations, may inform the marketing research ac-
tivity, but the market researcher in business has only limited interest in adding to
science. Because marketing researchers are interested in responses of classes of in-
dividuals, the random assignment of individuals from a representative sample is
far from the norm. Instead, we have seen and are more likely to see these research-
ers relying more heavily on an investigational paradigm that seeks to determine
how persons of a particular group respond as circumstances change. Statistical
control rather than random assignment will continue to rule the day.
For example, the rather sophisticated marketing studies, such as those offered
by firms such as Information Resources (Behaviorscan), that utilize dual cable
systems to test different commercials in different editorial contexts can determine
in which programs theiI commercials will be seen. What they are unable to do
most of the time is to increase the probability that individuals will see a desired
version of the program, if they choose that program in the first place. Thus, part of
the explanation for the response of particu lar consumers from particular demo-
graphic groups to particular versions of an advertisement for deodorant soap will
be found in the unmeasured and perhaps unmeasurable reasons behind each
viewer's choice to view program X rather than program Y. lnformation about why
people choose and report liking certain programs is sketchy at best. Patrick
Barwise and Andrew Ehrenberg suggest that it is in large part a function of cir-
cumstance that determines when a person is available to view. 54 However, large
samples and the detailed information that is gathered about participants in scan-
ning programs that link program viewing with shopping in participating super-
markets provide for quite remarkable analyses of the relationships among expo-
sure, context, and consumer demographics and life-styles.55
An important technological advance that facilitates the linkage of media expo-
sure to purchases of consumer goods is the laser optical scanning technology that
reads the universal product codes assigned to the great majority of advertised
commodities sold in supermar.kets and variety stores. These stores utilize simi-
larly coded cards to identify those customers who have requested the "privilege"
of cashing a check or renting videotapes. Use of this coded card at the time of
checkout provides a 1inkage of purchases, prices, and coupon usage that can be as-
sociated with information about the consumer that was supplied at the time of
application or acquired from one of hundreds of competing list vendors.
 OPERATING THE PANOPTIC SORT

Coupons are themselves an important resource in the analysis of consumer

behavior. In 1988 more than 252 billion manufacturers' coupons were placed in
prin t media alone, having increased dramatically in recent years.56 A marketing
service organization, the Advertising Checking Bureau, has a division, "Summary
Scan," that is involved in the monitoring of coupons and promotions for con-
sumer packaged goods. The bureau suggests that the increase in the use of cou-
pons can only reduce their effectiveness in the short run as, on the average, less
than 4 percent of all coupons are ever redeemed.
However, coupons are valuable as a tool in marketing research in that they can
be issued in varying denominations and they can be sent to specific consumers or
to target communities. Individuals may receive coupons by direct mail, they may
be imprinted or inserted in periodicals to which individuals may subscribe, or dif-
ferent coupons may be inserted into the weekend newspaper that is delivered
through subscription or for sale in neighborhood stores. Valuable marketing in-
formation is derived when and if coupons are redeemed.
Innovations in coupon technology have served to increase the information
that such redemption provides when the coupon is personalized. "Personalization
can be used to make variable offers. Since it is possible to create coupon offers
with variable discount amounts, different offers can be used, which enables flexi-
ble testjng and maximizes the opportunity for database-building." 57 One vendor
of a proprietary technology called "Softstrip" claimed that it could be encoded
with detailed information about the customer and printed on any form, including
a manufacturer's coupon.58 This information could then be revealed to participat-
ing vendors who would redeem the coupon or other token that carried the
Softstrip code. In one example of its application, targeted consumers received a
package announcing a contest in which the prize would be a trip for two to the Ba-
hamas. A two-piece form served to ctistract consumers from the data gathering as-
pects of the "contest." One form, which included a space for consumers' names
and addresses, aJso asked them to indicate whether they owned a video cassette re-
corder (VCR), where they shopped regularly, and on which day of the week they
usually shopped. This form was to serve as their entry in the grand prize drawing.
The second form, a tear off sheet, was supposed to serve as the token for an instant
sweepstake prize oflesser value. This card had no dearly identifying marks; all the
consumer would see was an arrow showing how to place the card into the reader
to see which of several smaller instant prizes they had won. The Softstrip bar was
described as the contest prize "key." In fact, the Softstrip contained detailed infor-
mation about the customer: " head of household's name and occupation, total
household income, number of children under 18 years old, number of employed
people in the home and numbet of vehicles owned by household members." 59
developers of the Softstrip technology also report that they have developed a set
printable colors that can be used to overlay, and make invi.sible, the Softstrip ·
case marketers are concerned that the public will react adversely to the presence
the unusuaJ strip. 60
OPERATING THE PANOPTIC SORT 71

DATA PROCESSING
Roger Clarke offers the term "dataveillance" as a device for characterizing the new
forms ofsurveillance that have been occasioned by the increase in the distribution
and use of computer-based technology. 61 His definition of dataveillance is indica-
tive of its value: "Dataveillance is the systematic use of personal data systems in
the i~vestigation or monitoring of the actions or communications of one or more
persons?'62 Clarke identifies two forms of dataveillance that have important dis-
tinctions and are worth considering in some detail. Personal dataveillance involves
identifiable persons who by their actions have attracted the attention of the
panoptic system; and mass dataveillance involves gathering data about groups of
people with the intention of finding individuals in need of attention by the sys-
tem. The techniques of personal dataveillance involve (1) the integration of data
regarding an individual that might have been stored at different locations within
the organization; (2) the screening or authentication of transactions by this per-
son by comparison against internal norms or the prescreening or front-end verifi-
cation of transactions that appear exceptional or problematic in comparison with
available data, either on band or gathered from other sources; and (3) the instiga-
tion of cross-system enforcement against individuals on behalf of other actors
who claim to have been harmed.63 Mass dataveillance techniques are similar to
personal forms except for the fact that an individual's behavior need not appear to
be exceptional; all individuals need to do is to be part of a subject population.
Mass dataveillance also includes a form of"single-factor" file analysis of all avail-
able data against some norm, derived from other data or from law. Profiling is
characterized as a form of multiple-factor file analysis, which might include a
form of"aggregative profiling of transaction trails over time."64
Clarke suggests that the integration of seemingly different computer-based
systems through telecommunications networks has generated a complex new
form of information technology. New developments tend to strengthen the mutu-
ally supportive relations between the systems, which facilitate the management of
production, travel, finance, security, and a host of ad~foistrative services. En-
hancements in these systems are expected in storage; in the efficiency and user-
friendliness of input and output devices (including advances in natural language
comprehension); in the speed, efficiency, and reliability of communications be-
tween systems; and in the ability of systems to integrate different data formats
(voice and inlage, symbols as well as numbers) and different processing logics
(fuzzy logic and stochastic processes). ·
In the fall of 1991. Amex reported its plans to acquire two supercomputers
noted for their ability to utilize thousands of processors operating.in parallel. It is
the nature of the information processing tasks that characterize the principal ac-
tivity that.explains the need for such machines. Although Amex was unwilling to
reveal precisely how it would use the systems, expert opinion held that the ma-
chines would be used to "refine its analysis of cardholders' purchasing habits."65
72 OPERATING THF. PANOPTIC SORT

Stewart Brand described a similar machine with 65,536 processors interconnected

in a "sixteen-di mensiona1 hypercube array," which was acquired by the Media Lab
at the Massachusetts lnstitute of Technology (MIT). 66 Multiple processors allow
large, complicated problems to be broken up into smaller chwlks, each crunched
by a single processor. The parallel processing approach is espciaUy important in
pattern recognition and behavior modeling of the sort that facilitates the develop-
ment and application of consumer profiles constructed from hundreds of differ-
ent kinds of personal and environmental information.

Information Processing Technology

In 1983, the Institute for Computer Sciences and Technology of the National Bu-
reau of Standards (NBS) published what it called an expert opinion forecast of
what we might expe~ in developments in information processing tedlnology
through i997. 6 7 It was clear then that increasing capacity and speed as well as de-
clining cost would support expanded use of the computer in military, commer-
cial, and consumer applications. It was also projected that advances in program-
ming applied to the development of applications software would reduce the need
for data processing professionals, in that the newer computers could be operated
more easily by nonprofessionals. The success of the Apple Macintosh with image-
oriented interface and point-and-click mouse seems to have justified its self-pro-
motion as the "computer for the rest of us."
NBS reported the common view that important strides would be made be-
tween 1987 and 1992 as the "development of specialized processors in the area of
database management and image processing areas will lead to major enhance-
ments in guery languages and image processing languages." 68 The experts pre-
dicted with remarkable accuracy that " relatively lmtrai.oed users will become able
to generate graphics (line Charts, animation, maps, histograms, etc.) on-line with
the systems in full color." These predictions included the ability of analysts to ex-
amine three-dimensional simulations of the behavior of models they were testing.
Relatively inexpensive workstations have more than met that expectation. The ex-
perts at NBS were similarly correct in their projections that analysts would be able:
to perform many individualized analyses independent of the mainframes to
whicl1 they might be connected. The accuracy of NBS projections into the furum
beyond 1993 will simply have to be awaited, but given the e."Xtent of aoalytical p
cisioo in their short-term projects, some confidence seems warranted.
NBS suggested widespread commercial availability of knowledge-based s
terns characterized b y a high degree of iinteroperability.09 They suggested
nonkeyboard interfaces will be commonplace md that knowledge-based syste
will be used reliably in decision support systems.
As the number of transactions increases, firms have to develop ways to incr
the rate of completion in order not to increase the waiting time for any custo
in a queue. Increases for manual systems mean increasing the number of lines,
number of workstations, and the number of transaction processors. Account
OPERATING THE PANOPTIC SORT 73

justments can add several mjputes to a process that could conceivably take a mat-
ter ofseconds. Automation of any subset of this process thus represents a potential
savings, and automation has been pursued intensively.
Barbara Elazari describes the development of an on-line transaction process-
ing (OLTP) system for Amex. 70 It became clear fairly early in the organization's
modernization phase that Amex's impressive growth could not be sustained if it
continued to rely on a manual authorization system. In 1976 Amex embarked on a
modernization plan in which the goal was "to automate all the clerical work per-
formed in support of customer accounts-activities such as account mrunte-
nance, ordering microfilm of statements and charges, writing letters, and finan-
cial adjustments." 71 Automation did not mean the replacement of clerical workers
with robots, but rather the replacement of the paper records with a set of com-
puter screens that could accommodate virtually all the transactions that reqwred
adjustments and that could not be performed entirely by software.
A variety of expert systems have been developed to increase the rate and effi-
ciency wi.th which information is processed to facilitate routine decision making.
Amex developed an expert system called Authorizer's Assistant, which facilitates
decision making about authorizing purchases for individuals who are at or be-
yond their approved charging limit. 72 The Amex system is a good example of the
process of "knowledge engineering," wruch transfers the knowledge, primarily
rules and a variable knowledge base, to a software system. In the case of Authoriz-
er's Assistant, the expertise of five of the top authorizer-sat Amex was incorporated
into the program. Even though since modernization the bulk of requests for au-
thorization at Amex were handled automatically, the small percentage of cases
that require the involvement of a human were still thought to be made more effi-
cient with the assistance of an expert system:

It evaluates credit authorisation requests in light of a cardbolder's usual spending

pattern. Its design reflects the problem that since the American Express card is not a
typical credit card with a preset credit limit, the question of when to grant
authorisation is not one of numbers, but one of judgement.... Before this expert
system was implemented, an authoriser would have to consult 12 to 14 data screens
on a card.holder's history and patterns prior to authorising or denying a charge, or
requesting confirmation of identification. Using the expert system to analyse the
history and pattern, an authoriser now needs to look at only two to three screens to
get a recommendation. Though the system could make decisions on its own, American
Express uses it only in an advisory capacity [emphasis added).73 ·

A similar process characterizes the credit card expert system developed for
Mitsubishi by its research institute and Diamond Credit group. Here we find an
expert system referred to as a Profile Analyser, which is used for scoring individual
applicants in relation to their membership in certain theoretically constructed
population groups. The 200 profile templates were developed through an analysis
of data from a sample of 2,500 "good" cardholders, and 1,800 ".failed" cardholders.
74 OPERATING THE PANOPTIC SORT

An original set of 340 patterns was reduced to the 200 "major profile factors." A
credit level was then assigned to each profile. On the basis of information pro-
vided by individuals on their applications, they will be assigned to one of these
profiles and wiU be denied or provided credit up to the limit specified by the pro-
,file.1•
Expert systems have also been developed to assist credit card firms in detecting
fraudulent uses of cards. Fraud detection is a complicated process that has to be
completed rapidly and without costly error. A great many considerations must be
made simultaneously: "the card user's track record, his or her ability to pay, any
possibility of fraud detectable from past records, what the effect of denying the
credit authorisation might have on the user's future use of the card: tear it up and
switch to a competitor, what may be the aftermath for merchants?'75 This ad-
vanced system is designed so that it can use several different databases.
A similar system was reported to have been in the works for the IRS to facilitate
automated tax examination. The efficiency rationale designed into this system is
clear. As the system operates, it seeks to identify filers with the potential for a high
payoff, and deductions are the principal focus. Ranges of deductions are com-
pared against many hundreds of rules, and a final "integrator module" produces a
result that can be compared against standards evolved through experimentation
and analysis of past returns and investigations.76

Matching
On its face, matching seems to be a simple and obvious analytical approach to
producing intelligence about individuals. Matches utilize a deductive logic, which
suggests that if an individual's claim of rights to a particular status would assign
her or him a place in list A, then that status might be validated by her or his ap-
pearance in another list B, which is causally or logically linked to either A or to a
third variable that affects them both. A person applying for an automobile insur-
ance policy (list A) is logicaJJyassumed to have a driver's license (list B). The nega-
tive or inverse model works in the same fashion. A person applying for a gun per-
mit is presumably an adult over the age of eighteen (list A), but may not be a
convicted felon (absent from list B). Persons who are claimants for unemploy-
ment compensation (list A), ought not to be found among the employed who
making contributions to their social security (listB). Matching is simply tbe co
parison of lists to note the presence or absence of an identified individual in sp
fied lists. The high-speed computer has made it possible for such matches· to
made at costs that continue to decline at a rate that has made what was once a
event a routine administrative requirement. Greater awareness of the possibility
savings associated with matches has created something of a challenge for bur
crats to dream up innovative new matches, which might be used to identify·
viduals engaged in actitivies leading to waste. fraud, and abuse of resources.
social security number is the de facto universal identifier that facilitates
matching of such lists. .
OPERATING THE PANOPT!C SORT 75

Matching has numerous uses as an aid to prescreening and qualifying appli-

cants for bureaucratic services, whether from government or private firms. The
logic is the same. The files of new applicants for public welfare numbering ap-
proximately 250,000 records might be matched against reported income and in-
terest earning in the IRS files, covering nearly i8o million persons, to determine if
the applicants are truly without resources as their application for assistance might
suggest. Such record linkage is not limited to single organizations, because the
failure to meet one's responsibilities with one member of the business commu-
nity, such as paying one's rent, may result in the denial of access to other services,
such as insurance. Matching has also demonstrated its potential as an aid in gain-
ing compliance with bureaucratic regulations, especially those related to the pay-
ment offees or fines. For example, in financially strapped Manhattan, new, porta-
ble computers called "mobile digital terminals" were put into service for sheriffs'
deputies to allow the officers to match an automobile's registration against a file of
w1paid parking tickets.
Increasingly, individuals are barred from enjoying one class of benefits or ser-
vices because they have not met some requirement or obligation in some other
area. At the University of Pennsylvania, students attempting to register through
the campus telephone-based automated system (PARIS) were surprised to find
that their registration was barred because they had not completed a required ap-
plication with the student health service. Similar examples are easy to find and
multiply with each passing day.
Federal and state governments utilize matching programs to identify and seek
payment from individuals with outstanding student loans by restricting individu-
als' access to other government services. With loan defaults in 1990 exceeding $z.4
billion, and with persons in default representing some 17 percent of the borrowers,
the Department of Education recommended a number of programs that would
involve matching of the lists on which students and recent graduates would be
likely to appear. The easiest targets were current or retired employees of the gov-
ernment. Because of the dependent relationship between numerous private orga-
nizations, including colleges and universities and the federal budget thro ugh con-
tracts and grants, those private institutions find themselves under increasing
pressure to perform such matches to demonstrate that their employees and clients
have met all government requirements. The requirement that contractors assure
the government that employees and clients are not scofflaws, deadbeats, or drug
abusers can conceivably be expanded almost without limit to include other state
interests in reproductive choice, traditional family values, and safe and responsi-
ble sexual practices. Employers already seem quite willing to pursue similar levels
of control over nonwork-related aspects of their employees' lives without the
compulsion of government regulation or contract limitations.
Although there is disagreement about the cost-effectiveness of matching pro-
grams, the Senate report on the Computer Matching and Privacy Protection Act
of i988 (S 496} described numerous matches that reportedly generated substantial
 OPERATING THE PANOPTIC SORT

savings for the governments that administered them. A match of social security
files with the unearned income files of the IRS identified some $117 million in
overpayments. Since tbe cost of the match and the follow-up was reported to be
only $6-4 million, the economic savings were assumed to be substantial. Claims
are also made that the use of matching in the form of"front-end verification" of
eligibility has reduced the number ofapplications for government benefits, such
as food stamps, presumably because fraudulent appJications were eliminated.77
The potential for matches to realize substantial savin,gs generates a kind of com-
petitive spirit among program analysts, who try to thi11k up novel matches that
might uncover actual or potential fraud, waste, or abuse of organizational re-
sources. Once developed and tested, the ideas for these new matches spread like
wildfire among organizations.
There are also "back-end" matches that serve as checks on the appropriateness
of payments relative to the measured status of claimants. The Medicare program
compares charges for medical services against the record of a patient's diagnosis to
see if there have been inappropriate charges. Similar screening matches are likely
to be pursued by private insurers.
Matching is routinely used for the verification of data about individuals, such
as their addresses, telephone numbers, places of employment, and assertions they
may have made about their income and indebtedness. There has been consider-
able controversy about commercial firms requesting that the federal government
verify the social security numbers that1ndividuals provide to those organizations.
There has been even more vocal public concern expressed about reports that the
IRS was considering matching its files against the estimates of personal income
developed by the commercial firms that serve the maTketing community.
Finally, in an era of increasing liability judgments for "negligent hiring;"'
prescreening of applicants for employment against a variety of government and
cornmerciaJ tiles is becoming the rule. 78 Background checks against criminal his-
tory records, as well as credit files, is a commonplace screening match that en-
hances the data gathered from applicatlons, paper-and-pencil tests, and medical
examinations. Because the provision of medical benefits is such a substantial pan
of the compensation package of most large businesses, information about health
status is particularly important, The demand for information about a potential
employee's histOiy of injury on the job has supported the emergence of several
firms that provide matches against files of workers' compensation claims, injury-
related lawsuits, and other indicators of risk. 79 Workers who appear on such li:sas
may find themselves excluded from future employment even though their clai.ns
would be considered to have been legitimate by any legal standard. A person's 31>'
pearance on a list is frequently sufficient grow1ds for rejection during periods
economic decline or stagnation when there is no shortage of available wor
with m1blemished records.
The possibility of matching applicants against federal and state databases of·
dividuals with genetic characteristics considered problematic is also a probl
that looms large on tbehorizon. Troy Duster notes that
OPERATING THJ:l PANOPTIC SORT 77

state and national registries for information received from newborn genetic screen-
ing programs arc already in place, collecting data on the chromosome and genetic
trait status of millions of infants. These data are collected for health and medical rea-
sons, and often deal with whole populations, not just those at the gl'eatest risk. ...
These registries, now in their beginning stages, are part of the machinery in place (or-
ganizational, institutional, legal and physical) which will slowly, subtly, sometimes
imperceptibly, help shift the refraction of human traits, characteristics, behaviors,
disorders, and defects through a "genetic prism."80

Clustering and Segmentation

The simple comparisons of two lists for ~e presence or absence of identifiable in-
dividuals are multiplied in complexity when organizations seek to identify groups
or segments of the population. The primary assumption underlying the pursuit of
market segments as a competitive strategy is that "segments actually exist;' and
that aJI that remains for the analyst to do is to find the data and the analytical ap-
proach that reveals their underlying structural composition.81 Because the collec-
tion of data is costly and time-consuming, the analyst is also concerned about the
development of the most efficient ways to predict the responses of consumers. The
assumption that segments ac.tually exist is one that is easily satisfied by the realiza-
tion that any market can be segmented in an almost unlimited number of ways,
with the only limitations being those associated with the measurement of relevant
attributes and those related to ~e power of the computer and computer software
to handle a large number of variables and cases.
The selection of variables is guided by a continually evolving theory of con-
sumer behavior. An early and still influential body of theory is centered on the
stages of the Life cycle through which an individual, family, or social unit is be-
lieved to pass. Individuals' needs, interests, and activities differ as they pass
through adolescence and young adulthood, est1tblish a household, become par-
ents and grandparents, and enter retirement. Significant differences have been
noted ill the amount of recreational tin1e that is spent viewing television as a func-
tion of age, marital status, and family size. The activities and opportunities that
are characteristic of people at those stages also vary with race, gender, age, and
membership in an identifiable subcultme, perhaps associated with a neighbor-
hood or geographical location.
Added to the traditional demographic identifiers are a number of personality
or life-style characteristics. Ronald Frank and Marshall Greenberg utilized a com-
bination of interests and activities to segment the television audience into four-
teen groups, which added significant power to the ability of basic demographic
measures to predict an individual's television program viewing. 82 Influenced by
Patrick Barwise and Andrew Ehrenberg,83 Ronald Simmons utilized a measure of
~ognitive style to predict television viewing preferences among African-American
college students.84 Finding s4ch linkages between cognitive style and television
program preferences within a relatively homogeneous population suggests that
such measures may have some potential utility in other media segmentation
 OPERATING THE PANOP'J'IC SORT

schemes. The limitation of most of these approaches to segmentation is their reli-

ance on voluntary responses from samples of the population. Economists have
traditionally been wary of individuals' self-reports of their interests and opinions
and have placed more confidence in indications of revealed preference-the ac-
tual behavior of individuals io the market. Transaction-generated information
thereby increases the "ecological validity" of any segmentation scheme to the ex-
tent that it measures behavior rather than reported attitudes, perceptions, prefer-
ences, or intentions.
AJ though there are a dizzying variety of approaches to the segmentation of
populations or markets, they share an underlying concern with maximizing the
similarity within groups or clusters and maximizing the difference between those
groups.as Within-group similarities may be based on a variety of measures of as-
sociation between individuals. One measure of association is the correlation that
expresses the sim ilarity in the patterning ofvalues among two or more individuals
or cases. Jonathan Gutman reminds us that a correlation coefficient describes
similarity in patterning, rather than similarity in attribute or behavior. 66 Thus.
two persons may share strikingly similar relations in th-eir height, weight, age, in-
come, and education, and yet one can be substantially taller, heavier, older,
wealthier, and more highly educated than the other. lf the primary concern is the
identification of group membership rather than the description of the character-
istic.c; of the segments or groups, then the preferred approach is one of cluster·
or multidimensional scaling.
Factor analysis is a popular approach for the identification of the similari ·
and differences between persons.87 Factor analytic designs .may explore the r
tiouships between persons as determined by their comparison across a number
variables. Alternatively, analysts may examine the relationships between varia
by comparing their values among a number of persons. The fast and gene
more popular application is called R-factor analysis. This approach allows the
alyst to indicate which people are similar in terms of their similar responses
questions or in terms of their similarity in attributes and behaviors. Such an
ysis cao be used to produce factor scores, or a summary index for each theore ·
complex that appears to differentiate between individuals in a sample.
Cluster analysis assigns individuals on the basis of the similarities be
them in terms of their scores on some scale or measure, including a factor
generated by a factor analysis.as The example in a popular manual for stati ·
analysis by computer is the search for "relatively homogeneous groups" of
which are eventually grouped on the basis of commonality and difference in
ries, sodium and alcohol content, and cost.119
An approach that reflects its instrumental purpose is called discriminant
ysis. The analyst who has identified two or more different groups or types can
lize discriminant analysis to identify measures that are the most useful or r
in assigning individuals to their nominal groups. 90 Such an analysis facilita
design of experiments, surveys, or other data-gathering activities in which
OPERATING THE PANOPTIC SORT 79

sions have to be made about the tradeoffbetween cost and productivity of the in-
formation gathered about individuals in the sample. Discriminant analysis in its
most direct appliqltion might be used to differentiate between those who will de-
fault and those who will nol on the basis of historical data. Where such models
have been developed on the basis of one set of data, they can be evaluated in terms
of their success in predicting the histories of an alternative sample.
Multidimensional scaling (MDS) provides the analyst with the possibility of
locating different products, including those of primary competitors in a multidi-
mensional space (such as that described by considerations of cost, durability,
safety, and style in automobiles).9 1 Clusters of potential customers for the product
can also be located in that same multidimensional space, thereby revealing which
aspects of the product's "positioning" are most important to which segments of
the population. The MDS approach, which uses consumer perceptions as well as
their expressed preferences, is preferred to one that uses either one or the other.
The perceptions are generated by asking consumers to indicate how similar or
how different two attitude objects are from each other on particular dimensions.
Frank, Massey, and Wind suggest that the MDS approach avoids the problem of
asking people to be introspective and to give reasons for their choices in that it
merely asks them to compare and to choose, leaving it up the the analyst to infer
the rules that are operational.92 Considerably more sophisticated approaches thal
seek a causal interpretation of consumer behavior include the increasingly popu-
lar structural equation models, which attempt to take into accow1t the influence
of the invariable errors in measurement that such research will involve. 93
Market segmentation, then, utilizes the information derived from these statis-
tical analyses to target messages to particular market segments. This segmenta-
tion has been demonstrated to work effectively even when the promotional mes-
sage is delivered through mass media channels. As Jong as the message is directed
to a particular audience segment and is designed to attract its attention, such as
through the use of models and circmnstances common to the target group, a req-
uisite level of efficiency can be obtained Of course, the most efficient approach
would be one that identifies those programs for which the preferred consumers
made up the majority of the audience.
The same analytical technology that produces the market segment is also uti-
lized to gene~ate conceptual profiles again.s t which individuals may be compared.
These profiles are used to identify individuals who represent a particularly high
risk, or conversely, a particularly attractive marketing opportunity. A profile is an
ideal type. Ideal types can be described statistically and confidence limits. can be
specified, which define the ranges for key variables that should be used to deter-
mine whether any particular individual should be labeled as a member of a partic-
ular group or not. Criminal profiles are common and familiar. The profiles are
used by a variety of government agencies to determine which vehicles ought to be
stopped, which suitcases ought to be inspected, and which tax returns ought to be
investigated more closely. Similar profiles are used by commercial firms to indi-
Bo OPERATfNG THc PANOPTlC SORT

cate when credit authorizations are to be questioned, when bankruptcies seem

likely, or when a particular offer might appear more reasonable.

A TECHNOLOGY OF POWER
The panoptic sort, which depends on ready access to personal information that
can be used in combination with information about the relevant environment is,
as I have suggested, a technology of power. It is a discriminatory technology, and
it is guided by an instrumental rationalism. Like the content analysis procedure it
imitates, the panoptic sort proceeds in stages and has component parts that vary
in importance, depending on the purposes and interests of the controller. In the
next section I will explore. aspects of the panoptic sort including identification,
classification, prediction, prevention and avoidance of risk, and allocation oflife
chances.

Identification
James Rule and his colleagues provide an analysis of the importance of a select
number of items of documentary identification that have become necessities for
an individual's successful negotiation of the bureaucratic maze. 94 They distin-
guish between the "documentary tokens:' which individuals possess and fre-
quently carry on their persons, and the data in files of organizatiollS that issue
those tokens or that require them to complete a given transaction. Moreover, they
also note that both forms work together, the use of one depends on the existence
and maintenance of the other. They suggest that personal documentation serves
the social function of "generating certainty about people" and helps these organi-
zations to "discriminate in their treatment of individuals!'
The first, and perhaps most important, of such documents is that which re-
flects the creation of a record of a person's birth. Wi.t.hout a birth certificate, it i5
difficult to establish the long chain ofdocumentary links that stretch from driver'
license to credit card and passport. The driver's license, as a convenient picture
card, is regularly demanded as a necessary adjunct to transactions involving
exchange of cash or commodities against personal or third-party checks, in ad
tion to its role in the-identification of a person presumed capable of driving an
tomobile. The document frequently contains information about sex, age, hei
and eye and hair color, i.n addition to a signature, an address, and perhaps a s ·
security number.
The social security card, possessed by "nearly every economically active a
in the United States," is rarely used as identification in ·the way that the driver's
cense might be. The social security number itself is the critical token used to¥
or validate other claims of personhood. The credit card, which Rule and his
leagues discuss, has joined the driver's license as an item of personal iden ·
tion, useful in other commercial transactions "because of the sophistication
surveillance and control achieved by the managers" of the most popular cards.'
OPERATING THE PANOPTIC SORT 81

At the core of their analysis is a question about the reliability of such docu-
ments when they are dependent on the individual to provide the evidence neces-
sary to verify self-identification. They note the relative ease with which individu-
als are able to acquire false documents and then to use those documents as
"breeders" to obtain a full complement of documentary tokens. They suggest that
the primary factor that serves to liinit the wholesale falsification of personal docu-
ments is the general uncertainty about the probability that organizations will
bother to check or verify the claims. This is the same uncertainty about surveil-
lance that made the operation of the Panopticon theoretically so efficient. Of
course, the presumption of surveillance is not sufficient. Rule and his colleagues
note that the commercial firms involved in authorizing credit transactions have
developed sophisticated means of surveillance that they use to limit the unautho-
rized or inappropriate use of the token. What Rule et al. described in i983 were the
early developments of a technology of "self-checking;' which is designed to re-
duce or eliminate organizational dependence on self-identification. Considerable
progress has been made since then.
Among those items of identification that increase the extent and reliability of
self-checking is the develop"ment of "smart card" technology. Early research had
revealed that the debit cards that are used by individuals to withdraw cash and
make other banking transactions through remote ATMs had losses that were
twenty to thirty times lower than transactions with credit cards.96 Use of the cards
requires a personal identifica.t ion number (PIN), and the transactions are gener-
ally conducted on-line, in real-time contact with the controlling records, thereby
increasing the surveillance capacity of the system. The debit card does not qualify
as a smart card because of its use of magnetic stripe technology. The stripe is lim-
ited in the information that it can carry, and, at least initially, it is limited in its
ability to record information reflecting any change in the status of the user. Still,
the 1987 estimates placed over one billion such cards in circulation. The smart
card alternative would add integrated circuit chips, which would add important
new functions, including "significant additional storage capacity, enhanced secu-
rity, the ability to capture transaction amounts and characteristics, internally vali-
dated PINs, user-specified logic, and a permanently recorded transaction jour-
nal."97 More sophisticated smart cards would replace the PIN with a ·biometric
identifier such as a fingerprint, but that would require greater intelligence and
storage capacity than is available at a reasonable price. Whatever the method, the
smart card nol: only would help to establish the identity of the user, but also would
contain the present balance or limit in the account. The smart card would com-
bine identification with classification.
Similar cards have been proposed for access to the health care system. An ex-
perimental "Life Card" under development for Blue Cross/Blue Shield would
contain some eight hundred pages of information about a person: their medical
history, multiple identification checks, and perhaps even digitized copies of chest
X-rays, scans, and other data that would facilitate the collection of medical histo-
82 OPERATING THE PANOPT!C SORT

ries in emergencies.98 This development would take identification and classifica-

tion into another sphere altogether.

Classification
Classification is a technology of control It is driven by the purposes or interests of
the actors who seek to take advantage of knowledge regarding the factors that pro-
duce or w1derlie µie similarities and differences between people. Michel Foucault
is not alone in characterizing classification as an activity that is linked intimately
with the exercise of power. 99 Eugene GaUahue's discussion of the history of market
standards provides a valuable insight into the means and justifications for the cre-
ation of classes, standards, and grades of marketable goods such as loaves of
bread. 100 In the eighteenth century, Gallahue found, it was the responsibility of the
monarch tb protect the common interest by regulating the weight and quality of
loaves of bread. Before the emergence of caveat emptor in the 1700s, grocers in-
volved in the sorting of spices bagged and labeled these goods with their own
marks as a guarantee of quality. These marks were very different from the brand
name labeling, used by the National Biscuit Company's Uneeda Biscuit brand,
which initiated an era of product differentfation in the early 1900s, serving pur-
poses of market control and producers' interest 101
The process of sorting and grading produce has, ofcourse, influenced the sort-
ing and grading of humans in similar fashion, as we have noted with regard to the
modern use of the term triage, which migrated from the sorting of coffee to the
sorting of claimants for medical care.1°2 Among the things that differ between the
sorting and classification of produce and the sorting and classification of human
beings is the fact that humans seem to have an interest in naming and sorting
themselves. Paul Starr notes that a great many factors may be involved in deter-
mining whether the state (or the commercial system) accepts the self-definitioll5
proffered by particular groups at critical moments in history. w3 The classificatioa
of people with Arrican and Latin heritage remains a fluid and often contentions
process at both official and social levels. The currency of the label African-Ameri-
can over the recently legitimized label "black" (whether capitalized or not), whicii
replaced "Negro," "colored," the more offensive "nigger," and the regional vari
"nigra," reflects a debate and struggle over self-deiin.ition that is far from co
pleted. This struggle plays itself out in the continually changing category sche .
for tbe census and other surveys that take note of racial and ethnic group me
bership.
The identification of an individual as being black or African-American is
particularly potent example of the arbitrary nature of many social classificatio
It is almost laughable to consider that the genetic materials that would
someone black 31·e such that they need only be present in o ne-sixtr fo ur:th part.
the classical "one drop of Negro blood," to assign a person to that racial group.
similarly absurd to classify the great vai-iety of cultures represented in Europe
OPERATING THE PANOPTIC SORT

Latin America under the bureaucratic label Hispanic. Yet, such classifications are
common and bear the force oflaw.
Star£ reminds us that the domains of social classification are social products,
historically determined, reflecting in large part the exercise of economic and po-
litical power. Mary Douglas shares Starr's view in that she argues that "no superfi-
cial sameness of properties explains how items get assigned to classes. Everything
depends on which properties are selected." 104 Starr suggests further that there is a
process of "labeling" through wruch the same individuals would be "framed"
quite differently for' purposes of public policy if they are labeled as "homeless"
rather than 1'vagrant:' The comparisons are simiiarly clear when we note the dif-
ferential response to the action of"terrorists" versus those of"freedom fighters:'
Douglas would add, however, the fact that the response of humans toward prof-
fered labels is not always one of rejection:

Jn the same way as sexual perverts, hysterics, or depressive maniacs, living creatures
interacting with humans transform themselves to adapt to the new system repre-
sented by the Labels. The real difference [between humans and bacteria reacting to
injections] may be that life outside of human society transforms itself away from the
labels in self-defense, while that within human sociery transforms itself towards
them in hope of relief or expecting advantage. 105

Classification generally involves some form of measurement or weighing. Dif-

ferences not measured, for all intents and purposes, are differences that do not ex-
ist; clearly, they are differences that do not matter. Measurement is both a techni-
cal and a theoretical process that involves an ability to recognize similarity and
difference. Foucault has suggested that classification is the "nomination of the vis-
ible."106 I might add that the statistic<il technologies that describe the associations
between hundreds of seemingly discrete variables, which would have been impos-
sible without the high-speed computer, facilitate the naming of types that such
analyses make visible in the same ways that the invention of the microscope made
other classifications possible within the .emerging natural sciences. Just as ad-
vances in technology make it possible to make distinctions between the visible
more clear, advances in theory presumably make it possible to characterize that
which is invisible (beyond vision), with more precision through inference from
that which can be seen. A technology such as factor analysis explicitly seeks to link
the relations between the visible and measured to the invisible (because unmea-
sured) variable or factor that those relations suggest. 107 The modern firm is ac-
tively involved in seeking to classify its present and potential customers through
sophisticated analyses of the never-ending .stream of data that electronic transac-
tions produce.
James Anderson's discussion of the different ways in which one might concep-
tualize an audience attracted to (or produced by) a particular cultural product
helps to emphasize the position of the analyst in determining the forms any classi-
 OPERATING THE PANOPTIC SORT

ficarion might take. 108 He differentiates between formal audiences imd empfrical
audiences. The formal audiences include what he caJls the "encoded audience:'
which is evoked in frequently politicized discussions about an industry that is
"serving the interests of the audience," or giving the audience "what it wants."
This audience can be distinguished from the "analytical audience:' which is con-
ceptualized by scientists or critical theorists as the object of study or theorizing.
Although potentially related, they are to be distinguished from the different "em-
pirical audiences," which Anderson also defines. The aggregated audience is that
which is represented by the Nielsen ratings. The categories (e.g., women, ages 18-
49) are purely arbitrary and surely do not represent·any purposeful self-identifica-
tion by women who see a common purpose in their role as such an audience. That
classification has, as Anderson notes, a clear purpose for the broadcasting execu-
tive who would like to charge advertisers for access to an audience with those
characteristics. Anderson's "strategic audience" is one that has relevance to dis-
cussions of feminist consciousness, in that analysts may be interested in how such
an audience, as a member of an interpretive community, would respond to a par-
ticular program or promotionaJ appeal. The categories that are most prominent
within the context of the panoptic sort are those that classify individuals in terms
of their potential value. Potential value, however, is dependent on behavior at
some time i.n the future. Thus, a major component of tl1e panoptic sort is classifi-
cation for the pl'ediction ofbehavioral response.

Prediction
The panoptic sort is a predictive technology. lt is even used to predict the likely
behavi.or of jurors, and it serves as an aid to the preemptory challenges used by at-
torneys to improve their clients' chances of acquittal. ln preparation for the high-
profile trial of a member of the Kennedy family on the charge of rape, the
representatives of a commercial firm, Trial Consultants, described the process of
gathering data through extensive interviews and then correlating age, sex, educa-
tion, and ethnic group membership with views on the guilt or innocence of a de-
fendant. Questions that prove to have the greatest power to discriminate between
those more or less likely to convict then become part of the voire dire. In advance
of jury selection, one consultant suggested that the factor that would prove to be
the best predictor was what she referred to as "the Kennedy love-bate factor." 109
Newton Minow and Fred Cate argue that the term "jury selection'' is a misnomer
in that what attorneys actually attempt to do is to "deselect," or select out, poten-
tial jurors in the interest of their clients. They suggest that _this process, aided by
social science, has the potential for serious distortion: "It is clear d1at if the me
bership of the panel is skewed by the selection process, then the fundamen
guarantee of fain1ess-the diversity and breadth of experiences and views -
likdy to be compromised." 110
If the panoptic sort is a predictive technology concerned with deselec ·
rather than including and if the panoptic sort is based on probabilistic rather _
OPERATING THE PANOPTIC SORT 85

exact predictions, then it is a technology concerned primarily with the assessment

and avoidance of risk. Nancy Reichman focuses our attention on the assessment
of risks as a conceptual marker underlying [he identification, classification, and
future interactions with individuals who engage the panoptic sort. She suggests
that we ought to conceptualize "the growing reliance on probability, opportunity
reduction, and loss prevention as part of a trend toward an insurance or actuarial
model of social control:' 111 The reduction of risk with regard to individuals in-
volves the development of information that increases the ability of the actor to an-
ticipate and thereby minimize losses by reducing contact with those who repre-
sent avoidable risk. The use of an insurance or risk avoidance strategy cannot be
applied universally because its utility depends in part on the repetitive nature of
the behaviors generating risk. The ability to predict a particular behavior depends
on the similarities in the circumstances as well as a finite limitation on the number
of relevant variables one can include in the predictive model. According to
Reichman, "Insurance based techniques of control are largely mass produced and
data dependent, and, thus, they require a level of routinization that has not been
typical of social control in the past. Information that is used to screen, sort, class-
ify, and exclude needs to be standard, and dearly defined so that it can be quickly
evaluated. " 11 2
Yet, it is in that direction that we appear to be moving at high speed. A boom-
ing industry in risk assessment is emerging in the wake of increased corporate lia-
bility for catastrophic events such as airplane crashes, oil spills, hotel fires, and ex-
plosions in chemical plants. One company, appropriately named the Failure
Group, is called into play after the fact to analyze what went wrong, assess respon-
sibility, and make recommendations about ways to avoid repetition. Defensive
risk assessment is in great demand in other less dramatic areas as well. Peter
Huber describes the recent spiral in negative judgments that corporations receive
in a variety of product liability suits. 113 The courts appear less willing to accept the
claim that consumers ought to accept the risks inherent in the contract to buy. In-
stead, providers of goods and services must engage in poorly informed specula-
tion about the kinds of uses to which some commodity might be put, which
might entail some risk of harm. So informed, the firms attempt to warn consum-
ers against such dangerous uses. And, in those instances in which the warnings are
not sufficient, they seek to limit access to such goods by persons less likely to un-
derstand or heed such warnings. Each new judgment raises the level of uncer-
tainty and the demand for more information about consumer behavior to exclude
them from the risk pool.
Similar risk management techniques can be seen in the judgments being made
by the courts regarding crimin al recidivism. When serving as the director of the
National Institute ofJustice, James K. Stewart noted:

To help avoid future crimes against innocent people, a judge or parole board neces-
sarily assesses the likelihod of future dangerous behavior by an offender. Indeed,
86 OPERATING THE PANOPT!C SORT

laws in many jurisdictionHequire consideration offuture danger to the community

as one fuctor in release decision . .. . Predicting which offenders will be high-rate of-
fenders remains as yet an inexact science. But continuing research is honing the ac-
curacy and usefulness of prediction methods. 11 4

Norval Morris and Marc Miller's review identifies three kinds of predictions that
are commonly made in the justice system: statistical predictions, based on the
comparisons of patterns of individual behavior with the behavior of others; an-
amnestic predictions, based on a person's repetitive pattern of behavior; and clini-
cal predictions, based on expert assessment of an individual's behavior. 115 They
suggest that "statistical predictions are the preferred method of prediction be-
cause they can be tested and are open to scientific challenge." They also note the
ethical concerns regarding questions of fairness and justice involved i.n using
group statistics to predict individual behavior: "The meaning of a prediction is
that the individual has a condition-membership in a group with certain behav-
ioral probabilities-and not that the individual has that likelihood of the pre-
dicted behavior." 116 But they conclude that such considerations do not matter
when the person being assessed has already been com>icted by a court . Outside the
criminal justice system, the questions of actuaria] or statistical justice are not so
clear. Indeed, it may be noted that the criminal justice system itself may be chal-
lenged in terms of the apparent inequity with which it operates in charging, con-
victing, and sentencing to incarceration whites in comparison with persons ofcol-
or.111
Reichman notes that in "the insw-ance conteX4 classification and exclusion
have been used to prevent individuals from joining risk pools. What insurers refer
to as 'selective discrimination' is the backbone of the industry~' 118 She suggests
that the same selection principles have become central to the screening of pro-
spective employees. Drug tests and honesty assessments are only part of the arse-
nal of devices that provide data for the assessment of risk.
The risk assessment/insurance model is also clearly at the heart of t he credit
and financial services industry. James Rule noted in i974 the importance of the
risk avoidance strategy for the consumer credit industry; i 19 Because there was very
little in the way of collateral for much of consumer credit, the costs of default
would be relatively high. "Thus the art and science of credit management lie in de-
termining, in advance, who will pay and who will not, and in screening credit ap-
plicants accordingly. This is, of course1 a problem of social control." 120 Rule sug-
gested that this form of social control would work through the "prevention of
default rather than the coercion of those who misbehave." Jn this way, the system
"acts to exclude the would-be delinquents from the opportunity to disobey the
rules." 121 What remained for the industry was the development of reliable tech-
niques for the identification of those who were likely to default if given the chana_
The response was the development of increasingly sophisticated "credit scor~
systems for estimating default risk.
OPERATING THE PANOPTIC SORT

Peter McAllister describes a form of"behavior scoring" that was being used by
Citicorp Retail Services as an "early warning system:' which reported "dramatic"
results. The reduction of delinquencies by nearly 30 percent was projected to
mean a io percent to 15 percent reduction in credit losses overall. 122 Early warning
systems differ from credit scoring techniques only in that the points assigned are
based on actual behavior rather than demographic information. Armed with the
identification of customer accounts at the highest levels of risk, collections de-
partments could focus their attention on them, rather than on those accounts of
customers who would continue to pay without any intervention.
Proposals for the implementation of new smart cards would utilize a similar
logic and would assess the risk of each transaction before it is completed because
the intelligence in the card would allow it to generate a continuous assessment of
the card and its user.

This assessment opens a new and important facility. Does this cardholder have an
"earned" credit line by paid performance? What is the "risk" as determined by the
actual and current economic condition of the cardholder? How rapidly is the full
credit line to be made available to this cardholder? Does the credit demand correlate
with the combined balance of all account relations? 123

The Target
It would be inaccurate to suggest that the panoptic sort has been restricted to the
elimination of risk. After poor risks have been eliminated in marketing sorts, the
classifications may then serve to guide specialized appeals to individuals or
groups for whom the probability of success is highest. Very early on, data from the
U.S. census were used to target commercial appeals to individuals on the basis of
the character of the communities in which they lived. As early as 1973, the technol-
ogy of "geocoding" had achieved a considerable degree of sophistication.
Geocoding was defined as the assignment of geographical codes to records of
events or other descriptive data. Geodemographic clustering was later to be as-
signed to the name of the procedure that linked extensive socioeconomic data to
postal zip codes, which had been classified into one of forty different kinds of
neighborhood types. 124 In 1973, however, the presentation of much more simple
analyses of census data was held in some awe: "This same system, commercially
applied, enables us to do penetration studies that will boggle your mind." 125 The
author was describing the plans of a large circulation magazine publisher who
wanted to send a special edition of the magazine to readers in high-income areas
to collect a higher advertising rate. Apparently the IRS had made available income
data by zip code. Using the zip code alone proved to be an imprecisely defined seg-
ment because there was considerable variability around the mean reported by the
IRS. The census tract provided a more precise basis for targeting the distribution
of the special edition because it revealed that the high-income households were
88 OPERATING THE PANOPTlC SORT

tightly clustered together. The names and addresses of the people in the chosen
tract were easy to obtain.
The same seminar presentation demonstrated the utility of linking computer
graphics with the census tract file. Maps showing income, ethnic mix, and com-
pethors were suggested as valuable aids to a hypothetical investor in pizza parlors
in deciding where to build a store. This mapping capacity was developed for the
1970 census by the Census Bureau at a cost of$22 million. The project called Dual
Integrated Map Encoding (DlME) was originally developed "to assist in the mail-
ing of census forms., but private firms soon obtained it at the usual bargain rate-
the cost of a copy of the computer tapes." 126 As an adjunct to the i990 census, the
Topologically Integrated Geographic Encoding and Referencing (TIGER) map-
ping serviee makes it even easier to generate readily interpretable maps of eco-
nomic, social, ru1d political geography at the level of census blocks, rather than at
the metropolitan scale of previous Census Bureau services. 127 A promotional flyer
from the Census Bureau referred to the database as "the Census Bureau's 2ooth
anniversary present to tbe Nation." 121! Numerous actual and potential uses of the
mapping capacity of the TIGER system were included in the pamphlet. Who
could argue with the use of the system by police in Baltimore County, Maryland,
to identify clusters of spousal abuse cases and to allow researchers to examine the ,
linkages between this behavior and other inde,xes such as income, unemployment,
and alcohol abuse? 129 More troublesome possibilities emerge when we consider
the use of TIGER resources to make economic and poUtical redlining more accu-
rate and efficient.
Jonathan Robbin, the founder of Claritas Corporation, utilized a geographical
logic to build a successfutbusiness devoted solely to targeting for commercial and
political marketing. Early in his entrepreneurial career, Robbin identified some
thirty-four descriptors that accounted for 87 percent of the variance in those mea-
sures across the neighborhoods in the United States defined by zip code. 130 These
neighborhoods wete then classified into forty different types or clusters and
ranked according to an underlying index of quality. The names assigned to th
clusters reflect the socioeconomic dimension that largely ruled the cluster·
scheme. The top duster was called "Blue Blood Estates," and the lowest clu
was called "Public Assistance." The system was a roaring success:

Magazines such as Time, Newswe~k and McCalls were among the first clients, sorting
their subscriber lists by cluster to pubLish upscale editions with ads hawking higb-
priced luxury cars and furs for the residents of Blue Blood Estates and Money &
Brains. When Colgate-Palmolive wanted to test-market a aew detergent for young
families, it sent minlboxes to Blue Collar Nursery, characterized by starter-home
131
neighborhoods teeming with yoling families.

The political applications that began in 1978 were JjttJe different. By targ ·
the prolabor households that had been identified using the Claritas system at
OPERATING THE PANOPTIC SORT

census block level, it was possible to reverse an almost certain loss in a battle over
~right to work." Public opinion surveys had indicated that the antiunion position
was favored by an overwhelming majority. On election day, the results at the polls
reflected a complete reversal resulting from the targeting of voters expected to op-
pose the measure. 132
Naturally, a great many imitators emerged with their own versions of targeting
methodologies based on census and other readily available data. One observer
suggests that the "scope of file enhancements offered by some firms is astounding
and, perhaps, a bit unnerving." 133 Kevin Kramer and Edward Schneider describe a
trademarked approach called Custom Targeting, which proved to be especially
useful in political campaigns. 134 They describe the technology as a "mechanism
for ordering priorities-which segments of the electorate should get what kind of
message, when, how and how often. Broadcast media buying, ad development, di-
rect mail, phone banks, door to door canvassing, and candidate scheduling can all
benefit from knowing who to target and who to avoid [emphasis added)." 135
More sophisticated political targeting includes "life-style" data, enhanced with
information derived from individuals' use of their credit cards. The comprehen-
sive nature of this information research leads one analyst to remark that
"geodemographic wizardry aside, computerized voter targeting is nearing a preci-
sion that suggests Orwellian individual monitoring and manipulation." 136 An ad-
vertisement in the trade publication for political consultants includes an almost
bizarre image: four photographic portraits-two males, two females; two whites,
nvo persons of color-each shown looking cross-eyed up toward their own fore-
heads, where the symbols of the Democratic or Republican parties have been
printed. The ad copy reads: "If you need to know who's who on Election Day, you
need to know about Conotabs. We'll locate your voters, check their addresses, find
their phone numbers, tell you all about them and produce their names for calling
and mailing [emphasis added]." 137
Commercial and political targeting moves back and forth from high levels of
aggregation to the identification of specific individuals based on an asssessment
of how they will respond to a particular issue, opportunity, or challenge. The
panoptic sort determines the extent to which individuals will be included or ex-
cluded from the flow of information about their environment. As applied to tradi-
tional print media, this approach is currently referred to as "target market pub-
lishing;' in which both the advertising and editorial contents of the published
magazines are targeted more directly to the perceived interests of individual con-
sumers. 138 Magazines or catalogs are specially bound by printers such as the Ko-
dak 4400, which is able to handle up to sixty-four different "signatures," or bind-
ing designs, for a single publication. An early innovator, Farm Journal published
using R. R. Donnelley's binding system, was able to print 8,896 different editions
of that journal in 1984. 139 As applied to other communication forms, electronic
media in particular, targeting is consistent with the notion of"narrowcasting;' in
which mass appeal messages are sent to increasingly homogeneous audiences.
90 OPERATING THE PANOPTIC SORT

Electronic systems promise the ultimate in narrowcasting or targeting, so it be-

comes possible to send an individualized message to each individual on the net-
work.
Amex makes good use of the data it collects about its cardholders from their
applications, and more important, from the transaction-based data Arnex collects
from its own sources and from commercial vendors. The segmentation within the
community of Amex customers is not limited to the surface appearances indicated
by the different colored cards (green, gold, and platinum), but "Amex card hold-
ers have been broken up into i5 distinct segments and each of them can be tar-
geted with a high degree of precision." 140 This facility is expected to serve Amex
well as it expands its publishing activities.

LIST VENDORS
The pan optic sort depends on ready access to information about the environment
as well as about the incLividuals who make their way in their multiple and inter-
secting roles as citizens, consumers, and employees. The emerging market in per-
sonal information includes a growing independent sector of firms that supplies
information that can be used in conjunction with data gathered internally. The
following are just of few of the leading firms that are helping to define this sector.
Donnelly Marketing Information Services, a cLivision of Dun and Bradstreet,
offers a number of specialized database products and services. Conquest/Direct is
described as a desktop marketing system that facilitates geodemographic market
analysis. Clients are assured access to a database covering 90 percent of all U.S.
households, which would allow clients to generate profiles of "customers" by de-
mographics, life-styles, and retail sales expenditures. The software resources
would allow them to generate customized color maps of target market areas. Re-
lated mailing list services include the possibility of making selections on the basis
of"mail responsiveness, crecLit worthiness, vehicle information, ClusterPLUS life-
styles, contributGrs, financial investments, hobbies, occupations, census demo-
graphics and more." 141 An on-line service (Express) allows dients to perform the
search and to order the mailing list resources by remote means.
In a letter to stockholders before the announcement of a merger agreement
with Amex, Epsilon's president, Thomas Jones, described database marketing as a
"household word among more knowledgeable marketers." 142 Epsilon Data Man-
agement counts not-for-profit organizations as among its more important cus-
tomers for information services. At one time, its largest account was the National
Rifle Association, providing nearly 25 percent of its income, but the list of clients
also included the National Multiple Sclerosis Society. The approach to these cus-
tomers reflects a view that social marketing is still marketing. In their offer of mar-
ket analysis services, Epsilon's brochure asks "Who are your best customers?
Where do they live? What charitable 'products' are they buying ... and why? We'll
help you discover the answers through a comprehensive analysis of your market-
OPERATING THB PANOPTIC SORT 91

place." 143 The special services they promise their customers include the expansion
of their donors list: "Our fully computerized Media Department analyzes more
than 3,000 lists and 65 million prospect names every year-giving us firsthand
knowledge of which lists are the best lists to help you acquire the most new do-
nors."144
Some firms enter the telemarketing business as an adjunct of experiences that
they have developed through the provision of services to their own organization.
The Gannett Company, a dominant force in the newspaper business, announced
the development of Gannett Telemarketing in t989 as a spin-off of the telemarket-
ing organization serving USA Today. The primary resource was the list of USA To-
day subscribers, but the lists were expected to continue to grow through the addi-
tion of names and addresses of entrants to the numerous sweepstakes run by the
newspaper as an aid to circulation. 145 Early clients included firms selling sports
videos, educational programs, and extended warrantee service plans. 146
The American Student List Company, a subsidiary of American List Corpora-
tion, offered a variety of lists of college students in the United States. In 1990, lists
by state were offered at $40 per thousand; adding zip code information would
raise the cost by $5, as would sorting by class year. Field of study would add an ad-
ditional $10 to the cost per thousand for the one-time-only use of these students'
names and addresses. Another list vendor, Best Mailing Lists, offered hundreds of
specialized mailing lists with prices ranging from $45 to $85 per thousand.
Whereas space scientists were going for $45 per thousand in the i992 catalog, soci-
ology department heads were on the block at $60 per thousand, with teachers of
high school mathematics priced at $65 per thousand. A number of the lists, such
as those that identified political contributions by party or provided the home ad-
dresses of prominent men, were unpriced, suggesting that within this category,
several lists of varying quality and price were offered and that the details would be
supplied to customers who inquired further. An analysis based on new consumer
lists published in the newsletter of the direct marketing industry, Friday Report,
suggested that the most valuable names were white, middle-aged, high-income
male consumers, especially when they have purchased high-cost consumer items
such as computers or automobiles. 147
Telesphere Communications (Telesphere) is a primary user of the geodemo-
graphic software and database of the Claritas Corporation. Among the more so-
phisticated services developed by Telesphere is its "Caller Profile Report;' which
would provide clients with an assessment of the character and "quality" of those
persons who called a 900 number or other service that utilized automatic number
identification (ANI) to capture the billing number identification of incoming
calls. For advertisers using broadcast media, an analysis of penetration by area of
dominant interest (ADI) represents one potentially useful product. Utilizing
Claritas's PRIZM life-style clusters, Telesphere would allow an analysis of penetra-
tion among each of forty different life-styles linked to residential character. When
combined with Telesphere's reverse telephone directory appending service, a
92 OPERATING THE PANOPTIC SORT

caller can be identified by name and address, calling frequency, and broad life-
style classification. The future of such linkage services will undoubtedly involve
the classification of incoming calls before they are answered by a human operator.
A subsidiary of Equifax, National Decision Systems, offered what was called a
higher level of precision in its MicroVISION targeting system because it was able
to specify a zip + 4 level of geography involving ten to fifteen households, rather
than the broader two hundred to three hundred households at the five-digit zip
code level. The service claimed to have classified every single household in the
United States into a market segment. 148 Computer-generated maps as well as cur-
rent and projected demographic figures were theoretically available to customers
on demand. Life-style information available to the firm allowed listing of individ-
uals as members of segments defined in terms of price sensitivity, coupon use,
brand loyalty, television use, and other characteiistics of interest to consumer
product marketers. Because it was a subisidiary of Equifax, a leading provider of
credit bmeau services, the personal information profiles offered by the firm also
contained information about consumer credit activity- an activity that generated
considerable negative response within the industry and in Congress. 149 National
Data Systems also provided a comprehensive training program for users of their
Infomark database management and marketing system, which would allow access
to consumer databases, such as their proprietary list of some one hundred million
employed individuals (Daytime Population), with only a desktop personal com-
puter.
Responding to mounting criticism that resulted in legal actions initiated by
more than a dozen states against TRW, the corporation announced the develop-
rnen t of its own "privacy risk assessment" scoring system, which would allow the
company and its clients to give due consideration to the sensitivity that certain
data might hold for individuals, who might react by pressuring their legislators to
take action. 150 The sensitivity scoring procedure developed by the company ~p­
peared to take a cue from the recommendations made by Raymond Wacks, who
offered a classification of personal information sensitivity that was based on his
assessment of the extent to which the collection and use of the data represented a
risk of serious harm to the data subject. 15 1 The TRW list and that proposed by
Wacks both placed information about health and race in the highest sensitivity
categories. However, the recognition that certain data are sensitive and the refusal
to sell that information to a buyer are quite different stories.
Ed Burnett Consultants' i990 catalog included a number of special lists that
could be acquired at rates higher than those usually charged for lists of business
establishments and executives. In addition to the higher prices, several of the lists
came with some rather uncommon restrictions, including the requiTement that
the Burnett organization be able to review the proposed mailing before the lists
would be provided. This caution seemed particularly appropriate for lists of
23,000 subscribers to .Exceptional Parent Magazine, a magazine "concerned with
children's disabilities and impairments;' or credit purchasers from Fashion Bug
OPERATING THE PANOPTIC SORT 93

Plus, serving women needing large-sized garments. This class of special lists also
included some 60,000 paid members of the Smithsonian Institution who also at-
tended the Smithsonian's seminars and tours.
The Database America (DBA) companies, of which Ed Burnett is a part, offer a
variety of custom data enhancement services. A client would provide a customer
file organized in a way that would facilitate its matching with DBA files of more
than 84 million households, and a new enhanced file would be generated with in-
formation about purchasing behavior, estimated income, credit status (including
credit cards), investments, and charitable and political contributions. Each en-
hancement would be provided at an additional one to two cents per record.
DBA claimed that the quality and completeness of their data about consumers
exceeded that of their nearest competitors, except in the area of credit informa-
tion. Whereas DBA'.s credit card data was limited to mail order records and ques-
tionnaire responses, credit agencies such as TRW had direct and privileged access
to the actual credit records of the majority of individuals in the database. 152 A
somewhat different kind of list is that provided by Nielsen Media Research, an-
other Dun and Bradstreet organization, known most widely as the firm that pro-
duces ratings of media programs. Nielsen will provide random samples of work-
ing residential telephone numbers at costs ranging from six cents to twenty-six
cents per number, depending on the size of the sample drawn.
One of the more well known efforts to introduce a consumer database product
is the Lotus Development Corporation's product, Lotus Marketplace: House-
holds. Developed jointly with Equifax and Apple Computers, the product was to
have been a sophisticated compact disk-read only memory (CD-ROM) database
of 80 million U.S. households and the uo million adult consumers who reside
there. For a fee of$695, a user with an Apple Macintosh could begin the process of
customer «prospecting" through a specially tailored list of 5,000 names. Addi-
tional names could be acquired on the familiar cost-per-thousand basis. Available
data would have included identification of the household's geographical location
and zip code, the sex of adult consumers, an estimate of household income, buy-
ing behavior, and estimates of the revealed preferences of members of the house-
hold for more than one hundred product categories. 153 The product was designed
for smaller businesses, which would presumably have a more limited need for
consumer information than the larger businesses, which contracted with on-line
services such as those offered by Dun and Bradstreet. Only a coordinated move-
ment among advocates of privacy and among computer professionals prevented
this project from being introduced.
On-line remote access to data represents a significant change in the nature of
the market for personal information. Students and researchers are familiar with
the bibliographic databases through which they may search for published and un-
published information on a variety of subjects. Journalists argue that the new
databases and on-line searching capacity promise to transform the practice of
journalism. Tom Koch identifies the professional journalist as the public's surro-
94 OPERATING THE PANOPTIC SORT

gate, who through the enhanced access to information that on-line searching pro-
vides is now able to offer an alternative vision to that which might be preferred by
"flacks;' or advocates, or other interested information sources. 154 Journalists pur-
suing corporate malfeasance can gain access to the financial reports that publicly
held firms submit to the Securities and Exchange Commission (SEC), and to sev-
eral other reports that the OMB still treats as allowable.
Highly profitable commercial services such as the DIALOG/Knowledge Index
generally serve as brokers for information gathered by smaller firms. A second
class of databases that may be searched through remote computers includes the
growing number of statistical services and the associated services with "adminis-
trative registers," which are the files that contain information about identifiable
individuals rather than aggregate statistics. The format of the data in some of
these individually identifiable files makes it possible to generate aggregate statis-
tics as well as to perform matches and sorts as part of the process of developing
profiles. 155 Meredith Corporation, for example, claimed to have one of the largest
of such databases in the fall of 1991. Their database of some 56 million customer
profiles was reportedly constructed from i50 million different records, many of
which were derived from Meredith's other communications and real estate opera-
tions.156
The growth in the number of firms competing in the market for on-line data
has been nothing short of phenomenal. Starr and Corson suggest that the growing
number of personal computers equipped with the modems necessary to access
these services ensures continued growth in the industry supplying the data. 157 An
increasing number of these personal computers are in private households, and
government reports in 1988 projected strong growth for both computers and mo-
dems, reaching 23.9 percent penetration of U.S. households by i992. 158 The launch
of the consumer-oriented videotex service PRODIGY by Sears and International
Business Machines (IBM) reflected this sense of optimism. For consumers un-
skilled or less sophisticated in the navigation of the hundreds of competing ven-
dors, another group of brokers called "gateways" provides the nonspecialist with
access and advice about gathering data. The approval in late i991 for the entry of
the Regional Bell Operating Companies (RBOCS) into the information business
marked what is likely to become a watershed in consumer usage of these ser-
vices.159 What should be kept in mind is that use of these services by consumers
will undoubtedly generate additional TTGI, which feeds back into the modifica-
tion and correction of the panoptic technology.
4
CORPORATE PERSPECTIVES
ON THE PANOPTIC SORT

IN PURSUIT OF THE CORPORATE VIEW

The panoptic sort is a technology that has been designed and is being continually
revised to serve the interests of decision makers within the government and the
corporate bureaucracies. The previous chapter focused attention on those organi-
zations that utilize personal information to segment and target consumers
through direct marketing. Of course, not all firms that are involved in direct mar-
keting see the operation of the panoptic sort in quite the same way. This is due in
part to the relative importance of database marketing to their overall income and
in part to the extent to which their business involves direct contact with the pub-
lic. Retail firms that provide services directly to the public are more sensitive to
the goodwill dimensions of a public issue such as privacy, whereas service firms
that deal with consumers only indirectly may feel less of a need to pay attention to
public sentiment in this area. However, as the panoptic sort becomes an essential
part of modern business practice, more firms will be forced to develop a corporate
position on the underlying questions about the legitimacy of its use. To under-
stand the dynamic process that underlies the evolution of a corporate posture,
three firms were selected for a limited case study. This study served as a back-
ground for an analysis of a survey ofbusiness leaders.
l examined copies of the annual reports for 1984 through 1988 from three cor-
porations that were active, or soon to be active, in the debates over personal pri-
vacy. My review of these corporate reports sought to characterize the different
approaches to the issue of privacy and the panoptic sort as it was described to ac-
tual and potential shareholders. The corporation's letter to their shareholders is
the primary instrument it uses to call attention to past successes and to problems
on the horizon that the corporation is mobilizing its resources to address. Com-
ments about information policy reflect the relative importance of these issues to
different corporations in the information business. The three firms I selected for
this study were the American Express Corporation (AMEX), TRW, and Equifax.
Of the three, Amex has traditionally had the most direct contact with consumers
and, therefore, it appeared to have made the greatest investment in generating
goodwill through its involvement in privacy policy.

95
 CORPORATE PERSPECTNES ON THE PANOPTIC SORT

American Express
Amex has had extensive experience in dealing with the complexities of the privacy
issue. This firm claimed leadership in the area of consumer privacy ·with its 1974
institution of a mailing list policy that would become a direct marketing industry
standard. Through its Warner/Amex Cable Communications division, the corpo-
ration became the first cable operator to establish a privacy code in 1981, only to
sell its share of the corporation in i985. 1
In its t984 report, as a strategic accomplishment Amex presented a corporate
policy that in the future will become a critical point in the public debate about
privacy. T his privacy sensitive year saw the introduction of Amex's One Enterprise
concept: "Under the One Enterprise concept, we provide our businesses consider-
able autonomy but at the same time we expect them to work together as a single
enterprise. At year end, approximately 140 One Enterprise projects were in opera-
tion or under development, including many where our companies 'cross market'
products and services created by other members of the American Express family." 2
As such corporations as Amex become increasingly involved in seemingly unre-
lated lines of business, telecommunications facilitates the combination of cus-
tomer data into a single marketing tool. Current restrictions on data sharing do
not cover most of the internal uses of personal information outside the banking
sector, but many observers recognize the potential conflicts that such practices
may generate between what customers expect and what firms do as a matter of
standard practice.
In i985 Amex increased its involvement with the panoptic sort by consolidating
its ownership of First Data Resources. In the view of Amex's nrnnagement, "infor-
mation processing has become more critical to all our businesses, in such areas as
point-of-sale electronic services and telemarketing."J The i986 report took note of
the expansion in One Enterprise efforts to take advantage of the "opportunities
inherent in their complementary strengths in markets, product lines and cul-
tures" and looked ahead favorably on strategic efforts to "increase the precision
with which we segment our markets-enhancing our ability to reach discrete
groups while offering additional value matched to our particular needs." 4 As part
of a corporate movement toward this strategic goal, Amex formed its Direct Mar-
keting Group in i986. One of its divisions, Merchandise Services, ranked as the
fifth largest direct merchandiser, and First Data Resources continued to grow
through acquisitions.5
Because 1987 bad been a troublesome year for financial institutions in general,
and especially for those with substantial stock market exposure, the Am ex annual
report focused more on the future than on the immediate past. Technological ad-
vances in artificial intelligence put the corporation's authorization and transac-
tion processing business "ahead of the curve" in this growing industry. This ex-
pertise supported expansion into health care marketing, where they utilized
simplified techniques for relational database analysis. 6 The i988 annual report re-
CORPORATE PERSPECIWES ON THE PANOPTIC SORT 97

vealed even greater involvement of the corporation in privacy sensitive activities.

The Data Based Services Group, already an industry leader in third-party authori-
zation and transaction processing for bank cards and cable television systems, ex-
panded further through two new projects. 0 ne project would join with Donnelley
marketing " to access a database of more than 89 million households for market-
ing purposes"; the other, taking advantage of the possibilities of 900-number ser-
vices, would "market services that may permit the simultaneous handling of up to
30,000 interactive telephone calls for marketing and entertainment purposes."7
Thus, we see that, over the years, Amex had become increasingly integrated
into the direct marketing industry, not only as a direct marketer through mail and
telecommunications but also as a dominant vendor of third-party authorization
and processing services. This increased risk provided a dear rationale for the cor-
poration's donation of an initial $so,ooo grant to establish the Direct Marketing
Association's (DMA) privacy task force8 and its support for the adoption of stan-
dards of information practice, which would be considered progressive from nearly
any position. 9 According to these standards:

i. The individual has a right to know that personal data about his life will not be
rented or sold against his wishes and a right to know exactly what information a
company makes available to others.
2. It is unethical for a company to collect information for one purpose and to rent
or sell it for another purpose against the customer's wishes.
3. An individual has the right to know who is the sender of a direct mail piece.
4. The consumer should be clearly and frequently advised of his right to be ex-
cluded from any and all lists. 10

Amex's long-term involvement with consumers through direct contact also pro-
vides an explanation for the appearance of more explicit discussions of corporate
concern for the privacy issue in the company's annual reports.
In the context of discussions of public responsibility, the 1988 annual report de-
scribed the expanded role of the company's Consumer Affairs Office. The office is
responsible for " major initiatives in consumer protection and education: ' and it
also "monitors the Company's Privacy Code of Conduct, which provides stan-
dards governing the collection, custody and! use of customer information." 11 Be-
cause of the international scope of the corporation's business and the potential
risks represented by the higher level of privacy protection in the European Com-
munity, participation by Amex's Consumer Affajrs staff in public forums in-
cluded many foreign sites i.n i987.
In the corporation's report to the SEC, Form 10-K, there was a more explicit
discussion of the potential regulatory restraints the corporation thought it might
face in its privacy sensitive lines of business. Regulations linked to credit access,
credit billing, and credit reporting were not thought to represent any particular
 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

threat to the company's card business, nor were there any specific regulatory risks
associated with either the Data Based Service Group or the Direct Marketing
Group. 12 ·

TRW
TRW, a major defense contractor and developer ofsophisticated computer and in-
formation systems, began in i984 to build on that experience to enter the informa-
tion economy more directly. 13 The corporation's consumer credit reporting unit
received only a slim paragraph's mention in 1985, but it was destined to become
more visible each year. 14 In i985, the information systems group, which included
consumer information services, grew by more than 20 percent. The annual report
claimed industry leadership for its credit reporting unit and announced the con-
troversial new service, TRW Credentials, which was "developed to help consumers
monitor and control the credit process" 15 but which was seen by critics as a means
to gather even more information for credit files at the consumer's expense. 16
In the 1986 annual report, TRW identified the information systems group as
the most profitable and fastest growing part of the organization. Their market
analysis suggested that the "growth of the information systems and services busi-
ness reflects a fundamental change in the way people increasingly obtain the facts
they need and want-from electronic databases." 17 The Credentials program was
described more completely: "The service enables people to monitor requests for
their c_redit histories and to apply for credit more easily by filling out a master
credit application, which is then stored in a TRW data bank for authorized use by
credit grantors." ts
The report also noted the expansion of the credit reporting service to all fifty
states and discussed TRW's controversial practices in which it "markets portions
of its credit database to financial services firms that wish to target particular
groups of prospective customers." 19 This same report underscored the integration
of defense and commercial applications of advanced TRW information systems
with significant implications for the enhancement of the panoptic sort. One such
system was described: "Called the TRW Fast Data Finder System, it will enable us-
ers to scan raw data for .nearly 600 different, complex search requests simultane-
ously at a rate of more than 7 million characters a second- the equivalent of six
500-page novels."20
ln the i987 report, information systems maintained its strategic importance for
TRW, and major expansion was planned for those areas that utilized advanced in-
formation systems to collate intelligence derived from proprietary databases. By
i987, TRW claimed to provide more credit reports than any of its competitors, and
1
its annual report noted practices that were later to make the corporation some-
thing of a pariah within the direct marketing community. The credit data in its
files on more than i38 million consumers would be used by other clients to market
financial products and services. This is a use of credit reports tho ught by many to
CORPORATE PERSPECTIVES ON THE PANOPTTC SORT 99

be barred by the Fair Credit Reporting Act (FCRA), but no mention of these po-
tential regulatory conflicts was mentioned in the annual report or in the Form
10-K report to the SEC.
Growth in the revenues and profits from the information systems group appar-
ently stalled in i988, but the annual report continued to identify it as "the key
building-block business" likely to be enhanced through the acquisition of Chilton
Corporation, establishing TRW as the unchallenged industry leader. Throughout
the five years examined, none of the corporation's annual reports or Forms 10-K
gave any indication of concern about the emerging privacy crisis.
It was not until spring 1989 that TRW began to take a more activist stance in re-
sponse to growing criticism. It hired a social scientist a~ a vice president to help
formulate its positions on information policy issues, and it funded a conference at
Georgetown University in June to help focus debate on the issues surrounding
technology and information policy. 21 It may be that TRW's decision to invest in a
more visible public posture was the result of increasingly negative coverage in the
press and negative comments from associates within the industry associated with
a TRW mailing of names on a credit list to a reporter from U.S. News and World
Report. 22 Reports that emerged in 1991 indicated that TRW was having second
thoughts about its role in the information business after the attorneys general of
several states had initiated suits against the corporation for its use of personal in-
formation from its credit files in ways these attorneys general considered to be
barred by the FCRA.23

Equifax
Equifax has been intimately involved with the panoptic sort because its principal
lines of business involve providing guidance to third parties about identifiable in-
dividuals. In the discussion of its corporate vision in i985, the annual report re-
veals this centrality:

With the emergence of the financial services industry in recent years has come a
broader perception ofEquifax's role as a provider of information for business deci-
sions. This perception brings us closer to the consumer in the sense that the con-
sumer's action creates the need for just about every service we perform. Stated an-
other way, every financial transaction requires information, and the consumer, at
the moment he or she initiates a transaction, triggers a process that can ultimately
involve Equifax.24

The report notes that growth in recent years involved not only the expansion of
their reporting network to owned and affiliated bureaus, but also the pursuit of
new service options "in the areas of credit promotion, marketing, applicant pre-
screening and others." 25
A special section in the report for 1984 focused on data protection, but the issue
was approached from the position of a data manager concerned with guarantee-
100 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

ing the security of that data from access and tampering by computer criminals.
Equifax indicated considerable corporate interest in and implied support for leg-
islation that would levy substantial fines for unauthorized access or unauthorized
use of information in a computer file. T he extended discussion of the problem in-
cluded the expression of concern that the press tended to present corporations as
the criminals, rather than the victims of computer crime.

The news media, which have an inestimable influence on the way people think about
issues, can do a great deal to promote public understanding. In coverage of com-
puter crime incidents, some media commentators have exhibited a disturbing ten-
dency to portray the institution rather than the criminal as the villain. This is some-
what like blaming a bank for being robbed, simply because that's where the money
was. 26

Perhaps because the press coverage ofcomputers in i984 contained a considerable

number of Orwellian references to corporate as well as governmental "big
brotherism;' this discussion in the annual report ended with an appeal for the de-
velopment of a new coalition of interests, "a unified effort by the business infor-
mation industry and its customers to arrive at solutions that protect data while
allowing the unfettered use of technology necessary for the legitimate gathering,
storage, transmission and use of business information." 27
For Equifax, i984 was a year in which the firm gained from increased corporate
use of information in the panoptic sort as a means of risk avoidance. Equifax ser-
vices were used to evaluate credit risks, as well as to aid in screening potentiaJ em-
ployees. The report noted that "marketers of goods and services increasingly re-
lied upon selective information, especially demographic information targeted to
population segments." 28
The expansion ofEquifax's involvement in the panoptic sort in 1985 included
the addition of drug screening, as well as the highly specialized screening of po-
tential employees that the company performed for organizations in the nuclear
energy field. lts credit and marketing-related services were expanded to include
sophisticated and largely automated routines for identifying potentially fraudu-
lent applications or purchases. By i987, Equifax claimed to be the "world's leading
provider of information for consumer-initiated financial transactions," involving
more than i.5 million transactions each business day. 29 In a somewhat novel way
of presenting the corporate position on policy issues, the i986 annual report pre-
sented a panel discussion in which senior managers answered questions posed by
a variety of interested parties, including a consumer-affairs specialist.
Richard Bullock, executive vice president of the National Council of Better
Business Bureaus, asked the "privacy" question, presumably in the public interest:
"With your increasing use of information, what are you doing to protect the pub-
lic and maintain individual privacy?" 30 The published response was narrow and
repeated the cloaking phrase, which excuses all and limits little: "There are Fed-
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 101

eral and state laws to protect privacy and to ensure that information is used prop-
erly and not released to anyone without a legitimate business purpose." 31 Io addi-
tion, the response attributed to chairman W. L. Burge included notice ofEquifax's
efforts to increase computer security and emphasized the value of its automated
systems, which provide the corporation with historical records of access to partic-
ular files. Equifax's corporate management saw no threats on the horizon arising
from pending legislation that might have any adverse effect on the information
industry or its role in the pan optic sort.
The rest of the annual report presented business sector reports that were glow-
ing in their promise of continued growth in the information intensive aspects of
the corporation's business. In its discussion of the consumer credit and marketing
services, the report noted continued expansion of fraud detection and risk avoid-
ance services: Through the use of "sophisticated statistical modeling, we help cus-
tomers monitor credit portfolios to provide early warning of potential problem
accounts:' 32
In 1987, Equifax continued to expand the number and variety of services to in-
surers and other bearers of risk. One new service with the informative acronym,
CLUE, facilitates exchange of information, including motor vehicle records, to as-
sist insurers in making underwriting decisions. 33 The corporation continued to
claim a competitive advantage over other database management firms that were
unable to enhance their target marketing profiles with credit information. This
use of credit information for marketing purposes represents one of the more sen-
sitive aspects of the practices ofeach of the three firms examined in this case study,
yet Equifax gave no sign of recognizing this threat in its i987 report. Equifax exec-
utives were heartened, perhaps, by their recent success in an appeal before the
United States Court of Appeals for the Eleventh Circuit.34 The court's analysis ar-
gued for a narrow definition of "consumer report;' which would involve restric-
tions and penalties under the FCRA. Thus, market gains in that year, in the con-
text of a judicial all-dear signal, led the firm to pursue au even more aggressive
expansion in the following year.
The year 1988 saw the creation of a new division, the Marketing Services Sector,
the acquisition of fourteen companies, and a significant common stock offering
of nea.rly 2.9 million shares. Equifax's segmentation services were precisely the
kind of profiling efforts that represented the leading edge of panoptic technology
and were efforts about which privacy advocates had been complaining most vo-
cally:

Sophisticated models anaJyze various combinations of data using complicated

mathematical equations and then construct indices to help predict consumer inter-
est and buying potential for particular goods and services. The key to success in pro-
viding marketing information and services is to integrate all the relevant transaction
activities and information available from Equifax to help businesses answer a variety
of marketing questions. 35
102 CORPORATE PERSPECTIVES ON THE PANOPTlC SORT

Equifax seemed apparently unconcerned about any kind of consumer backlash.

Indeed, growing competition from other firms in the information industry
seemed to be their prime concern. In a dedication to W. L. Burge, who retired as
Equifax's chair in 1988, tbere was a clear sense tbat the troubled times associated
with consumerism of the 1960s were past. Burge had become chief executive offi-
cer (CEO) in i967, and the late 1960s presented substantial hurdles for the corpo-
ration because of the presumed influence of corporate practice on consumer well-
being.

During that period, which preceded what bas now become the Information Age,
Equifax, along with all who gathered and used essential information for decision
making, came under serious challenge and examination by consumer groups and
political activists. Some even questioned the right of business to evaluate risks. Mr.
Burge led Equifax through it all with dignity and with calm assurance rooted in the
knowledge that the Company was filling a vital role in society. 36

However, by the end of 1989, Equifax had contracted with Louis Harris and Alan
Westin to administer a national survey of public opinion regarding the kinds of
business practices that many consider to be invasions of privacy. These Equifax
surveys continue the series of such studies funded by Sentry Insurance and New
England Telephone that not only inform industry public relations, but that also
have been critical components of the public debate on privacy. 37
Amex, TRW, and Equifax are three very different corporate "citizens:' Their
differences can be explained in part by the differences in the visibility that is cre-
ated through their direct contact with consumers. Amex, the corporation with the
most direct contact, appears to have been the organization most sensitive to con-
sumer reaction about apparent threats to privacy or to abuses of personal infor-
mation. Yet the need to expand sales, profits, and market share has led each of
these firms to explore the instrumental use of personal information in ways that
will continue to attract the harsh glare of publicity and the risk of sanction. For
these and other firms like them, it is the use of personal information for the sup-
port of telemarketing that is most attractive and most risky.

THE TELEMARKETING VIEW

The telephone will increasingly become the focus of the debate about privacy and
the panoptic sort. Not only is the telephone the instrument that provides an an-
noying intrusion into the privacy of the household, but, as has been noted, the use
of the telephone for a variety of transactions is an important source of data that
enhances a consumer's profile. This enhanced profile will generate still more an-
noying calls. Mark Nadel discusses the issues surrounding the threat that the tele-
phone presents to "the right to be left alone."38 The recognized privacy interest is
seen to be the strongest in the home, yet the technology of the telephone allows for
penetration of the barrier of the closed door and the pulled curtain. The insistent
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 103

ringing of the phone is difficult to ignore, and ·even though devices such as an-
swering machines or passive displays such as Caller-ID represent technological re-
sponses, they cannot completely protect against unsolicited and unwekomed in-
terruptions.
v\Then the telemarketing industry introduced the automated technology that
had the capacity to dial hundreds of numbers and play recorded messages (auto-
mated dialing recorded message players, or ADRMPs), many citizens and their
legislative representatives argued that the invasion of privacy had gone too far. A
variety of restrictive bills were introduced and passed in state legislatures across
the United States. The patchwork nature of telephone regulation generates con-
siderable difficulty for direct marketers and market researchers who utilize ce;,_
tralized phone pools to place calls around the country. Thus, the industry has had
a powerful incentive to work toward uniform legislation at the federal level that
would support unrestricted use of the telephone for "legitimate business pur-
poses." Even when the congressional response to telemarketing reflects the general
sense of public annoyance, the Federal Communications Commission (FCC) acts
to protect the interests of the marketers. 39
Limitation on the use of the telephone to gather information about consumers,
generally without their informed consent, represents a second concern for the
telecommunications industry and their primary clients in telemarketing. After di-
vestiture and the seemingly interminable process of removing the competitive re-
straints on their participation in other areas of the information business, AT&T,
the Bell operating companies, and their competitors in long distance have begun
to offer a variety of information services that promise not only to expand revenue,
but also to elevate the temperature of the privacy debate. The premium 900 num-
bers represent one such business venture with great business potential. The tele-
phone company as carrier shares the fee collected on behalfof users of the service.
A$ part of its service, the telephone company will forward the numbers and per-
haps the names and addresses of the calling party. This information was recog-
nized for its potential market value in that it could become part of a marketable
list of-persons demonstrably interested in a particular service. Arrayed against
these corporate interests in the collection and sale of personal information are the
claims made by privacy activists that individuals have substantial property rights
in the information they generate through their transactions. 40
Telephone companies see the provision of "privacy enhancing technologies:'
such .as an option that would allow individual consumers to block the passage of
their telephone numbers when they make specific calls, as a threat to the profit
potential ofan entire line of enhanced telecommunications services. One industry
analyst suggested that the industry's expected income might be reduced by as
rnuch as 50 percent if blocking were allowed.41 One AT&T vice president has been
identified as linking these privacy questions to the entire future of what the phone
companies call "the intelligent network." From the industry perspective, "that
network relies on the ability to recognize and utilize the calling party's number.
104 CORPORATE PERSPECTNES ON THE PANOPTIC SORT

And what effect some very broad, Luddite type of policy would have on the devel-
opment of that network is of concern to us:' 42
Telephone-based services, such as the teletext service offered by Sears and IBM
(PRODIGY), involve similar privacy issues. The PRODIGY system was designed
from the start to be a marketing tool. After four years in development and some
$JOO million in investment, the service moved into test marketing with a number
of hopeful vendors of shop-at-borne services.43 Advertisers, or "information pro-
viders" as they are called, are to be charged on the basis of tbe number of users that
actually view their messages on the screens of their home computers. Thus, for the
system to operate efficiently, it must provide a record of which "messages" have
been viewed, although not necessarily by which consumers. That additional infor-
mation is a valued enhancement that may or may not emerge as a separate prod-
uct.44 The PRODIGY service users are asked to provide demographic information
at the time they initially "log on" to the system, and they add to that database each
time they "jump" or "zip" or "look" or take some other action in response to a
new screen. Item seven in the service agreement notes this fact: "One of the valu-
able and unique features of the PRODIGY service is its ability to personalize in-
formation and transaction services to each Member's interests. Personalization is
based on data provided by the Member (or Membership Holder) to Prodigy, data
derived from the Member's use of the PRODIGY service, and from the Member's
responses to Prodigy's questions and surveys."45
The agreement indicates that aggregate information about members can be
disclosed for any purpose the company chooses. Rules regarding the use of indi-
vidually identifiable information about members seem to make it available for the
marketing purposes of present or future "information providers:' And thus, ques-
tions about the use of transaction-generated data are botmd to emerge as privacy
concerns for PRODIGYand its imitators.

DIRECT MARKETCNG ASSOCIATION

The DMA has the abnost impossible task of bringing together the vastly different
organizations (more than 3,500 in the United States) that share little in common
beyond their use of a common marketing technology. The difficulties that these
organizations have in finding a common policy stance that satisfies the needs of
all of the members of business associations frequently serve as the basis for the
formation of single-issue coalitions, which then contract with outside lobbyists to
present their particular views on an issue.46
In support of its members, and as something of a foil against public outrage,
the DMA sponsors mail and telephone preference services that allow consumers
to indicate that they would prefer not to have their names included in national
consumer lists. DMA officers participate actively in the policy process through
their formal testimony in hearings and through their submission of language for
proposed legislation. Their efforts to "educate" the public about direct marketing
CORPORATE PERSPECTNES ON THE PANOPTIC SORT 105

involves the publication and revision of pamphlets, booklets, and guides for good
business practices. Their efforts to influence their membership include the estab-
lishment of numerous councils, task forces, and special seminars, which bring to-
gether industry leaders to discuss critical issues and concerns. The monthly news-
letter, Washington Report, provides a regular update on regulatory and legislative
activities.
The DMA is periodically forced to revise its "guidelines for ethical business
practices," as technology and industry practices generate new problems. A recent
version of the guidelines includes a policy statement regarding unlisted telephone
numbers. The ethical position of the DMA is that "telephone marketers should
not call telephone subscribers who have unlisted numbers unless a prior relation-
ship exists." 47 Discussions within the DMA's Privacy Task Force, however, suggest
that questions about automatic calling number identification, which threaten to
reveal unlisted numbers and therefore make them available in published lists, may
require revision of the guidelines.48 Another association whose members are pri-
marily in the business of providing "yellow pages" information services agreed
that they would not release information about calls made to their services without
prior consent. Some providers that would presumably be bound by the DMA pol-
icy seem determined to circumvent the policy by including the "electronic con-
sent'' (using the star button on a touch-tone phone) as part of the number unsus-
pecting customers would be asked to call.
The issue of transaction-generated data has the potential to stimulate consid-
erable debate within the DMA as organizations such as TRW and Equifax pursue
corporate strategies through activities likely to bring public wrath down on the
shoulders of the industry. In one issue of the DMA's bimonthly newsletter, Direc-
tions, Ed Burnett, an active list vendor, discussed the privacy issue in the context of
"problems and abuses." It seemed that different state legislatures had responded
to growing public concern about privacy by placing restrictions on use of the pub-
lic databases, such as automobile registrations.49 Problems for direct marketers
identified with the list industry arise most significantly when the commercial
business ofDMA's members becomes closely identified with the activities of gov-
ernment. In 1984, the association became alarmed when the IRS sought to use
commercially available lists to identify persons who should have, but who appar-
ently had not, filed tax returns. 50 In his testimony before a House committee,
DMA chair Alexander Hoffman expressed the collective concerns quite clearly:

They will come gradually to understand that the IRS is using census data to overlay
on the basic mailing lists. And we believe that an inevitable consequence of such a
chain of events carried out broadly and nationally would be a tendency of the people
to view this as just one more invasion of privacy; j ust one more step in Government
intrusion in our lives; and they would gradually tend to conclude that it is not a very
good idea to have your name on a mailing list. 51
106 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

In 1992, the direct marketing industry was threatened once again by the efforts of
the FBI to acquire compiled lists from leading firms. The FBI was reported to be a
steady customer of Metromail and used its MetroNet service to provide remote
identification of individuals identified by telephone number, but the firm and its
competitors claimed that they would oppose FBI requests for compiled lists.si
The government/indus.try link alone does not raise concern within the indus-
try about its vulnerability to growing public concern about privacy and the conse-
quences of the panoptic sort. In a forum organized for its readers, Target Market-
ing asked somewhat rhetorically, "What do you think is the most important issue
direct marketers face today?" Katie MUidoon, a direct marketing CEO, answered
very much on point:

Privacy is my major concern because the consumer is becoming much more aware
of the databases·we are collecting-files of names of bankruptcies and the exact
credit on individuals' ch;rge cards and so forth. The consumer doesn't really under-
stand how we use suppress files. They see. this as a real invasion of their privacy. I be-
lieve this attitude will cause legislation. It is very frightening. s3

Muldoon thought that the problem was exacerbated because it was not just con-
sumers talking among themselves, but it was newspapers and other periodicals,
and from her experience, "when publications talk about it, consumers talk about
it more." 54
TRW's business strategy produced many raised eyebrows and no small amount
of concern within the DMA when it asked the IRS to validate the social security
numbers of its more than 143 million data subjects. TRW would be likely to cause
even greater concern if more people come to realize that it is a contractor with the
U.S. Postal Service and it helps to process the millions of change-of-address forms
that consumers submit when they move.ss Members of the DMA have been
openly critical ofTRW in print and in industry seminars, but these do not exhaust
the bases for disagreement. There are basic differences in philosophy and strategy
that are still being ironed out within the DMA and within the corporations in-
volved in direct marketing. TRW's response has· been that much of the criticism
that explodes periodically in the press is the result of competitive jealousy. "If I've
got an airplane and you've only got a car, you don't want me to use the plane" was
the reponse attributed to Dennis Benner, a vice president ofTRW's Target Mar-
keting Group.56 Differences between firms in terms of their resources, their so-
phistication in the use of panoptic technology, and their understanding of the
emerging privacy debate will be explored at a later point in this chapter.

ROBERT POSCH
AND CORPORATE OPINION LEADERSHIP
Robert Posch, vice president and counsel for the Book Clubs Group of Doubleday
and Company, is a productive and highly visible defender of the interests of data-
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 107

base marketers. Posch publishes a regular legal column in the trade magazine Di-
rect Marketing and forcefully represents his views at conferences and seminars or-
ganized by the DMA. Posch's approach may be characterized as a frontal attack.
He argues that there is no constitutional basis for a privacy claim. In addition, he
minimizes the extent of the intrusion that telemarketing or direct mail represents.
For Posch, "telephone marketing not only isn't an invasion of privacy but it isn't a
nuisance either." Thus, he explains, "there can be liability for a nuisance only to
those to whom it causes significant harm, of a kind that wollld be suffered by a
normal person of ordinary sensibilities in a community."57 Rather than respond to
escalating claims for what he sees as a nonexistent right, Posch counsels the indus-
try to push instead for its constitutionally guaranteed rights of free speech, which
have been extended by the Supreme Court to include truthful commercial speech.
"Free speech is the industry's public posture high ground and winning constitu-
tional argument. Reliance on a 'cut our losses' argument on privacy means our
opponents continue to define the issues and win the states:'58 Posch concludes by
noting that "privacy is the Achilles' heel of database marketing law and consumer
relations. How well we finesse this issue will determine whether non-store mar-
keting prospers or follows other U.S. industries into decline as a result of their re-
fusal to recognize changing public policy."59 Posch has been especially vigilant
with regard to the importance oflists to the direct marketing industry, suggesting
that every list is important, and he has stated that to allow the state to restrict the
industry's use of any list would eventually lead to their being restricted from using
every list. "If we continue to lose non-commercial. lists (e.g., motor vehicle, li-
brary, etc.) WE SHALL LOSE ALL LISTS:' Posch emphasized, and then he
warned, "already there are attempts to ban 'residence' and 'geographic discrimi-
nation.' Tpis will increase as insurance companies unable to prescreen for AIDS
will consider screening applicants by ZIP."1>0
Of course, not all sectors of the industry have been willing to adopt Posch's po-
sition. In a debate organized by the Long Island Direct Marketing Association,
Posch argued his position against the more conciliatory, "corporate responsibil-
ity" views of Roy Schwedelson, CEO of WMI!Worldata. 61 Posch warned that rec-
ognition of any privacy claims in corporate data begins to take the industry down
a slippery slope:

There is no legal difference between a library list, a motor vehicle list or any other
list. It's all one composite of information. lfyou believe we're right, and I do, then it's
our free right to have this. But if you believe some Lists are invasion of privacy, how
do you distinguish them? We can be consistent for free speech, or we can get bogged
down in privacy. Ifwe get bogged down in privacy, we cannot win. 62

Schwedelson's position is one that might be considered constrained rationality.

He argues that privacy claims cannot be ignored or merely "wished away." From
his analytical position, existing legislation restricting corporate activity under the
FCRA will be brought to bear against the interests of direct marketing because of
108 CORPORATE PERSPECTIVES ON TH£ PANOPTIC SORT

the behavior of several organizations, including TRW, that use credit data for
marketing purposes in ways that generate public alarm. It is argued that because
of their marginal position in the direct marketing industry, such credit reporting
firms as TRW and Equifax do not share a full commitment to the industry's posi-
tion regarding privacy. Schwedelson argues further: "IfTRW loses on their data-
base, they're going to go home [to their primary business] and leave us with the
problem. Frankly, I don't think the DMA should have put Benner [vice president
and general manager of TRW Target Marketing Services Division] on the Ethics
Committee. That's like letting the fox run the hen house." 63
Few question the fact that the privacy issue is destined to become a significant
aspect of consumer affairs legislation in the i99os. Attitudes toward business simi-
lar to those that characterized the turbulent 1960s might accompany the reemer-
gence of ecological concerns on the national policy agenda. A lack of agreement
on the issues within firms centrally or peripherally involved in the panoptic sort
will continue to challenge associations such as the DMA quite severely. A survey of
U.S. business leaders provides some evidence of cleavages within the corporate
sector that may widen as the issue develops.

THE MAIL SURVEY

David Linowes reported on a survey of 126 Fortune 500 companies that sought to
describe the nature and extent of corporate efforts to protect the privacy of their
employees. 64 There had been no comparable surveys of corporate practices pub-
lished with regard to corporate collection and use of consumer information, espe-
cially with regard to their use of such information for the purposes of marketing.
This section describes the results of a survey that I designed to help erase part of
this knowledge gap. 65
Early in i990, I sent letters to nearly one hundred corporate representatives
who had been identified by the direct marketing newsletter Friday Report as hav-
ing introduced some innovation in di1·ect marketing.Twas especially interested in
those companies that had new products or services that developed, enhanced, or
in some unique way, expanded the ways in which consumer lists could be used for
marketing purposes. The letter of inquiry identified the sender as an academic in-
terested in marketing technology, specifically in the use of customer data, and re-
quested available material about the company's list products. About 25 percent of
the companies responded, and their responses ranged from cursory to most gen-
erous. One secretary forwarded a copy of her employer's popula.r book on direct
marketing. 66 Others provided marketing packages, with examples of list manage-
ment and promotional services that the companies provide. These responses, in
addition to the materials gathered from the academic and trade literatures, pro-
vided the basis for the questions that were pursued through a mail survey of cor-
porate leaders.
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 109

The questionnaire reflected an attempt at classifying firms in terms of the orga-

nization's primary and secondary lines of business. Key variables included the size
of the corporation, measured in terms of the number of employees. I had a partic-
ular interest in firms with multiple establishments, including separate divisions in
quite different lines of business, because I thought that such firms were likely to
move information across internal boundaries for the purpose of cross-
marketing.67 An establishment was defined as an economic unit, generally at a
single physical location. With increasing conglomeration, mergers, and the ex-
pansion of the franchise as a form of business organization, definitional uncer-
tainty about reporting level remains a serious analytical problem. Because an es-
tablishment may actually be identified as part of a division, an establishment is
not necessarily the same as an enterprise or company, which may consist of one or
more establishments.
Another measure linked to but not a substitute for any measure of size was an
index of technological sophistication. In a survey by Lance Hoffman and Alan
Westin, participants at a computer security conference in 1984 were asked to class-
ify their organizations in terms of their sophistication in computing.68 My survey
followed that approach, and respondents were asked to classify their organiza-
tion's general status as an information technology user. Related questions assessed
the company's use of computers and telecommunications to store and access cus-
tomer data.
A series of questions with ordinaUy scaled categories asked respondents to in-
dicate the frequency with which they used customer information for marketing,
used credit information to screen potential customers, or used customer informa-
tion acquired from other internal units or from external vendors. A question re-
lated to the measure of the importance of customer information to the marketing
function asked whether the organization had a special administrative unit or a
particular individual assigned the primary function of managing the customer
database. Another question inquired about the amount of attention that the orga-
nization paid to maintaining the security of their lists or files "to protect them
against unauthorized use:'
A checklist was used to allow respondents to indicate which of a variety of op-
tions they would consider to be a "fairly routine" technique that the organizatio11
used for marketing. The ten listed options included "use of a customer's name
within a mailing"; use of"life-style lists to classify potential customers"; and cus-
tomizing "catalogs to customer/prospect profiles:'
A series of questions sought to characterize these corporations by their use of
telecommunications, through outbound or inbound telemarketing. More impor-
tant, I sought information about their use of the automatic number identification
(ANI) service provided by long distance telephone carriers. Factual as well as hy-
pothetical questions asked about the extent to which they used or might value ser-
vices that could provide additional information about customers that might be
gathered on the basis of their telephone numbers.
llO CORPORATE PERSPECTNES ON THE PANOPTIC SORT l
Several items asked about the nature of the corporation's policies regarding the
release of customer information to third parties. These questions asked about
whether customers were provided with the option of restricting the sharing of
their names for use in mailing lists, and if there were such options, they were
asked to indicate how frequently the customers were informed about such options
and the manner in which customers might exercise them.
The final set of questions addressed the use of customer information as an is-
sue of public policy. Respondents were asked to indicate the extent to which they
believed that their business was being hampered by current restrictions on the
sale or exchange of customer information. After indicating the extent to which
they thought the use of customer information would emerge as a political issue in
the next five years, respondents were presented a checklist that they could use to ·
indicate what kinds of people would bear the greatest responsibility for the escala-
tion of this issue. The List of nine potential actors included "antibusiness activists;'
"disreputable firms," "telemarketers;' and "Congress."

Sample Selection
The source of businesses included in the sample was the Dun and Bradstreet
Dun's Market Identifiers database, accessed through the DIALOG Information
Retrieval Service. This particular database includes both private and public cor-
porations with five or more employees, or companies with sales in excess of $1
million. The corporate files provide current addresses and financial and market-
ing information for nearly 2.4 million business establishments. In addition, the
corporate record identifies all known businesses within the corporate family. For
the purposes of this survey, I restricted this sample to the universe of listed corpo-
rate headquarters of public and private firms with more than fifty employees.
Standard Industrial Classification (SIC) coding was selected as the initial crite-
rion used to define the relevant universe of firms. SIC codes are a uniform and
commonly utilized system for classifying business establishments by their princi-
pal economic activities. For the purposes of this survey, "primary" was based on
the value of corporate output associated with a particular activity, and outputs
were defined as the value of receipts, sales, or revenue. If the company was an ag-
gregate of several establishinents, primary activities were defined on the basis of
the relative value of aggregate outputs. Because of my interest in companies with
major business activities in the marketing of consumer goods and services, I elim-
inated several SIC codes from the research population. With all exclusions based
on size and SIC classification, a smaller population of some 97,172 organizations
was initially identified in the search of the Dun and Bradstreet database.
A systematic sampling interval was utilized to select a more manageable group
of 750 headquarters. Because of the ratio of small to large firms, the initial sample
was thought to contain too few representatives of the larger firms involved in con-
sumer sales. A subsidiary sample of 109 ·of these large firms was included in the
group to which surveys were mailed.
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 111

Of the 859 establishments included in the sample, responses of some sort were
received from 238. Seventeen were returned as undeliverable by the U.S. Postal
Service. Some 26 were either blank or the executives indicated that they had no
consumer sales and were not eligible. Only i81 questionnaires were received that
could be qualified as complete or partial responses. Although the covering letter
promised anonymity, there was evidence of considerable mistrust of the project
within the target community. Identification numbers on the return envelopes that
had been assigned to facilitate record keeping were obliterated by more than a
dozen respondents. The sample was further reduced by the elimination of any
questionnaires with a substantfal number of missing responses. The sample used
for the analysis that follows totaled 139 cases.

CORPORATE POLICY AND PRACTICE

Because of the relatively small sample and high rate of nonresponse, descriptive
statistics are relatively meaningless as anything other than a description of the
tabulated sample. That is, we can have no real confidence that they accurately rep-
resent the broad segment of U.S. business most directly involved in the collection
and use of personal information for marketing purposes. As with any survey of
the general public, we expect that those firms most sensitive about the privacy is-
sue were more reluctant to respond. However, information about the nature of the
sample provides some background against which to assess our analysis of the as-
sociations between characteristics, orientations, and the information practices of
those who did respond.
The twenty-six different SIC categories were combined into six comprehensive
classes: Retail Sales; Banking/Insurance/Real Estate; a Services group including
Transportation/Lodging/Recreation/Restaurants; Automobile Sales and Services/
Personal Services/Miscellaneous Repair; Telecommunication Services/Mass Me-
dia; and finally, Health/Legal/Social Services. The largest group was Retail Sales,
containing approximately 30 percent of the sample. The smallest groups were
Telecommunication Services/Mass Media, with less than 7 percent, and Health/
Legal/Social services, with approximately 10 percent.
Although there is a tendency to treat corporations as fictional persons and fur-
ther to consider that the organization has a personality, or a corporate style, it is
important to keep in mind the fact that corporate practices reflect decisions made
by individua.ls. To the extent that chief executives can impose their opinions on
others in the organizations tl1ey direct, one might argue that the executive speaks
for the company. That relationship is never perfect, and, in any event, is rarely ap-
parent to an outsider. Still, I felt it was important to question the executive at the
top about the firm and its relationship to the panoptic sort. Sixty percent of the
respondents indicated that they were the CEOs of their establishment. Another 23
percent of the surveys were completed by executives in general managerial posi-
112 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

TABLE4.1 Corporate Tendencies (in percentages)

Low High No
1 2 3 4 5 Response
Technological sophistication 5.0 20.1 40.3 26.6 7.9 0.0
Use customer information
to market 5.8 17.3 42.4 25.2 7.9 1.4
Use internal sources 27.3 19.4 25.9 14.4 2.2 10.9
Enhance with purchased
data about customers 45.3 28.8 12.2 7.9 4.3 1.4
Depend on vendors for
customer information 23.7 43.9 16.5 8.6 2.9 1.4
Get customer information
electronically 57.6 17.3 9.4 5.0 5.8 5.0
Use outbound telemarketing 57.6 15.8 17.3 3.6 0.7 5.8
Use inbound telemarketing 45.3 15.9 16.5 10.0 4.3 7.2
Would find added
demographics useful 41 .9 23.7 20.9 7.2 0.0 7.2
Add data to phone #s 54.7 21 .6 15.1 5.0 0.0 3.6
Estimate of customer's
concern about data use 8.6 8.6 19.7 23.7 19.4 20.9
Attention to list security 7.9 6.5 15.1 19.4 34.5 16.5
Limited by laws 39.6 23.7 7.9 0.7 3.6 24.5
See lists as an
emerging issue 4.3 14.4 39.5 26.6 9.4 5.9

tions. Less than 3 percent of the forms used in the analysis were completed by em-
ployees in public relations, although nearly 7 percent were from sales or market-
ing.
We note a nearly normal distribution of responses describing the corporation's
status as an information technology user (Table 4.1). The greatest number of re-
sponses identified the organization as being mainstream, with slightly more than
30 percent identifying themselves as advanced users or even trendsetters in some
areas. A similar proportion indicated that they kept customer data in a centralized
database, which could be accessed remotely through a corporate network. With
regard to the use of customer information for marketing, the tendency within this
group was toward more rather than less frequent use of data in this way. Approxi-
mately 40 percent of the respondents reported that they used credit or financial
information in selecting new customers. As many of the firms were small, single-
unit establishments, it might be expected that nearly a fourth never utilized cus-
tomer information from other units, divisions, or affiliates in their marketing ef-
forts. Less than 20 percent of those responding to this question indicated that they
used such information very much or extensively.
Neither outbound nor inbound telemarketing was a method relied on by many
of these respondents, and outbound was slightly less popular than inbound as a
marketing tool. Given the relatively limited use of the telephone for marketing
purposes, it was also not surprising that respondents tended to see little value in a
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 113

TABLE4.2 Marketing Techniques (percentage using routinely)

Practice Practice
Use customer's name in a mailing 56.9 Use customer profile for mailing 33.8
Classify customers demographically 34.5 Use zip code-based profile 40.3
Use surveys to gather customer Target marketing with
data 28.1 statistical models 19.4
Classify prospects by life-style Classify prospects by
indicators 9.4 psychological profiles 3.6
Prescreen customers with credit Customize catalogs to match
lists 20.1 customer profiles 5.0

service that might provide demographic information with incoming calls (Table
4.1).-
With regard to the attention they paid to the security of their customer lists, re-
spondents tended to claim that security was a very high priority. On a five-point
scale, five being the highest, the mean was 3.8, with 5 as the modal (or most com-
mon) response (34.5 percent). Respondents tended not to think that their busi-
ness practices were currently being limited by laws governing the sale oflists. The
mean on the same five-point scale was i.7, with i.o as the modal response (39.6
percent).
With regard to marketing techniques that respondents believed were used
"routinely" within their firms (Table 4.2), the use of the customer's name within a
mailing was the most popular, used by more than 56 percent of the respondents.
The most sophisticated techniques, those facilitated by computer processing of
customer transaction information, were used by only a few firms. Psychological
(3.6 percent) and life-style (9.4 percent) profiles were used routinely by less than
10 percent of the respondents. Although it was not the most popular technique,
zip code-based profiles were used routinely by more than 40 percent of the re-
spondents' organizations.
The responses to a question about customer lists as an emerging public issue
were almost normally distributed, with a mean of 3.2, although only six respon-
dents (4.3 percent) indicated that it would not emerge as an issue at all.
A series of questions asked respondents to indicate which individuals or
groups would be most responsible for the escalation of the use of customer rec-
ords into a public issue. The questions were coded so that responsibility could be
assigned values, ranging from none to major, in three steps, with three indicating
major (Table 4.3). On the basis of the mean ratings, list vendors are seen as the
most likely culprits, followed by telemarketers and disreputable firms.
Antibusiness activists were assigned major responsibility by approximately 28
percent of the respondents, whereas list vendors were so labeled by nearly 63 per-
cent. Disreputable firms were assigned major responsibility by more than 56 per-
cent of the respondents, but aggressive competitors were generally seen as blame-
less.
114 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

TABLE 4.3 Actors Seen as Responsible for the Escalation of the Customer Information Issue
(percentage identified as major, N= 139)
List vendors 62.6
Telemarketers 57.6
Disreputable firms 56.1
Complaining customers 54.0
Federal regulators 48.9
Congress 4·6.8
State legislators 43.9
Aggressive competitors 23.0
Antibusiness activists 27.3

Underlying Patterns in Corporate Responses

In the next section, l examine the corporate orientation to the panoptic sort
through the use of a simple method for differentiating between types of firms.
This differentiation is made not only on the basis of. their characteristics such as
size and technological advancement, but in terms of corporate orientations to-
ward customers' claims of privacy interests in the company records that identify
them. V\'hat is sought here is evidence of some patterns or relationships that
might suggest that there is an underlying linkage be.nveen the nature of the firm
and the postures of management toward its customers. Three primary variables
are explored: the size of the establishment; its sophistication with regard to infor-
mation technology; and its primary line of business. A nonparametric correla-
tional approach is used as a first step in estimating the relations between these
variables. .
Table 4.4 presents the correlations between eight predictors and thirteen types
of corporate practices. To facilitate int.erpretation of the coefficients, a superscript
a is used to indicate a relatively conservative criterion of significance. Coefficients
so marked may be interpreted as indicative of a genuine, even if weak, association
between these variables. ·
The size of the corporation is seen to be positively linked to five of the thirteen
comparisons. The strongest association (r = .52) is with a measure of the extent to
which the firm gathers information about customers from vendors electronically
(Electron). Recall that more than half of the respondents indicated that they never
gathered information in that way. Thus, the association of this behavior with the
size of the corporation seems reasonable on the face of it The somewhat lower
correlation (r = .39) with a variable indicating that the firm enhances customer
files with information purchased externally (Exac;quir) underscores the distinc-
tion benveen the technologies perceived to be available and those deemed appro-
priate to different sorts of firms. Somewhat lower still, but in the same direction (r
= .29 ), is the tendency of the larger firms to depend on vendors for customer data
(Depend)". Although the number of respondents actively involved with inbound
telemarketing (Inbound) is small, there is still a positive, or direct, relationship
between company size and such an approach to sales.
 115

TABLE 4.4 Correlates of Business Practices

Correlations
with Predictors
Practice 2 3 4 5 6 7 8
Usemktg:
Extent of use of customer
information for marketing .14 .33a -.07 .15 .02 -.01 .02 - .13
purposes
.lnacquir:
Extent of use of customer·
information acquired internally .20 .41a - .12 .02 .04 .03 -.02 -.01
Exacquir:
Extent of use of customer
information acquir~d externally .10 -.02 .09 - 11
Depend:
Extent of dependence on
commercial vendOr$ for CUS·
to mer
information .29a .27'1 -.12 .10 - .10 .02 .05 .05
Electron:
Extent of electronic means
to acquire customer information .52a .34a - .18 -.10 .05 .06 .243 .00
Outbound:
Extent of use of
outbound telemarketing .31a .31 8 - .16 - .09 .16 - .01 .36a - .15
Inbound:
Extent of use of
inbound telemarketing .25a .18 .04 -.16 .16 - .09 .23a -.12
Callinfo:
Extent of use of
information about callers .20 .17 - .22 - .07 .04 .17 .16 .01
Telnos:
Uses information based on
telephone numbers of callers .05 .03 - .03 - .08 - .05 .11 .13 -.10
Concern:
Extent of customer concern
about sale of name .14 .08 -.01 . .16 -.19 -.21 -.01 .333
Security:
Extent of attention paid to
customer name security .1 4 .13 - .20 .263 - .263 -.18 .04 .33a
Lawlimit:
Extent of legal limitqtions
on saie of customer names .11 .10 -.12 .09 -.11 .00 .03 .16
Issues:
Belief that customer information
will escalate as an issue -.09 .07 .07 .04 -.09 -.15 -.08 .20
116 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

TABLE 4.4 (cont.)

Predictors
1. Size (number of employees)
2. Technological sophistication
3. Retail Sales SICs
4. Banking/Insurance/Real Estate SICS
5. Services/Transportation/Lodging/Recreation/Restaurant SICs
6. Automobile Sales and Service/Personal Services/Miscellaneous Repair SICs
7. Telecommunications Services/Mass Media SICs
8. Health/Legal/Social Services SICs
8
p = .05, or less, two=tailed

Technological sophistication with regard to information technology emerged

as the most important correlate of the business practices explored. Although this
was statistically significant in six of the comparisons, there are some departures
from the associations with size worthy of comment. Technological sophistication
appears to be more strongly associated (r = .33) with the use of customer informa-
tion for marketing (Usemktg) than may be explained by corporate size. Although
one would think that the more technically sophisticated firms would be more
likely to have the capacity to gather and process information received
electronically from external sources (Electron), the association is not as strong as
it is with regard to corporate size. Instead, there is a more consistent association
with the tendency to utilize internal (Inacquir, r = .41) and externally purchased
(Exacquir, r = .42) data.
"When we examine the associations in Table 4.5, it appears that company size is
the primary characteristic that explains the use of particular analytical techniques
or different classes of data or information. The use of surveys (Survey), targeting
with statistical models (Targstat), life-style (Life-style), and psychological profiles
(Psych) is positively associated with the size of the firm as measured by number of
employees. The only significant associatioo with technological sophisticati.on is
with the use of surveys, which suggests that computerization is an aid and perhaps
a necessity for interpreting survey data.
Overall, the identification of the firm's SIC group failed to provide any reliable
indications of the likely orientations or practices of :firms within that group, in-
cluding the use of marketing techniques. This may be due, in part, to the substan-
tial diversity within the groups as constructed for this analysis. Only a larger sam-
ple, utilizing the two-digit SIC code, will allow us to answer this question more
conclusively. Membership in the group representing firms involved in telecom-
munications services and mass media is significantly associated with three ten-
dencies, all logical correlates (Table 4.4). These firms tend to gather customer in-
formation electronically (r = .24) and to engage in outbound (r = .36) as well as
inbound (r = .23) telemarketing.
 117

TABLE 4.5 The Association Between Organizational Characteristics and Marketing

Techniques (N = 139)
Correlations
with Predictors
Practice 2 3 4 5 6 7 8
Custname:
Uses customer's name
in a mailing .02 .18 - .09 .15 - .10 .07 .11 - .09
Custprof:
Uses customer profile
for mailing .16 .07 - zf' .13 .04 .02 .12 .01
Demograp:
Classifies customers
demographically .11 .07 -.02 .09 - .05 - .18 - .01 .21
Zipprof:
Uses zip code- based profile
for customers .10 .18 - .16 .10 - .01 - .00 .02 .12
Survey:
Uses surveys to gather
customer data .35a .24a -.10 - .11 .05 .08 .16 .00
Targstat:
Pursues targeted marketing
with statistical models .37a .08 -.09 .07 .08 -.23 .09 .14
Life-style:
Classifies prospects by
life-style indicators .40a .11 - .05 -.10 .10 - .15 .12 .06
Psych:
Classifies prospects by
psychological proflles .34a .11 .04 -.10 .04 - .09 .11 .06
Screen:
Prescreens customers with
credit lists .05 .21 - .02 .11 - .14 - .00 .09 - .11
Catalog:
Customizes catalogs to
match customer profiles -.01 .20 .14 - .03 - .09 -.02 .07 - .07

Predictors
1. Size (number of employees)
2. Technological sophistication
3. Retail Sales SICS
4. Banking/Insurance/Real Estate SI Cs
5. Servicesrrransportation/Lodging/Recreation/Restaurant SICs
6. Automobile Sales and Service/Personal Services/Miscellaneous Repair SICS
7. Telecommunications Services/Mass Media SICs
8. Health/Legal/Social Services SI Cs

8
p = .01, or less, two-tailed
118

TABLE 4.6 Rotated Factor Matrix (varimax) Business Information Orientations

(N = 121 cases, loadings > .35)
Loadings for Factors
Variables 2 3 4 5 6 7 8
Fedreg:
Federal regulators would be
responsible if customer records
become an issue .93
Stateleg:
State legislators would be .91
responsible
Congress
Congress would be responsible .79
Activist:
Activists would be responsible .56
Compet
Aggressive competitors would
be responsible .55 .55 .36
Telemkt:
Telemarketers would be .81
responsible
Disrep:
Disreputable firms would be
responsible .76
Listvend:
List vendors would be responsible .72
Catalog:
Customizes catalogs via
customer profiles .75
Usemktg:
Use of customer information
for marketing .71
Exaquir:
Enhancement of files with
externally purchased information .63 .39
Zipprof:
Use of zip code-based
customer profiles .81
Demograp:
Classification of
customers demographically .80
Psych:
Classification of prospects
by psychological profile .80
Life-style
Classification of prospects
by life-style .72
Targstat:
Use of statistical models
for target marketing .36 .55
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 119

TABLE4.6 (Cont.)
Loadings for Factors
Variables 2 3 4 5 6 7 8
Custname:
Use of customer name
in mailings .79
Survey:
Use of surveys to collect
customer data .60
Custprof:
Use of customer profile
for custom mailing .40 .57
Telnos:
Add customer data to
telephone numbers .36 .71
Issues:
Belief that customer information
will grow as an issue - .68
Screen:
Uses credit-based
prescreening of customers
Customer.
Complaining customers would be
responsible if customer records
become an issue .75

With regard to the extent that the organjzations pay attention to the security of
their customer records (Security), the strongest link is with membershlp in the
=
Health/LegalJSocial Services group (r .33), followed by the Banking/Insurance/
=
Real Estate group (r .26), and there is a striking reversal (r =-.26) with regard
to the Services group, for whom customer records are of only limited concern.
None of these measures of corporation type are reliable predictors of the re-
spondent's identification of the individuals or groups that their leaders believe to
bear the greatest responsibility for any escaJation in concerns about the panoptic
sort.

An Alternative View
In an attempt to reduce the number of variables, while exploring the underlying
structure of information practices that characterizes these firms, a factor analysis
was performed with the twenty-three variables with the highest response rate. Ta-
ble 4.6 presents the eight factors that account for 65 percent of the measured vari-
ance. Only variables with factor loadings greater than .35 are included in the table.
The first two factors, explaining i6.6 and i3.o percent of the variance, respectively,
reflect the loarung of eight of the nine actors identified as being responsible for the
120 CORPORATE PERSPECTNES ON THE PANOPTIC SORT

TABLE 4. 7 Multivariate Analysis of Variance

Factor Scores with Three Independent Variables (significance of F, N =121 )
Level of Significance for
Interactions and Univariate Sources
Dependent Variables 2 3 4 5 6 7
Multivariate (8 Factor Scores) .65 .28 .15 .02a .32 .02a .16
Factor 1 .86 .60 .74 .52 .71 .16 .47
Factor2 .49 .70 .95 .81 .90 .95 .92
Factor 3 .07 .16 .87 .68 .48 .32 .73
Factor4 .79 .18 .38 .09 .68 .23 .71
Factor5 .87 .91 .03a .o1a .07 .31 11 .
8
Factor6 .50 .59 .07 .22 .50 .04 .74
Factor 7 .98 .46 .20 .76 .18 .44 .29
Factor 8 .29 .048 .058 .01 8 .03
8 ,01a .o1a

8
Pillais test of significance ( p< .05)
Employ = Number of employees
Usetech = Technological sophistication
lndgroup = Industrial group membership
Interactions and univariate sources:
1. Interaction of Employ, Usetech, and lndgroup
2. Interaction of Usetech and lndgroup
3. Interaction of Employ and lndgroup
4. Interaction of Employ and Usetech
5. lndgroup
6. Usetech
7. Employ

escalation of the privacy issue. The variable representing aggressive competitors

loads substantially on three factors and is therefore not particularly useful as an
aid to defining the factors. With the elimination of this variable, factor i appears
to be a political factor, whereas factor 2 is a business or corporate factor.
Factor 3, its clarity marred by the loading ofExaquir on factor 8 as well as factor
3, reflects a common tendency for those that use customer information, including
that acquired externally for marketing, also to use that information io specially
tailored catalogs. Factor 4 has high face validity, linking the use of zip code profiles
with the use of demographic classification. Factor 5, which shares the tendency to-
ward using statistical models with factor 4, ·includes a further tendency toward the
use of life-style and psychological measures.
Factor 7 is somewhat difficult to interpret. It suggests a tendency to blame ag-
gressive competitors for an issue respondents do not see as likely to escalate in the
near future. Both of these tendencies are associated with fairly extensive use of
data to enhance corporate files based on captured telephone numbers. Perhaps
this factor reflects a bit of wishful thinking, rather than being an accurate reflec-
tion of current corporate practices. The final factor (8) reflects a tendency to use
information acquired externally to enhance customer files, and to fear that if these
CORPORATE PERSPECTIVES ON THE PANOPTIC SORT 121

practices escalate into a major issue in the future, it will be because of the actions
of these disgruntled customers.
Table 4.7 presents the results of a multivariate analysis of variance approach,
which takes into account the linkages between tl1e underlying tendencies that
these factors represent. Factor scores for 121 cases were used to estimate eight dif-
ferent dependent variables (factor 1 to factor 8). The table presents an evaluation
of the importance of the explanatory variables, singly or in interaction with other
predictors, as the source of different patterns in the distribution of factor scores.
The multivariate scores reflect the contributions of the predictor variables to the
variance of the eight dependent variables taken as a group, whereas the eight
univariate scores reflect the contribution of the predictors to the variance in each
of the factors independently.
With regard to the explanation of the variance in the set of factors, only two
measures appear to have any demonstrated explanatory power: technological so-
phistication (Usetech) and company size (Employ). The interaction between size
and sophistication is a significant factor (p = .02). At the same time, company size
is not a significant factor by itself (p =.16), although technological sophistication
is (p = .02). Whatever influence corporate size bas with regard to the multivariate
question, the univariate analysis suggests that it is limited almost entirely to factor
8. The distinction between these two influences may be pursued further by exam-
ining the univariate analyses associated with factor 6. Where technical sophistica-
tion is a source of influence witli regard to V;ariance in the use of customer names,
profiles, and survey methodology (p = .04), company size seems to matter not at
all (p = .74).
With regard to factor 5, reflecting the terndency to utilize sophisticated model-
ing, including the relatively less popular psychological and life-style profiles, the
interaction of corporate size and technological sophistication is the most impor-
tant source of measured variance (p = .01). There is a significant interaction effect
=
(p .03) of company size and industry group (Indgroup) membership, by which
tlie larger firms in particular groups are more likely to use computer profiles.
Factor 8, which reflects the peculiar tendency of some firms to rely on external
sources for customer infonnatio11 but to fear that disgruntled customers will be a
major source of the escalation of public concern about marketing practices, is
seen to be affected by every measure except the most complex interaction of all
three variables. Examination of the univariate alphas would suggest that industry
group identification is the least useful in regard to this measure. Size, sophistica-
tion, anci tlieir interaction are all significant sources of influence. The correlation
data suggest that, especially for those in the Health/Legal/Social Services sector,
the relationship between bigness and anxiety about customer complaints is posi-
tive. Unfortunately, at least for the purposes of interpretation, firms in this sector
are less likely to be dependent on externally acquired customer information. One
interpretation that solves the apparent paradox in the data is tlie suggestion that
firms in this sector expect the problems to emerge not because they acquire cus-
122 CORPORATE PERSPECTIVES ON THE PANOPTIC SORT

tomer information, but because someone might release this information without
proper authorization.
Some support for this explanation can be found in the tendency of these firms
to have customers who are very much concerned about personal information (r =
.33) and for the firms themselves to report paying considerable attention to these-
curity of their customer data (r = .33) while also maintaining a substantial belief
that customer information will emer ge as a major policy issue (r = .20). Support
for this interpretation is also found in the Equifax (1990) survey of business lead-
ers in «privacy intensive industries." Respondents from the human resources and
insurance sec,tors were most likely to believe that their customers were genuinely
concerned about the practices associated with the compilation and use of mailing
lists.69

BUSINESS PERSPECTIVES ON
THE PANOPTIC SORT
These data suggest that the use of panoptic technology is just beginning to make
its way into the pool of organizational resources. The association with corporate
size and technological sophistication is not surprising, but it underscores the in-
fluence ofdominant firms in the establishment of the corporate culture. Linowes's
claim that "major corporations are standard setters of business practices"70 de-
serves to be taken seriously, although my analysis suggests that there are impor-
tant differences between organizations, many of which are reflections of the ori-
entations of their chief executives as individuals, rather than strictly a function of
structural demands. 7 1
These data also suggest considerable awareness of and concern about the po-
tential consequences of a consumer backlash. Business executives seem to share a
common tendency to see problems as being caused by otl1ers outside their own
organization. Because of this, less attention is paid to their own corporate prac-
tices. Similarly, as we will see in me next two chapters, individuals have an aware-
ness of the existence and operation of the panoptic sort but they tend to see it as
affecting the lives of others, rather than themselves.
5
RELATIONSHIPS
AND EXPECTATIONS

IN SEARCH OF PUBLIC PERCEPTIONS

The development and implementation of the panoptic system are conditioned by
the level of awareness and acceptance of practices and consequences as normal, le-
gitimate, and consistent with the broadly held pub}jc understanding of the rights
and responsibilities of institutional actors with regard to the collection and use of
personal information. The dynamic interaction among technology, market, and
nonmarket relations governing the exchange of valu e and the laws and social cus-
toms that condition their development defies any simple categorization. The ma-
terialist view generally bolds that the laws and customs of a society-the super-
structural realm of culture-are determined in the final analysis by the changes in
the realm of production and exchange. Although it is difficult to establish the pri-
macy of the technical over the exchange relations, because technological innova-
tion is generally understood to have been a response to stagnation or con.stramt in
the operation of the market or government sector, we tend to accept the view that
sees changes in attitudes and opinions about information practices as having been
occasioned by changes in technology and its application.
What people feel about a particular social practice depends on wbat they know
about it and how they understand its use, its justification, and its consequences
for them or for others. The relat ionships among the cognitive, affective, and be-
havioral spheres of social opinion are more closely linked in theoretical models
than they are in empirical social research. 1 Not all of the difficulties have to do
with problems of measurement. Almost insurmountable difficulties lie in the
contextual or situational variety that qualifies opinion and constrains action.
Thus, confidence in what people say they believe and intend to do has to be tem-
pered by a recognition that circumstances always intervene.
Anthony Giddens's exploration of practical and d iscursive consciousness un-
derscores the methodological difficulties that speak for caution as we approach
the assessment of social opinion. 2 Giddens distinguishes three levels of conscious-
ness, which bears an unfortunate association with the Freudian levels of id, ego,
and superego. The unconscious is presumed to be simply unavailable to research-
ers and analysts because it is also unavailable to Giddens's knowledgeable actor.

123
124 RELATIONSHIPS AND EXPECTATIONS

The connections and distinctions between the levels of practical and discursive
consciousness matter most to our ability to access and understand social opinion.
Practical consciousness refers to the knowledge, insights, logical structures, sche-
mas, rules, and cognitive resources that individuals have access to while they are
engaged in their social routines. Yet, not all of this knowledge is available to an in-
dividual at the level of discursive consciousness, where they might report or dis-
cuss their reasons or rationales for taking a particular action. Tacit knowledge is a
resource at the level of practical consciousness. Giddens suggests that there may
be something ofa "negative bar," or restriction on an individual's ability to moni-
tor actions reflexively or to represent the complex logics that underlie seemingly
contradictory preferences and values. People simply cannot tell you all they know,
especially with regard to their reasons for the actions they take.
These difficulties raise, or at least should raise, a number of red flags when
social scientists claim to be able to represent individuals' mental structures or
schemas, their value systems, or their preferences for goods, services, or social
outcomes and policies. Amos Tversky and Daniel Kahneman have used creative
experiments to provide devastating critiques of rationalist assumptions about
consumer behavior. 3 Their examples have great force because of the gaps they
reveal between what people say they value and what they demonstrate they ac-
tually value when they are forced to make a choice. This difficulty is
complicated further when the forced choices required by a standardized ques-
tionnaire or survey instrument bear only a limited but probably unmeasurable
relationship to the complex attitudes, opinions, or preferences that indiv iduals
may actually hold.
Social researchers are noL unaware of the problems of reactivity that compli-
cate their efforts to measure knowledge and social opinion. 4 Questions asked in
the early part of an interview help to prime or establish a frame or orientation that
is likely to constrain or limit the range of responses to questions asked later in the
interview. Understanding that a survey has something to do with "privacy" is
bound to color a person's responses to questions about personal attributes, such
as income and educational attainment, in much the same way as such a recogniz-
able frame might influence responses to other questions about views on genetic
screening, drug tests, wiretapping, or surreptitious monitoring. Efforts to modify
the influence of question order through randomization can make only a modest
contribution to solving these and related problems of reactive bias.
An additional problem is found in the nature of sampling and in the untenable
assumptions that have to be made about samples as representatives of attitudes
and opinions that matter. Clearly, it is impossible to ask everyone about every-
thing. Yet each step we take away from this conceptual ideal widens the hole
through which bias, distortion, and misrepresentation creep into models of social
opinion. The truly random sample has never been taken because a complete sam-
pling frame does not and cannot exist. Approximations of randomness depend on
RELATIONSHIPS AND EXPECTATIONS 125

-
the quality of the sampling frames that are available. Even sophisticated
approaches such as Random Digit Dialing, which is thought to overcome the dif-
ficulties represented by nonpublished numbers, are unable to overcome the fact
that telephone service itself is not randomly distributed to all households.5 Uni-
versal telephone service is a fiction, and there are significant differences associated
with gender, age, race, and marital and employment status reflected ·in the distri-
bution of households without telephones.6
As noted in the discussion of the expanding market in personal information,
the increasing pressure on individuals to provide information about tastes, pref-
erences, and opinions has resulted in a growing reluctance to participate in sur-
veys or to complete questionnaires. It is also likely that persons who are particu-
larly concerned about privacy are even less likely to agree to participate in such
surveys. Tb us, any survey that is concerned about opinions related to privacy and
personal in formation will be nonrandom and systematically biased.
Even if a truly random, and the reby representative, sample of individuals could
be constructed, it would still be necessary to sample opinions and the experiences
on which they are constructed. And, because this sampling is ultimately done by
the participants in the survey, its scope thereby depends on their memories, their
interests, or their motivations, and a variety of other factors beyond the research-
er's control. The fact that these distortions are themselves nonrandom, but may be
linked systematically to the question at hand, opens the hole still wider.
All this suggests the need for great caution in interpreting the results of surveys
and interviews as reflections of what people understand, believe, feel, or intend to
do with regard to the panoptic sort. One response to the problem of measurement
has been the attempt to utilize multiple measures and a variety of means through
which to assess social opinion. Focus group interviews are frequently cited as a
means th.rough which researchers can begin to appreciate the great range of re-
sponses and constructions of the subjects of concern. These groups help the re-
searchers to recognize the variety of ways in which individuals have come to label
or talk about events, relationships, and emotional reactions. Of course, focus
groups are not without their own problems. Because groups are rarely naturally
occurring, much of the time and energy of group participants may be spent in
coming together as a group, and the chemistry of a particular group may simply
not be conducive to exploring a particular issue.
Before a national telephone survey was developed and administered in January
and February 1989, a set of five group interviews was conducted in summer i988. 7
Each of the two-hour sessions was tightly formatted so that the experiences of the
five groups would be directly comparable in terms of the questions that were
asked and the procedures that the participants were to follow in moving toward
what they saw as the best expression of the group's thinking abou t privacy and the
panopticsort.8
126 RELATIONSHIPS AND EXPECTATIONS

THE GROUP INTERVIEWS

After explaining the purpose of the group, the interviewers elicited personal intro-
ductions by having participants talk about their favorite food, favorite color, fa-
vorite subject in school, favorite teacher, and the like. Next, the interviewers intro-
duced the subject of privacy by asking participants to indicate bow much
attention they had paid to the issue of privacy before they had been contacted by
the members of the project. The subjects were next asked to name an invasion of
privacy and to indicate why they thought it was an invasion.
To focus the group's attention more directly on the nature of surveillance in ev-
eryday life, the interviewers showed a five-minute videotape that was produced
especially for the project. The tape, conceived as a "day in the life" of an ordinary
person, provided examples of the ways in which transaction-generated informa-
tion might be combined to create a detailed image or profile of an individual. Ex-
amples included use of a credit card to purchase a train ticket, employment appli-
cation forms, computer monitoring on the job, monitoring cameras in the
supermarket,' automatic teller machines, and even a videotape kiosk that made a
record of a rental transaction. The connection between transaction information
and targeted communication was made through references to direct mail pre-
sumed to have come from lists gen erated by magazine subscriptions. On the vid-
eotape shown to the group, a closing image appeared on the television screen that
was being viewed by the principal actress as she relaxed at the end of her busy day.
The picture was of herself as she had appeared at the beginning of her day, just as
she had entered the train station.
The questions that the project interviewers pursued in the remaining time were
directed toward the gathering of assessments of three key components of an orien-
tation toward informational privacy: (l) what participants knew about the tech-
nologies that are used to gather information to create profiles of consumers; (2)
who they thought was likely to use such techniques and what they saw as the
boundaries oflegitimate use of the·se techniques or the information that they pro-
duced; and (3) what they thought about the sharing of information produced
through these surveillance methods.
As a final opportunity for the groups to organize their thinking about these is-
sues in the context of this highly focused discussion, they were asked to act as
though a representative of the news media was waiting outside the room to record
their thoughts. Each person was asked to compose a statement for the television
cameras on "privacy as the issue of the 1990s." All sessions were tape recorded and tran-
scribed, and those transcriptions provided the basis for the analysis that follows.

THINKING ABOUT PRIVACY

Our uncertainty about the nature of public concerns about privacy is due in part
to the limitations inherent in the questions asked in public opinion surveys. Sur-
RELATIONSHlPS AND EXPECTATIONS 127

vey evidence suggests that there are no universal definitions of privacy, and indi-
viduals are likely to be responding to quite different things when they indicate the
presence or absence of concerns. I thought that it was important to understand
the salience of privacy as an issue by understanding both the level of concern and
the basis for that concern. Participants in each group were asked to indicate the
extent to which they had given much thought to the privacy issue before being
contacted.
The general response was that privacy per se was not something most partici-
pants bad given any thought to. In every group, one or more.individuals suggested
that the questionnaire had really forced them to think about privacy in ways they
never had before. Some indicated that what they understood to be the domain of
privacy concerns was quite different from the broad range of issues addressed in
the questionnaire.
For some, an external event, highlighted in the media and usually concerned
with a politician, had caused them to think about privacy in the recent past. For
others, there were personal experiences, usually related to work and often trig-
gered by questions they encountered on an application form. And for others, con-
cern about drug testing was the stimulus for their beginning to think about pri-
vacy. One man worked i.n a shipyard that had recently instituted a drug-testing
policy. As he understood the new policy, any accidental injury would make him
subject to a drug test: "Like you just cut your finger or bang your finger, or fall, : ..
you go to the dispensary for an aspirin or bandaid, [and] you have to take a urine
test. They got 60-year-old men down there taking a urine test for drugs."
One participant decried the lack of privacy on the job because of the constant
press of people in a cramped office. This example was countered by someone with
the opposite concern-the isolation felt by a newcomer to a c_:ommunity where
people rushed home in the evenings and, in the pursuit of their privacy, closed
their automatic garage doors, not to be seen again until the next day. .
Another person reported having been sensitized to the issue by discussions
with friends who also worked with computer databases. She shared their concern
about the amount of"garbage and misinformation there was" in those files. This
concern with privacy, linked to the inaccuracy of data in files, was heightened for
one participant whose own credit records had been confused with those of his
parents because they shared the same last names and had similar addresses.
One group was composed of persons who had recently been called to serve on a
jury. Perhaps because the members of this group were all registered voters (that was
the basis on which they were selected for jury duty), they expressed considerable
concern about the role of government in invasions of privacy. The examples they
provided began with an expressed concern about banks sending the government
information about their personal savings and the fact that they were linking that
information with social security numbers. Another example described experiences
with FBI surveillance in the 1960s. One man's wife had sought access to her file
through the Freedom oflnformation Act, and when the report was received, he was
128 RELATIONSl-IJPS AND EXPECTATIONS

alarmed by tl1e nature of information and the variety of methods that had been
used by the FBI to gather information for his wife's file. Another member of this
group expressed concern about the potential impact of information jn his own FBI
files, which indkated that he had led a demonstration some eighteen years ago. He
had been told that this information could not be erased from his file, and he was
thus worried that he might be barred from future employment or rrught otherwise
suffer because he would be perceived as a threat to national security.

DEFINING PlUVACY INVASIONS

A fol low-up question asked participants to provide explicit justifications for call-
ing something an invasion of privacy. Very few of tbe responses had actually iden-
tified the reasons underlying their selection of particular examples as invasions of
privacy. Television programs and recent news stories were a ready source of exam-
ples that defined a privacy invasion as an act that stepped outside of some zone of
legitimate action that participants had drawn in their personal conceptual
scheme. This zone was frequently drawn on the basis of an assessment of the pre-
sumed relevance of the information being collected. For one participant, the lim-
its were clearly drawn: "I know when I was going to school, ... they would always
emphasize family orientations, such as marriage, legitimacy, things of that nature.
They always wanted to know whether your parents was married or what ... just
didn't have anything to do with being helpful to go to school." A similar response
reflected a concern with the structural conditions of disclosure that might involve
embarrassment. There was an apparent distinction between being asked an irrele-
vant question on a form and being asked the same question in a personal inter-
view: "I was once asked face to face in an interview if I was married, and I felt the
implication was that it was a question of whether or nol l was stable. I felt that it
didn't have much to do with my qualifications for the job."
One person thought it "outrageous" that somethin.g someone did while they
were in college should be relevant to their consideration for high office. This
oblique reference to the charges of marijuana use by nomjnees to the Supreme
Court reflected a widely held belief that public figures deserved more privacy than
the media seemed willing to allow. One participant's comDlents suggested that the
victim was not merely the public figure, but the members of the audience who
were continually being subjected to these stories: "It seems like once they get a
hold of something, you know, that happens in their life ... you know they drag it
all out, and I don't want to know every time they go to the doctor, I don't want to
know every time they go to the bathroom. You know, some part of their life is pri-
vate." Another continued the critique of the press: "They're taking things that are
not crucial to the issues or yom judgement of the individual. They're blowing it all
over the place as if implying it should be:'
Participants frequently mentioned negative consequences as the criterion on
which a determination of an inquiry as an invasion of privacy could be made.
RELATIONSHIPS AND EXPECTATIONS 129

Some participants highlighted the loss of control that is involved in invasions of

privacy. The consequences that might flow from a loss of control over personal in-
formation were frequently linked to the use of derogatory information in personal
credit histories. Individuals were seen frequently to have to suffer the negative
consequences that flowed from the use of false, or at least contestable, information
in their credjt files- information that they did not even know existed but that,
nevertheless, was being treated as factual by lenders.
The practice of sharing personal information for marketing purposes was also
seen as au invasion of privacy. One participant noted that the act of getting mar-
ried generated telephone calls from insurance agencies; visits to an obstetrician
generated calls from diaper services. There was general agreement among the
groups that the widespread practice of sharing subscriber lists was an invasion of
privacy, but it was the consequence of receiving unwanted solicitations, rather
than any inherent unfairness or impropriety, that seemed to be the primary basis
for concern among the participants.

PROFILES
The interviewers determined the extent to which participants were familiar with
the concept of profiling. After tl1e participants had focused their tllinking by
means of the "day in the life" videotape, the interviewers asked the members of
each group to help identify some terms that might describe the "pictures of our-
selves" that are formed ilirough the records or traces we leave as we go about our
daily lives. The notion of a profile was familiar to most members of the groups.
More sophisticated observers included other technical language, describing the
method as one of modeling, a relational database, or a computerized biography.
One creative member offered a suggestion tl1at described tlle profile as an autopsy
"'cause it could be death by the examiner depending on whether you filled an in-
terest." One observer's critical response labeled profiles as "preconceptions based
on circumstantial evidence:'
A matter of interest was to determine whether people could identify the differ-
ent circumstances in which these profiles could be used in ways that would help
make a person's life easier or, alternatively, could harm a person or make his or her;
life more difficult. Some participants readily assumed the perspective of business .
or industry users of personal data. Information about customers or potential cus-
tomers was seen as a potential source of savings or a means to avoid wasted effort:
"I think it probably saves money. Like those stores can calibrate their orders of
milk or cheese or other perishables depending on how much exactly tl1ey sold
from those bar codes and there's less waste ... then the prices, in theory, could be
lowered somewhat."
The business or organizational purpose in acquiring lists of potential clients or
donors was readily identified as a legitimate use of personal transaction data. As
one participant suggested:
130 RELATIONSHIPS AND EXPECTATIONS

I was about to say that the obverse of your, what you object to [is] ... buying a book
and getting notice of other books. That probably does make it easier for cultural,
maybe valuable cultural enterprises and w1dertakings to get started.... If you're
starting a book club or a dance group or a theater subscription series, you want to
buy a list of people who go to the theater, and those are likely to be of value to the
commun.ity and possibly even to you, so it seems to me that there are some legiti-
mate areas where this is useful.

Participants agreed that consumer profiles might actually serve to reduce the
nwnber of telephone solicitations or the amount of junk mail they receive if the
senders actually knew more about them- for example, that they did not like to re-
ceive telephone solicitations: "l was thinking if people knew what type, . .. that
you don't like those kind of calls ... the phone company could maybe filter out
those kinds of calls, if th~y knew. You see, if they could get to that extent, the
phone company could . .. [be] monitoring your phone calls." Another suggested
that "maybe the junk mail would become only junk mail of things that you would
be interested in, rather than a thousand things you're not interested in."
People tended not to include the government in their set of compilers who
might gather data in a way that would make life better for the individual. One par-
ticipant .did include the census as "patterning the country to see where we're
going, so they can plan for the future." ln general, however, the government, espe-
cially the police and the IRS, were more readily identified when it came to the par-
ticipants' suggestions of compilers of data who may cause some negative conse-
quences of profiling.
When asked to think about the problems that might be associated with pro-
files, the overwhelming tendency was to focus on the consequences that would
flow from having false, inaccurate, or outd~ted information in the profile. Several
participants suggested that this problem was made even worse with the use of the
computer because data takes on a kind of permanence, and data managers are
slow, if not actually resistant, to make corrections. Computers are seen to mal-
function, automatic teller machines to run out of money, and "computer error"
becomes a ready excuse for poor service or carelessness: "It's ineptitude wh.ich
we've always had with us and we always will, but the computer makes it so much
more difficult because people go ... well, the computer says so. You can't argue
with the computer. And, the idea that the computer could be wrong is completely
alien to these peoples' culture:'
Several participants identified the negative consequences that would flow from
the development and use of a profile or model to discriminate against individuals
because they shared characteristics with others who might be credit risks or who
might have a lower estimated potential as customers. One participant identified
the practice of redlining as an abuse of profiles. "This has a self-fulfilling prophecy
because it is impossible to get home repair loans and then all the houses go .. .
turn into slums and then obviously people live there who are not middle class be-
cause that's all they can afford."
RELATIONSHIPS AND EXPECTATIONS

To pursue participants' understanding of the problems associated with faulty

profiles or decision models, the moderators relayed the story of a young woman
who had applied for a credit card, responding just like the thousands of others
who had been solicited as the dates of their college graduations approached. 9 This
English major was notified that her application had been rejected, and after sub-
mitting a request for the reason, she received a note indicating that her "area of
study (literature)" disqualified her. As anticipated, the general response in the
groups was one of incredulity. Many consumers assume that credit decisions are
based entirely on an assessment of one's honesty, re}jability, or record of prompt
payments in the past. Few considered that the selection criteria might also include
some estimate of a person's economic potential-the kind of life-style, including
the llSe of other banking services, that an engineering graduate might eventually
purchase, but that an English major was seen to be unlikely to choose.
"I don't agree with it Just because he's an English teacher doesn't mean he's ...
you could make $s an hour and still pay your bills on time, or you could make
$70,000 a year and not pay the bills:· Many simply refused to believe that such a
thing really happened. They knew of youngsters who were just graduating from
high school who were given credit cards for an upscale department store. Others
thought that it was simply a case of a poor decision made by an individual, rather
than a reflection of company policy: "Well, some low-level person made a bad de-
cision. In my opinion anyway. Because I don't believe a higher-level person would
have done that. It's just too absurd."
A rather sophisticated critique of the use of profiles was presented in the jurors'
group. The intelligent consumer of news and entertainment was seen to be in-
creasingly denied a vital information service by the the mass media because of a
marketing logic that pursues the largest possible audience. This "least common
denominator tendency toward most marketing and entertainment decisions"
means that quality content disappears. And "you lose the freedom of choice....
They're dictating what you're going to watch, so you lose the freedom."
The participants' criticisms of profiling also emphasized the extent to which it
was impersonal. Decisions were not based on the particular circumstances of the
individual; instead, a person is merely a statistic. Insurance companies were fre-
quently criticized for their dependence on actuarial statistics and probabilistic
models in making decisions about coverage or rates: "Insurance is a big one. I
mean they have statistics, just built up on guesses .... They know that ... most
males under the age of 25 drive recklessly, so they boost their insurance premiums
... it makes you seem like you're not really a person, you know, because they base
everything in statistics."

TECHNOLOGY OF SURVEILLANCE & LIMITS ON DATA GATHERING

The interviewers asked a series of questions to determine the extent to which par-
ticipants could identify people who collect personal information, the variety of
132 RELATIONSHIPS AND EXPECTATIONS

methods used to acquire this information, and the range of uses to which the in-
formation is put. The "day in the life" tape and the discussions to this point had
introduced a great variety of examples of information gathering, but group par-
ticipants continued to add examples of data-gathering activities. Surveys, ques-
tionnaires, and applications were combined with transaction records-a category
in which some participants included the records generated by people who read
the gas and electric meters. The fact that most applications for insurance, espe-
cially medical insurance, involve a blanket provision of consent to the provider to
gather information about your present health and past claims was taken by some
participants to be common knowledge. Credit reporting agencies such as TRW
were aJso identified as active gatherers of personal data. These responses together
would compose a tl1ick catalog of legitimate business or government purposes
justifying the collection of personal information.
As a way of triggering fuxther discussion to more clearly identify what partici-
pants saw as limits-the boundaries beyond which the legitimacy of the collection
ofinformation would be questioned-the interviewers offered me groups a some-
what embellished representation of hair analysis techniques. Around the time of
these interviews, hair analysis was being promoted as a reliable way of testing for
drug use. 10 There had also been considerable speculation that oilier information
about the individual could be gained from an examination of the genetic material
in the hair follicle. Genetic analysis was described as able to provide information
about potential susceptibility to illness, stress, or workplace hazards, and thus
such information might be useful as a job screening aid.
Again, a great many participants accepted the blanket justification of data
gathering associated with a legitimate business need to know. For them, almost
any information could be justified if the organization could demonstrate that it
was relevant to a business decision. Others sought to find the limits in the conse-
quences of ignorance. That is, some kinds of jobs or decisions have what some
participants considered to be serious consequences, which served to outweigh the
interests of the individuals in meir privacy. "Things to do with like public safety
type jobs, you know, I'm definitely against ... drug testing and lie detector tests,
but you know, ifsomebody's going to be flying an airplane, they gotta be straight."
A few participants sought to establish limits based on the invasive natme of the
technology that would be used to collect tlie information. It was clear that many
participants identified the body as being the ultimate dividing line between the
private and the public realms. In the words of one participant: "Bodily informa-
tion. Your hair, your drug test. J mean that is absolutely absurd to me, and to me
tliere'sjust no cause. If they're hiring you to type all day long, let's say, what does it
matter about all that stuff? They can find out your health records from your refer-
ences. They don't need your bodily info, the way you were describing."
Participants offered oilier, related responses: "I don't think anything having to
do with your physical self should be available"; "on the form they asked me to fill
out, they asked my height, my weight .. and all I did was put a question mark
RELATIONSHIPS AND E-XPECTATIONS 133

next to it and didn't answer it, because I was applying for a job in the skill areas.
What do you need my height and weight for'?"
An important variant on the concern for relevance in data gathering was the
concern that only negative information be retained. Such a view gives the individ-
ual the benefit of the doubt. It presumes innocence, rather than assuming that ev-
ery employee is a thief or every applicant is a hjgh risk:

In other words, they shouldn't record every little thing that you do, only if you do
wrong. If you take a loan, aad you pay a loan, it shouldn't be recorded. Nothing
should come up. If you mispay a loan, then it should go on your record .... In other
words, when you look into a person's record, if you want to look for a credit check
and you find nothing, that means he didn't do anything wrong.... I should not be a
risk because I never had a chance to do something. l should only be a risk if I did
something wrong.

For this participant, anything other than negative information is irrelevant

and ought not to be recorded in anyone's file. Such a policy would have made life
much simpler for a participant in another group. It seems that, within a short
period, she had made three different credit applications related to the purchase of
a house that needed remodeling. Her application for an additional credit card was
rejected because there had been "an excessive amount of investigation into my
credit, willch of course I initiated." Her attempts to correct tills situation only
added to her frustration: "And of course, I wrote tills furious letter. You know, to
these anonymous people. What you do is then you write a letter and then you wrut
30 days, in order for them to check it out. And you can't yell at anybody, because
they don't answer the phone, and so you're left with this tremendous frustration
of the total irrationality of this."
When the interviewers probed people's views about what restrictions should be
applied to data gathering by govenunent agencies, there was considerable uncer-
tainty about these limits. The consensus was that the IRS should be able to gather
any financial information that would help it to decide whether one was paying
one's "fair share of taxes." At the same time, the respondents felt that FBI investi-
gations of library records to identify foreign spies were a threat to the freedoms
that characterized U.S. democracy. Again, the jurors' group offered some of the
more detailed justifications for their concern:

There's kind of an inverse ratio .. . between power and offensiveness. That the gov-
ernment and the FBl in particular .. . any form of the government being very power-
ful (and if you don't like this government yot11 can't pick another government), they
should be most severely restricted because you don't know where to go ... except
into exile. ... Only just slightly less [restricted] ... should be those places like TRW,
computer banks, and the insurance companies that share files. There should be a
hedging of inquisitorial rights of individuals and firms with whom you can either
choose to do business or not ... or [those) whose findings you can chaUenge or not.
134 RELATIONSHIPS AND EXPECTATIONS

Similarly, a person might not even know the criteria being used for an evalua-
tion: "You know ... you don't even know what to tell them. You don't Jiave no ex-
cuse. You can't even defend yourself. I think it's areas of not being able to defend
yourself ... in an employer's situation, that's interviewing you.... That's basically
it, you're not able to defend yourself."

DATA SHARING LIMITS

Although the relevance criteria within the limits of invasiveness associated with an
individual's body might apply with regard to the initial gatherer of information,
the interviewers asked the participants if they would use a different set of criteria
to establish limits on the sharing of personal information.
Whereas one participant would limit the collection and storage of anything
other than negative information in response to the question of sharing, another
participant suggested that any data that would help the data subject should be in-
cluded. Several respondents suggested that organizations should seek a person's
permission before they share any information about that person with others. At
the very least, the participants thought that those who gathered or provided any
data from a person's files to others had a responsibility to inform that individual
that the sharing had taken place.
Several participants distinguished between public and private, or corporate,
rationales for the sharing of personal information. There was greater latitude ex-
pressed for the gathering of data for collective or public purposes, as well as an ex-
pression of a greater willingness to restrict sharing by private concerns. One par-
ticipant expressed it in terms of individual differences in tolerance for the appeals
from commercial or nonprofit organizations that come from the sharing of infor-
mation about purchases or economic status.
Underlying these responses is the instrumental concern ultimately with the
benefit or harm that might flow from the sharing of personal information. People
would like to restrict the sharing of harmful information and facilitate the sharing
of beneficial information. In one view, "I don't care about what people know
about me personally, unless it's used against me, in some negative fashion. Then it
becomes a problem." In the view of some, sensitive information is information
that, if released or shared, could cause some harm to the subject of the informa-
tion:11

When I think of the census and what they collect, it's not harmful information. How
many people live in your house, what sex they are, what your income is. That's not
harmful. Harmful is when you start getting into what diseases they have, or what's
their psychological orientation or sexual orientation. What the hell you gonna do
with that? ... The law could define this type of information before it is used ...
[They] must have some clearances; [with] this kind of information, you don't need
clearances.
RELATIONSHIPS AND EXPECTATIONS 135

Participants, however, noted the difficulties in differentiating between what is

genuine harm and what is simply an annoyance. The general sense was that infor-
mation shared for marketing purposes was not harmful; it merely had the poten-
tial to increase the extent to which a person would be annoyed by solicitations.
There were very few hard-liners, or privacy fundamentalists, who took the posi-
tion that no unsolicited material should be sent to the home: "IfI want insurance,
I'll ask for it."

THE .CAPSULE VIEW

At the end of nearly two hours of talk about privacy, the interviewers asked the
members of each group to formulate a personal statement for a local television
station that summed up their feelings about privacy as an issue. Most of the par-
ticipants were quite reflective in their summary statements. Several returned to
points they had attempted to make earlier but apparently thought that a particu-
lar aspect had not been given its due weight. Among the jurors, two responses
stand out. One participant sought to bring attention to what he saw as a critical
aspect of Western social philosophy-the high value we place on individual au-
tonomy. The loss of privacy means the loss of the "space in which to be an individ-
ual." The source of the threat to this space was found in the immediately. follow ing
response: "It occurred to me that as long as people can make a living on invasion
of privacy, then privacy is going to be invaded. It's money in the back of it. As long
as they can make money from it, or gain power, then, they'll use it:'
The relevance criterion was summed up !best by a statement that suggested that
"every ·person should have access to information on a need to know basis. No in-
formation should ever be used to restrict or limit qne's pursuits, happiness, or jc;>y
of life as long as this doesn't impinge on others." One participant offered a rather
eleganr statement of the complexity of the issue, suggesting that "information
used to inform and protect can also be used to invade, discriminate and harass.
The questions of who should know, what should be known, and why they should
know it are questions that must be answered either by a consensus or a majority of
the people involved." Another person gave a more terse ultimatum: "If I haven't
applied for a loan, if I don't affect national security, and I'm not a criminal ele-
ment, then stay out of my business."
It was clear from this survey that there are several different levels of meaning
that are called into play when people are asked to think about privacy. Also, differ-
ent realms of activity are involved. Participants apparently utilized quite distinct
schema when thinking about privacy in relation to the government as compared
to their employers, to organizations providing them with goods and services, or to
their friends, neighbors, or coworl<ers. These differences have to do, Ln part, with
their recognition and their acceptance of the relative power of the actors in each of
those relationships and their faith or trust that such power will not be used for il-
legitimate ends. The overall tendency within the group as a whole was to trust or-
 RELATIONSHIPS AND EXPECTATIONS

ganizations not to gather more information than they need to make necessary
business decisions. This tendency to trust was reinforced by a belief that there was
a fairly reliable body of law that restricted serious invasions of privacy; or, if there
were no laws, it was felt that it was only a matter of time before such protective leg-
islation would be passed.
For this group, personal exper iences apparently played a critical role in helping
to crystallize a person's orientation to a particular aspect of privacy. Readily pro-
duced examples of problems with government files or credit files or mailing lists
support the view that personal experience tends to make certain attitudes more
salient and more stable than others. The assumption tl1at persons of a certain race,
gender, level of education, or work status will share orientations toward privacy as
a result of their shared experiences appears to have been supported in the discus-
sions and in the written responses to the questions.
In general, the members of these focus groups had a high degree of awareness
of the techniques of marketing that are already in widespread use. A substantial
majority fotmd such uses appropriate and, therefore, not a serious invasion of
ilieir privacy. Only a small minority expressed strong criticism ofsuch approaches
toward ilie rationalization of ilie marketing function.
The next chapter explores the issues that were raised in the focus groups
through prinlary and secondary analysis of data from several national surveys.
6
THE SOCIAL ORIGINS
OF VIEWS ON PRIVACY

INTRODUCTION
The purpose of this chapter is to explore the insights into the nature of social
opinion that can be derived from an analysis ofsurvey data. The data used for this
analysis have been collected from avariety of sources, and they differ considerably
in the extent to which I have been able to look behind the numbers and search for
the underlying meaning of the responses captured by the questionnaires. The pri-
mary and most valuable sources of data have been three datasets that I have ac-
quired from the Louis Harris Data Center. These files contain data from the public
samples of surveys administered by the Harris organization for which Alan
Westin served as faculty adviser. Each survey was financed by a company that was
involved in some way with the insurance industry, and many of the questions re-
flect that corporate/institutional interest. 1 The surveys also reflect Westin's long-
standing interest in privacy and computer technology. The W83 survey focused on
public perceptions of the computer and computer-based technology and their in -
fluences on the quality oflife. Privacy was a subset of that larger area of interest.
Primary survey data come from a study I conducted with support from a grant
from AT&T. 2 These data include questionnaires completed by participants in five
focus groups and 1,250 adults interviewed by phone through AT&T's prime con-
tractor, Maritz Marketing, In 1988 and 1989, respectively.; Each of these surveys
will be referred to primarily b y their dates of administration, 1978, i983, 1988, i989,
and i990. Additional data about the nature ofsocial opinion come from published
studies or from searches oo the subject of privacy through the Roper Center. 4

THE PENUMBRA OP FEAR

One of the first considerations that must be addressed is the potential confusion
that is introduced when the complex of opinions releva11t to the panoptic sort are
compressed under the single rubric of concerns about privacy. There are a great
many definitions of privacy, and one can never be clear about which aspects of
that construct are being evoked when an individual is asked to express an opinion.
The sustained publicity surrounding the Senate hearings to confirm the nomina-

137
 THE SOCIAL ORIGINS OF VlEWS ON PRIVACY

tion of Judge Robert Bork to the Supreme Court heightened public awareness of
privacy as a right that was not mentioned specifically in the Constitution but that
can be seen as emanating from a penumbra reflecting interests protected by the
First, Fourth, Fifth, Ninth, and Fourteenth Amendments. Questions of reproduc-
tive rights, sexual relations, and a variety of other activities that might be q ues-
tioned or even forbidden by law have become attached to the overly broad concept
of privacy through its association with constitutional protection of fundamental
liberties.
The notion of"informationalprivacy" associated with the formative contribu-
tions of Alan Westin refers to the "claim of individuals or groups or institutions to
determine for themselves when, how, and to what extent information about them
is communicated to others." 5 This definition, although cenu-al to my interests
here, is rather strictly limited to the sharing or distribution of information and
does not address the related questions about the collection of that information.
There has been much said and written about the indignities involved in the collec-
tion of information about drug use. Concerns about the invasiveness of surveil-
lance techniques, the impropriety of involving male officers in the search of fe-
male suspects or prisoners, or the use of mass screening and random testing are all
linked to conflicts about the legitimacy of an inquiry into particular spheres of
personhood, which are complicated by aspects of manner and place. The point I
wish to make here, however, is that we are never quite sure about the underlying
dimension of this complex that is being tapped when individuals are asked to re-
spond to a question.about privacy per se.

Concern About Privacy

When we inquire about the public's concern about privacy, we operate from an as-
sumption that privacy is something that is universally valued and, more impor-
tant, is perceived to be at risk. Privacy in this regard refers to some attainable status,
generally discussed in the context of an individual's control over the access that
others might have to regions, including regions of personality, that the person
holds as private. Thus, in the area of informational privacy, the concern about a
potential loss is based on the fear that individuals will no longer be able to exercise
control over access by others to information that defines them. ln a variety of rela-
tionships with others, and especially through interaction with representatives of
institutions, individuals may provide access to·information about themselves vol-
untarily to facilitate interaction, in exchange for information of presumably equal
value, or more reluctantly, as part of a nonnegotiable requirement for access to
other desired goods or services. Each of these circumstances or relationships is
marked by a particular set of expectations or assumptions about the uses to which
the information will be put, including the extent to which the information will be
shared with others. Increased concern about privacy can reflect a declining confi-
dence that others, especially institutional others, will maintain personal informa-
tion as confidential. Increased concern about the loss of privacy can also reflect a
THE SOCIAL ORIGINS OP VlEWS ON PRIVACY 139

common perception that interactions with organizations increasingly involve less

than voluntary surrender of personal information of varying degrees ofsensitivity.
The concept of sensitivity is one that has been explored with great effect by
Raymond Wacks, who has attempted to classify broad categories of personal in-
formation in terms of its sensitivity.6 For Wacks, sensitivity has two related di-
mensions: the personal, as distinguished from the public aspects of being, and the
consequential, which implies the potential of harm to the individual that might
result if unauthorized access or unauthorized use is made of that information.
Sensitivity might be thought of in terms of the "information value" of the "bits"
or "chunks" or insights that are gained, and the linkage of that information to
other complex chains. Sensitive information is highly linked to other critical as-
pects ofa person's economic and political status. Thus, sensitivity is a continually
evolving, socially constructed sphere of consequentiality. Its points are not fixed,
but must change as behavior, interests, orientations, tendencies, and the like move
in and out of favor within different spheres of activity in a society. Individuals may
differ markedly in terms of what they consider to be personal and consequential.
And, as I will discuss in the next chapter, these differences make the formation of a
legal regime quite difficult because the standard of a "reasonable person" as a
component of an "expectation of privacy" necessarily reflects dominant cultural
norms. Thus, information about what a young man thinks about sports cars
would not be considered highly sensitive; however, information about what that
young man thinks about young women wouJd be sensitive, because information
about sexual orientation and identification is highly linked in contemporary dis-
course and culture to considerations of health status, intimacy, and even suitabil-
ity for employment and training.
Increased concern about information privacy also reflects the belief that the so-
phisticated approaches to matching and profiling, which facilitate the use of per-
sonal information stored in unrelated files, are being used to produce sensitive in-
formation from that which was not sensitive in its original form. Such a concern is
not simply paranoid fantasy, as some would lb ave us believe, but has found a place
at the highest levels of government. At one point, the Reagan administration pro-
posed to create a category ofsensitive information and to ban the use of database-
searching software by foreign intellectuals because such software would facilitate
the production of intelligence that might be damaging to the national security in-
terests of the United States. Use of the technology, not the basic information, was
to be banned because the searching algorithms were too efficient. It previously
had been thought that it would have been too costly to produce an equivalent level
of useful information by processing the data in published reports manually. 7

Changes over Time

As technology changes, as its applications in the panoptic sort expand, and as
scholars and the press report on the consequences of its use, concerns about pri-
vacy are likely to change. James Katz and Annette Tassone provide one of the few
 THE SOCIAL ORIGINS OF VlEWS ON PRNACY

analyses of trends in public concerns about privacy.8 Although the trends are not
clear, reflecting in part the differences in the formulation of the questions and the
constructs they tap, Katz and Tassone identify several studies that suggest that the
proportion of the public that is very concerned about "threats to, or invasions of,
their privacy" has increased substantially. In addition, people seem to believe that
the loss of privacy will become an even greater problem in the future than it was in
the i98os. Jn the Appendix to their report, Katz and Tassone include data from
several Harris and Roper surveys regarding privacy, and it can be seen that the
proportions fluctuate from year to year. It is not clear whether these fluctuations
are reflections of variability in the samples, or whether they reflect the salience of
the privacy issue in the public discourse at the time. Public policy deliberations
about privacy in Congress, or the spectre of the much feared "1984" and the domi-
nation by "big brother:' can be seen to be linked closely to increases in the num-
bers of citizens who are concerned about privacy. A question that asked respon-
dents to indicate how close we had come to the society that George Orwell had
described in his book 1984 found the proportion who thought that we had already
arrived at such a society to have more than doubled between i983 and i988 and to
have tripled between i983 and i989 (from 6 percent to 19 percent).9 The Harris
survey published in 1979 reported that 31 percent of the public was "very con-
cerned" about "threats to ... [their] personal privacy"; the Equifax survey in i990
found that the proportion had increased to 46 percent.

A QUESTION OF TRUST
A key dimension underlying the levels of concern that individuals might have
about threats to their privacy is the relationship of trust they may have developed
with individuals or organizations with whom they interact and on whom they
may depend for access to goods and services. The more we trust an individual or
an organization, the less concerned we are likely to be that they will abuse that
trust and use the information to cause us harm or be careless in their sharing of
that information with others who might cause some harm. A number of surveys
over the years have asked respondents to indicate the extent to which they trusted
or had confidence in the information practices of public and private organiza-
tions.
One can compare the rankings that respondents assigned to institutions that
are deeply involved in the collection and use of personal information (Table 6.1).
In the 1978 data, rankings are based on the proportion that agreed with the state-
m ent that the organizations or individuals were "doing enough to keep the per-
sonal information they have on individuals confidential." In the i990 data, rank-
ings are based on the mean ratings of trust in a scale ranging from i (no trust) to
10 (complete trust) when trust means that they will "collect and use information
about people like you in a responsible way. " 10
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 141

TABLE 6.1 Organizational Rankings by Trustworthiness

1978 1990
Least Credit bureaus Credit bureaus
Internal Revenue Service Insurers
Insurers Internal Revenue Service
Social Security Administration Telephone companies
Census Bureau Employers
Employers Social Security Administration
Telephone companies Hospitals
Most Hospitals Census Bureau

Perhaps most notable are the decline in the trustworthiness of the telephone
companies and the improved status of the Census Bureau. In response to a ques-
tion about their confidence that the Census Bureau was not sharing personal in-
formation with other government agencies, only about 13 percent of the respon-
dents indicated great confidence in this agency in 1978.
Of course, the 1990 survey was taken at a high point in the national debate
about Caller-ID, as well as during the administration of the i990 census. Both
were the subject of considerable press coverage; that for the Census Bureau was
quite a bit more laudatory than that for the telephone companies. A similar ques-
tion asked in i983 allowed for a ranking of some of these institutions in terms of
their perceived tendency to maintain the confidentiality of the personal informa-
tion they had gathered. The order, from most to least trustworthy in this regard,
was the IRS, telephone companies, the Census Bureau, insurers, and credit bu-
reaus. In i983, the loss in public confidence that accompanied the breakup of the
AT&T monopoly had only just begun to emerge. All of the rankings computed
from these national surveys identify the credit bureaus as the least trustworthy of
the information intensive organizations, although by 1990 a newcomer had been
added to the list. Companies that were involved in direct mail or telemarketing
sales anchored the least trustworthy end of the scale with a mean rating of 2.9,
with their nearest competitor, the credit bureaus, earning a mean rating of 4.5 in a
scale that reached to 10.

HOW DID YOU GET MY NAME?

I have suggested that the level of public opposition to direct mail and telemarket-
ing is considerably below that which would emerge if respondents to surveys actu-
ally understood the countless ways by wbjch their names, addresses, and other
personal information were added to the files of the list brokers. Not even people
we might expect to be aware of such practices have an understanding of the extent
to which personal information from transactions becomes part of the panoptic
sort. One colleague, a business school professor, reports that she expanded her
professional interests from management information systems to questions of pri-
vacy and personal information after she began receiving direct mail solicitations
142 THE SOCIAL ORIGINS OF VlEWS ON PRJVACY

after her purchase of gifts for a baby shower. 11 Another colleague (who asked to
remain anonymous) related the story of an invitation he had received to speak be-
fore an organization. On asking about compensation or an honorarium, he was
told that the organization had not yelt decided. Their decision was to be based in
part on his record of donations to certain Jewish charities. He had given to some,
and not to others, and they were as yet undecided how this record was to be evalu-
ated. He was first shocked and surprised that they had access to information
about his charitable contributions, and furthermore, he was annoyed that his re-
cord of giving should in any way determine whether he should be compensated
for his lectures. Of course, his experience suggests that there are probably a con-
siderable number of other scholars whose records of donations would have dis-
qualified them from speaking at all!
In the 1990 survey, respondents were asked three questions that reflected differ-
ent levels of awareness and concern about organizations' dat;<i gathering and use
that might affect them through the marketing of personal information. Nearly 58
percent of the respondents suggested that it was a major problem that consumers
were being asked to provide "excessively personal information" by organizations
that gathered information about consumers. A slightly smaller proportion (55
percent) felt that "inaccuracy and mistakes" were similarly problematic. However,
a considerably smaller proportion thought that the sharing of personal informa-
tion between companies in the same industry was something to be concerned
about (39.8 percent). Yet, in the same survey, 97 percent of the respondents indi-
cated that they thought it was a "bad thing" that companies could "buy from
mailing list companies information about your consumer characteristics." When
asked further about how concerned they actually were about this, a relatively small
percentage (28.2 percent) indicated that they were very concerned. It is perhaps
this level of response that convinced Equifax, the survey's sponsor, to move for-
ward with its partners on the development of the disastrous Lotl.\S Marketplace
CD-ROM project A decade earlier, Equifax representatives had indicated to Con-
gress in testimony regarding a Fair Financial Information Practices Aq that they
placed more confidence in their contacts with the subjects of their files through
disclosure interviews than they did in the results of "highly touted opinion sur-
veys."12 It is especially ironic that the survey criticized in i980 was administered by
the same team that produced the analysis for Equifax in 1990.
The fact that there is widespread agreement that a particular business practice
is wrong on the face of it but is not at the same time apparently worthy of great
concern, points to the underlying sense of faith that people have in the operation
of the marketing system. That a corporation would share information without
first gaining permission is seen as an indication of poor manners, an abuse of the
trust that is at the heart of the relationship between a customer and t!ie organiza-
tions that individuals may come to depend on. However, businesses' use of infor-
mation about their purchases and transactions does not represent a particular
threat or an identifiable risk to a great many adults.
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 143

In a factor analysis of responses by focus group members to questions about

the technology of surveillance, the strongest factor identified was one termed
Mybusiness. 13 This factor reflected a general feeling that what people read, the
films they viewed, the numbers they called for information, and, more generally,
how they spent their money were their business and no one else's. Another factor
(Trust) served as an index of the confidence people had that organizations would
use information appropriately. This factor included agreement with the follow-
ing: "The more organizations know about me, the better they can meet my
needs"; "organizations have to have the best workers, and testing helps them to
choose"; and, "you simply have to give up some privacy to enjoy the conveniences
of the modern world."
In 1989, however, 61.3 percent of the respondents indicated strong agreement
with the statement that "companies should seek your permission before they tell
anyone else about the products you buy or the services you use:' Yet, when asked if
they believed that "the more businesses know about me, the better they can meet
my individual needs;' nearly 25 percent agreed strongly and another 43.1 percent
agreed somewhat. Presumably, it should not really matter how businesses gather
this information as .long as they are gathering it to meet the consumers' needs
rather than to exclude them from markets and opportunities. When the correla-
tion between these two measures is examined, the relationship is near zero and
tending .toward the inverse (r = -.05, p = .09). That is, people who think that a
business ought to seek permission before sharing information about its custom-
ers are somewhat less likely to believe that having such information helps the
business to serve its customers any better. This correlation suggests that there is
more than courtesy at issue.

CAN THEY REALLY DO THAT?

Part of the construction of social opinion is linked to the public's understanding
of the technologies that are used to generate intelligence through various forms of
surveillance. The literature of alarm that is thought to have stimu.lated public con-
cern about privacy highlighted the role of the computer as a device that made it
possible to collect and store tremendous amounts of personal information in files
that people knew little about but that coW,d be used to control their lives. When
asked in 19&3 if they believed the statement that "personal information about
yourself is being ~ept in some files somewhere for purposes not known to you,"
some 66 percent indicated that they believed that it was. When asked in i989 if
having a social security number makes it easy for organizations to gather personal
information from different sources, nearly half of the respondents (49.6 percent)
were confident that this was true, and an additional 32 percent tended to agree.
The use of information stored in unknown files and the use of a social security
number as a reliable token for matching those files are the ingredients necessary
for the production of detailed profiles of consumers.
144 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

One aspect of profiling is the assumption that past behavior can be used to pre-
dict future behavior. Indeed, the assumptions underlying predictive models are
relaxed considerably in that predictions need not be made on the basis of infor-
mation about the past behavior of any particular individual, but on the basis of
the behavior of other individLtals in the class or group to which a person may be
assigned on the basis of one or more of that individual's attributes. The 1989 sur-
vey asked respondents to indicate the extent to which they agreed with the state-
ment that "how a person behaved in the past is a good indicator of how they'll act
in the future." More than 60 percent of tbe respondents indicated at least some
agreement with that statement, although only about 20 percent indicated that
they agreed strongly. A related question asked if they agreed that "psychological
testing helps employers to select the best workers for the job." A somewhat smaller
proportion agreed (52.2 percent), with only about 14 percent agreeing strongly. Jf
the correlation between these two measures is examined, a relatively strong associ-
=
ation (r .16) is found, which indicates that those who have faith in tests also have
faith in projections based on the historical record. Thus a substantial proportion
of the public believes that profiles represent meaningful intelligence about indi-
viduals, and information in such profiles, therefore, can rise to the level of"sensi-
tive" and thereby become the basis for concern.

THERE OUGHT TO BE A LAW!

There is a distinction to be made between what people believe is technologically
possible and what they feel is justified, either morally or under the law. These per-
ceptions are susceptible to change as conditions change. John Detweiler's analyses
of changing perceptions of social entitlements provided an index of changing ex-
pectations regarding "legal protection from unauthorized use of confidential in-
formation." In i979, 88 percent of respondents indicated tliat they felt so entitled,
with 70 percent expressing optimism regarding that status. By 1984 only 69 per-
cent felt so entitled, altliough slightly more expressed some optimism about an
improvement (72 percent). 14 One interpretation of the change is based on the in-
creased level of surveillance initiated by the Reagan administration's war on
"fraud, waste, and abuse" in the provision of government services and a general-
ized policy toward creating "downward expectations" about social entitlements.
There are also differences that emerge in terms of comparative legitimacy. In a
Harris survey in 1978, respondents were asked to indicate whether they thought
tliat tlie maintenance of certain files by public and private organizations was justi-
fied. Observed differences reflected confidence in the organizations as well as
evaluations of the sorts of persons who might find themselves in particular files.
Whereas more tlian 87 percent of the respondents thought that the government
was justified in using matching to compare welfare rolls against employment rec-
ords to "identify people claiming benefits they are not entitled to:' only about 47
percent thought that the IRS was justified in comparing "tax records against
THE SOCIAL ORIGINS OF VlEWS ON PRIVACY 145

credit card records." Nearly 52 percent thought that it was acceptable for the in-
surance industry to maintain a central file of individuals who were suspected of
making fraudulent claims, but only 22.8 percent thought that it was similarly ap-
propriate for a central file to be maintained for the use of employers that would
contain the names of individuals who had been "treated for mental health prob-
lems."
Questions asked of the participants in the i988 focus groups provide some in-
sights into the underlying dimensions that help to organize the public's highly dif-
ferentiated sense ofjustice regarding data collection.15 Respondents were asked to
indicate the extent to which they believed a list of persons "should be entitled to
privacy rights as you understand them." They were asked to assign scores ranging
from i (no rights at all) to 10 (absolute rights). Factor analysis was used to identify
the underlying constructs with which the scores for particular classes of persons
were most highly correlated. 16 Not surprisingly, one factor that emerged was com-
posed of three measures Jinked to crime (Criminals): criminals, persons who have
been convicted of violent crimes, and persons who have been convicted of nonvi-
olent crimes. There is some comfort, perhaps, to be taken in the fact that persons
who had only been arrested and not yet convicted were evaluated differently from
the convicted criminals. Persons who were only suspected of crimes but who had
not been arrested were classified as having rights similar to government workers,
political activists, politicians, minors, and workers responsible for the safety of
others (Politicos). A third class of individuals seen to have similar privacy rights
was composed of people who deal with the public, persons who apply for credit,
workers who handle large sums of money or sensitive data, and people who travel
by air (Risks). The fourth group included people who m ight knock on your door
or who might call your home telephone number (Sales). One way to assess the
benefit of the doubt that these respondents assigned to these groups is to compare
their average scores. On the basis of the means, or average scores, for the groups
defined by the factor analysis, the members of the focus groups would assign pri-
vacy rights, from greatest to least, to members of the following conceptual groups:
Politicos (mean = 7.76), Risks (mean= 7.08), Sales (mean= 6.21), and Criminals
(mean = 4.01).
There is a distinction to be made between the regulation of information gath-
ering and the regulation of its sharing or exchange. Katz and Tassone report a
Cambridge Repor ts question regarding whether responden ts favored "laws that
would restrict the exchange of information between government and p rivate in-
stitutions." The proportion favoring such laws increased from 56 percent to 69
percent between 1986 and i988. 17 The increase in concern may reflect the level of
publicity that followed reports of government use of commercial data to enhance
the efforts of the IRS. But it is also this aspect of information sharing that makes
the panoptic web so difficult for individuals to escape. The extent to which indi-
viduals are concerned about sharing by one organization and not by another is an
indication of their understanding of the risks that such sharing might entail. In
 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

TABLE6.2 Orientation Toward a Legislative Response

Policy Area 1989 Mean (S.0.) 1978(%)
Medicine and health 7.3(2.6) 63.7
Social security numbers 7.3(2.8)
Credit cards 7.1 (2.7) 60.8
Mailing lists 6.9(3.0) 60.9
Telephone call records 6.9(3.0) 51.1
Insurance 6.8 (2.6) 64.9
Employment 6.8(2.6) 61.4
Political activity 6.6 (3.0)
Consumer purchases 6.0 (2.9)

the i989 survey, respondents were asked "to indicate the extent to which they
thought that there was a need for strong laws to "control the sharing of personal
information." The scores ranged from 1 (no need) to 10 (very great need). Table
6.2 presents the areas ranked in terms of the mean ratings assigned by respondents
for particular classes of information.
The range of responses was quite narrow, as indicated by the means and stan-
dard deviations. The estimated need for a regulatory shjeJd suggests that respon-
dents seem little concerned with the ·use of most transaction-generated informa-
tion. By ranking information about consumer purchases at the bottom of an
ordered list, weUbelow the positions assigned to mailing lists, telephone call rec-
ords, and credit cards, respondents appear to see less of a relationship among the
four areas than my analyses would suggest. 18
It is clear, however, that this concern is continually evolving. In 1978, interview-
ers assessing the perceived need for regulation ofsome of the same organizations
asked respondents to indicate how important it was "that Congress pass legisla-
tion" to control their information practices. The second column of figures in Ta-
ble 6.2 indi<:ates the proportion of respondents in i978 who thought that it was
iniportant for Congress to act regarding these information spheres. Telephone call
records were clearly seen as the least problematic area.
A somewhat different approach was taken in the i983 survey. Interviewers
asked respondents about whether or not they would favor specific forms of regu-
lation governing particular kinds of offenses. By 1983 the public's orientation to-
ward regulatory or legal action appeared to have grown more aggressive and puni-
tive. In none of six regulatory areas identified did Jess than 66 percent of the
respondents indicate that they favored regulation. The least popular regulation
was that which would have specified "just what kind of information about an in-
dividual could be combined with other information about the same individ-
ual"-in other words, rules governing the construction of profiles (66.5 percent).
The most popular legislative option was a law requiring "that any information
from a computer that might be damaging to people or organizations ... be
double-checked thoroughly before being used" (90.6 percent).
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 147

The overwhelming majority of respondents in a survey performed for the Na-

tional Consumers League in 1989 indicated that they thought that it was inappro-
priate for employers to base hiring or firing decisions on the basis of information
about "off-the-job" activities, especially those activities that might reflect life-
style. 19 This region of privacy included several behaviors, such as smoking or poor
diet, as well as status considerations, such as obesity, which could easily be dem-
onstrated to be relevant to job performance or to employee expenses in the sphere
of health insurance or medical costs. Yet adults fairly consistently wish to deny
employers the right to even inquire about these matters.

OPPOSITIONAL TENDENCIES
There is a distinction to be made among what individuals report they understand,
how they feel about practices and the need for a regulatory response, and what
they report they do on their own to protect their interests. During the focus group
discussions, several participants discussed the ways in which they responded to
what they thought was an improper invasion of their privacy or the ways in which
they acted to disturb the operation of the panoptic sort. Options ranged from the
active refusal to supply the information or the provision of false information to
the rather passive avoidanc.e of situations in which the request for personal infor-
mation would be likely.
Individuals have to be extremely confident about their rights, their value to the
institution, or the availability of.suitable options before they resist the panoptic
sort by refusing to provide·requested information. When asked in 1989 if respon-
dents had ever refused to provide information or had left questions blank that
they thought were inappropriate, nearly 35 percent indicated that they had. When
asked a similar question in 1990,'46 percent indicated that they had refused inap-
propriate information requests from business, but only 15 percent had refused
such requests from government.
Several studies have reported the increasing resistance of people to requests for
them to participate in public opinion surveys. Some of this resistance is their un-
derstandable reaction to what the industry calls "sugging"- sales efforts made
under the guise of research. When asked in 1989 if they had ever been asked but
had refused to participate in a survey, 28 percent indicated that they.had. This re-
sponse was from a sample in which approximately half had never before partici-
pated in a survey.
· A more passive response is the simple withdrawal from or avoidance of situa-
tions or circumstances when one might be required to provide information that
might conceivably be disqualifying or embarrassing or might threaten the loss of
other relationships currently enjoyed. That is, an individual might not apply for
welfare or insurance benefits because completing the application might require
updating -information about one's status in an agency file, providing information
about living arrangements, or transforming extremely complex circumstances
 TllE SOCIAL ORIGINS OF VIEWS ON PRIVACY

into a simple yes or no answer- any of which might serve to alter the status quo in
a negative way. There has been a rather dramatic increase in the number of per-
sons who report that, at one time or another., they did not apply for something be-
cause they did not want to provide information. The proportion increased from
° 2
14 percent in 1978 to 30 percent in i990. Curiously, only u percent of the respon-
dents to the i989 survey indicated that they had "failed to apply" for something
for the same reason.
It is always difficult to assess the extent to which individuals will attempt to cir-
cumvent or oppose the operation of the panoptic system by the introduction of
false or misleading information. Lying is not a socially acceptable activity, and in-
dividuals hesitate to admit to lying, even to an anonymous telephone interviewer.
It is not clear that a projective measure is always more accurate than a direct in-
quiry in this regard either. When interviewers asked people directly in 1989: Have
you ever "provided false or misleading information on an application that you felt
they had no right to ask?" only 8.5 percent of the respondents indicated that they
had. Yet, when asked in i990 if they thought that "most people" were likely to mis-
represent the facts often or "a lot," 33 percent agreed with regard to employment
qualifications, 35 percent agreed with regard to health insurance applications,
nearly 42 percent agreed with regard to income taxes, and another 35 percent
agreed there was frequent misrepresentation on loan applications.
In the 1990 survey, interviewers asked about several items that related to spe-
cific concerns linked to the introduction of new technologies. There was an ongo-
ing public debate sun-ounding the introduction of Caller-ID. This technology
would enable a device to display the telephone number of incoming calls before
the first ring. Opponents claimed that the device represented an invasion of the
privacy of the calling party. The telephone companies argued that it potentially
increased the privacy of the called party because it would allow them to screen in-
coming calls. The Equifax Report indicated that 43 percent of the respondents
thought that the service should not be allowed to be sold. When interviewers
asked respondents a follow-up question, which included an efficiency and privacy
enhancing rationale for the service, 25 percent of the respondents suggested that it
should be banned completely and another 55 percent thought that if it were of-
fered it should be regulated by law. When they were asked a third time, with the
notice of a potential added benefit in the form of a reduction in obscene and ha-
rassing telephone calls, 27 percent of respondents still favored banning the tech-
nology, indicating it as too intrusive, and another 48 percent would allow the ser-
vice to be offered only if individuals could block the display of their numbers. The
Commonwealth of Pennsylvania intervened by ruling that the Caller-ID technol-
ogy was barred by Pennsylvania's wiretapping statutes. Yet, in the context of the
increasing press coverage of the "war against crime" or the "war on drugs," na-
tional survey data suggested a significant increase in public acceptance of wiretap-
ping, from a low of 16 percent in 1975 to a high of 26 percent in 1989.21 Thus, again,
an increasing share of the public seems to feel that it is appropriate to use a privacy
THE SOClAL ORIGINS OF VIEWS ON PRIVACY 149

invasive technology in pursuit or control of criminals, as long as its use is re-

stricted in its reach into their personal lives.
In the 1989 survey Caller-ID was not described by name. Instead, interviewers
asked respondents to indicate the extent to which they agreed with the statement:
"It would be good to know who is calling before I answered the phone." Some 38
percent agreed strongly, and 35 percent agreed somewhat. Thus an overwhelming
majoriry liked the advantage of knowing the identity of callers in advance, but a
substantial majority also believed that such a technology needs to be limited or
controlled.
There is a clear ends/means justification underlying the public response to this
technology and its use. There is also an issue of reciprocity and fairness that has
been raised in the debates about Caller-ID, which will be explored more fully in
the next chapter. Because information about another brings advantage and power,
it is understandable that individuals would like to know who is calling, would like
to know what is on their mind, and, if it is a sales call, they would be pleased to
know in advance whether the first offer the caller makes will be their final or best
offer. At the same time, these respondents would not like to reveal any information
about themselves, especially any information that might change the balance of
power against them. The value of such a technology as a bar against the annoy-
ance of sales calls is reflected in the fairly high level of interest in a potential appli-
cation of a fictional enhanced Caller-ID service. Approximately 72 percent of the
respondents to the 1989 survey agreed that "someone should invent a telephone
that would automatically screen out calls from people trying to sell you things."
It should be noted that Caller-ID as proposed, and then as offered in several
markets, does not fully meet the informational requirements of this majority, in
that only the number of the calling party is delivered. Supplemental services
might display additional information about the calling number, but it is only
when the personal calling number, such as the PIN or the social security number,
is de.livered that the called party will really be confident in knowing who is calling
and the likely nature of the call.

STRUCTURATION, OR THE SOCIAL ORIGINS

OF SOCIAL OPINION
It is important from a critical perspective to identify the social origins of the views
on the panoptic sort that come to characterize identifiable groups in society. So-
cial origins matter from a materialist theoretical position. Social origins also mat-
ter from a social policy perspective. If we understand how it comes to be that indi-
viduals accept what some of us may recognize as an oppressive and limiting
relation of power, we stand a greater chance of intervening to change those rela-
tions. This orientation resists the common analytical tendency to explain atti-
tudes and perspectives by using other attitudes and perspectives. Although the
correlations or predictive ability of such models may be high, there is very little
150 THE SOCIAL ORIGlNS OF VIEWS ON PRIVACY

contribution to our understanding of the processes that may have given rise to
both sets of opinions. Those involved in marketing research are less interested in
explanations than they are in reliable predictions of consumer behavior. If an atti-
tude will predict an opinion, that is more than enough. If that attitude can predict
a behavior with any degree of confidence, that is an insight worth some money!
The critical scholar, however, is interested in understanding the social processes
that underlie both attitudes and behaviors.
One broadly identifiable approach toward the development of explanatory
models is referred to as the social categories approach. It suggests that members of
identifiable social groups will share common attitudes, opinions, and reactions to
novel circumstances because they have had similar experiences and have been
subject to similar socializing influences during formative periods. The underlying
assumption behind the classification of respondents by race, class, and gender is
based on the expectation that the socialization of African-Americans differs from
that of whites; that young women have different experiences than young men, and
that they have those experiences interpreted to them in ways that differ substan-
tially from those of young men. The same logic applies to differences that reflect
the variance in social experience among persons from different social classes.

IN SEARCH OF SOCIAL EXPLANATIONS

A secondary analysis of the 1978 Harris dataset provides our first broad look at the
social basis for privacy orientations (Table 6.3). This database, with more than
i,500 cases and more than 200 variables, provided an opportunity to examine the
influence of race, gender, employment, political ideology, age, and education on a
multitude of constructions of the panoptic sort. Variables were all dummy coded
so that agreement with a privacy enhancing view would equal one; otherwise, the
variable would be coded as zero. Factor analysis 22 was used for the purpose of data
reduction; estimates of reliability2 3 are provided as an index of the additivity of
scales suggested by factor solutions, and one-way analyses of variance were con-
ducted for an assessment of the magnitude of the differences among groups iden-
tified by the social categories as measured.
Of the seven predictor variables, only political ideology can be characterized as
an attitudinal construct that might be specified as a result of a complex of social
experiences, rather than as a contributing cause of a person's orientation toward
the panoptic sort. It is an important variable in this secondary analysis in that it is
a significant discriminant in each of the eight factors presented in the table. How-
ever, it is not directly causal, even though it appears to be highly correlated with
other measured views. 24
Factors 4 , 6, and 8 appear to reflect the most s ubstantial differences among the
social typologies identified by race, gender, employment status, nature of employ-
ment, political ideology, and education. For factor 6, the categories described by
each of the variables are associated with highly significant mean differences. All of
 151

TABLE 6.3 Social Origins of Orientations Toward the Panoptic Sort (1978)
Factor 1. Practices (Use of psychological tests, lie detectors, eavesdropping, monitoring via
television, monitoring of speed, and so forth; 5 variables, alpha = .6942)
1. Race Sex Job Jobcat Polview Age Educat
2. p> .05 p< .01 p<.01 p> .05 p<.01 p<.01 p< .01
3. W/M SIR R/N 25/65 AA/0
Factor 2. Autoinsur (Automobile insurance companies should not have the right to obtain
information about applicants' drinking, credit, health, criminal record, and the like; 6 variables,
alpha = .755)
1. Race Sex Job Jobcat Polview Age Educat
2. p<.01 p>.05 p< .01 p<.01 p<.01 p< .01 p< .01
3. W/O SA/A PF/F UN 21/65 G/7
Factor 3. Finances/1 (Banks, finance companies, insurance companies, credit card
companies, and credit bureaus should do more to keep information confidential; 5 variables,
alpha = .895)
1. Race Sex Job Jobcat Polview Age Educat
2. p> .05 p<.01 p< .01 p<.05 p< .01 p< .01 p<.01
3. M/W S/R R/N 25/65 AA/12
Factor 4. Government (e.g., IRS, CIA, FBI, government, SSA, Congress should do more to
keep personal information confidential; 6 variables, alpha= .839)
1. Race Sex Job Jobcat Polview Age Educat
2. p<.05 p< .01 p< .01 p< .05 p< .01 p< .01 p< .01
3. W/H M/W SA/A PF/O UN 40/65 G/12
Factor 5. Finances/2 (Banks and other financial institutions ask for too much information; 5
variables. alpha= .843).
1. Race Sex Job .Jobcat Polview Age Educat
2. p>.05 p<.01 p<.01 p>.05 p<.01 p< .01 p< .01
3 M/W S/R R/N 18/65 1217
Factor 6. Employers (Employers should not be allowed to ask personal information, such as
questions about race, age, marital status; 5 variables, alpha= .798).
1. Race Sex Job Jobcat Polview Age Edu cat
2. p< .001 p<.001 p<.001 p < .001 p< .001 p<.001 p>.001
3. W/B W/M · SAIA PF/SL UN 25/65 G/7
Factor 7. Passivity (Failed to apply for job, credit, or insurance because respondent did not
want to give information; 2 questions, alpha = .972)
1. Race Sex Job Jobcat Polview Age Edu cat
2. p>.05 p>.05 p<.01 p> .05 p<.01 p<.01 p<.01
3. SA/R UN 25/65 BA/7
Factor 8. Matching (Computer used to predict fraud, drug use, mental patients, government
files, identity cards, and so on; 5 variables, alpha = .697)
1. Race $ex Job Jobeat Polview Age Educat
2. p<.01 p< .05 p<.01 p< .01 p<.01 p< .01 p< .01
3. W/O M/W SA/A PF/F R/N 30/65 G/O
Key: 1= variable name; 2= significance level; 3= highest over lowest group mean, Scheffe test
of difference
Race: W=white; B=black; O=Oriental; H=Hispanic
Job: HR= hourty; SA=salaried; SE=self-employed; R=retired; U=unemployed; S= student;
H = housewife; D = disabled
Jobcat: PF=professional: M= rnanager. official; PP= proprietor; C= clerical; SL= sales
worker; SK= skilled crafts; O= operative; SE= service; F= farmer
Polview: C=conservative; M=middle-of-the-road ; L=liberal; R=radical; N =not sure
Educat: O= no formal schooling; 7= 1-7 years completed; 8= 8 years completed; 11 = 9-11
years completed; 12=12 years completed; C=1-3 years of college; AA=AA degree;
BA= BA degree; G = one year or more of graduate school
152 THE SOCIAL ORIGINS O F VIEWS ON PRIVACY

the differences are significant at the 99.9 percent level of confidence. With regard
to the differences among groups evaluated by the Scheffe test, whites more than
blacks wish to restrict the employer's collection of irrelevant personal informa-
tion. They may think that this information is irrelevant because information
about marital status, race, height, and the like is thought to have no meaningful
link to job performance. Women were more adamant than men, salaried and pro-
fessional workers more so than other employees. This appeared to be a liberal po-
sition. Indeed, in all eight factors, self-proclaimed liberals or radicals were found
to anchor the most extreme privacy seeking end of the distributions. Age was also
a significant component in each of the factors, with the oldest cohorts anchoring
the most conservative, or least concerned, position on the scale. Education was
also a significant influence in each of the factors, and the general tendency was for
those with the most education, frequently those with some graduate training, to
be the most concerned with the preservation of their control over their personal
information.

Gender, Race, and Ethnicity

Being female is a relatively important influence on the way one comes to under-
stand, accept, or reject particular aspects of the panoptic sort. When the correla-
tions between being female and holding privacy related attitudes as measured in
i989 are examined, several significant relationships emerge. By using a conserva-
tive standard of greater than 99 percent confidence, it was found that women re-
ject the statement: "The more people know about you, the more control they have
over your life" (r = - .11) . This is consistent with a tendency for women to reject
the statement: "How a person behaved in the past is a good indicator of how
they'll act in the future" (r = .09). These data suggest that women are less likely
than men to be selective users of the media in that they view more hours of general
television (r = .n) than they attend to public affairs on television (-.08) or in the
press (- .10). This greater exposure and the presumed dependence on mainstream
television would suggest that women are more likely to be subject to the hege-
monic rather than the oppositional influences of the medium.
Race and ethnicity influence individuals through the same kinds of socializing
experiences that generate differences between men and women. In the i989 sur-
vey, the strongest association with minority ethnic status, being African-Ameri-
can or Hispanic, is revealed in the tendency to agree that corporations gather only
what they need in order to make good business decisions.25 African-Americans
and Hispanics apparently also share a tendency to accept the narrow specification
of privacy interests as being linked to having something to hide. Being African-
American (r =.14) is more closely associated with this view than is being Hispanic
(r = .05) with regard to one's estimate of the extent of public concern about pri-
vacy. The data indicate a greater tendency for Hispanic respondents to want to
know who is calling before they answer the phone. One executive at AT&T sug-
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 153

gested that this preference among Hispanic individuals may be related to the
greater commonality in surnames, which results in more unwanted calls from
people with "wrong numbers" that they gathered from directories or from direc-
tory assistance. African-American respondents tended to reject the view that past
behavior was a reliable predictor of the future (r = .u), a perspective that, as we
have seen, was apparently shared bya large number of the female respondents (r =
.09). Overall, the number of differences linked to gender and ethnicity was not
large, and many do not survive statistical controls designed to remove the influ-
ence of social class.

Income and Social Class

Increasingly, respondents to surveys appear unwilling to indicate their household
income. Twelve percent failed to answer the household income question in i989,
with near ly 10 percent refusing outright. A similar proportion (10.9 percent) did
not provide the information when requested by Harris interviewers in i990. It is
not unreasonable to assume that those who refused to answer this question are
more protective of the financial aspects of their personal identification. If those
respondents are excluded from analyses that utilize income as an explanatory
variable, the influence of that particular sensitivity is underestimated. By dividing
the respondents in the 1990 survey into those who did and those who did not pro-
vide an estimate of household income, it is possible to compare the means of those
two groups on core questions. For example, there was no significant difference be-
tween the two groups in terms of the proportions of persons who said they were
very concerned about threats to their personal privacy. There was no difference
between the groups in terms of their belief that consumers had lost control over
how information about them was circulated. There was only a marginal difference
benveen the nvo groups in terms of the proportions of persons who reported a
great deal of exposure to information about the potential misuse of personal in-
formation. However, there was a highly significant (p = .005) difference between
the two groups in terms of the proportions of persons who reported being very
concerned about the sale of personal information to mailing list firms. Consistent
with our assumptions about sensitivity, those who did not report their income
were more likely to report such concerns.
Alternative measures of social class included the level of education attained
and the nature of work performed by the respondent or the principal wage earner.
The 1990 survey included a measure of social class that introduced an ideological
spin that is difficult to untangle in that the categories included the labels "lower
class" and "working class," which may have served as representational or affil-
iative categories rather than as more objective indicators of status.
By using the nature of the work of the principal wage earner as an index of
class, an ordinal scale was constructed that ran from hourly, to salaried, to self-
employed. This measure was positively correlated with education (r = .21). Several
154 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

weak correlations emerged when ranks on this scale were compared with the
ranks on the twenty-five constructs in the 1989 survey. The strongest association
noted is with regard to what might be called a victimization measure. The higher
one's social class, the less likely respondents were to feel that people with power try
to take advantage of people like them (r = - .10). Those in the lower classes were
also more likely to report an interest in knowing who was caUing before they an-
swered the phone (r = -.09), perhaps to avoid the pressure of creditors seeking
payment. Class, as measured, was also a factor in explaining the preference of
some respondents for policies that might limit the kinds of information that em-
ployers might gather about job applicants. Curiously, those in the lower classes
tended to be more accepting and trusting of business, as evidenced by the correla-
tion of class with the beliefs that businesses rarely gather more information than
they need and that the more information they gather, the better businesses can
meet their individual needs (r = -.08).

Education
Education has always been a powerful factor in the explanation of social and be-
havioral differences. It has been seen to produce differences in two ways: first,
through common experience or socialiw.tion; an·d second, through the transfor-
mation of an individual's capacity to process information and ideas. Education is
the door through which individuals enter into experiences that further broaden
and condition their existences. Education is part of a process that helps to deter-
mine the social contacts that a person will find enjoyable and beneficial in work
and in recreation. It is usually thought to be a liberalizing influence, but, as dis-
cussed later, not all aspects of one's concerns about the panoptic sort can be un-
derstood as a reflection of a political continuum.
Educational attainment, such as that associated with the completion of specific
stages-high school. college, and graduate degree programs-represents a mark
of personal achievement. On the one hand, it is not surprising to note that those
vvith more education had the strong tendency to reject the view that "there is very
little that an individual can do to improve the quality of their life" (r = -.18). We
noticed that there was an even stronger tendency toward the denial of power as in-
dicated in the statement: "Most people with power try to take advantage of people
like me" (r = - .24). There was a simjlar rejection of the suggestion that "comput-
ers give big organizations an unfair advantage over the average person" (r = -.15).
On the other hand, those with more education appeared to be more accepting
of the logic and rationales behind the panoptic sort. For instance, there was a ten-
dency for those with more education to agree with the statement: "How a person
behaved in the past is a good indication of how they'll act in the future" (r = - .15).
However, their acceptance of the logic of the models does not mean that they
found s uch business use legitimate.. Those with more education were found to be
more willing to restrict data gathering by insurance companies (r = .u), in part
THE SOCIAL ORIGINS OF VlEWS ON PRIVACY 155

because they apparently do not believe that "companies rarely gather more infor-
mation than they need to make good business decisions" (r = .11). And they
tended to reject the view that "the more businesses know about me, the better
they can meet my individual needs" {r = .09).
Overall, those with more education tended to share a much lower estimate of
the general public's concern about privacy than did those with less education (r =
-.21). Significantly, their view of the public's concern was less cynical in that the
more highly educated tended to reject the view that "the only people who are con-
cerned about their privacy are people with something to hide" (r = .15).

Age Cohort Analysis

In nearly every analysis of each available dataset, age has emerged as an important
source of variance in opinions. There are several ways to think about the influence
of age on one's policy orientations. We might consider age to be an index of matu-
ration. In general, we have come to characterize a particularly liberal or progres-
sive social and political orientation as youthful-revolution is sparked by the
youth. It is generally accepted that we grow increasingly more conservative as we
age, as our liberal ideals repeatedly confront the cold hard realities of economic
and social responsibility. Thus we might expect political, economic, and social
conservativism (even though they are clearly different concepts) to increase with
age. This view might be labeled a life cycle model of value change to the extent that
different roles and requirements influence changing attitudes and orientations.
Such a structurationaJ view suggests that there are common demands on people
that characterize particular life stages, which explains the greater sin1ilarity
witllin, rather than between, broad age cohorts.
There is a second approach to understanding the influence of age that does not
see it as a continuously occurring process of maturation. This approach suggests
that certain orientations, political and social values in particular, are established
fairly early in life and tend to reflect the dominant values being expressed within
the cohort when such values are formed (between eight and twelve years of age).
With tbjs theoretical position, we would expect to see age cohorts remaining rela-
tively homogeneous and, more important, relatively stable in orientation as they
age. Thus adults who reached po]jtical maturity during the Roosevelt period of
state activism should be quite different from tllose who matured during the
height of McCarthyism or the U.S. cultmal revolution of the 1960s. Ronald
Inglehart argues that the most powerful influences on cohorts are economic, and
among the economic influences the most powerful are changes in educational at-
tainment. 26 He notes that in the United States, the rate of access to higher educa-
tion doubled from J950 to 1965.27 He suggests tliat education, especially college,
changes students dramatically, making tllem more liberal, less authoritarian, less
ethnocentric, less dogmatic in their views, and more interested in political mat-
ters. He notes that, although the U.S. population almost doubled between 1920
156 THE SOCLAL ORIGINS OF VIEWS ON PRIVACY

and i970, "more than sixteen times as many college degrees were awarded in 1970
as in i920." 28 Inglehart's "post-materialist values" are the result of changes in the
economic resources available to individuals in different industrial countries,
which are enhanced through education and the expansionary vista provided by
television.
Inglehart's approach has much in common with that developed by Milton
Rokeach.29 Rokeach focuses on human values as relatively stable core constructs,
which have been demonstrated to have predictable links to attitudes and opinions
about various things, including automobiles and political candidates and televi-
sion programs. 30 Whereas Inglehart's approach seeks to identify the common core
values of a nation as a reflection of its level of economic development, Rokeach's
approach involves the pursuit of difference. It is the differences in the core values
that predict and explain the observable differences in attitudes, opinions, and
behaviors. Unlike Inglehart, Rokeach makes no effort to specify the complex of
"cultural, societal and personal experiences" that contribute to an individual's
value structure. However, his analytical framework pursues a social categories ap-
proach and demonstrates significant differences in structure between whites and
blacks, men and women, and social classes defined by income and education. He
concludes that "the many value differences found between the very poor and rich
almost suggest that they come from different cultures:'31 At least thirty of the
thirty-six values Rokeach uses reveal significant age differences. "The general im-
pression gained from inspection of the data is one of continuous value change
from early adolescence through old age with the presence of several generation
gaps rather than just one." 32 The differences between Jnglehart and Rokeach at
one level have to do with the different values measured. Inglehart's postmaterialist
index represents only a small subset of the complex of values that are measured in
relation to each other in the Rokeach scheme. The second problem is in the unit of
analysis. Inglehart seeks to make comparisons among nations; Rokeach's analysis
seeks comparisons among groups.
The third approach is one that suggests that there are important differences
within cohorts that are associated with the concrete experiences individuals may
have, experiences that might be common to a group based on race, gender, or class
but that may also represent orientations based on cumulative experience that is
not predicted or explained by the other influences, independently or in combina-
tion. This view recognizes the complex nature of in Auences on social orientations
but assumes that these influences are dynamic, rather than stable, and reflect
learning to accept or to resent and resist institutional demands.
The first two positions are incompatible with each other, and the third risks the
frustration of unmanageable complexity. Longitudinal data would be required to
answer the question with any confidence. Such data are not readily available, and
questions about the panoptic sort were certainly not on the research agenda early
enough to make such comparisons possible. Two less satisfactory alternatives in-
volve the comparison of samples taken at different points in time while making a
THE SOCIAL ORIGINS OF VIEWS ON PRJVACY 157

generally untenable assumption that the samples are truly representative of the
same populations, just separated by a given number of years. Such an approach is
also troubled by the nature of cohort measurement and the irregularity of survey
administration. Not all surveys ask for age in years, in part because some respon-
dents refuse to provide a precise answer. Instead, respondents are asked to indi-
cate if their ages fall within a given range. The differences in the size of the cohorts
so described and the length oftime between surveys make it impossible to assume
that the same cohort has simply become five years older, that is, merely shifted
from ~me cohort to another. The third difficulty is common to all cross-sectional
analyses: the need to assume part of what you seek to discover-that personal at-
tributes indicating group membership allow the assumption of similarity in expe-
rience within group or cohort.
Because of all of these difficulties, no single approach is satisfactory, but if the
evidence from multiple approaches points in the same direction or toward the
same conclusion, our confidence should increase. For example, if we examine the
distribution of ideological self-identification by age and focus on the proportion
of the samples in 1978 and 1990 who claim to be conservative, there is a curvilinear
pattern (Figure 6.1). In the 1990 survey, the low point is reached by age twenty-
five, whereas the low point was no t reached until age thirty-five in the 1978 assess-
ment. The younger cohorts are considerably more conservative in 1990 than their
counterparts were in 1978. Thus, although there is evidence of a traditional "mat-
urational" move toward greater conservatism with age, the increasing conserva-
tism of political discourse may have had a direct effect as well. 33 The challenge of
separating the complex of influences called maturational from the common expe-
riences that are associated with the life cycle but that vary with race, class, and
gender is confronted in this section.
A similar but potentially troubling pattern is seen in the curves (Figure 6.2) re-
flecting the relationship between age and political cynicism. Although the pattern
is similar to that of a growing conservatism, overaU the level of cynicism has de-
clined in i990 in comparison with that in 1978. Because the measure of cynicism is
the proportion of the cohort who agree that how one votes bears little relationship
to what government does, the fact of increasing conservatism in the whole popu-
lation, linked with a declining cynicism, suggests a greater confidence in the polit-
ical process associated with the Reagan and post-Reagan era. Such a conclusion is
clearly not warranted as a general model because there are different responses in
different cohorts. For the three oldest cohorts, there has been a substantial decline
in trust in government (to look out for one's interests), but a substantial increase
for the younger cohorts, which may reflect a changing structure of reward in the
society (Figure 6.3).
What is most striking about the public's trust in institutions is the great shift in
the level of trust in business (Figure 6-4). Although respondents in the older co-
horts are more satisfied with business, the gap between the 1978 and 1990 cohorts
increases dramatically and suggests that a considerable improvement in business
 Age Cohorts
Younger Older

FIGURE 6 . 1 Conservat ism and age

0.5-- - - - - - - - - -- -- -
1990
1978

0.4

. .... .f
•••
0.3
.
·... . .···

o.z...________________,

A~e Cohorts
Younger Older
FIGURE 6.2 Political cynicism
o.s
....
c
e
...c
~ 0.4 1978
"'"'
..
·! 1990

..
~
g 0.3
·s
Q.
0
~

0.2 ~---------------'
Age Cohorts
Younger Older

FIGURE 6 .3 Trust in government by age

 .... ...........
0.80
1990
..t:
c:
0.78

l 0.76
.g"'
.. 0.74
•••
g 1978
c:
0 o.n
~
~

.....
.8' 0.70

0.68

0.66

Age Cohorts
Younger Older

FIGURE 6.4 Trust in business by age

2.6

....
E

..8 2.4
0
] 2.2 .• 1990
2.0 ..· •······ • • •. .-. ....
·-..

Age Cohorts
Younger OJder
FIGURE 6.S Extent of concern by age

0.4~--------------.

~
Q.

.• 0
0.3
.... . . .. ......... \.

"'c:
• •··...
! 0.2 •• 1990
8

i
ct. 0,1

1978
0,0
Age Cohorts
Younger Older

FIGURE 6.6 Failed to apply by age

160 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

practices (or in public relations) had occurred in twelve years. This pattern sug-
gests that there has been a dramatic improvement in the interactions between
businesses and older adults. It may also mean that the level ofexpectation has also
changed, which may contribute to a higher level of annoyance when particular
businesses step out ofline.
As is clear from Figure 6.5, the level of public concern about privacy had in-
creased dramatically between 1978 and i990. The shapes of the age distributions
change most noticeably in the younger cohorts, in whom the level of concern in-
creased more dramatically than with any other cohorts. Also, the greatest level of
concern occurred in an earlier cohort in 1990 than in i980-the opposite of what
we would have seen if those with the greatest concern had merely gotten older in
twelve years.
An upward shift in the proportion of age cohorts who have decided not to ap-
ply for something needs to be interpreted cautiously (Figure 6.6). The form of the
question produces a cumulative estimate. That is, those who answered in the af-
firmative in 1978 would presumably have answered in the affirmative in i990, and
thus any increase in the proportions would be attributable to individuals who had
never been passive before i978, but who had been so at least once since that time.
As a cumulative measure, it should increase but shift to the right. The apparent
failure of the curve to shift to the right suggests that the increases did not occur
equally across all cohorts. Indeed, we would have to suggest an inverse relation-
ship with age: As people get older, they are less likely either to face the choice or to
take a passive response.
Alternatively, the measure of technophobia, the belief that technology has got-
ten out of hand, reflects a relatively gentle shift in opinion consistent with a cohort
aging interpretation (Figure 6.7). Those in the fourth cohort in 1990 are not very
different from those in the third cohort in 1979. The pattern is similar between
other lagged cohorts except for the final cohort in whom there is a dramatic rever-
sal. Such a reversal must indicate a changed perception in that cohort. Both static
and lagged comparisons would be statistically significant with samples of this size.
A different approach has been taken in the exploration of the influence of age
in the 1989 Survey. In a factor analysis of the unweighted cases, the twenty-four at-
titudinal variables produced eight factors accounting for nearly 50 percent of the
variance (47.8 percent). 34 Factor scores were generated for each factor, and the fac-
tor scores were subject to multivariate analysis of variance with gender and age co-
hort as variates. Gender emerged as a significant main effect on the set of factor
scores. Univariate tests revealed that gender mattered most with factors 3, 7, and 8.
Factor 3 might be characterized as an antilist factor. Three measures loading most
heavily on this factor were: ( i) a belief that individuals had "a right to have their
name removed from any mailing list;' (2) agreement with the statement that there
"should be a way to keep your name off certain mailing lists;' and (3) agreement
with the statement that companies should "seek your permission before they tell
anyone else about the products you buy or the services you use:'
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 161

0.6 ~-----------~

1990

0.5 1
: 1978

0.4

Age Cohorts
Younger Older

FIGURE 6. 7 Fear of technology by age

Factor 7 was primarily concerned with technology and the predictive utility of
data. Three measures loading heavily on this factor were: (l) the belief that past
behavior was an indicator of the future, (2) the belief that psychological tests
would be helpful for employers involved in choosing the best employees, and (3) a
tendency to reject the assodaled belief that companies rarely gather more infor-
mation than they need to make good business decisions. Respondents with high
scores on this factor believed corporations could gain useful intelligence from per-
sonal information, although they did not believe its use was always justified. The
third factor (8) for whid1 gender-linked differences were the strongest, involved
agreement with the statement that you " have to give up your privacy to enjoy the
convenience of the modern world" and rejection of the claim that "information
provided to the Census Bureau is held strictly confidential." None of the tests of
the gender/age interaction were statistically significant.
However, the classification of respondents into five age cohorts reveals age to
be a highly significant main effect. The univariate tests identify age as a significant
factor in seven of the eight comparisons. The only factor in which age did not play
a major part was factor 4, which involved the desire for a device like Caller-ID that
would identify the calling party but would also serve to screen out calls from sales
people. These two measures are highly correlated (r = .32).
Trust in business and government data handlers also varied significantly with
age in the i990 survey. An analysis of variance using the means of the trust scores
for each of eight cohorts reveals a highly significant relationship (F = 8-487, p <
.001). The relationship was significantly linear, with a negative corrdation (r =
-.08). However, it is also clear that the curve is U-shaped, with tJ1e greatest trust
among t11e youngest cohorts, the least trust within the middle groups, and the se-
niors reflecting a somewhat greater degree of trust than the middle group (Figure
More 66

64

..~ 62
0
60
j
,!j
58

56

Less 54

Age Cohorts
Younger Older

FIGURE 6.8 Trust index by age

More 6

q
O Ufe insurers

...~
0 Health insu(ers
5
Auto Insumrs

Less

Age Cohorts
Younge.- Older

FIGURE 6.9 Trust of insurers by age

7
More

..
·~
2

....
0

..
c
5

.:1 4

Less
2

Age CohortI
Younger Older

FIGURE 6. 10 Trust of data handlers by age

THE SOClA.L ORIGINS OF VIEWS ON PRIVACY

6.8). However, by examining the age-linked distributions of trust for a subset of

the organizations, an experience-based interpretation of this relationship presents
itself.
The most consistent curvilinear pattern is with the insurers (Figure 6.9). The
decline in trust reverses or levels off at the fourth cohort, involving respondents
thirty to thirty-four years of age. The younger cohorts' trust in these insurers is a
reflection of their level of interaction with these organizations that require forms,
tests, and blanket permission for investigations into their background and status.
The continuing need for new automobile insurance subjects persons of this age
group to the full weight of surveillance and the risk that something in their files
will operate to deny them this service. This analysis, of course, assumes that, on
balance, all interactions with insurance providers are negative experiences. The
improvement in trust is not related to any change in insurer practice but is rather
a tendency of respondents to forgive and forget negative experiences further in
their past.
The same is not true for most of the other data handlers identified in Figure
6.10. The level of trust of direct marketers declines steadily with age, reflecting in-
creased marketing pressure organized in pursuit of the increased wealth of the
older cohorts. The explanation for the results for the telephone companies and
credit bureaus is similar to that offered for the insurance providers. The older co-
horts have limited experience with making applications for telephone service or
seeking credit for major purchases.

The Media and the Information Environment

One of the most powerful sources ofexplanatory power .in the t990 dataset is to be
found in the response of individuals to a quest.ion about their exposure to infor-
mation about the panoptic sort. The Harris/Equifax survey asked respondents to
indicate how much they had "heard or read during the last year about the use and
potential misuse of computerized information about consumers." Responses were
coded into five categories ranging from i (a great deal) to 4 (nothing a tall). A frac-
tional proportion of respondents (eight people) indicated that they were not sure,
and, for the purposes of this secondary analysis, the responses were recoded to as-
sign those eight cases to a middle category of uncertainty or indifference. Then re-
sponses were reverse coded so that greater exposure received the higher score.
The relationship between exposure and a generalized concern about personal
privacy is highly significant, linear, and direct. The more people have heard or
read about the use and potential misuse of computerized information about con-
sumers, the more concerned they are about threats to their privacy (r = .13, p <
.001). The relationship is even stronger (r = .15) with regard to concern about the
sale of personal information by the list industry for use by businesses marketing
goods and services. A similar relationship appears with regard to the summary in-
dex of trust discussed earlier, except that the relationship is inverse: the less you
More 66

64

..e
t:

0
62

60
i.....
"' 58

56

Less 54

Age Cohorts
Younger Older

FIGURE 6.8 Trust index by age

More 6
0
0 Life fnsurers

Health Insurers
5
AuJo Insurers

Less

Age Cohorts
Younger Older

FIGURE 6.9 Trust of insurers by age

7
More

6 ! ·..-•. 6
Telephone companies
.....6 ' ."
··b.:•·...-·•·;tt-·•
...!
Empla)'ers

....
0 5
o...o
..·a..
&·&- •
~

..
c
.!l 4
·o ···o..
'·d···
lJ'··D Credit bul'eaus

Less
3

2
~"'"'~"'~
Age Cohorts
Youn_ger Older

FIGURE 6. 1 O Trust of data handlers by age

 THE SOCIAL OfUGINS OF VIEWS ON PRIVACY

have heard or read, the more you trust organizations that collect and use informa-
tion about consumers (r = - .14).
Alan Westin proposed an index measure that combined responses to four
questions: extreme concern about threats to privacy; a belief that businesses seek
excessively personal information from consumers; a belief that the federal govern-
ment still invades its citizens' privacy, even after Watergate; and a belief that con-
sumers have lost all control over the circulation of their personal information. ~5
This measure is even more strongly linked to exposure (r = . 18). There was no ap-
parent relationship between being exposed to stories about information use and
abuse and respondents' opinions regarding Caller-ID. This suggests that the rela-
tively high level of opposition must have been based on an opposition to the tech-
nology more than an understanding or acceptance of the claims made by privacy
activists that the technology would be used to gather TTGI in the same way that
AN1 services gather information generated by calls to 800- and 900-number ser-
vices.
There is a tendency for those with the greatest exposure to information about
the panoptic sort to also be more likely to work with computers on their jobs.
Along these same lines, those who have the greatest exposure to this information
also tend to agree that privacy activists play an important role through their
"work to protect people's privacy by exposing abuses, bringing lawsuits, and
sponsoring legislation."
In the 1989 survey, several questions were asked that allow us to conduct an in-
vestigation of the influence ofexposure in somewhat more detail. George Gerbner
and his colleagues, working within the tradition of cultivation analysis, have
tended to focus on television as the primary source of media influence. 36 The cul-
tural indicators perspective, which underlies much of this work, assumes that a
common representation of social reality pervades all that television presents.
Thus it matters very little what particular programs a person views; it matters only
how much television is viewed. Interviewers gathered a measure of overall expo-
sure to television in response to a question that asked: "What is the average num-
ber of houJ'S that you watch television on a typical weekday?" Responses were
measured to the quarter hour, with the mean response computed as 3.01hours.37
Because there remain good logical and empirical reasons to believe that differ-
ences in people's und~rstandi.ng of some of the complex issues related to the
panoptic sort will be derived from, or at least associated with, their exposure to
discussions of issues not common to entertainment fare, interviewers asked addi-
tional questions about more specific viewing. They asked respondents to indicate
how often in the last six months they had watched "programs on television having
to do with politics, government, or public affairs:' Responses were placed into one
of four categories from i (never) to 4 (frequently). Interviewers asked a similar
question with regard to newspapers. 38
An analysis of the correlations suggests that, although there is little media-re-
lated difference in respondent positions on the need for banks, finance compa-
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

nies, and credit bureaus to do "more to keep personal information confidential:'

the tendency between the selective and nonselective measures is diametrically op-
posed with regard to respondent beliefs about business practices.39 These data
would appear to support an assessment of television's hegemonic role in that
greater exposure is associated with greater agreement with the statement: "The
more businesses know about me, the better they can meet my individual needs:'
Those who pay more attention to public affairs on television and in the press are
more likely to reject that view. Similarly, the influence of general television view-
ing is even stronger with regard to agreement with the statement that "companies
rarely gather more information than they need to make good business decisions."
Those reading more about public affairs tend to reject that view.
A similar divergence is seen with regard to the claim that "the only people who
are concerned about their privacy are people with something to hide." Apparently,
the more people watch television in general, the more they are likely to accept
such a view; the opposite is true for those who pay more attention to public affairs
in the press, or even on television. At the same time, it should be noted that those
who view more television estimate the extent of public concern about privacy to
be higher than those who watch less, and this tendency is opposite to that which
characterizes those who watch public affairs programs. 40
A causal interpretation of these data associates general television exposure with
a hegemonic view of business as good, efficient, and fair. Recognition of a wide-
spread public concern about privacy is modified by an interpretation of that con-
cern as being unjustified- a reflection of guilt. The only association that is incon-
sistent with such a business-oriented view is the tendency among the heavier
viewers of television to express an interest in a device that would screen out tele-
marketing calls. One interpretation, of course, is that heavy viewers resent the in-
terruptions that such calls produce in their viewing, rather than reject business or
marketing in general.

Political Interest, Involvement, and Ideology

The literature in recent years has begun to question the utility of traditional mea-
sures of ideological self-identification.41 A bipolar scale ranging from radical to
conservative may be seen to have been influenced by long-term trends in the over-
all political climate, which reflect an increasingly conservative tendency, as well as
by particularly salient assaults on one pole or the other.42 The attack on "liberal
ideas;' which characterized the Bush/Dukakis campaign, may have influenced the
willingness of respondents to identify themselves as liberal, despite John
Fleishman's finding that such influences were difficult to demonstrate. 43 In the
1989 survey, responses to the question that asked. for the identification of the re-
spondent's political orientation indicated that the current tendency was indeed a
movement toward conservatism. On a 10-point scale, in which i was conservative
and 10 was liberal, the mean was 4.8, with nearly 70 percent scoring s or below.
166 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

Three measures, derived from the terminal values inventory developed by Mil-
ton Rokeach, were in duded in the 1989 survey as an alternative index of political
orientation. 44 Interviewers asked respondents to indicate how important they
thought each of three values was to them; 1 was equal to being not at all important
and 10 was equal to being very important. The most important of the three for the
persons in this "conservative" sample was Freedom, with a mean of 9.6. Not sur-
prisingly, Equality was less important, with a mean of 8.7. Equality is a tradition-
ally liberal value, and it continues to provide a basis for distinguishing respon-
dents in terms of their privacy preferences-especially with regard to their views
on the need for legislative action.
Interviewers measured people's interest in politics and public affairs by a ques-
tion utilizing a simple 1-10 scale, in which 10 represented very great interest. The
interviewers asked respondents to "rate your interest in politics, government, or
public affairs." The mean of 7.1 suggests a moderately high interest in politics in
this sample. By the use of two additional questions interviewers sought to deter-
mine if that interest was reflected in the use of the mass media for information
about politics. One question asked about the frequency with which people
watched "programs on television having to do with politics, government, or pub-
lic affairs" in the past six months. Somewhat less than 25 percent indicated that
they never or rarely watched such programs, whereas nearly 40 percent indicated
that they did so frequently. When asked to indicate how much "attention they pay
to stories about politics, government, or public affairs" when they read the papers,
21 percent indicated that they paid little or no attention to such stories. ·
Interviewers asked questions about seven different items, each reflecting a dif-
ferent form of involvement or commitment to a "social problem, political issue, or
political candidate." These items were; marched or demonstrated; signed a p~ti­
tion; wrote a letter; attended a meeting; joined an organization; donated money;
and sought additional information. Interviewers asked respondents to indicate
whether they had or had not taken each of these actions. The activity reported
most frequently was the seeking of additional information. But only 55.1 percent
of the respondents indicated that they ever made such an explicitly purposeful
search. The signing of petitions and the donation of money in support of causes
appeared to be the next most popular form of involvement. Only 4.4 percent of
the respondents indicated tliat they recently marched or demonstrated, and.only
18.7 percent indicated that they had joined an organization. Only eleven respon-
dents, on the one hand, could be identified as very active, in that they indicated
that they took each of the seven actions in the past year. On the other hand, 198, or
nearly 20 percent of the sample, indicated complete inaction, in that they had not
taken even one of tliose actions in recent memory. An activism score, computed as
the simple sum of the activities reported, will be discussed as an important corre-
late of privacy perspectives.
Table 6-4 presents the relationships between the explanatory variables I have
discussed in terms of their bivariate relations witli particular orientations toward
 167

TABLE 6.4 Correlations Between Explanatory Factors

Variable 2 3 4 5 6 7 8 9
8
1. Watchpol .48 .01 .308 .14a .20a - .08 -.05 .01
2.Readpoli -.11 8 .41a .21a .23a - .09a -.04 .10a
3. Avghrstv - .178 - .27a .07 .11a .128 - .03
4. Activist .368 .06 - .06 - .05 .09
8

5. Education - .06 - .098 - .06 .01

6. Cohort .04 - .088 .01
7. Female .04 .128
8. Black .04
9. Equality
8
p< .01

Key:
1. Watch politics and public affairs on television
2. Read politics and public affairs in newspapers
3. Average hours of television viewing
4. Index measure of activism
5. Educational attainment
6. Age cohorts, younger to older
7. Being female, dichotomous variable
8. Being black or African-American, dichotomous variable
9. Rating of importance for equality, low to high

TABLE 6.5 Multivariate Influences on Privacy Orientations (hierarchical regression)

Criterion Variables
Predictors Public Concern Privacy Laws Privacy and Guilt
1. Class - .016 .004 .016
2.Gender .0708 .0638 .098b
3.Age .004 .084b - .041
A-sq change .022c .018c .Q10a
4.Education - .1 45c - .0738 .0668
A-sq change .028c .002 .017c
5.Rokeach -.0738 -.041 -.082b
6.ldeology -.004 .033 -.025
A-sq change .005 .003 .011b
?. Interest .017 .028 - .033
8.lnvolved .022 - .038 -.109b
A-sq change .001 .oogb .019c
8
9. TV hours .080 .005 -.0798
8 8
A-sq change .005 .000 .006
10. Watch pol TV -.067 - .008 .061
11. Read politics .033 .122b .065
A-sq change .003 .D10b .oo8b
A-squared (adjusted) .045c .033c .061°

Beta = Standardized betas, all variables in the equation

8
p < .05, bp < .01, cp < .001
168 THE SOCIAL ORIGJNS OF VIEWS ON PRIVACY

the panoptic sort. The relatively high correlations within the table help to identify
structures of similarity and clifference within the sample. For example, those who
tend to watch political or public affairs programs on television are also more likely
to pay attention to political issues in the press (r = -48). However, while there is no
clear relationship between watching public affairs and watching television in gen-
eral, those who do attend to politics in the press would appear to limit their televi-
sion viewing to politics as well, as the relationship between such reacling and gen-
eral television viewing is negative (r = - .11). Those who watch political television
also tend to be relatively more involved, or activist, than others (r = .30), but the
relationship between media use and activism is considerably stronger with regard
to print (r = .41). The finding that this selective media use also characterizes the
older and more highly educated cohorts is consistent with the contemporary liter-
ature. It is also unfortunately the case that women and African-Americans are less
closely attuned to politics and public affairs and tend to spend more of their spare
time with general television fare. The value placed on equality was generally quite
high, and the variance was slight. There is, however, a solid basis for understand-
ing the tendency for women, activists, and avid readers of politics to favor equal-
ity. The absence of an association with being African-American is a bit more
problematic.

Multivariate Analysis: i989 Survey

Table 6.5 presents the results of an effort to examine the combined influence of
variables that had been demonstrated to be related to several different orienta-
tions toward the panoptic sort. A hierarchical regression design allows an analyst
to enter variables into an equation in ari order reflecting underlying theoretical
assumptions about causal order. Variables may be entered singly or in groups rep-
resenting a class of variables or influences. The coefficients labeled "R-sq change"
estimate the proportion of the variance in the criterion variables (ranging from o
to 100 percent) that is "explained by" or statistically associated with the variables
just entered. At each step, the change in the R-squared value inclicates an addition
to explanatory power beyond that provided by variables already in the equation.
Of particular importance in such an approach is the fact that the presence of
variables in the equation statistically removes or partials their influence from that
of other variables in th~ model. That is, such an approach, for example, allows an
estiinate of the importance of educational attainment, after the common influ-
ence of gender on both the criterion variable and education bas been removed sta-
tisticaily. Although this is not an explicit estimation of the importance of the
interaction between education and gender, it increases our confidence in the
R-squared value as a measure of explained variance.
in using the regression model, in which the influence of gender can be statisti-
cally controlled, we can more confidently utilize all t,250 cases without weights.
This is especially important when meaningful interpretation would require the
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

elimination o(all cases that did not have complete data for all variables used in the
analysis.
The following were selected for use as independent, or predictor variables:

i. Class: An ordinal measure, treated as interval, based on the employment

status of the main wage earner. Work for hourly wages was coded 1, sala.
ried compensation was coded 2, and self-employment was coded 3.
2. Gender: A dichotomous variable; male= 1, female= 2.
3.. Age: A cont~n.uous measure, the respondent's reported age at last birth-
day.
4. Education: An ordinal measure, treated as interval, ranging from " no
formal school" = i to "postgraduate education" = 8.
5. Rokeach: A computed ratio score, expressed as the value of Freedom di-
vided by the value of Equality.
6. Ideology: An interval measure, ranging from 1 to 10; i = conservative, 10
=liberal.
7. Interest: An interval scale, measuring reported interest in politics; i =not
at all interested and 10 = very great interest.
8. Involved: A computed factor score. The output of a factor analysis of the
seven variables utilized earlier in the simple additive index of Activism.
The initial variables were dichotomous, with i =having taken the action
and 2 = not having taken the action. Thus, in this analysis, utilizing fac-
tor scores, lower scores represent greater involvement, or activism.
9. TV hours: An interval scale, m.easuring average daily television viewing
to the nearest quarter hour.
10. Watch political TV: An ordinal scale, treated as interval, measuring view-
ing of public affairs programs on television; 1 = never and 4 =frequently.
u. Read Politics: An ordinal scale, treated as interval, measuring the atten-
tion paid to public affairs material in the newspaper; i = none and 4 =
quite a bit.

Three variables were selected for use as criterion (dependent) variables:

i. Public Concern: This interval measure is the projective index of respon-

dents' concern, reflected in their estimation of the extent of concern
among average Americans about "threats to their privacy." Alth·o ugh
this, as with the other measures, is highly skewed, with 33 percent of the
respondents choosing io as their response, there is substantial variance
to be explained, as is suggested by a mean of 7.56 and a standard devia-
tion ofz-42.
2. Privacy Laws: Th.is is a simple additive index composed of the sum of the
scores for each of the nine areas in which respondents could indicate the
170 THE SOCIAL ORIGINS OF VIEWS ON PRIVACY

extent to which they thought that there was a need for more strong laws
to control the sharing of personal information (alpha = .85).
3. Guilt: This is an ordinal measure, treated as interval, that reflects respon-
dents' agreement with the view that privacy concerns are limited to peo-
ple with something to hide. As the responses are coded, 6 = strong
agreement and 10 = strong disagreement. This measure is also severely
skewed toward rejection, with only 15 percent of the respondents report-
ing agreement.

In each of the three models in Table 6.5, the equations are statistically signifi-
cant, but the amount of variance explained is quite small. Because there is less
variance in the criterion variable, it is not surprising that more of the variance in
Guilt is explained than is one's estimate of Public Concern or one's preference for
Privacy Laws. What is more important at this point is the importance of the pre-
dictor Yaria'bles in each of these models.
In each model, the coefficient for gender is statistically significant. This can be
interpreted to suggest that after we control for age, education, political interest,
involvement, and even television exposure, there remains an independent contri-
bution of gendered experience to an indiYidual's orientation toward privacy as it
has been measured.
Age, so important in many of the bivariate relationships discussed previously,
remains a significant factor only with regard to the preference for a regulatory re-
sponse. Because the variables are not independent but are significantly correlated
with each other, caution is needed in interpreting the magnitudes of the beta coef-
ficients as indicating their relative importance as explanatory factors. This is even
more of a problem when the low R-squared measure 'suggests that the model is
underspecified. Still, the sign of the coefficient for Education suggests that less
rather than more education leads one to seek protection of one's privacy interests
in the legislature. The larger, positive coefficient for Read Politics can be inter-
preted to mean that in general, expos we to press coverage of political issues lends
support to a view that the legal system can meet the public's needs in this area.
The negative sign for Education emerges again in the model with estimates of
Public Concern about privacy. When all the measured influences are controlled,
those with less education tend to estimate the level of public concern to be higher
than those with more education. Thus, those with less education are more likely
to see privacy as a widespread concern and to see regulation as a viable option.
The smaller but positive coefficient for television hours suggests that a representa-
tion of the life of the average American as one in which privacy is of great concern
may be provided by exposure to teleYision in general. Thus, with all other mea-
sured variables considered equal, including Education and Class, the more time
respondents spend watching television, the greater is their estimate of the threat
to privacy.
THE SOCIAL ORIGINS OF VIEWS ON PRIVACY 171

The importance of television viewing is also reflected in the model evaluating

acceptance of the claim that privacy reflects Guilt. The negative sign for the coeffi-
cient reflects the wording of the question and therefore should be interpreted to
mean that the more one watches television, the more one accepts the linuting
view. Whereas more education is associated with rejecting that view, when educa-
tion and other measured variables are held constant statistically, increased televi-
sion viewing is associated with the acceptance of guilt How might this relation-
ship be interpreted?
In both the Concern and the Guilt models, the Rokeach var iable is statistically
significant. Recall that the Rokeach variable is the ratio of Freedom to Equality.
Thus the higher the value of the Rokeach variable, the more freedom is valued rel-
ative to equality. The negative sign of the coefficient for Rokeach in both equa-
tions should be interpreted to mean that those favoring freedom more than equal-
ity tend to see that privacy is less of a problem than those who value equality
somewhat more, and that those favoring freedom also tend to agree that the pur-
suit of privacy is an attempt to keep others from knowing about activities that are,
at the very least, not in public favor. The political involvement factor suggests that
those less involved in political activities are also more likely to accept the narrow
view. Thus television viewing is associated with a socially conservative posture,
but it may also be seen to contribute to a view that so many people are concerned
about their privacy because they are doing so many things about which they share
a sense of guilt. These are things that they would not be doing if others could easily
find out about them. This is a powerful indication that the lower social classes
have agreed to play according to the rules of the game as long as this game is being
played under the watchful gaze of a panoptic system. And, noted earlier, heavy
television exposure is associated with a mistrust of power (r = -.12) and a sense of
powerlessness (r = - .08). The heavy viewers are in a "mean world"- they perceive
that their personal freedom is limited, and they apparently have little hope of do-
ing anything about it.
Although the previous analyses make some assumptions about the nature of
the experiences that individuals have in their interactions with various organiza-
tions, the data gathered in i990 allow exaniination of the relationship between
negative experiences and measures of trust or concern. Interviewers asked three
questions about respondents' experience with automobile insurers, health insur-
ers, and credit bureaus. One question asked whether respondents had ever "been
refused health insurance or been told that you had to pay a higher rate because of
your health or the amount of your medical bills." Both refusal and being asked to
pay a higher rate were coded as i; otherwise, the response was coded as o. A similar
question asked if the person had "ever been refused auto insurance or been told
that you bad to pay a higher rate because of your accident and claims record."
Again, refusal and higher rate experiences were coded as 1. The final question was
a simple indication of whether the respondent had ever been refused credit. This
variable, called History, was then used in combination with other predictors. Only
172 THE SOCIAL ORIGINS OF VIEWS O N PRNACY

thirty-eight respondents reported having had all three kinds of negative experi-
ences. Some 261 reported having two, and 714, or approximately 32 percent of the
respondents, reported having had at least one such experience.
Analyses of variance indicate that History is a significant factor in explaining
differences in Trust, with the least trust expressed by those with the most negative
history (r = - .10). It is also the case that those with the most negative experiences
express the greatest general concern about the panoptic sort. Indeed, the data sug-
gest that those who have been "burned" most frequently are also most likely to try
to avoid further rejection,.as is indicated by their greater tendency to report hav-
ing failed to apply because they did not want to be asked more questions (r = . 15).
When the differences in level of trust among members of different social
classes are examined, there is a critical difference between the influence of knowl-
edge from third parties and knowledge from direct experience. The cultivation
hypothesis associated with George Gerbner contains a variant called "main-
streaming:' which suggests that although concrete social experience may set peo-
ple apart in terms of their understanding of the world, sharing the common expe-
rience of television brings them together in the mainstream. 45 The data in support
of this hypothesis reveal that groups, such as whites and African-Americans who
differ about perceptions of crime in the low television condition, will share a simi-
lar view of crime in the high television condition. This similarity is usually the re-
sult of one group's reducing the extremism of its views.
Exposure to information as measured in the 1990 survey is not precisely a mea-
sure of media exposure. We cannot assume that respondents were exposed to the
same information in the same way that we can if we ask about their viewing of
television. Contrary to the expectations of the mainstreaming hypothesis, the four
social classes are not brought together through increased exposure to information
about the panoptic sort (Figure 6.11). Greater exposure is associated with a lower
level of trust for all social classes, but convergence does not apparently occur.
Concrete social experience appears to be operating more forcefully, as is indicated
in Figure 6.12. Where the classes are relatively close together in the condition of no
negative experiences, there emerges a dramatic class division as negative experi-
ences cumulate. First, no one from the lowest class was counted among the thirty-
eight respondents who were so privileged as to have had three negative experi-
ences. For the poorest of the poor, two strikes are all one is entitled to. However,
the class divisions are dramatic, which may indicate that the absence of ready al-
ternatives suggests that the actual costs of rejection as experienced by the poorer
respondents are much higher than we find in the simple measure of History as
event.
Figures 6.13 and 6. 14 show that neither exposure nor direct experience serves to
bring together groups differentiated by life cycle stage. Three age groups, young
adults (ages 18-29), economically active adults (ages 30-49), and mature adults
(ages so+) respond to different levels of exposure as well as to their negative expe-
riences with service providers. For all groups, exposure and experience are associ-
 173

More

··- .-.. ""·

...
"'
:I
60
,
-, ··. . __.......... .. "'
· '·- ·~
··.. .. .. ... . .....
...j!:· ···" ·· ~ .... :..~ . ~ Lower
0
...c ....···.• . """:.: n ..... ...
~
..
··..... . Working

.
··&· Middle
...
so ···.
··
-···- ~per

Less 40-1--~~~~~~~-....~~....-~~-.-~~..--~--...~~~
0 z 3 4

E.x tent of exposure ~o information

FIGURE 6.1 1 Trust, class, and exposure

70
, ....
... '·
"'
:I 60
:i
~
0
....
..
c
................. ··..
·... ... ····• ···
Lower

Working

....~ so
·······.., • • ·.l!.
•• '11·. Middle

40
··..·.
··. . ..
-···- Upper

Less 30~~~~~~~~~~-...~~~~~,...-~~~~~

0 2 3

Number of negative experiences

FIGURE 6.12 Trust, class, and history

 174

..
....
~
2
0
60
·-·-·-· .... ~,

...c .................... :.............·. ...... ,~~ • - • - • - • - ·t:,. Young adult

e..
.... ·.·.. Mature adults

·\·.•..
50 ·.·.
···. . ..
Ecoilomi<:'lllY active adults

Less
404---~------~----~----....-----~----.-----.-----~
0 3 4

leve l of exposure to Information

FIGURE 6. 13 Trust, life cycle •. and exposure

Mono

··-. ...-·-·-·--
60
tJ. ..... ••• .. ·····t:x
t:
::l .............
• . .__ . ""
u···
;
0
.... Maturi adults
.,c 50
\

e Young adults
.!l \
..
\

40
9 Economically active adults

Less 30.l.---~--~--------~--~--~--~----~--~--~
0 2 3 4 5

Number of negat ive experiences

FIGURE 6. 14 Trust, life cycle, and history

THE SOCIAL ORlGfNS OF VlEWS ON PRIVACY 175

ated with declining trust in institutions. The anomalous behavior of the mature
adults, in whom the extremes of experience and exposure are not associated with
the same loss of trust as the other cohorts, might be explained in two ways. The
first, as noted earlier, is that class position provides some security and access to al-
ternatives. The second, which is difficult to test with the data as measured, is that
the qualifying experiences occurred earlie1· in their economic lives; this influence
has dissipated, and the exposure to the horror stories in the press has less )Veight
because they are no longer subject to the risks so described.

SUMMARY
The factors that are involved in the development ofan individual's response to the
panoptic sort are many, and their relationships are complex. The data analyzed in
this chapter describe a dramatic increase in the public's awareness and concern
about what are considered to be invasions of privacy. The differences among
groups of people identified on the basi~ of their race, gender, age, or social class
position appears to be related to the differences in relative power with regard to
organizations that those positions represent. These differences appear to change
dramatically with the relationships that become most salient at particular stages
in an individual's life cycle. Yet not all of the attitudes seem to reflect direct per-
sonal experience. The mass media play a critical role in informing people about
the risks and dangers that people like themselves face in their relations with the in-
stitutions of business and government. A critical role played by the mass media,
especially television, is the reinforcing of a compliance with the dominant values
that legitimate the operation of the panoptic sort.
7
A DATA PROTECTION REGIME

NEED WE INVOLVE THE LAW AI ALL?

It is important to recognize the difference between the law as an ideal, a moral
structure, and the law as a weapon, a resource. lt is the strategic use of the law and
the legal system by goal-oriented actors that introduces contradictory forces,
which is part of the structuration to which Giddens refers with regard to social
systems in general. Far too frequently, we tend to talk about the law as the even-
handed instrumentality that ensures that the WJiversal values claimed for any so-
ciety are protected and guaranteed. This idealized system is no less an illusion
than the ideal of the competitive marketplace, which is thought to operate auton-
omously to produce optimal social results. lt is important to understand those
circumstances in which assumptions about the law cannot stand. Many of the
problems with the law as a system are similar to the problems that have been iden-
tified with regard to the market. They are problems rooted in the vast inequality in
power and influence that operates in a system that depends on the assumption of
equality.
The law can be, and has been, used as a weapon, a machete with which to hack
away enough brush to prnvide some breathing room, a protected space ·within
which to pursue personal interests. This use of the law is to be distinguished from
a notion of the law as an expression ofa basic moral vision, a set ofguiding princi-
ples that serve as foundations supporting relations among persons. The adver-
sarial nature of legaJ practice and the resultant ends-means rationale have con-
tributed to the disruptive, contradictory, and nearly unmanageable morass of
theory, regulation, and case law that applies to the definition and preservation of
an individual's rights in relation to the operation of the panoptic sort.
Understanding the dialectical evolution of the law requires a recognition that
the process of structuration always.reflects the actions of persons acting in pursuit
of tJ1eir interests, and it always reflects the significant differences in power among
these actors as they act and react to the unexpected and unintended consequences
of actions taken or restricted in the past.
Argumentation in the context of the pursuit of tort relief and compensation
must be understood as strategic communication. As Habermas reminds us, stra-

177
178 A DATA PROTECTION REGIME

tegic communication is not pursued in the interest of understanding, but rather

in the interest of private gain.
However useful the pursuit of private gain might be in explaining the behavior
of individuals and their associations, it does not provide a sufficient explanation
of the behavior of judges, legislators, and administrative bureaucrats whose daily
decisions are also involved in the reproduction and transformation of a legal re-
gime. Ongoing debates about the relative autonomy of the state and the pursuit of
self-interest by policy actors provide more questions than answers about the
forces that determine the structure of laws and regulations that define privacy. 1
Explanations of the development of the legal regime with regard to personal infor-
mation and privacy are made even more difficult because of the specific nature of
the issues to be addressed. It is hopedl that this chapter contributes to our under-
standing of these issues by breaking them into parts that reflect the competing and
contradictory interests involved.

PERSONAL PRIVACY AND DATA PROTECTION:

A TANGLED WEB
According to Richard Hixon, "privacy, in concept and by tradition, is a basic hu-
man right; as a legal entity based in social reality, however, it is ambiguous at best.
The concept is important, yet it has acquired different, at times contradictory
meanings. "2 In this chapter, my emphasis will be on the development oflaw and
regulation with regard to informational privacy, and within that general category,
I will pursue an interest in the regulation of the behavior of private firms. It is in
the area of private corporate action that the law is most in need of attention.
The interests that are evoked by the concept of privacy are many and confused.
The challenge in understanding how particular forms of social practice have de-
veloped to protect these competing and perhaps contradictory interests involves
an attempt to represent the similarities and differences between the underlying
factors in ways thatare most helpful in understanding the laws, the machines, and
the bureaucratic systems that have emerged to protect those interests. The analyti-
cal challenge and logic are similar to those leading to the use of factor analysis. I
seek to identify the underlying factors that define a conceptual structure in which
these factors relate to each other and to the ideas, actions, and technologies that
help to define them. Just as with factor analysis, there are different rationales for
choosing different solutions, based on assumptions about the underlying struc-
ture, such as whether they are correlated with each other or are orthogonal. These
assumptions force a certain kind of thinking about the construct
One line of tension that emerges clearly in this conceptual factor analysis is that
between the individual and the collectivity, between the public and the private. Let
us ask ourselves if aU the rationales for providing or guaranteeing the private are
antagonistic or opposed to the public or collective interest; or, can we find in the
rationale for the protection of privacy the same underlying assumptions and in-
A DATA PROTECTION REGIME 179

strumental rationality that justify the protection of individual interests in intellec-

tual property. With regard to copyright, patent, and other forms of protection for
individual rights to intellectual property, a contribution to public knowledge is
the underlying goal. for the public interest in knowledge and creative works to be
served, it is argued, individuals must believe that they will receive just compensa-
tion for their efforts at invention and creative expression. Similarly, the operation
and sur.vival of a vibrant democracy may be specified as a goal that defines active,
informed participation by its citizens as a necessary prerequisite. The autono-
mous development of that citizenry can be argued to depend on the protections of
personhood and individuality that privacy describes.
Alan F. Westin has been credited with making the most comprehensive analysis
of the variety of interests that are part of the complex generally referred to as pri-
vacy. The most central concerns about what has been called informational privacy
were explored in his Privacy and Freedom. 3 Essential to Westin's definition is the
emphasis on the desire, claim, or right of individuals to decide for themselves the
extent to which they will be exposed to others. This exposure includes their physi-
cal person, their attitudes, their opinions, and their activities. Since i967, literally
hundreds of commentators have sought to amplify, extend, and explain the com-
plex of rights that form this cluster. David Flal1erty offers a list of thirteen different
privacy interests, which I suggest would form two identifiable but conelated fac-
tors. Factor 1, the control of access, would involve the right to be let alone; the
right to a private life; the right to limit accessibility; the right of exclusive control
of access to private realms; the right to minimize intrusiveness; the right to enjoy
solitude; the right to enjoy intimacy; the right to enjoy anonymity (the right to
lin1it identification); the right to enjoy reserve; and the right to secrecy. Factor 2,
th~ right to control disclosure or representation, would involve the right to con-
trol information about oneself an<\ the right to expect confidentiality. 4 The chal-
lenge in understanding the importance of these rights is to explore the contribu-
tions that control over access and representation make to the development of an
autonomous individual.

Autonomy
Hixon's discussion of the long-term evolutionary changes in the meaning of the
term privacy places special attention on the changes that emerged between the
sixteenth and nineteenth centuries.5 In this period, privacy came to be associated
most clearly with a sense of individual, personal privilege, not altogether unre-
lated to the privileges associated with class position. The importance of this privi-
lege in relation to the demands or requirements of the state on an emerging bour-
geoisie is a distinction that ought not be lost. Barrington Moore, Jr., suggests that
the desire for privacy emerges fundamentally when society places obligations on a
person that this person cannot or does not want to meet. He suggests that until
the end of the nineteenth century, only the propertied and employing classes were
able to claim and enjoy any mea~ure of privacy.6
180 A DATA PROTECTCON REGIME

Part of our understanding of the meaning of autonomy involves an individual's

freedom to determine how she or he will respond to options and challenges. Au-
tonomy is an ideal. There are always Limitations and constraints. Many of these
limitations are sociocultural. The extent to which ati individual conforms to the
expectations of others is the extent to which she or he has given up some auton-
omy. The panoptic sort threatens the autonomy of the individual by increasing
the range of activities that may be brought under the watchful eye of significant
others. If every action generates a record and if this record is available to a wider
and wider sphere of interested, or potentially interested, others with the power to
make decisions about a person's options, the freedom of action becomes self-lim-
iting. The limitation on autonomy that derives from the belief that one is being
watched is not restricted to formally illegal behaviors. Depending on the individ-
ual, the constraint spreads to include behaviors that may be mere.ly questionable
or out of fashion. The limitations on :autonomy that characterize the panoptic sort
are evermore broad because it is unclear just who is watching and what their in-
terests and standards might be. -
Individual autonomy may be realized in part through the maintenance of dis-
tinguishable personae that a person may adopt as the situation may require. These
varying personae will be reflected by differences in dress, language, posture, and
dynamism that are understood to be appropriate to a setting or an interaction. We
can find a basis for understanding the importance, for an individual, of the ability
to keep spheres and personae relatively independent of each other by considering
the concept of cognitive dissonance. 7 The anxiety that people feel when they are
forced to confront conflicting or incompatible beliefs, values, or opinions oper-
ates to restrict their effective functioning. Yet it is dear that individuals have devel-
oped complex mental structures or schemas that allow logicalJy incompatible un-
derstandings to coexist, primarily because they are rarely called into play at the
same time.
The operation of the panoptic sort threatens the ability of the individual to
maintain these structures of personality independent of each other. Their repre-
sentation in profiles is structured externally by others according to the others' in-
terests rather than their own; any apparent incompatibility within the structure is
resolved through a process of classifi.cation, which is necessarily limiting because
it cannot approach the flexibility with which individuals can shift between their
operational selves. Restrictions on the freedom of indivi9uals to make these shifts,
to adopt the personae they believe to be appropriate to the circumstance, are thus
linked to a loss of control over information about themselves.
Frequently, the level ofautonomy in the postmodern age is compared with that
which was available in preindustrial times or in isolated rural communities. It is
not that the extent of constraint is necessarily any greater than it has been at any
point in the past, but that it is very different in terms of the visibility of constraint.
The power is diffused invisibly, and is often exercised without the knowledge of
the persons so constrained. It is this diffused, distanced, and semiautomatic qual-
A DATA PROTECTION REGIME

ity of panopticism that makes it so difficult for individuals to challenge power on

the basis of an analysis of its operation.
Tt is also important to recognize that limitations on individual autonomy are
not entirely imposed from without. The efficacy of control is determined in part
by the extent to which it has been internalized. It is a deeply vexing issue that sur-
rounds the difference between external and internal constraint. A constraint is no
less limiting if people apply it to themselves. If we truly believe that autonomy-
freedom of action and freedom of choice-is a primary value, even internalized
restrictions on the range of options must be seen as a limitation on that value, if
the process through which the internalization took place was one in which op-
tions have been limited by the exercise of power. As noted, differential awareness
of the panoptic system will be reflected in different responses to its constraint. Tbe
growing number of persons who have decided not to apply for some product or
service to avoid revealing information about themselves represents only one of
many forms of response that reflect an internalized constraint.

Related Interests
All of us are not equally proud of all that defines us as a person. Our evaluation of
ourselves is based in part on comparisons with externally deveJoped standards.
We are able to enter into interaction with others to the extent that we can believe
that those aspects of our person that would not be looked on with favor may be
held in reserve, away from the public gaze. For many of us, the display of our bod-
ies represents the potential risk of embarrassment. No less is true of our thoughts.
Elisabeth Noe!Je-Neumann has identified a construct that suggests that people are
less willing to express their opinions in a manner reflecting the extent to which
they perceive their opinions to be at variance with the majority of those around
them. 8 We are concerned tbat we not reveal weakness to actual or potential ene-
mies. Access by others to those things we hold in reserve involves the potential loss
of control and the transfer of power to the other. Part of the definition of privacy
interests includes the ability to limit access by others to areas that we would prefer
to hold in reserve, protected perhaps as in "sanctuary."
David Bazelon, echoing the comments by Alan Westin made ten years earlier,
identified the nurturing of individuality, expressed as the right to be different, as
an aspect of autonomy.9 This individuality is realized through experimental self-
discovery, which cannot operate when a person is always under surveillance,
always being evaluated, always at risk. The pursuit of autonomous self-determina-
tion, therefore, requires a zone of privacy, which frequently means an actual phys-
ical space wherein experimentation might take p lace. This developmental process
depends on some degree ofseclusion, or isolation, areas in which a person can re-
tire with some confidence that her or his privacy will be respected. Gary Bosnvick
defines repose as "freedom from anything which disturbs or excites." 10
Another aspect of this process of autonomous development involves the ex-
pression of emotions and opinions and the exploration of intimate relationships
 A DATA PROTECTION REGIME

without the pressure of having to consider the opinions and assessments of others.
This space for relatively unhindered release is believed to be essential for human
survivability-part of the reproduction of the individual's capacity for labor and
interaction in the public sphere. It is, perhaps, the identification of the sexual with
this intimacy that privacy allows that has occasioned the problematic definition of
reproductive rights as privacy rights per se.
Reproductive rights, as part of the realm of sexual behavior, may be seen to be
part of the "zone of intimate decision," which is currently protected under the ex-
pansionary interpretations of the Fourteenth Amendment, which suggest that
only due process can restrict the right of individuals to make intimate decisions.
The decision of the U.S. Supreme Court in Roe v. Wade has fundamentally altered
how we think about privacy. 11 The sexual arena is clearly an area in which there
are greater, and perhaps more legitimate, "expectations of privacy, " although
there is danger in establishing orders of priority that are susceptible to change and
are responsive to influence. The issue of information privacy, however, should not
be focused on the limitations exercised by the state on what one might legitimately
do within this z.one of intimacy, but on how the state, or other uninvited party,
would come to know what is or is not being done. 12

THE HERITAGE OF WARREN AND BRANDEIS

Samuel D. Warren and Louis D. Brandeis are credited with having published the
seminal law review article in i890 that began the chain of theoretical agglomera-
tions that make up the privacy regime in the United States today. Entitled "The
Right to Privacy:' this article proposed an addition to the body of laws that pro-
tected person and property. As they argued, developments in the law involve a
continued expansion through the identification and specification of new rights,
such as the right to "enjoy life" as enabled by the "right to be let alone." They ar-
gued that the process of expanding the coverage of a regime of rights was inevita-
ble and was part and parcel of the expansion of civilization. Indeed, they sug-
gested that "the elasticity of our law, its adaptability to new conditions, the
capacity for growth, which has enabled it to meet the wants of an ever changing
society and to apply immediate relief for every recognized wrong, have been its
greatest boast." 13
One environmenta1 force demanding expansion of the law was thought to be
intellectual reasoning, which compelled recognition that protections for property
needed to be extended beyond the physical to the intangible-hence, the develop-
ment of rights to intellectual property through copyright, patent, and trade secret
protections. An opposite force was associated with the emergence of needs occa-
sioned by the development of technologies that threatened to limit the private en-
joyment of personal pursuits. ln the circumstance at hand, it was the populariza-
tion of instant photography and its use by photojournalists that Warren and
A DATA PROTECTION REGIME

Brandeis saw as pressing for the expansion of intangible property rights to include
this new right of privacy.
Warren and Brandeis did not limit their criticism to journalism per se, but they
clearly sought to challenge the commercialization of the private sphere of the indi-
vidual: "Modern enterprise and invention have, through invasions on his privacy,
subjected him to mental pain and distress, far greater than could have been in-
flicted by mere bodily injury. " 14 This hyperbolic claim was made in the context of
a discussion of the transformation of gossip into a trade "which is pursued with
industry as well as effrontery."
In their argument that developed their position on this new right, we can find
the seeds of the narrower right of information privacy developed by Alan Westin.
Warren and Brandeis reached to the recordl of common Jaw for the right of an in-
dividual to determine "to what extent his thoughts, sentiments, and emotions
shall be communicated to others." 15 This right of publicity thus formed the core of
their newly formed right of privacy. But unlike the exploitative rights that are in-
herent in the copyright and are realized on publication, the rights envisioned by
Warren and Brandeis were negative rights: the value of which lay in the ability of
the individual to prevent publication. It is the distinction between the economic
value derived from the exploitation of personal information and the personal,
sentimental, emotional, or other less pecuniary dimensions of value that is at risk
of being lost in contemporary discussions of a market for personal information
with compulsory licensure akin to that in the area of broadcast retransmission.
But more about that later. Itis sufficient to make it clear that Warren and Brandeis
thought that "the protection afforded by the common law to the author of any
writing is entirely independent of its pecuniary value, its intrinsic merits, or of
any intention to publish the same:' 16 The right, they argued, is therefore not a
right of property but a right of" inviolate personality:'
Warren and Brandeis also suggested a distinction between the protections and
property interests that might inhere in intellectual property by virtue of the
amount of labor expended in its production. This distinction would privilege the
products of artistic and intellectual labor. They suggested that if forced, one might
argue so as to demonstrate that the effort involved in conducting oneself properly
is at least as substantial as that involved in creating a work of art. Thus the claim of
substantial labor should also be set aside as making little contribution to a theory
of difference between claims of a right to restrain publication of information
about an individual from information produced by that indjvidual.
They suggested that a photographer who makes additional copies of pictures
taken of a client from a glass negative that she or he has is guilty of a breach of an
implied contract. 17 Although the courts have moved far from this view since 1890,
the argument still rings true-the photographer may have performed the labor to
produce the negative and may have come to possess the negative, but she or he has
done so only in the context of providing a service to the client. Subsequent use of
the image ought to be limited by law as well as by courtesy to those uses explicitly
 A DATA PROTECTION REGIME

noted or unambiguously implied by the initial contract The same, it might bear-
gued, should apply to any records (akin to negatives) that are generated by trans-
actions involving long- or short-term contracts. It cannot even be argued that the
client need specify the nature of the records kept. The subject of a portrait need
not know how glass negatives are to be made or even that creation of a photograph
requires a negative at all. The contract is for an image, the choice of the technology
is that of the professional, but the fact that one process produces a potentially use-
ful artifact should not by itself assign any rights to the professional to use the as-
pect of personality captured therein.
The cases of photographic invasion by journalists identified by Warren and
Brandeis differed from those in the case of portraiture, in that there is no longer
any necessity of participation, contractual or not, between the photographer and
the unwilling, and increasingly unknowing, subject. It was the ability of this tech-
nology to capnue images at will that required, in their view, the entry of a raised
legal fence or the threat of successfull tort action.
It should not be forgotten that Warren and Brandeis did not propose a righ t
without limit. Indeed, several limitations were suggested in the analogies they dis-
cussed in the common law of slander, libel, and intellectual property. They in-
cluded a broad exclusion for matters of "public or general interest, " which on the
face of it seems an opening large enough to drive a carriage through. If the distinc-
tion comes down to a definition of what it means to report or make public in the
interest of the general welfare, it could be argued quite easily today that t11e public,
which is the commttnity of credit granters or health insurers or automobile insur-
ers or perhaps the community of employers, has a legitimate interest in knowing
about potential customers or employers. Thus the only restrictions that would be
appropriate would be on knowledge or information that is unrelated to their ac-
tivities as members of that particular "interested public:'
The core distinction that is noted but not fully developed by Warren and
Brandeis is the interest in protecting the privacy of "private life." 18 The realm of
private life includes things that "all men alike are entitled to keep from popular
curiosity," and this realm is to be distinguished from the more public aspects of
life that are not the subject of publicity because the individual is not worthy of
public attention.
The underlying emphasis on publicity of the sort accomplished by the press
and other media of mass communication implied a multiplication of images-if
not tangible images, then images in the minds of the observers or readers of the
published report. Because of the technical limitation on the reach of the oral rep-
resentation of those same facts and impressions, Warren and Brandeis would not
find a cause for action if the publication were made through unaided speech. The
injury from oral gossip was thought to be so slight as to be unworthy of concern.
Some eighty-nine years after Warren and Brandeis made their historic contri-
bution, James Barron wrote to explain why the legal protections against the pre-
sumed excesses of the press never developed as the authors had intended. 1 ~ Rather
A DATA PROTECflON REG IME

than developing as a limitation on the "public disclosure of embarrassing private

facts" by the mass media, the tort of privacy had developed so as to emphasize
three other torts: (i) intrusion into a person's seclusion or private affairs, (2) pub-
licity that places a person in a "false light in the public eye," and (3) the appropria-
tion of the person's name or likeness for commercial or other advantage. After
noting the preference of the courts for the guarantee of a "free press" capable of
defining newsworthiness as broadly as it wished, Barron turned his attention to a
demonstration that Warren and Brandeis's initial representation of the severity of
the harm was a broad overstatement, reflecting not a principled defense against a
ruthless press but a hostile response of a hypersensitive individual (Warren) who
enjoyed the benefits of a skillful and willing second in the person of Brandeis (who
bad actually written the article out of friendship and indebtedness).20 Barron may
have succeeded in finding, in pique and petulance, a reason for the o riginal article
and its emphasis on the press, but he left for others the task of explaining how
other aspects of the privacy regime have evolved so expansively over the years.

A QUESTION OF DIGNITY
AND OTHER INTANGIBLE INJURIES
In a gallant, if only partially successful, effort to "put t he straws back into the hay-
stack," which the law of privacy had become by the mid-196os, Edward J.
Bloustein seeks to find the core rationale for the legal defense of privacy. 2 1 In the
process, Bloustein turns the authoritative statement of torts from Dean William
.Prosser inside out, by picking up the thread of theoretical difference that most had
dropped or missed entirely in their reading of Warren and Brandeis. Bloustein re-
minds us that the "concept of'property' was put forward by the courts as a fiction
to rationalize a form of legal relief which was really founded on other grounds of
policy." 22 As we are just beginning to learn with regard to the problem of corpo-
rate speech, legal fictions such as "corporate personhood" take on a life of their
own and do great damage to the underlying values, which at one point seemed to
require such a fiction for their preservation. 23 Unlike the courts and scholars who
must have seen the point as incidental, Bloustein argues that the starting point for
the explication of a privacy interest was to be found in Warren and Brandeis's ref-
erence to "inviolate personality," which deals with "the individual's indepen-
dence, dignity and integrity; it defines man's essence as a unique and self-
determining being." 24
Now defining the essence of humankind is heady stuff. But that is indeed
Bloustein's self-imposed charge. He understands Warren and Brandeis's motiva-
tion to be derived from a fear that "a rampant press feeding on the stuff of private
life would destroy individual dignity and integrity and emasculate individual
freedom and independence:'25 On his way toward the explication of this interest,
Bloustein mounts an assault on Prosser and the received wisdom of the American
186 A DATA PROTECTION REGIME

Law Institute by finding the dignity interest in each of the torts that have come to
define privacy since 1890.
With regard to intrusion, Bloustein argues that the injury is not mental or
emotional distress, but rather the fact that intrusions are "demeaning to individu-
ality ... an affront to personal dignity." If a person is not able to determine who
does and does not have access to the realms that they define as their own, they are
by definition less than whole. Another individual who is able to enter unbidden
must be seen to have superior power, and if the entry is pursued with apparent le-
gitimacy, they must also be assumed to be of higher status and worth. In
Bloustein's words, "He who may intrude on another at will is the master of the
other.'' 26 Bloustein also suggests that there is a distinction without a difference
when the intruder is an agent of the state rather than another private citizen- the
wrong, a challenge to the liberty of the individual, is still the same even though the
power of the actors may differ substantially.
With regard to the disclosure of private facts, Bloustein rejects the distillation
by Prosser of the privacy interest into a concern with reputation. An injury to rep-
utation evaporates if the public response is sympathetic and understanding. Yet
Blaustein argues that the affront to dignity remains. The essence of the affront is
to be found in the fact that publication allows not one but perhaps thousands to
join the intruder at the window peering in on the private experience of either joy
or suffering. Both are intrusions; only the means are different. "The wrong is in
replacing personal anonymity by notoriety, in turning a private life into a public
spectade.''27 A public figure loses much independence and becomes subject to the
claims and demands of the public. People on the street address movie stars and
demand the attention of public figures as if they knew them personally. The kind
of face-to-face intimacy that films and television provide explains this forward-
ness, in part. The majority of public figures may be understood to have exchanged
part of their individuality for what they believe to be greater financial rewards.
However, public disclosure, by mass communication or networked virtual mem-
ory, creates a public person without the benefit of prior consent; indeed, such
transformations of personhood frequentJy involve coercion.
Commercial exploitation, the third tort identified by Prosser, has been the
most damaging point of departure from the underlying thesis of Warren and
Brandeis. When one's image, profile, or personality is appropriated and used by
another without warrant, it is the victim's liberty or freedom to decide that it has
been abridged. Rather than focus on the negligible proprietary interest that might
inhere to such a threat, Bloustein argues that commercial exploitation represents
tJ1e same insult embodied in the disclosure of private facts. "In the public disclo-
sure cases what is demeaning to individuality is being made a public spectacle by
disclosure of private intimacies. In these cases what is demeaning and humiliating
is the commercialization of an aspect of personality." 28 The threat to liberty that is
at the heart of this insult is that some part (intangible though it may be) of a per-
son is taken and then used by another without permission. To focus the tort on pe-
l
4

1
j
A DATA PROTECTION REGIME

cuniary value essentially eliminates the possibility of pursuing successful claims

on behalf of average citizens because the value of their images in the market is
insignificant-they have no market power. The cases of public figures, in which
there may be identifiable, measurable, indeed substantial, economic rents associ-
ated with their claims of proprietary interest, should not be allowed to distort or
distract attention from the primary concern with misappropriations that generate
harm in the taking.
Jonathan Graham cites the case of Shilbey v. Time, Inc. as an example of a case
in which the tort of appropriation was inadequately pursued, but he concludes
that its impact would have been minor in any event because the definition of the
tort had been limited to commercial exploitation in relation to product endorse-
ment. 29 In Shilbey, the petitioner claimed that the sale of subscription lists consti-
tuted an invasion of privacy because the information about the plaintiff's person-
ality that might be inferred from such a list was the equivalent of an image, or
likeness, not unlike that provided by a photograph or an artist's rendering. The
fact that it was transferred to another party without permission made it an action-
able appropriation. Unfortunately, the argument in the case clouded the issue by
suggesting that the harm came by way of resultant appeals and solicitations from
others interested in reaching a consumer with such apparent qualities. An earlier
decision in Lamontv. Commissioner ofMotor Vehicles had introduced this distrac-
tion by arguing that the solicitations that flowed from the sale (by New York state)
of information from automobile registrations represented the privacy loss (seclu-
sion), rather than the loss of control over personal information (appropriation
and transfer), which is at the heart of the issue. 30 Although subsequent harms may
flow from an unauthorized disclosure of private facts, the courts have erred by fo-
cusing on this, rather than on the initial action, which is the primary breach.
In a most insightful passage, Bloustein argues that the "right of publicity;'
which has emerged to respond to the proprietary interests of those with saleable
images, is itself dependent on a prior recognition of a right of privacy. In his view,
"Name and likeness can only begin to command a commercial price in a society
which recognizes that there is a ... right to control the conditions under which
name and likeness may be used. Property becomes a commodity subject to be
bought and sold only where the community will enforce an individual's right to
maintain use and possession of it against the world."31
Thus, following Bloustein's argument, there is truly no right of publicity, but
only the right to demand a fee for not enforcing one's right to privacy. 32 Just as the
disclosure of private facts may bring sympathy rather than scorn, Bloustein notes
that commercial use may bring fame, compensation, and other benefits that the
subject may actually value and enjoy. Many drive.rs who h ave been convicted of
driving whiJe under the influence of alcohol actually appear to have been pleased
that a firm that makes portable "breathalyzers" was able to identify them through
court records and sell them a device that would help them to determine when the
alcohol in their system exceeds the legal limit. 33 Yet, this arguably positive out-
188 A DATA PROTEC I'ION REGIME

come does not overcome Bloustein's objection that, however beneficent the mo-
tive or successful the result, the "touching" is considered wrongfuJ because with-
out consent, it represents a threat to the freedom of the individual to decide.
Perhaps we have been led to consider with amusement the fears of native peo-
ples that a camera would steal their spirit, yet there is much here to treat seriously.
It is a peculiarity of information that I have discussed, that one can take or possess
information from someone without denying that person further use of that infor-
mation. There is still a taking that has to be considered and evaluated. The empha-
sis on exploitation is focused on how that which was taken bas been used. The pre-
sumption is that some "takings" do not give rise to injury unless the harm is to
reputation. Yet I join Bloustein in arguing that that which is taken by the "peeping
Tom" or the intruder is inherent in the taking itself. This theft of personal dignity
inheres in the right to decide, the right to control access to the person.
Clearly, such taking is commonplace in the perception and apperception of
others whenever we encounter or observe them in public spaces. Such taking
moves to an invasion of privacy when the observation occurs in a space defined as
private by the indjvidual. Yet developments in technology that allow the experi-
ence of such taking to be repeated, to be reviewed at leisure, and perhaps to be ex-
tended to others change the nature of the contract that is implied when individu-
als enter and pursue their interests in a public space. What is suggested here is that
if people thought that more tha11 fleeting and imperfect memory might be in-
volved in the taking of their image through the unaided glance, then greater re-
serve would characterize behavior in so-called public spaces.
The "false light" cases represent the fourth and perhaps most easily disposed
class of tort~ that Bloustein recasts through the lens of individual dignity. It is sim-
ply the case that the use of any representation of a person without consent is an
abuse of personality, whether the representation is accurate or distorted.
Bloustein rejects Prosser's economistic emphasis on the assumption of a propri-
etary interest that is threatened by distortion. As noted, the economic value of the
interests of the broad mass of the public is treated as individually insignificant by
the market, and because of this, its force in the regime of tort injury is negligible.
Here, as with commercial use, the law of defamation is seen to benefit from the
law of privacy once it is understood to be based on an interest in human dignity.
After dispensing with the distortions inherent in the limitation of privacy to
torts, Blaustein turns his attention to the substantial area of privacy law that is
found in state constitutions, statutes, and nontort decisions under common law.
He notes the statutory prohibitions agaiust peeping Tomism, including the use of
more sophisticated technical means, and the prohibitions on the disclosure of
confidential information. In noting that the very collection of personal informa-
tion that is required for access to the benefits of government represents an impair-
ment of privacy, Bloustein concludes, perhaps a bit too hastily, that "most of us
have agreed, however, that the social benefits to be gained ... are worth the price of
diminished privacy." 34 He seeks to retrieve part of his lost ground, in my estima-
A DATA PROTECTION REGIME

tion, by suggesting that the expectation in such exchanges of information is that

the use of the information will be limited to the purposes for which it was gath-
ered. Although a popular demand, this is a naive hope. We recognize that the pur-
pose of the state is comprehensive, and therefore., all information becomes rele-
vant if it can somehow, now or in the future, be used to further a legitimate state
interest. The corporate interest is no less expansive.
To conclude Bloustein's contribution to our understanding of informational
privacy, I note his emphasis on the question of means. If the fundamental value
that is to be preserved is captured by the concepts of dignity and individuality,
what makes privacy different from other actions that threaten those values are the
means that are used to gain access or weaken control. The "character of the inter-
ference" with individual liberty helps to define invasions of privacy. 35

ALL OF THIS AND DEMOCRACY TOO?

Spiros Simitis makes it clear that the pursuit of privacy as an index of individual
autonomy is incompatible with egalitarian or communitarian forms of democ-
racy.36 In his view, the claims made for a legally guaranteed protection of privacy,
at least with regard to protection from the state, are of a kind with the liberal p ur-
suit ofpersonal, p rivate benefit, without con.cern or responsibility to any collectiv-
ity. "The more, therefore, that privacy is equated with a deliberate and legally pro-
tected seclusion of the individual, the more the right to be let alone develops into
an impediment to the transparency necessary for a democratic decisionmaking
process. As long as the data required to understand and evaluate the political and
economic process are withheld, suppressed, or falsified, participation remains a
pure fiction." 37 Yet Simitis is fully aware that the routinized systems of bureau-
cratic surveillance operate not to produce a kind of transparency that facilitates
democratic participation, but instead to produce a legitimate demand for more
privacy. The rational desire of inclividuals to avoid disclosure is based in part on
the limitations they face in gaining the information they need to evaluate the rele-
vant environment, including the environment in which they are expected to pro-
vide informed consent for the collection and use of personal information. In this
environment, a privacy regime does not guarantee participatory democracy but
rather enhanced protection of the interests of the elites who are better able to ne-
gotiate the pa.noptic maze.

A COMPLEX OF INFORMATIONAL RIGHTS

One way in which the goals of autonomous self-development may be realized in
the context of a formal social structure is through the development of a regime of
rights that establishes the limitations on freedom that are necessary so that the ex-
ercise of one person's freedom does not unjustly limit the freedom of another.
190 A DATA PROTECTION REGIME

This emphasis on informational privacy leads quite natuxally to a consideration of

a regime of informational rights as guaranteed by a democratic state.
Anne Branscomb has made an important contribution to understanding the
competing claims and rationales that complicate efforts to establish such a regime
of rights. 38 Each of ten broad rights she identifies can be seen to involve a matter of
"balancing equities and sensibilities that often defy codification."39 The "right to
know" may be seen as primary, beginning with the right to know oneself. The
right to know about one's environment includes the right to know about the risks
and opportunities one faces and the qualities of goods and services that one might
acquire. This right to know is believed to be essential for the smooth and efficient
operation of the market and the political system as well. However, this right is
clearly limited.
One limitation is to be found in the procedural limitations on a "right to col-
lect information." In order to know, someone must have a right to collect informa-
tion. Yet Branscomb suggests that the right to collect information is reserved pri-
marily to government, although we would do well to follow Foucault and note
that a great many professionals have also been "qualified" and privileged by the
state to collect information. The distinction that clarifies this right is related to the
extent to which coercion or force can be used to compel disclosure. By withhold-
ing services until approval is given for a particular investigation or search, organi-
zations restrict individual liberty to some degree. The difference between public
and private investigations is that agencies of government exercise those access
rights by means of the rule of law and the ultimate threat of force, rather than
through the mechanism of an explicit or implied contract governing the exchange
of value in the market. Private actors exercise these investigational rights through
the exercise of economic or social power-the same power involved in shaping,
defining, and extending or limiting rights as circumstances change.
Branscomb's distinction between the right to collect and the right to acquire
information is subtle but is worth noting. The right to acquire information is the
right of equal access to information that has been collected by others authorized
to do so. The right to education might be interpreted as a right of access to accu-
mulated knowledge as information. Underlying her conception of access as a right
is a specification of a certain class of information as a public resource. The public
nature of this information derives in part from the use of public funds or legis-
lated advantage in some stage of its production or in the protection of its produc-
er's market position. 40 The privatization of government information represents a
reduction in the public's right to demand access to information that may have
been produced at public expense. 41
Branscomb defines a "personal right to privacy" as a right of individuals to
withhold information about themselves, that is, to restrict access to personal in-
formation. This right is a limitation on the right of others to collect or even to gain
access to information. Thus the right to privacy may be seen to limit the right-to
know when knowledge of the other is necessary if one is to pursue an interaction
A DATA PROTECTION REGIME

or relationship with that person. Although Branscomb indicates some doubt that
this right is an expression of some long-standing and fundamental instinct, its
position as a right with both moral and legal status cannot be denied. In
Branscomb's view, the creation of formal rights or limitations on inquiry follows
very closely the changing assessments of the sensitivity or the potential for harm
that would follow from disclosure. Presumably, restrictions would evaporate as
soon as consequentiality was diminished beyond some socially determined level
of tolerance. It might be argued, for example, that questions about the natUie of
one's sexual preference would no longer be restricted once society achieves some
level of tolerance of homosexuality and gay ljfe-styles. 42
Branscomb notes that businesses are able to protect their trade secrets under a
broad blanket of legislative and judge-made law. I agree with tnose who suggest
that corporations are not truly persons and thus ought not to be considered to
have rights of privacy. Yet similar instrumental concern with the consequentiality
of disclosure is at the heart of trade secret protections. Branscomb includes within
this category of trade-related protections the shield laws, which allow journalists
to resist attempts by others to learn their sources. Yet we also should note the
growing exceptions to corporate privacy that are inherent in the privileging of the
public's right to know about the dangers that corporate operations represent. 43
If individuals exercise a right to prevent access to personal information, they
may also choose to provide access but reserve the right to limit the sharing or dis-
closure of that information to others. Branscomb includes in this specification of
rights the problematic sharing of personal information that is produced through
commercial transactions. Another right, the "right to protect information," is
more directly linked to an effort to define and classify different kinds of informa-
tion and the limitations that might be placed on disclosure. Government docu-
ments provide a model in that they can be stamped or classified as sensitive, confi-
dential, secret, top secret, and so on, and their distribution can be restricted to
persons qualified through classification as well as by "need to know." Similar clas-
sifications of personal information are made within organizations that handle in-
formation of varying sensitivity.
However, it is not until she begins to define a "right to control the release of in-
formation" that we begin to see the emergence of the concept of information as
property. In United States v. Miller, the U.S. Supreme Court held that individuals
had no legitimate interest in their bank records, which were required of the banks
by the Bank Secrecy Act.44 The government's interest in identifying criminal activ-
ity, including tax avoidance through its examination of bank transaction records,
had resulted in the passage of a federal record-keeping requirement. Because the
bank was keeping. the records to meet the requirements of the government and
not at the request of the depositor, the court argued that, without an interest,
there could be no expectation of privacy in those records. Individuals were put on
notice that they place themselves at risk when they enter into a business relation-
ship that involves the surrender of personal information, that this information
192 A DATA PROTECTION REGIME

might be shared with the government. The dissent by Justice William 0. Douglas
in a related case is notable for its prescience, if not for its utility in limiting access
to information. Douglas recognized that "a person is defined by the checks he
writes. By examining them the agents get to know his doctors, lawyers, creditors,
political allies, social connections, religious affiliation, educational interests, the
papers and magazines he reads, and so on ad infinitum."45 Persons are similarly
known by their telephone call records, their videotape rentals, their magazine
subscriptions, and their accumulated and detailed credit card transactions. Dem-
onstrating a legitimate interest and a privacy expectation in such data pools is the
goal that Miller helped to frustrate.
Branscomb's recognition of a commercial property interest emerges primarily
through her notice that actors who have been given the right to collect personal
information for one purpose have sought to transform the informational prod-
ucts of that transaction into marketable goods in a way that exceeds the rights of
access. She extends the definition of this right of control to include a corollary
right to "release information at a time or place of one's choosing."46 Presumably
the exercise of this right involves a legally binding constraint on any agent that has
been given the right to make the information public. The timing of the release of
information can be seen to have significant economic consequences. The success-
ful exploitation of property interests in entertainment products such as theatrical
films depends in part on the ability of distributors to control and coordinate their
release in theaters, via cable, via videotape, and finally through broadcast televi-
sion.
The property interest in personal information is made explicit in the right she
defines as "the right to profit from information:' However, her discussion in this
area is almost entirely limited to questions of intellectual property, except for the
novel discussion of the commercialization of personal histories or memoirs. The
fact that exceptions have emerged to limit the realization of profit from populari-
zation of illegality suggests a variety of social interest rationales on which such
limitations might be based.47
Concerns about personal information emerge again in l3ranscomb's discussion
of"the right to destroy or expunge information" in the records or files of others.
Concerns about "freeze dried" criminal records that provide a representation of
individuals as they were, not as they are today or as they might become in the fu-
ture, are at the basis of this claim. It is similar to the claim of a "right to correct or
alter information" in the files or records of others.
FinaUy, the right to know is linked to the right to speak, or in her words, the
"right to disseminate information." The right of access to information about one-
self and one's environment is argued to be limited if there is not also a right of ac-
cess to the means through which that understanding can be communicated to
others. Making this link between the right to know and the right to speak links all
the informational rights into a system of governance that we have come to call
democratic. Conflicts about the extent to which these rights are to be limited to
A DATA PROTECTION REGIME 193

natural persons and citizens or will extend to fictional persons and others granted
rights of national treatment remain the souxce of conflict with regard to informa-
tional privacy in ways similar to conflicts about intellectual property. 48

INFORMATION PRIVACY AND CORPORATE SPEECH

As Branscomb's review of rights suggests, there are a number of tensions between
rights that have no inherent priority, except those that might privilege the rights
of individuals over the rights of corporations. Robert Posch, as a representative of
the interests of the direct marketing industry, claims that the industry's most
powerful legal stance in the debate over informational rights is one that argues
that the primary constitutional issue is one offree speech, rather than one of indi-
vidual privacy.49 Posch takes as a given that which is still controversial in some ad-
mittedly limited circles, that corporations have First Amendment rights. Limita-
tions on corporate speech are seen in any restriction imposed by state action on a
firm's ability to sell customer lists, which these firms wish to define as speech.
Given that the information in lists is primarily factual, First Amendment c~aims
represent a powerful argument, as Posch and his colleagues have come to recog-
nize. Tort restrictions on the commercial disclosure of private facts have been lim-
ited in practke by both extent and public interest exceptions. It is also the case
that the distribution of lists may not be seen as publication, limited as they are to a
select group with a legitimate business interest. It might be seen as a different mat-
ter, however, if feminists, for example, rented space on a bulletin board and pub-
lished a list of local citizens, public figures, and "average Joes" who had.subscrip-
tions to Penthouse or were frequent callers to a sexually oriented phone service. so
At the moment, the courts still have not erased all distinctions between com-
mercial and other forms of protected speech. C. Edwin Baker discusses commer-
cial speech as a problem in the theory of freedom that helps us to understand the
remaining gaps.st Although we might wish to make a distinction between com-
mercial speech that seeks to inform or influence consumers with regard to a prod-
uct or service and the forms of speech involved in the sale of personal information
in the form of lists, it is not a distinction th.at serves well. Baker argues that com-
mercial speech, as it relates to advertising, is at least as important as other forms of
speech in terms of its influence on how we understand and act in the world. Such
a position with regard to commercial and other forms of speech makes sense, he
axgues, in the context of a marketplace logic where ideas are allowed to "compete;'
and superior ideas, like superior products, win in the long run. If, however, com-
mercial speech is viewed through the lens of a "liberty theory" of the sort he pre-
fers, a rather different conclusion is seen to emerge.
The liberty theory, as Baker presents it, focuses on the interests of the speaker
and her or his choice to speak, rather than on some assessment of the usefulness of
the content of the speech to a listener or to listeners in general. Whereas usefulness
is relative and broadly conditioned, individual liberty is more broadly absolute.
194 A DATA PROTECTION REGIME

Because the pursuit of profit is a purpose imposed by the market, commercial

speech within the marketplace is distinguishable from the speech of individuals,
which is used for expressing an almost unlimited variety of perspectives on indi-
viduality and self. The e}..i>ression of self is, of course, preferred to the pursuit of
profit within markets. 52
Baker's posture on human liberty thus finds commercial speech wanting be-
cause of its limited instrumental purpose. Commercial speech seeks to influence
the realization of profit by modifying individual preferences. This purpose is in
conflict and is perhaps even incompatible with the value of self-determination.
Therefore, the "very values of respect for human autonomy and self-deter-
mination that require the first amendment necessitate that commercial speech be
subject to collective, political control." 53 He argues, further, that when a given ac-
tivity is subject to regulation and the conduct of business is clearly and_:Qecessarily
regulated, it follows that speech that is an integral part of that activity should also
be subject to regulation in pursuit of the same public interests. 54 Thus the regula-
tion oflists as commercial speech might reasonably be pursued on behalf of the le-
gitimate and demonstrated state interest in preserving individual privacy.

WHO'S THE ENEMY HERE?

Privacy, rather than being seen as a state that one might seek to attain, may be seen
as a protection against unwarranted access or disclosure. We invest in protections
against threats. The character of the protections must reflect in some way the
character of the threats. The nature of the defense is determined in part by the na-
ture of the attack. The nature of the attack, although about this I am less certain,
may reflect the nature of the attacker and the means that are chosen. The choice of
means may reflect the power and the power differential between the actor who
seeks access and the individual who seeks to limit it. Kenneth Laudon suggests
that "privacy is a value which describes a power relationship between individuals
and organizations. This relationship can be seen as a continuum marked on one
side by complete informational moral supremacy of the individual, and on the
other by complete supremacy of the organization and its needs for efficiency and
survival."55 I cannot, however, envision a single continuum, because organiza-
tions differ from each other in important ways in terms of the nature of their
threats to privacy. Comparisons of industrial nations suggest that aspects of their
people's history have influenced how the citizenry of different nations depart
from each other in terms of their fear of the state, rather than of the private firm,
with regard to personal information.
Priscilla Regan suggests that ordinarily we might expect that a distinction
would be made between public and private bureaucracies and that private bu-
reaucracies would be regulated more severely because these actors would not en-
joy the same level of legitimacy that agencies of the state are assumed to com-
A DAD. PROTECTION REGIME 195

mand. 56 She notes that, instead, the exact converse is true-the bureaucracies of
the state have been constrained more completely in most Western democracies.
John Bennett suggests that part of the difficulty that characterizes the develop-
ment of a privacy policy in Western democracies is the fact that there is "no iden-
tifiable external group which receives a cost or benefit from the production of the
policy"; because of this, there is no clientele or constituency for such a policy.57 No
politician sees her or his long-term future to be dependent on satisfying a demand
that is so poorly articulated. Flaherty also sees this problem of the identifiability of
interests at the base oflegislative confusion and inactivity because it "is very diffi-
cult to demonstrate need in the commonsense way that Congress requires. This is
the particular bane of the sectoral approach to sectoral protection, since it is too
hard to document the necessary horror stories without holding full-scale investi-
gations of a particular sector."58 The problem of "identifying the culprit," on the
one hand, is reflected in the development of legislative protections against
excesses, rather than an affirmative effort to protect fundamental rights and free-
doms. On the other hand, the bureaucracies of the modern state, both public and
private, find that privacy, however defined, represents an obstacle to the pursuit of
their organizational missions.

Protections Against the State

The passage of the Privacy Act of i974, the primary instrument for regulating the
information practices of federal government agencies in the United States, ·is
thought to have been a response to the excesses linked to President Richard Nixon
after Watergate. David Linowes reports that there were some 3.5 million personal
files that federal agencies maintained on U.S. citizens,59 and he cites a National
Bureau of Standards report that outlined the informational abuses that would
shock the nation further when the extent of the abuses was revealed in the Senate
intelligence hearings coordinated by Frank Church.60
Political surveillance is a readily understandable concern. Much has been writ-
ten about the need to Limit such uses of state power. The administrative use ofsur-
veillance by the state is less closely studied, although the interests of the state are
barely distinguishable from the interests of corporate bureaucracies in this regard.
Flaherty finds it ironic that the Reagan administration paid so little attention to
privacy, in that "the protection of the individual from various forms of govern-
ment intrusion is surely a conservative issue." 61 What Flaherty misses in this
ironic glance is that conservative concerns about privacy were easily subordinated
to concerns about efficiency and the administration's program to "eliminate
waste, fraud, and abuse" in government through the use of matching, profiling,
and a host of other privacy-threatening technologies that wouJd be focused on the
poor and at the lower reaches of the middle class.
The ability of the government to avoid the restrictive intent of the Privacy Act
through its use of administrative exceptions, such as those that defined restricted
data sharing as "routine;' simply exemplifies the problems inherent in self-reguJa-
 A DATA PROTECTION REGIME

tion. 62 As Flaherty noted in 1984 with regard to the OMB, the executive branch or-
ganization charged with oversight of the Privacy Act: "OMB is doing nothing on
the policy side of privacy. It would not be unfair to say that OMB is and has been
essentially uninterested in privacy issues. It views its role under the Privacy Act as
one of balancing interests, not on behalf of privacy but in favor of the needs of
government. This is a very limiting factor from a privacy perspective, since OMB
is hardly an impartial arm of the executive branch." 63
The potential threats to privacy that were inherent in the administrative use of
matching were identified early in their history.64 Eventually these issues made
their way through a series of congressional hearings only to result in the legitima-
tion of the very practices that early critics had claimed were barred by the Privacy
Act. In her testimony before the House committee hearings on matching, ACLU
staff attorney Janlori Goldman noted that the bill "does not impose any
subtantive limits on matching programs."65 Yet she expressed support for the pro-
posed bill because it would "enhance privacy and due process rights of citizens
who are the subject of matching and verification programs." Thus a program of
questionable efficiency in terms of dollars saved for dollars expended, but of un-
questionable (although unmeasurable) costs in terms of the loss of privacy to the
millions who would be matched routinely, was supported by one of the chief ad-
vocates for consumer privacy rights. It seems naive to believe that the hundreds of
thousands of citizens subject to surveillance through matching of their records
would be regularly scanning the Federal Register to note that a match was about to
occur. That they could do little about it an}'\vay would make the awareness of an
impending match even more costly. To compound the error, the ACLU recom-
mendations sought to further define as a legitimate justification for a match, "evi-
dence" from a benefit/cost study that only projects likely outcomes. That is like ex-
pecting investor-seeking producers to project next year's earnings. Referring to
this testimony in particular, Flaherty appears to share its logic and its measure of
the legislative pulse: "The new law emphasizes due process and administrative
goals, including analysis of costs and benefits, rather than concentrating on pri-
vacy and surveillance issues. T his reflects a shrewd political assessment of how
best to persuade Congress to act." 66 As we shall see, this second best solution is
replicated time and time again as privacy advocates have attempted to control pri-
vate bureaucracies.

Protections Against the Corporation

The critical and perhaps most problematic aspects of the balancing of interests that
set the interests of the individual against the interests of the organization are to be
f0tmd in the expression and development of the concept of a "legitimate business
interest." As discussed in the previous chapter, there is a general willingness within
the population to allow business to gather the information it needs in order to
make business-related decisions. There is considerable variance, however, in the
belief that business will linlit its information gathering to that which is truly neces-
A DATA PROTECTION REGIME 197

sary. Differences begin to emerge when inquiries are made about whether employ-
ers, insurers, vendors, or government agencies should be limited in the kinds of in-
formation they can ask for directly. Researchers have yet to fully explore what peo-
ple see as limitations on the kinds of investigations that these organizations might
pursue to collect information from third parties. It is clear, as I have noted, that the
relative ease and marginal expense that allow organizations to search, sort, match,
and acquire information from discrete, independent, and remote databases serve
to increase the amount of such information demanded. The only constraint ap-
pears to be the definition of what is a legitimate business interest.
The development of rights and privileges for corporations, including the de-
velopment of the fiction of the corporate person, can be seen to have had a justifi-
cation in the broader common, or public, interest. For society to enjoy the eco-
nomic benefits that the corporate form could provide, it was believed that
incentives had to be provided to individuals to encourage the use of the corporate
form. Herbert Hovenkamp suggests that a concern with growth was a dominant
force in the early days in U.S. history, an~ this concern can be seen in the develop-
ment of laws involving contracts, debts, and eventually the rights associated with
the corporate form. 67 The American model, which Hovenkamp refers to as classi-
cal political economy, was vintage trickle-down theory: Make the society wealthy,
and the wealth of individuals will thereby be achieved in due course.
Hovenkamp suggests that the ideological position of the leadership in the U.S.
government changed from the preclassical federalist orientation of Alexander
Hamilton, which supported an. active role for the state in the development of the
economy, to the classical laissez faire vision of Andrew Jackson, which warned of
the dangers of monopoly power. The government might avoid such dangers at the
same time it realized the benefits of incorporation through an emphasis on "tl1e
public purpose" as a rationale for whatever privileges a monopoly grant might en-
tail.
Some of these advantages included the privilege of limited liability. Morton J.
Horwitz suggests that a tension between the logic of the law and the logic of the
natural sciences with regard to causation and responsibility was at its height
around the turn of the century. The challenge for the law was to establish a basis
for understanding the extent to which a chain of causality could indicate a line of
responsibility for injuries. Entrepreneurial energies might be diminished if there
was an absence of "foreseeability" in the limits of liability.68 Contemporary risk
analysis can be seen to be a further response to these issues, and such analyses in-
crease the demand for information. 69
Although the U.S. Supreme Court in Tureen v. Equifax suggested that there
might be abuses of privacy through unreasonable intrusion by corporations, it
was still believed that "in order to make informed judgements in these matters, it
may be necessary for the decision maker to have information which normally
would be considered private, provided the information is legitimately related to a
legitimate purpose of the decision maker. In such a case, the public interest pro-
 A DATA PROTECTION REGIME

vides the defendant a shield which is similar in principle to qualified privilege in

libel:' 70 Neither this court nor others have fully considered the tension between
business interests and what might be claimed as more basic individual interests in
privacy.
Susan Gardner71explored the privacy interests inherent in the use of a variety
of techniques, many of them of questionable reliability, to assess the honesty of
employees or persons who were candidates for employment.72 Such examinations
are criticized because they are intrusive and because they are an affront to the dig-
nity ofan individual. The definition of"honesty" is categorical rather than empir-
ical. That is, a score is assigned to an individual on the basis of a test or measure-
ment exercise, whether an electronic analog, such as provided by a polygraph,_or a
more theoretical abstraction of the sort generated by a paper and pencil test. An
individual's score is then compared with the distribution of some normative scale,
and the location of that score is the basis of an employment decision. Individuals
who want a job or who want to keep the one they have are without a meaningful
option. Their refusal to take a requested but voluntary test may be interpreted as
evidence of dishonesty or disloyalty. They are, therefore, compelled to surrender
their privacy. They are made to suffer an indignity at the same time that they are
compelled to answer questions that frequently bear no identifiable relationship to
their ability to perform a job.
Restrictions on the use of polygraphs or other tests of employees raise critical
questions about the possibility of establishing general standards that would spec-
ify which questions and which means of gaining access to personal information
ought to be barred as a matter of policy. If we accept, as a matter of course, that
employers have a right to know the qualities of persons whom they consider for
employment (and indeed, we may have imposed on employers a requirement to
know to avoid legal liability for irresponsible hiring), then we face a contradic-
tion.73 By restricting the collection of information that employers consider to be
relevant by form or content, we increase their need for information that reveals
these qualities only indirectly but may in consequence provide insights into other,
previously uninteresting aspects of person hood.

SENSITIVE INFORMATION
Raymond Wacks's contribution to this problem is to identify what he calls "sensi-
tive information." From his position, the general failure of the law of privacy is a
failure to identify the types of information that are most in need of protection.74
Thus, Wacks's emphasis is on content rather than means. He notes that the defini-
tions of these areas of sensitivity are culturally determined and therefore vary sig-
nificantly within and between nations. For example, income appears to be much
less sensitive in the Scandinavian countries than it would be in Great Britain or in
the United States.75
A DATA PROTECTlON REGIME 199

The notion of sensitivity has great value if we consider the concern about link-
age. Identification by itself is usually not problematic. It is the linkage of reliable
identification with a sensitive measure, such as sexual orientation or political af-
filiation, that raises privacy concerns. The most difficult aspect involved in devel-
oping a measure of sensitivity is a recognition that there are also individual d iffer-
ences in sensitivity or concern about aspects of personhood.
Even though Wacks claim.$ that his scalar is based on an assessment of the "po-
tential for serious h arm to the subject:' it must be recognized that even this i~ a
·relative measure.76 Standards of reasonableness, or average person standards, are
inadequate by definition because no person can be precisely average; they must be
at least fractionally more or less than average. It is not at all clear what kind of
rules of thumb, or even empirically validated standards, we might develop. It
should be clear, however, that an explicit, socially developed standard would be
more democratic than standards that currently evolve in the narrow vision of elite
jmists.
Wacks uses the kinds of data normally held by public and private agencies in
Great Britain to create twelve different categories of information. The categories
include: biographical information and data about the home, family relationships,
employnient, finances, health, education, ideology, police, habits, leisure activi-
ties and travel, and communication. Within each of these categories, Wacks"then
assigns ratings indicating one of three levels of sensitivity: low, medium, or high.
Although Wacks is to be applauded for the diligence with which he developed his
list, I am sure that many will find much to argue with in his ratings. For example,
none of the t_vavel and communication categories receive a rating of high sensitiv.
ity. Telephone numbers contacted, newspaper subscriptions, periodical subscrip-
tions, and books borrowed from libraries or purchased from book clubs are all
classified as being similarly sensitive, and all of them less sensitive than informa-
tion about the "frequency of sexual intercourse with spouse." If the basis for such
a sensitivity rating is the potential for harm to the data subject, the nature of the
harm in this category escapes me, at least when compared with the potential harm
from inferences drawn from reading records. Similarly, data about past voting
behavior are identified as being highly sensitive, as are sixteen out of thirty items
listed under medical information.
In addition to the probkms in assigning ratings to individual items of infor-
mation, which Wacks recognizes as problematic, the use of such a scale would also
need to explore the possibility that empirical sensitivity would·vary as informa-
tion from different categories is combined.77 Thus an image of an individual that
might be generated from information about communication, address, and em-
ployment could easily exceed the sensitivity level of a single item of information
about the regularity ofsexual activity with one's spouse. Furthermore, the correla-
tions between measures suggest the very real possibility that limitations on access
to one source of information do not truly limit access to the knowledge or intelli-
gence that information would provide. Consider the following argument.
200 A DATA PROTECTION REGIME

Richard Posner's analysis of privacy rights emphasizes the interests of individ-

uals in concealing "information about themselves which others might use to their
disadvantage." 78 On the face of it, this is a concern with sensitive information as
Wacks defines it. The neoclassical posture that Posner reflects seems to assume the
absence of any moral or social limits on the pursuit 'o f those intereSts. 79 Thus
Posner assumes that "people want privacy in order to manipulate other people by
concealing from them aspects of their character, prospects1 or past that would if
known reduce their opportunities to engage in advantageous market or
nonmarket transactions."80 H~ claims that discrimination on the basis of race and
gender reflects little more than a rational response to the costs of acquiring more
reliable information about performance. Because the mean of a population is the
best estimate of any individual's score, selection among groups is thought to be
made on estimates of mean group performance.sL
As the association between disadvantageous information and life chances is
explored further, Posner's analysis leads him to conclude that privacy legislation is
designed primarily to limit discrimination against identifiable groups who !)hare
in common several disqualifying attributes or qualities. Privacy legislation, as he
sees it, forbids the use of such information in specific economic decisions. Privacy
legislation limits the panoptic sort and is therefore inefficient on its face. Posner
reports finding a correlation between a state's presumed interest in limiting racial
discrimination and the number of statutes in that state that protect individual pri-
vacy.82 Posner qmcludes, with some insight, that the ban on discrimination by
race or gender will have little effect so long as those who wish to discriminate can
identify correlates of race and gender that predict mean performance just as well.
It is this insight, that the panoptic sort is capable of finding analogues or indexes
that serve the same function as more politically or socially sensitive indicators,
that weakens the long-term utility of any definition of classes of information as
being more or less sensitive. Robert Posch actually defines efforts to restrict the
use of credit histories as being contrary to the interests of historic civil rights legis-
lation, precisely because it would force the industry back to apolicy of geographi-
cal redlining. 83 When barred from using one form of information, decision mak-
ers will pursue the second best solution, which might involve using additional
information sources.
However, when the socially conditioned association between indicators is quite
strong, barring the use of a particular indicator may, in the short run, increase the
benefits enjoyed by a particular group. That is, barring the use of information
about bankruptc.ies or prison records may allow more African-Americans to 'gain
employment and credit because of the high proportion of black people with such
problems.84 However, Posner concludes further that any restrictions on the nature
and amount 9f information that employers or creditors might gather represents a
negative-swn game for society as a whole because "the effect of such statutes on
the public as a whole is to increase the amou11t of fraud in society, raise interest
rates, and reduce business productivity?'85
A DATA PROTECTION REGIME 201

This final analysis of Posner's ignores the social costs that result from the real
discrimination that is allowed to continue, even in the face of restrictions on the
use of some categories of personal information. Empirical evidence of racial dis-
crimination continues to mount, yet analysts claim to be unable to identify which
inputs into the panoptic sort are responsible for the resultant pattern. An exami-
nation by the Federal Reserve Board of the limited success that African-
Americans and other minorities experienced in obtaining home mortgages in
1991 revealed a striking pattern of discrimination. 86 Minorities were denied mort-
gages two and three times more often than whites. The logical assumption that
differences were explained by income did not stand up to scrutiny. In fact, when
income was held constant; the apparent discrimination actually increased, pro-
ducing even greater disparities at the upper-income levels. Not even credit-
worthiness might explain the differences, because credit reports are being used
increasingly as prescreening tools, and therefore applicants with poor records are
not even shown a home in the first place. Because utilities and other service pro-
viders have begun making reports to credit agencies as soon as bills become delin-
quent, a great number of African-Americans find themselves shut out of the hous-
ing market even before they begin. Thus, something other than income or
creditwor thiness predicts or explains the failure of African-American consumers
to acquire mortgages: It is clear for any who would care to look that race, or an in-
dex of race, whatever its form, was being used to disqualify any but the most ex-
ceptional applicants for loans. _

REASONABLE EXPECTATIONS
AND THE DESTRUCTIVE GALES OF TECHNOLOGY
Among the many insights that Spirqs Simi tis has provided, one of the most influ-
ential and troubling is his linkage of privacy to the realization of an active demo-
cratic state. "Neither freedom of speech nor freedom of association nor freedom
of assembly can be fully exercised as long as it remains uncertain whether, under
what circumstances, and for what purposes, personal information is collected and
processed:'87 The realization that such uncertainty might be resolved through the
recognition that there-may be no limits, because there are no longer any reason-
able expectations of privacy, does not, of course, engender great joy. Yet, as wiU be
seen, the U.S. Supreme Court has moved actively and aggressively to narrow the
limits of reasonableness.
Laurence Benner suggests that the decline in the scope o.f reasonableness in re-
cent years has been dramatic in comparison with most ·of the years th4t have
passed since the ratification of the Fourth Amendment. 88 Although the Fourth
Amendment was concerned fundamentally with the conLrol of-the federal govern-
ment, the right formalized a principle that might apply to each and every institu-
tion with the power to impose its will on individuals. The amendment defined a
right of security against unreasonable search and seizure and specified further
202 A DATA PROTECTION REGIME

that warrants governing such searches must be specific and demonstrate probable
cause. However, the reasonableness clause soon became separated from the war-
rant clause, and each became weakened on its own. 89 In its early stage.s of demise,
the reasonableness clause ~ecame balanced against an indication of state interest.
"If the government's need to intrude oulweighed the citizen's privacy interest in-
fringed by the intrusion, then the instrusion was 'reasonable' and did not violate
the Fourth Amendment." 90 ·

Benner charges the Rehnquist Court with making the most far-reaching excur-
sions into the privacy realm through its rulings on the appropriateness of drug
and alcohol tests of employees without the requirement ofan assumption of prob-
able cause. All that was necessary for the search of persons or property was the
successful claim that there was a legitimate public interest, such as the safety of rail
passengers or the smooth opera ti on of public agencies. Furthermore, it appeared
that little more than the concern with "efficiency" and "convenience" was needed
to justify a search. If this same court allows the search of one's trash, when a truly
reasonable expectation would be that it becomes mixed with the trash of hun-
dreds of others and that the link to a particular household is forever lost, then
there is little basis for expecting that such a court would keep the government
from searching through the residue or "trash" that is the computerized record of
the hundreds and thousands of transactions we make as we go about our daily
lives. 91
In the case of Skinner v. RailwayLabor Executive's Association92 this hardening
court suggests that even the most invasive of techniques, involving the taking of
blood or the supervised collection of urine, raised privacy interests that could be
subordinated to "the government's 'special interest' in ensuring rail safety." 93 But
invasiveness as a test does not hold against the assault of reasonableness of expec-
tations, which crumbles in the wake of technological advance.
Benner credits the ruling in the case of Katzv. United States with developing the
line of argument that he believes has resulted in the total loss of Fourth Amend-
ment protection.94 If it is possible to argue that no "search" actually took place,
then there can be no claim of a Fourth Amendment issue. No search under the
Fourth Amendment has occurred if the court determines that no reasonable citi-
zen could have expected privacy interests to survive inspection. In Ka.tz, the court
concluded that an individual had no reasonable expectation of privacy in a publk
telephone booth, aware as he or she must be that the police have access to and are
likely to utilize surveillance devices that would allow them to listen in on at least
one side of a conversation made from a public telephone. Similarly, a California
court would follow this logic and rule that a citizen could not have a reasonable
expectation of privacy in his backyard, io-foot fence or not, because it was clear
that airplanes and helicopters fly overhead and could take pictures of things, like
marijuana plants, growing in a backyard. 95 By leaving it up to the judges to deter-
mine the reasonableness of expectations and by placing the burden on the defen-
dant to convince them otherwise, the requirements of the Fourth Amendment
A DATA PROTECTION REGIME 203

simply evaporate. Thus, Benner argues, "The Court's refusal to conceptualize pri-
vacy as a power to control who has access to information about ourselves has led
to diminishing expectations of privacy, and thus to diminished protection under
the Fourth Amendment:' 96 Benner is joined by James Tomkovicz, who suggests
that we ought to determine the need for Fourth Amendment protections not on
the basis of reasonable expectations, but on the basis of an individual's need for
informational privacy. For him, "courts should attempt to ascertain whether the
government's actions deprived an individual of informational privacy and
whether the individual's freedom to exercise and enjoy constitutional rights re-
quired some degree of informational privacy vis-a-vis the government."97
As the technological means to gain or facilitate access to personal information
about individuals continue to develop and to become all the more broadly avail-
able as the cost, complexity, and skill requirements necessary to use them are all
diminished, it will soon be the case that no expectation of privacy at all could be
reasonable. Just because the courts have based the erosion of this constitutional
right on the need to reduce crime, it would be unreasonable to expect that it will
not spread on the basis of some pressing need to eliminate waste, inefficiency, or
·even bad taste.

FIGHTING BACK
One response to this assault against civil rights that has come to characterize the
U.S. Supreme Court is the creation of a statutory right through legislation at the
federal and state levels. Loretta Murdock argues that a "statutory approach would
provide both certainty and a clear legal basis from which individual and organiza-
tional rights and obligations could be determined."98 Murdock has a national pol-
icy in mind, and her recommendation echoes a common call for the creation of a
national agency with the responsibility for protecting privacy interests. Preemp-
tion of state options in this regard is based on a recognition of the high level of in-
tegration of commerce in the United States, which is facilitated by telecommuni-
cations. The likelihood that states would develop inconsistent or even conflicting
legislation can be inferred with confidence on the basis of a glance at any of the re-
cent compilations of privacy laws published by Privacy Journal. 99 As will be dis-
cussed further on, attempts at federal preemption in the area of Caller-ID under-
score the consequences of piecemeal legislation, which does not involve a clear
definition of the underlying privacy interests.
Colin Bennett asks us to consider the distinction between an approach that
°
emphasizes technology as compared with one that emphasizes civil rights. 10 Fail-
ure to recognize the interests and imperatives of private and public bureaucracies
has led technologically oriented scholars to suggest technological solutions. David
Chaum's proposal for a "public key/pri~ate key" solution, which would allow in-
dividuals to adopt different personae and identities as they interact with different
204 A DATA PROTECTION REGIME

bureaucracies, fails to recognize the pressures that would deny individuals the
right to use such personae in their economic and political lives. 101
If the problem with information privacy is simply one of the inappropriate use
of technology, then Bennett suggests that specific restrictions can be developed
that might control the use of computers for the matching of independent data-
bases. However, it is also recognized that the problem of data linkage emerged as a
result of an independent change in the nature of technology. Previously, privacy
advocates were concerned about the threat inherent in the creation of a single,
massive database, which would contain information about every citizen and
would perhaps be subject to the control of some enemy or to an unexpected shift
in the democratic orientation of the government. Today, however, the technologi-
cal reality is one in which there is not any single computer, but hundreds and
thousands of computers that can be accessed remotely and linked on demand for
searches enabled by use of a common identifier like the social security number. It
seems unlikely that a policy response will forbid the development or use of a tech-
nology with a demonstrated value. By the same token, efforts to regulate and Limit
the use of technologies seem fine in the abstract, but in the light of experience
have been demonstrated to be ineffective.
The civi.I rights approach focuse~ on the rights of the individual, and the "con-
cern is essentially to protect or promote the individuality, dignity, or integrity of
each and every one of us." 102 The result of this' approach would be the develop-
ment of a set of"fair information practices," which supposedly would guide infor-
mation handlers. The difficulties inherent in the coexistence and differential com-
mitment to botb of these approaches are based on the recognition that the actors
who, on the one hand, have an interest in utilizing the most efficient technologies
and may only begrudingly respect the claims of individual rights must, on the
other hand, also be trusted to guarantee those rights. As David Flaherty suggests,
"It cannot be emphasized too strongly that the incentives for the government and
the bureaucracy are in the direction of invading, or at least ignoring or neglecting,
privacy interests rather than protecting them." 103 Flaherty concludes that "with-
out a privacy protection commission, it will be of dubious utility to continue to
rely 011 individuals protecting their privacy through their own initiative in the
courts and on shaping data protection legislation on a sector-by-sector basis. The
processes are simply too expensive and complicated to be accomplished without
continuing input by the specialists working for a data protection agency." 104
We should have no illusions, however, about the chances of success that a na-
tional policy and a national agency would enjoy. Flaherty's exploration of the dif-
ferential success and common difficulties faced by data protection officials in Eu-
rope and North America suggest that the problems are inherent in an underlying
conflict of interests. For example, one essential requirement of any data protection
agency if it is to actively pursue the goal of privacy protection is independence
from the political process. The OMB, which is responsible for the administration
of the Privacy Act, has historically been unconcerned about privacy and did not
A DATA PROTECTION REGIME 205

feel any pressure from the Reagan administration, because privacy was not to be
found on its agenda. It is only the occasional hearings and investigations orga-
nized by Congress, such as the investigation of matching programs in 1982, that
have forced the OMB to pay any attention to questions of privacy.

THE MARKETPLACE SOLUTION

The problems involved in maintaining appropriate control over personal infor-
mation are little different from any other social problems that are on the national
policy agenda in that we are likely to hear the same response: Let the market de-
cide. As the argument goes, the state is incapable of determining rules that should
govern the allocation of goods and resources; that is the role of the market. Per-
sonal information, to the extent that it has a value to more than the individual it
describes, can be controlled by the operation of the marketplace, with the involve-
ment of the state limited to.providing an assurance that contracts will be honored.
Even Alan Westin, whose earlier work underscored the linkage between control
over personal information and individual autonomy, has come to extend that
logic to suggest that individual freedom is best served when property rights are
determined and individuals can sell this information to the highest bidder. 105 At
one level, this marketplace approach·appears to be consistent with a notion of in-
dividual liberty- individuals choose the level of privacy with which they are most
comfortable. But not all those who support a liberty theory equate freedom in
markets with other expressive freedoms.
In an interesting and instructive discussion of the differences between the ra-
tionales behind exchange or market activities, which should be regulated, and the
private uses of property, which are expressive of freedom and individuality and
should not be regulated, C. Edwin Bake'r links market exchange with the exercise
of power, one actor over another. Because this exchange takes place in the context
of power relations and involves attempts to produce influence, regulation in the
public interest is justffied. Baker is explicit about this. "In an exchange, each party
conditions the availability of a resource on the other party doing something be or
she would otherwise not choose to do . .. . This exercise of power is inherent in ex-
change." 106 The compulsion need not be complete or absolute for us to recognize
that there is some degree of power or influence. Recall, if you can, any experience
you have had in bargaining with another over the price you will pay for some craft.
Replay the dialogue. "Sorry, I just can't take· any Jess than that." You respond,
"Well, this is my last offer." He sighs, and says, "All right, I don't want to carry all
this stuff back home with me, but you are getting a real bargain here, and I am not
making a cent!" This example is not about truth, but about a struggle, in which
each negotiator is reluctant to give up value freely. This is the nature of exchange.
Depending on the nature of the market, some are price makers and others are
price takers- a point to which l will return.
206 A DATA PROTECTION REGIME

.For Baker, these power-seeking, instrumental uses of property would thereby

include the sale or exchange of rights in personal information. Because it is ex-
change, rather than expression, a liberty theory does not require that it be privi-
leged. Furthermore, Baker's assessment of the allocative function of exchange sug-
gests that the society can and should evaluate and presumably choose between
systems of allocation "in terms of fairness, general welfare, and societal self-
definition:'107 It is under such a model that questions would be raised about the
development of a market in personal information and the need for regulatory
oversight of such a market.
What Westin and other proponents of the marketplace solution (and their
numbers are legion) seem to forget, or choose to ignore, is that all markets for in-
formation are characterized by substantial inefficiency and inequity. These mar-
kets fail almost by definition because of the nature of information as a commod-
ity. 108

Contract Problems
in the Market for Personal Information
Meheroo Jussawalla and Chee-Wah Cheah discuss the substantial inequalities that
lead to market failure. 109 They define privacy invasions as externalities that are the
products of other transactions involving individuals and corporations. An exter-
nality may be positive or negative, seen as a cost or a benefit, but its critical char-
acter is that it is not the primary focus of production, consumption, or.exchange,
but is incidental. We are most familiar with pollution as an externality that is asso-
ciated with the manufacture of industrial products. The loss of privacy can be seen
in the same way as an external consequence of the sale of those products. If we
consider, for example, a man who wishes to buy a suit to attend a wedding, to be
properly fitted,'he must disclose information about his height and, perhaps, his
girth. If he is overweight, he may not wish to make such a disclosure, but it is a re-
quirement of the transaction. He may even have to submit to the "touch" of the
tailor, who in her or his own self-interest, wants to ensure that the measurement is
accurate. Paying for the suit will involve other requests for information that would
not ordinarily be disclosed. These losses of privacy are externalities, incidental to
the purchase of a suit.
Rohan Samarajiva and Roopali Mukherjee provide additional examples of
such external costs in their discussion of the requirements of personal informa-
tion that was thought to be necessary to limit access by minors to sexually explicit
communications. 110 The loss of privacy that would occur if an individual were
forced to use a credit card or to provide a telephone number for a return call be-
cause carriers no longer wanted to provide billing services for sexually explicit
material is unquestionably an externality. The difficuities with externalities,
which underscore their importance as signs of market failure, are to be seen in the
limitations on the ability of parties to control them.
A DATA PROTECTION REGIME 207

An additional problem in the functioning of markets in which privacy losses

are involved is the notion of transaction costs. Jussawalla and Cheah suggest that
because the number of "potential 'providers' of privacy invasions" is so large and
"it is too costly foca data subject to engage in private contracting with all those
who might violate his privacy, the collective enforcement of privacy rights appears
warranted on grounds of economic efficiency:' 111
An even more important contribution to understanding the problems inherent
in market solutions is the recognition that individuals are contract takers rather
than contract makers. Unlike formal contracts and regulations governing intellec-
tual property, many of the contracts that involve the transfer of personal informa-
tion are implicit, and these standard forms "tend to place data subjects in an unfa-
vorable position" in relation to the data gatherer. Although individuals may see
the cost (in terms of value foregone) of not providing the information as substan-
tially higher than the cost of providing it, individuals frequently do not realize that
the costs associated with the loss of control over personal information occur long
after the initial transaction has taken place. The individual is not able to monitor
or control the use of that information after it is disclosed. The problem with stan- ·
<lard forms and implicit contracts is that the average consumer does not have the
time, knowledge, or skill needed to engage in negotiation of the contracts govern-
ing each transaction with privacy implications.
Jussawalla and Cheah recognize that the bureaucracies or data users might find
or claim that it is too costly to gain the consent of individuals. Indeed, the courts
have acted on the side of commerce in assuming that consumers' rights have been
waived, rather than subjecting corporations to the expense of gaining consent
through formal contracts. The challenge, which would have to be faced if either a
market or an administrative solution were sought, would be to develop a more ex-
plicit recognition of the problems inherent in gaining meaningful consent

Informed Consent
It is unlikely that individuals fully consider the consequences that might flow
from their disclosure of personal information, their agreements to permit back-
ground and status investigations, and their implicit consent for an organization to
share or exchange this information with other organizations that claim "legiti-
mate business or government" interests in such information. It is or it should be
clear that no one can be fully informed about the consequences that might result
from such blanket authorizations. Similarly, there is a question about whether
consent is ever freely given if the alternatives are far too costly. The costs of alter-
natives rise dramatically in the face of monopoly supply. But the essence of mo-
nopoly supply is realized in the context of a virtual network in which information
known to one is known to all who have access and have a "need to know." In such
an environment of informational transparency, such as that which characterizes
the market for health insurance, refusal by one insurer is a reliable predictor of re-
fusal by most others because each shares access to the same information.
208 /\ DATA PROTECTION REGIMJl

There is no right to informed consent comparable to the right that has been
recognized for privacy in the U.S. Constitution. However, to the extent that we
recognize the right to privacy as a right that enables the development ofan auton-
omous individual and involves freedom of choice, then informed consent is di-
rectly linked to the exercise of such choice. It is argued that "to respect an autono-
mous agent is to recognize with due appreciation that person's capacities and
perspective, including his or her right to hold certain views, to make certain
choices, and to take certain actions based on personal values and beliefs." 112
If the right to privacy involves the right to control access to aspects of the per-
son, then the right to informed consent has similar roots. Edward Bloustein ar-
gues that the tort in privacy is more appropriately considered to be akin to that de-
veloped for battery, in that it involves access to the personality without
permission. Informed consent has similar roots in that "the battery theory oflia-
bility protects the right to choose whether to permit others to invade one's physi-
cal integrity, and thus is based on the general right of self-determination in the
law." 113 The injury is not dependent on any physical harm that might ensue, but it
is the touch itself that is the affront. Justice Benjamin Cardoza is credited with the
first formal expression of the right in the context of a physician's responsibility,
which emphasized the liability a physician would face. "Every human being of
adult years and sound mind has a right to determine what shall be done with his
own body; and a surgeon who performs an operation without his patient's con-
sent commits an assault,.for which he is liable in damages." 114 Although there is
room to debate which standards ought to be followed in determining the suffi-
ciency of the explanation-those standards of professional convention. or those
that recognize the differences between individuals in terms of their knowledge,
situation, and responsiveness to authority-the parallels between the claims are
striking.
There is, however, little room for optimism that fully informed consent will
ever become the rule governing interactions between individuals and cor?orate·
bureaucracies, any more than it is the rule with regard to the state. Martin Gard-
ner describes the variety of ways in which the coll!ts have acted to fictionalize con-
sent, including the rise in the standard of police reasonableness, which finds suffi-
cient consent when no actual consent was received. 115 This reasonableness
standard even extends to consent that is given in response to deception. Gary
Marx describes the elevation of deception to the level of high art. Deception is jus-
tified by some as being ethical on the basis of an ends-means rationale. "If the in-
tent is noble, then the action is justified, even if it has some bad effects." 116 Marx's
comments reflect the difficulty that is inherent in the balancing of social and
moral values: "In dealing with such moral dilemmas, the problem is not only
whether we can find an acceptable utilitarian calculus, but that the choice always
involves competing wrongs. The danger of automatically applied technical, bu-
reaucratic, or occupational subcultural formulas lies in their potential for gener-
 A DATA PROTECTION REGIME 209

a ting the self-deluding and morally numbing conclusion that a costfree solution is
possible!' 117 ·

PUBLIC POLICY AND THE PANOPTIC SORT

As I suggested at the beginning of this chapter, the law is both a product and a pro-
cess that reflects substantial differences in power. Most of the time, the develop-
ment of rights under the law must be seen as a zero-sum game; in order for some-
one to win, to improve her or his lot in life, someone else must lose. This view of
the law is far removed from the naive assumptions about the operation of an effi-
cient market, in which trades are not made that worsen any trader's condition.
The negotiation of rights under the law is an adversarial process because rights are
in conflict. All too often, when the regime of rights is changed, many who lose in
the process have no way to make their voices heard. This -problem is most severe
when those changes are made in the courts and the process involves a kind of ac-
cretion or disintegration without a focal point or critical event.
The negotiation of statutory rights is no less characterized by distortions of
power and influence. Edward Laumann and David Knoke suggest that organiza-
tions differ significantly from individuals in their ability to mobilize resources, in-
cluding informational resources, to achieve goals in influencing public policy. 118
Industry lobbyists attempt to convince legislators that their interests, linked as
they are to the interests of powerful funders, depend on forging a legislative re-
. sponse that does the least harm while at the same time giving the impression of
ensuring the greatest good. The power of the Direct Marketing Association
(DMA), representing the combined interests of retailers dependent on mail and
telephone access to customers identified by computerized prospect lists, is re-
vealed as substantial in the example of the Video Privac;y Protection Act of i988
and the continuing debate over telephone caller identification.

The Video Privacy Protection Act ofi988

In the record speed with which this legislation was passed there is a bitter lesson
about power and influence in the outcome of this policy debate about personal in-
formation. Policy activists increasingly find themselves taking solace in what they
see as small victories snatched from a sea of almost certain defeat. I would suggest
that instead of victory, these measured retreats actually have served to r~ify and le-
gitimate practices that were previously carried on behind a screen of uncertainty. I
recall asking Janlori Goldman, a privacy specialist with the ACLU, to explain her
apparent change of heart with regard to a fw1damenta1 issue of privacy rights. It
appeared from her comments that she now accepted the institutionalization of the
DMA's negative option as the norm for information practice. She indicated that
no other option was politically feasible. In the realm of necessity, in which the
ACLU would have to face the DMA on bill after bill that threatened privacy inter-
210 A DATA PROTECTION REGIME

ests, Goldman felt that some ground had been gained, instead of a major battle
having been lost. 119 You may be the judge.
In June i988, Robert Kastenmeir, citing the release of U.S. Supreme Court
nominee Robert Bork's video store rental records to the local press, introduced the
Video and Library Protection Act of i988. The bill, as initially introduced, would
have prohibited the release of individually identifiable information about patrons
of video stores and libraries (library related issues would later be separated from
this bill). It would permit disclosure of information about these patrons under
rather limited circumstances, such as a court order or the "specific and informed
consent" of the user. 120
The hearings provided the occasion for the expression of strong public advo-
cacy positions from the American Library Association and the ACLU. A repre-
sentative from the Video Software Dealers' Association also testified in support of
the bill and underscored the need for limitations on government access to rental
data and apparently accepted the need for "dearly expressed written, informed
consent" on the part of the consumer. 121 However, Richard Barton, senior vice
president of the DMA, took a somewhat different position. First, he argued that
mailing lists or the use of mailing lists per se was not an invasion of privacy. The
wording of the proposed legislation, which required an affirmative consent on the
part of the consumer, would, in the eyes of the direct marketing industry, guaran-
tee that such lists could never develop. That is, in his view, requiring affirmative
consent is "frankly tantamount to prohibition for use of these lists." 122 Questioners
did not pursue Barton to determine whether this effective prohibition would occur
because individuals valued their privacy or because they were generally suspicious
of business and would not give their consent voluntarily. Industry representatives
regularly claimed that the failure of people to make use of mail and phone prefer-
ence options was evidence tliat they were not concerned about privacy.
As part of his formal testimony, Barton included numerous examples of con-
sumer options that tbe direct marketing industry would prefer to see instituted. In
effect, Barton was suggesting tliat tlie legislation formalize and provide a legisla-
tive justification for the negative option as the standard form of implied consent.
If consumers would prefer that their names not be included on a list and sold to
list vendors and other direct marketers or perhaps to government agencies or po-
litical consultants, the consumers would have to exercise the option of indicating
that preference. Otherwise, the firm would assume that such sharing of rental in-
formation had been agreed to through informal consent. · ·
Two days after Barton's testimony, the lead article in an industry newsletter,
Friday Report, highlighted the industry's position on the proposed legislation and
criticized the .Congress for introducing a potentially harmful distinction between
mail order and retail purchasing lists. In the view of the editors,

the real distinction should be the release of categories of purchase behavior for legit-
imate marketing purposes versus the disclosure of any consumer information re-
A DATA PROTECTION REGIME 211

quested on an individual for non-credit authorization purposes. Unless the re-

questor is a potential credit grantor, strict fines are warranted for release of
information on an individual. But if this legislation passes, will magazine subscrip-
tion lists be next? 123

The DMA was making a distinction between economic or financial information

and information·:a bout consumption patterns, a,s though this difference reflected
a fundamental specification of a privacy zone as beginning at the point of finan-
cial information. There was also a distinction being drawn between marketing as
a legitimate business purpose and credit authorization as a legitimate butdiffer-
ent purpose. It was at the very least disingenuous for Barton to suggest that finan-
cial information was not a highly valued component in the marketing mix. His at-
tempt at distancing is understandable, perhaps, in the light of an emerging debate
about just this issue that involved mounting public complaints against the sale of
credit information as a part of the services being offered to marketers by TRW.
On October 3, the day before the bill was scheduled for mark up, the president
and chairperson of the People for the American Way wrote to Kastenmeir to ex-
press the organization's strong support for the requirement for informed affirm-
ative consent. Yet, on November 11, Friday Report claimed a "great win" for the in-
dustry in that the bill that had just been signed by President Reagan was one that
would stand as "a congressional endor,sement and trade-off which should carry us
into. the next century." What could this be? With the leadership of Ronald Plesser,
former general counsel of the U.S. Privacy Protection Study Commission, the
DMA's Privacy Group was apparently able to engineer the insertion of two crucial
·paragraphs into the bill. 124 Under the law as signed, video operators could disclose
the names and addresses of consumers as long as they had "provided the consum-
ers with the opportunity, in a clear and conspicuous manner, to prohibit such dis-
closures"; and equally important in terms of the practices it legitimated and re-
ified, the bill specified that "the subject matter of such materials may be disclosed
if the disclosure is for the exclusive use of marketing goods and services directly to
the consumer." 125 Two years later, representatives of Blockbuster Video, one of the
nation's largest providers of video rentals, indicated that they had not even dis-
cussed the inclusion of the negative option in their membership forms, even as
they were reportedly about to embark on a program to sell their customers lists to
direct marketers. 126

Calling Number Identification

The debate over the introduction of calling number identification services pro-
vides another example through which to examine the relations between technpl-
ogy, market, and culture that are being reproduced over and over as the panoptic
sort develops into its intermediate forms. Calling number identification is just
one name for a complex of related technologies and services that provide auto-
matic delivery of calling party identification (CPI). Potentially, there are no tech-
212 A DATA PROTECTION REGIME

nological limits on the identification of the calling party. Initial proposals for a
Caller-ID service involved the fowarcling by a central office of the billing or ac-
count number for the calling party to a device, owned or rented by a residential or
small businesss consumer, that would capture and display this incoming number.
Because the calling party number was delivered during the first silent interval of
the signaling cycle, it would be possible for a person to identify ari incoming call
before the first ring was heard. Differences in the sophistication of the devices that
subscribers might use in conjunction with such a service ranged from the simple
units capable of display, capture, and storage of a limited number of calls into a rec-
ord that included the time of the call to sophisticated systems integrated into
home computers that could trigger a display of the caller's profile. It soon became
clear that an associated market of caller identification services and system en-
hancements was likely to develop. These services might supply names, addresses,
and different amounts of the information about the calling parties that were cur-
rently available through the hundreds of database vendors marketing personal in-
formation. Telephone companies or their competitors iii the enhanced services
market promised to develop and maintain files, reverse directories, and call man-
agement systems to meet the needs of small and large businesses.
As will be discussed, a related service escaped m4ch of the initial controversy
generated by Caller-ID. Perhaps because it had been presented as a business-to-
business offering, it bad somehow avoided the watchful gaze of privacy activists.
ANI refers to the delivery of the calling party identification to businesses that have
contracted for long distance 800- or premium 900-number services. The num-
bers, gathered initially for billing purposes, remain as a transaction record, and
their delivery to the receiving or called party makes them different only in form,
rather than in kind, from the delivery of the number through a Caller-ID ser-
vice. 127 AT&T's INF0-2 service, an early use of the company's integrated services
digital network (ISDN), delivered the incoming number in a way that promised to
facilitate the handling of incoming calls., as well as to contrioute to the develop-
ment of telemarketing databases. As the debate over Caller-ID matured, several
observers asked a critical question: "Why should ANI, which already is used by
'non-telephone companies' for purposes other than billing, be granted even a
'temporary' exemption from the blocking requirements imposed on Caller-ID? Is
there really a significant difference between the fact that Caller-ID is marketed as a
residential/small business service, while ANl generally is sold as a pure business
offering?" 128
The Public Service Commission of the state of New York, responding to an ini-
tiative of Commissioner Eli Noam, issued a call for comments on the general issue
of privacy in telecommunications that helped to identify several of the core issues
in the calling number identification (CNI) debates. 129 One contribution that
Noam thought an investigation by New York state might make was the establish-
ment of "reasonable expectations" of privacy in telecommunications. The hope
that discussions organized under the administrative gaze of an activist state com-
A DATA PROTECTION REGIME 213

mission would establish the bounds of reasonableness and thereby help to define
the limits oflegal protection moves to a point just short of hubris. However, the
arrogance of the claim is tempered only by the recognition that the telephone
companies favored a policy of establishment by publicity. 130 That is, if the public
could be made aware that a technology exists and is in widespread use, then rea-
sonable persons could no longer have any expectation of privacy regarding their
identification whenever they made a call. The trade and privacy press provided
numerous examples that suggested that the telephone companies were not above
creating fictional public demand for Caller-ID by requiring their employees to
write letters to the Public Utilities Commission. t 3 1
For Noam, the complex of privacy concerns could be reduced to two interests:
(1) the interests of the called party in controlling unwanted intrusions into the
private spaces individuals have defined for themselves, and (2) the interests of the
calling party in maintaining control over information about themselves. Both in-
terests are potentially affected by the technology of caller identification, and both
interests can be seen to be inextricably linked but contradictory at the point at
which commercial interests emerge. The interests of the called party in controlling
intrusion are recognized as a part of a broader privacy interest in seclusion; the in-
terests of the calling party are recognized more directly as the interests in auton-
omy that r have joined Westin in calling information privacy, or the right of deter-
mination.
Tn addition to discussing the privacy interests of calling and called parties,
Noam's memorandum identified the economic interests of the complex of busi-
nesses that utilize CNT to coordinate sales and marketing and the telecommunica-
tion service providers themselves who are interested in the income from CNl de-
livery as well as the profits to be derived from related enhancements. In addition,
interested parties include the police and emergency service providers that find
their activities both enhanced and threatened by the widespread adoption of CN[
technologies. 132 lndeed, even the growth of international businesses that depend
on telecommunications is seen to depend on CNI and an absence of regulatory
barriers to the flow of information. Thomas McManus examines the competing
interests in the information generated by telephone-based transactions in a way
that reveals the tension between the interests of individuals and the interests of
commercial firms.' 33 The importance of transaction processing for comparative
advantage in business is also explored in some detail by the U.S. Office of Tech-
nology Assessment. 134 The primary focus in this section, however, will be on what
has been misrepresented as an opposition between individuals rather than be-
tween individuals and organizations.

Called Party
As already noted, the interest in sedusion does not demand an absolute prohibi-
tion but involves exercise of the right of self-determination in that individuals
claim the right to determine which informational stimuli they will receive and
214 A DATA PROTECTION REGIME

which they will deny access. CNI, as a technology, can be seen to be privacy en -
hancing in this regard in that it provides information that people can use to
decide if a caller should or should not be granted access. Developments in the con-
trol of signaling information by digital telecommunications networks allow tele-
phone companies to offer a variety of enhanced services, some of which can be
seen to support privacy interests. 135 Bell Communications Research refers to this
category of services as CLASS, and Bell Atlantic markets its offerings as IQ Ser-
vices. Many of the proposed services promise to facilitate the identification of the
caller. For example, .calls from family members or for particular members of the
household could be given a distinctive ring. Most of these services involve some
measure of consent, or at least prior knowledge, of the calling number of persons
to be so identified. Caller-ID differs significantly in its method and in its proposed
use.
In their promotion of Caller-ID, representatives of the telephone companies
have emphasized the uninvited "intrusions" into the private sphere by individuals
making obscene or similarly harassing calls. The technology would serve a screen-
ing function by allowing the consumer to identify an incoming call by name,
number, or other means of identification and thereby avoid these unwanted mes-
sages. Opponents of Caller-ID have suggested that alternative techniques are
available that would allow screening but that would not require the same level of
identification. In several ways, the alternative technologie~ are considered to be
superior to Caller-ID in that they are potentially more efficient, and do not raise
competing privacy claims. For example, a blocking feature (Call Block) would al-
low a residential user to notify the telephone company, through the use of a sim-
ple procedure, which numbers the central office should bar from ringing through.
By itself, the Caller-ID feature would not eliminate the disturbance represented by
the ringing of an unwanted call. Unless that number could be blocked or barred
from ringing through, a form of harassment could continue in a manner that is
clearly disturbing of solitude and repose, especially if the calls were placed late at
night. An individual's knowing the number of the incoming call would not, by it-
self, eliminate such harassing calls.
Still another available service would allow the person being harassed to for-
ward the number of the last incoming call to a special unit of the telephone com-
pany or authorized agency (Call Trace). This forwarding would provide docu-
mentation of the call, and, in addition, harassment specialists would be able to
counsel the consumer about how to proceed to take formal legal action against the
harasser. For parties willing to confront harassing callers directly, another avail-
able feature (Call Return) would cause the number of the last incoming call to be
redialed and a warning of legal action or some other threat might be delivered.
None of these particular alternatives would require that the harassed party actu-
ally see a display of the number or other identifying information about the alleged
harasser, although some proposed services would allow consumers to record the
incoming number for future use. 136
A DATA PROTECTION REGIME 215

A different kind of harassment, generally not mentioned by the telephone

companies in their promotion of Caller-JD, is thatwhich describes the annoyance
and distraction that accompanies calls from telemarketers or organizations seek-
ing contributions. The level of annoyance from such calls increased dramatically
with the use of automated systems that are capable of calling hundreds of num-
bers, delivering a scripted appeal, and collecting order information, all without
the requirement of a Live caller. Those systems that would not disconnect, and that
would thereby bar outgoing calls until the prerecorded message was completed,
generated several state and federal initiatives seeking to outlaw automatic dial-
ers. 137 Caller-ID could provide only linllted control to the individual because the
thousands of numbers from which such calls might be made were unlikely to be
known to the called party.
Options that have been explored to increase the ability of individuals to control
access via telephone have been limited in part by concerns about the protections
afforded to commercial speech under recent interpretations of the First Amend-
ment. 133 Whereas a telephone preference service, such as that offered by the D MA,
that would allow individuals to indicate that they did not want to receive any tele-
marketing calls at their homes would support industry self-regulation, any regu-
lations that barred an entire class of callers because of the presumed content of
their messages immediately raise objections on constitutional grounds. 139
The alternative some courts have taken is to argue that they are regulating con-
duct rather than speech. Such an approach is pursued in the regulation of auto-
mated calls as a class and in Limitations as to times at which unsolicited calls might
be barred. The primary concern is to increase the ability of the individual to deter-
mine which calls they welcome and which they would prefer to avoid. 140 The
problem is one of the gathering of information that the caller may not wish or
may not be able to provide in advance of gaining entry via the telephone. A key as-
pect ofthis desirable information is concerned with intent. A salesperson presum-
ably calls with tl1e intention of making a sale, not with the intention of invading
privacy and disturbing the peace of each home entered electronically. Attempts to
distinguish between commercial and noncommercial or charitable solicitations
raise the similar problem of establishing any form of blanket specification of the
kinds of calls that might be limited by law.
Of course, enhancements and more sophisticated telephones might display
more detail about the incoming call, such as tllat derived from reverse directories,
but the cost of such enhancements would severely limit the usefulness of the
screen for the average consumer. Easily imaginable are technological options that
could identify the level of privacy desired by the called party, issue warnings about
that person's own definition of intrusion or harassment, and then finally leave it
up to the calling party to choose if, when, and in what manner information will be
exchanged to gain entry. The same options that lead callers with touch-tone
phones through branches and chains of automated call routing systems installed
by several large organizations might allow calling parties to identify themselves
216 A DATA PROTECTION REGIME

and the nature of their call before they could make any progress through various
levels of access. This interactive screening might approximate the kind of control
in the home that executives traditionally realize through their secretaries who fre-
quently ask for name, organization, and the nature of the call before telling you
that Ms. Highpower is unavailable at the moment. Of course, all of these options
represent substantial costs that individuals must bear if they wish to maintain
whatever level of privacy tliey enjoyed before the electronic roadway increased ac-
cess to their homes. 141

Calling Party
The connection between the interests of the calling party and those of tlie called
party is simple. The parties are often the same. When making a call to a business
or government office, individuals may, or perhaps should, have an interest in con-
trolling their identification or, at the very least, in limiting the extent to which that
information can be shared witli other parties. Generally, a person calling a doctor
would have no hesitation in providing the doctor with information about where
she or he can be reached. The doctor's office might need to call back if, for some
reason, an appointment had to be changed. There are any number of other imag-
inable reasons for the doctor or oilier service provider to have ready access to a
customer's telephone number. Emergency service providers have an easily j ustifi-
able reason for knowing the number (and address through a reverse directory) of
an incoming call. Any number of circumstances may prevent the cailer from pro-
viding this information. There is no expectation, however, tliat the doctor would
supply that number without explicit permission to a number of vendors who
might offer the caller addjtional goods and services.
Many of the discussions of Caller-ID have introduced a distraction by suggest-
ing that callers have 110 privacy interest in their telephone number. 142 In approv-
ing the offering of Cailer-ID in South Carolina, Judge Thomas Hughston argued
that if there were any interest in tlie number, it rested with the telephone company.
Consumers do not ordinarily select their own telephone number but have one as-
signed for the convenience of the telephone company. 143 By itself, the telephone
number is oflittle interest to anyone. It is only the fact that it provides a means of
access to a person identified with it that makes its disclosure to others a matter of
privacy interest. The large number of individuals who pay a fee for ensuring that
their mtmbers are not published or revealed by directory assistance suggests that
the level of privacy interest in the number as a means of access is quite high for a
great many individuals. Estimates published by the Library of Congress put the
proportion in California at 55 percent, 144 with the prize going to Las Vegas with
more than 62 percent of telephone subscribers opting out of the telephone direc-
tory.145

But the number itself is not the concern. The concern arises when the number
is associated with an address and, through the address, w ith additional informa-
tion that serves to define an individual. 146 There were apparently enough cons um-
A DATA PROTECTION REGlME 217

ers who wished to avoid leaving an audit t1rnil that a firm charging $2 per minute
for domestic calls and $5 per minute for international calls opened for business
with great fanfare in 1990. By calling a central number before completing the in-
tended call, all records of the telephone contact between the calling and the called
party would be eliminated.147 Furthermore, when the name, address, and tele-
phone number are associated with a particular activity, the concerns about the
panoptic sort reach a more serious level. Finally, when the telephone call itself
serves to add to the profile of the citizen or consumer, the call for a legislative re-
sponse begins to be heard. Although residential users of Caller-ID are unlikely to
collect and share information about persons who call their homes, it is not so
clear that small businesses wiU be similarly disinterested, and there is ovenvhelm-
ing evidence that larger businesses are making use of ANI information to develop
marketing databases that aid the panoptic sort.
It is possible to think of telephone transaction-generated consumer lists as
guides to the homes of the stars. Persons who would be included on a prospect list
for a telemarketer because they either made purchases by phone or made an in-
quiry about some product they saw advertised on television have not acted know-
ingly to invite these calls. Although, as a matter of policy, the providers of some
lists might not include persons with unlisted telephone numbers, a great many
others are not so constrained. But focusing on whether the phone number is un-
listed or not misses the point. The fact that the telephone number has been listed
at all does not mean that it has been listed in relation to a particular call or class of
calls the subscriber might make. It is the association of the number called with the
calling party identification that produces an indication of interest or orientation
that an individual may or may not be willing to disclose or have disclosed to a
third party.
With an answering machine, the calling party has the option of heeding the
usual request and leaving information about the date, time, and purpose of the
call. The cajoling cre-ativity you hear in many of those recorded messages suggests
that only a proportion well short of ioo percent of the callers willingly comply. The
Caller-ID technology automatically leaves a record of each call fod the time at
which it was made unless the callers have been allowed to block the forwarding of
their calling number or have placed the call from a public telephone. Although we
might not expect the courts or the legislatures to actively support the privacy in-
terests involved in the social lie that is threatened by Caller-ID-no more calls to
one's spouse claiming to be working late in the office or no more calls to the office
claiming to be in bed with the flu-other privacy interests are very much at risk in
an era of CNI. 148
There is little doubt that real-time CNI can be used by business to improve the
efficiency and quality of the services delivered. Automated billing for premium ca-
ble television services could be facilitated by some form of CNI; billing for other
services facilitated through telecommunication seems to be a natural gain in effi-
ciency. The identification/authorization link represents potentially valuable sav-
218 A DATA PROTECTION REGIME

ings of time and effort. The number of potential enhancements in the efficient
provision of service to customers grows with every passing moment. According to
one source, "an inbound telemarketer can pull up a customer record based on a
telephone number, personalize the greeting to the calling party and associate the
calling number with an existing account record, such as ordering information:' 149
The improvement of relationships with existing customers through this use of
CNI by businesses provides little room for complaint. 150
However, critics of CNI have suggested that there are other uses of this techni-
cal capacity that are more problematic. In addition to providing information that
adds to a consumer profile, the involuntary identification of the calliJJg party can
facilitate economic redlirung or the provision of differential quality service based
on an assessment of the quality of the neighborhood from which the call was
placed. Economic discrimination on the basis of race is a fact oflife that we cannot
simply wish away. African-Americans who do not "sound black" on the telephone
when they call to ask about a job but who find all the jobs have miraculously been
filled when they come in to apply in person might see the number of such disap-
pointments decline in an era of CN1. CNI increases the possibility that callers
from communities identified as likely to be poor and African-American may be
routed to long queues or to messages indicating that loans, insurance, apart-
ments, or jobs of interest are no longer available. They will save a trip downtown
or out to the suburbs, but they still will not get that job.
Innovative suppliers of information services are already providing a broad ar-
ray of caller information to add to the billing number passed by AT&T and other
carriers. Although not initially available in real time, lnfomedia Corporation pro-
vided "800 ID" services, which included "complete name and address, with up to
24 demographic characteristics about the caller:' 151 A representative from MCI
described a variety of justifications for using real-time CNI functions that in-
cluded security and custom answering; a variety of database "lookup" activities;
specialized routing options that would forward calls to specialized locations, such
as to operators able to provide special skills needed by customers; and a form of
triage or prioritization, in which the "platinum" customers could receive the
highest standard of service. 152 James Rule and Paul Attewell have described the use
of computerized systems to discriminate among incoming telephone calls to a taxi
service. The possibilities for applying this form of the panoptic sort are almost un-
limited:
Previously, the approximately ten telephone operators/dispatchers on duty would
perform these discriminations themselves, answering each call and then deciding
how rapidly to dispatch a cab. Now this process is computerized. The computer
classifies each incoming call according to its potential profitability; lowest priority
calls from the general public and ascending priorities to corporate subscribers ac-
cording to the fees they pay. 153

With enhanced CNI, some consumers from "low profit/high risk" neighborhoods
will find that they have great difficulty in "getting through" when they call for tax-
ies, pizza, or other services.
A DATA PROTECTION REGIME 219

The Need for Caller-ID

The carriers who have made substantial investments in the installation of Signal~
ling System Seven (SS7) enhancements believe that much of the expected value to
be derived from its advanced digital signaling features are placed at substantial
risk if consumers are provided with free blocking. According to one analyst, "With
blocking, the take rate is expected to be reduced by 50 percent, which would make
the billion dollar investment much more risky and therefore much more pro-
longed and delayed in terms of its roUout." t 54 Pacific Bell estimated that expected
revenue from Caller-ID with blocking would be as much as 30 percent lower than
if the service were offered without blocking. 155
In his testimony before the Pennsylvania Public Utility Commission, Gary
Marx raised a critical issue regarding limitations on the lines of business that tele-
phone companies might profitably pursue. "If Bell can establish the right to give
out now private information about a caller-their telephone number- the prece-
dent may very well permit the Telephone Company to transmit with a call all sorts
ofother data.... If the calling party gives up aJJ rights when they make a call, as the
Telephone Company appears to argue, then there would not appear to be any log-
ical restraint on the transmission of such data:' 156 Rohan Samarajiva, in his com-
ments submitted to the New York Public Service Commission, called attention to
evidence that the FCC had been concerned about the telephone companies' use of
their privileged access to customer network proprietary information (CNPI),
which would place them at a competitive advantage over other providers and ven-
dors.157
A year later, the American Newspaper Publishers' Association (ANPA) raised a
similar cry of alarm under the banner of pr ivacy concerns following the removal
of restrictions on telephone company entry into information businesses from
which they had been previously barred. Cathleen Black, CEO and president of the
ANPA, argued that "up untiJ now, the Bells have had no incentive to fully exploit
this information ... but the very second they get into the content end of the infor-
mation business, you can bet that they'll take advantage of every fact they know. If
you call an auto repair shop, the Bells will have that information in a computer.
They'll be free to sell that information to local car dealers .... Call a marriage
counselor, and the next thing you know a divorce lawyer may call you.... I hate to
even think about what they might try to sell you after you call that '976-LUST'
line." 158

Negative Versus Positive Option

To date, the DMA has been rather successful in convincing policymakers that the
negative option is the most efficient and socially responsible way to regulate busi-
ness information practices. The negative option assumes that reasonableness is
defined by openness to commercial, charitable, and political speech. The only
time that limits on such speech should come into play is when individuals have in-
dicated that they are not interested in receiving such information. Individuals are
220 A DA'D\ PROTECTION REGIME

expected to indicate this status by choosing the negative option, indicating, for ex-
ample, that they do not wish to receive any direct mail solicitations. The negative
option is a highly inefficient technology: It is all-or-nothing, it is relatively nonse-
lective, and it is fairly nonresponsive. That is, by notifying the DMA, you have
presumably notified those of its members who will invest in the added cost of
purging their lists of all such individuals that you do not wish to receive any solici-
tations from organizations with which you do not already have a relationship.
This is not a status that changes without action on your part, and to the extent
that its operation depends on routine checking by all telemarketers, it is likely that
calling lists will be out of date if you diange your status periodically. It is nonselec-
tive because it is a blanket denial. There are few opportunities for the individual to
specify the class of solicitations that might, under certain circumstances, be ac-
ceptable. It is nonresponsive in that it is time-consuming for the consumer as well
as for the DMA to maintain the up-to-date status of the system.
An alternative approach would be based on the assumption that individuals
generally prefer solitude. When they wish information or an information-based
service, they will seek it out. It is not unreasonable to assume that individuals
would be the best judge of when they are the most interested and therefore most
receptive to information ofa particular kind. Others with information to provide
ought to assume that, unless requested, no information is desired. This would be
the positive option. Through a va!iety of means, individuals would provide a pos-
itive indication that yes, I want to learn, hear, see more about this subject at this
time. Individuals should be free to choose when they are ready to enter the market
for information.
There may be an implied contract in which it is assumed that individuals ac-
cept, as that part of the cost they must pay for access to information and entertain-
ment, having to sit through advertisements between segments of a television pro-
gram.159 People also have apparently come to accept, perhaps begrudgingly, that
the price they must pay for information through "talking yellow pages" or other
advertiser-supported audio- text services is exposure to a brief message. But there
is also considerable evidence that people pay less attention to messages of limited
interest, which adds to the inflationary spiral as advertisers attempt to find new
ways to attract and maintain the pottential consumer's attention. lt simply is not
the case that we assume that the price we pay for having a telephone is a blanket
invitation to a commercial appeal.
Segmentation and targeting is a marketing strategy pursued by business to re-
duce the likelihood that a message is received by an individual who has a near zero
chance of responding affirmatively to a particular appeal. In reality, by allowing
individuals to choose messages in which they are interested, the probability of a
successful relationship is much higher. This aspect of interest-determining infor-
mation seeking explains the value of the yellow pag~ as an information service.
The options are alphabetized and indexed or are searchable by subject of interest.
A DATA PROTECTION REGIME 221

Th e value in the positive option is its preservation of the individual's right to

choose. Information about past purchases of chocolate, combined with informa-
tion that the person is a repeat buyer of"clothing for the big woman:' might pro-
vide the marketer of chocolate (or other vendor ofsweets) with a basis for betting
that there may be an interest in their product. Information that indicates that this
person has also been a customer 9f several diet and exercise programs and has ~r,­
rolled in these programs in the late spring of most years may also suggest that her
power to resist an appeal is not great. Information from coupon promotions may
even provide data about her price elasticity of demand for chocolate at different
times of the year. All this information suggests vulnerability. But if this woman is
trying to get through this particular winter with her weight under control, she is
not likely to seek out this information on her own. With unhindered telemarket-
ing, fed with information generated from caller-identified transactions, the posi-
tive option is not hers to choose.
The fact that so few people call or write to have their names included on a no-
call list is not an indication that people enjoy receiving telemarketing calls, any
more than the small number on the mail preference list is proof that direct mail is
a response to consumer demand. Processing the reams of direct mail many people
receive is less intrusive and less annoying than a telemarketing call. However, if the
technology of mail delivery were such that a person would have to answer the
door for each piece of mailed appeal, the public response would surely be at least
as hostile, if not even more so, than the reponse to telephone solicitations.
Full line blocking is the telephonic equivalent of the positive option. It assumes
that the normal state of teleph one calls is that which obtains at the present time,
anonymity with regard to the incoming call. It should be up to individuals to de-
termine if and to what extent they are willing to identify themselves in response to
requests from those on the other end. However, success in defining the positive
option as the preferred option, more consistent with my vision of individual au-
tonomy and the privacy that protects and nurtures it, seems unlikely.

FAIR INFORMATION PRACTICES

It is not that there is no legal safety net that limits the operation of the panoptic
sort. The ACLU has published a second edition of their handbook, Your Right to
Privacy, which is quite comprehensive in its coverage of constitutional, statutory,
and common law protections. 160 An impressive review that focuses on computer-
based information systems has been provided by Meredith Mendes. 16 1 What we
have is a safety net full of holes, one in need of maintenance and repair, yet all we
see on the horizon are signs of a storm, a tsunami with the potential for tearing the
net to shreds. Numerous sources describe the principal areas in which there has
been success and identify those areas in which technology and common practice
have eroded gains seemingly overnight.
222 A DATA PROTECTION REGIME

Critical events theorists like Laumann and Knoke 162 would have us emphasize
key decisions by the U.S. Supreme Court, such as that in U.S. v. Miller, which serve
to accelerate change through the establishment of some new standard or princi-
ple.163 Others would have us look more closely at the power of interest groups in
limiting the reach of privacy supports at the point at which they threatened collec-
tive interests.
Although success by privacy advocates in the legislative arena has been difficult
to claim, the threat of adverse publicity remains a potent weapon. The Lotus
Marketplace case represented an example of a grass-roots (if elite) mobilization
that received timely and supportive press coverage. 164 The Wall Street Journal has
continued to play a curious role in publicizing challenges to the legitimacy of the
information practices of the credit and direct marketing industry. In December
i990, Michael Miller, who has been the source of a number of articles critical of
business information practices, adopted an alarmist stance in reporting on al-
leged plans by Blockbuster Entertainment to market customer data. The article
began with an unvarnished threat: "The next time you pick up a James Bond
movie at the world's biggest video chain, the spying may start long before you
turn on the television." 165 The article continued to describe the potential threat
that was inherent in the corporation's plan to sell information to direct marketers
and reported the simiJarities between Blockbuster's plans and practices already
common to other retail chains. Within days, the press published denials from
Blockbuster's management that they had ever made any such plans.
Although the press has played an important role from time to time in raising
publk awareness and calling attention to particularly egregious departures from
the ideals that we maintain regarding privacy and the autonomous individual,
David Flaherty has been clear in his warning that the press is an unreliable guard-
ian of those values because its attention is so easily distracted. I have already pro-
vided a daunting amount of evidence to suggest that the judicial system is incapa-
ble of holding on to the thread that runs through the various pieces of the puzzle
of torts. There is evidence that the U.S. Supreme Court has done all in its power to
strip away what remained of the thin veneer of protections that could be claimed
within the shrinking domain of reasonable expectations. It is also clear that those
legislators who might pick up the mantle of privacy in defense of constituent in-
terests would find themselves immediately under attack from a rapidly mobilized
and fully armed phalanx of corporate lobbyists, in a coordinated assault made
easy by the narrowness of any single biU's reach. This same fear of corporate revolt
has left the FCC all but powerless to do more than to form advisory committees,
announce i~vestigations, and schedule proposed rule-makings, all the while se-
cretly hoping that Congress would act on its own.
A great many commentators on the question of privacy and the law have sug-
gested the need for a comprehensive privacy initiative, 166 an appeal made almost
annually siuce the publication of the report from the U.S. Privacy Protection
Study Commission. 167 They have recommended the establishment of a code of
A DATA PROTECTION REGIME 223

fair information practices that would govern the collection, use, and exchange of
personal information, and most have called for an independent body with the
power to ensure their effectiveness. The European response to privacy concerns
has been seen to be enormously threatening to those businesses in the United
States that have come to depend on the control and the revenue that an unfettered
panoptic sort has meant. 1111990, after the publication of a draft proposal to gov-
ern the collection and sale of consumer information within the European Com-
munity, members of the DMA and their counterparts within the European adver-
tising community began to mobilize to oppose the restrictions that such a
compact would put into place. What is striking is the fact that the proposed rules
would do no more than make explicit the data protection guidelines that had been
established in the Council ofEurope in 1980, and the Organization for Economic
Cooperation and Development (OECD) in i 982. 168
David Flaherty identifies twelve principles that he suggests should apply to all
personal in formation systems under government control. 169 I see no compelling
reason to limit these principles to agencies of the state. The first principle calls for
openness, or transparency, in the sense that there should be no files containing in-
formation about individuals that are secret. This principle of openness might be
reasonably expanded to include a responsibility of notification, which informs in-
dividuals that they have become the subject of a ftle. Although this might seem, on
the face of it, to represent an administrative burden and expense, it should serve
the second and third principles well.
The second, third, and fourth principles- necessity, minimization, and final-
ity-call for limitations on the collection, storage, and use of personal informa-
tion to the maximum extent possible, primarily through limiting such collection
to that which is necessary and relevant. Simitis has suggested making the data col-
lectors responsible for demonstrating the necessity for all the information they
collect, and Flaherty's principle requires that the purposes need to be established
in advance of their collection. 170 The rationale underlying this principle appears
to be linked to Flaherty's sixth principle involving the control of linkages, trans-
fers, and interconnections that involve personal information. In that further in-
formation or intelligence about individuals is produced through such linkages,
which clearly were not specified in advance, each linkage represents, at least con-
ceptually, a new collection of personal information.
The fifth principle would require the identification of persons who would be
responsible for ensuring that personal information is maintained consistent with
these principles. Presumably this principle would also require the provision of re-
sources and the necessary autonomy that would allow this person or persons to
pursue the interests of privacy without fear. .
The seventh principle, that requiring informed consent, is absolutely funda-
mental to the understanding of privacy as an aspect of individual autonomy, yet it
is a principle that is rarely honored in practice. We must assume that informed
consent means consent freely given, by which an individual has meaningful op·
224 A DATA PROTECTION REGIME

tions. This principle of informed consent is an expressed preference for the posi-
tive rather than the negative option. Consent cannot be assumed in the absence of
an expressed denial. Instead, the assumption ought to be that of reserve, until
such time as panopticism no longer represents the risks that are so apparent to-
day. Consent is especially important with regard to the sixth principle, controlling
linkage and exchange. The rationale and the risks that flow from consent, as well
as the costs of refusal, ought to be made clear before any transfers or linkages pro-
ceed.
The eighth principle, that requiring accuracy and completeness in personal in-
formation systems, is potentially contradictory. To ensure accuracy and complete-
ness, a bureaucracy will demand more information, more often. It is not clear that
the principles of limitation and finality will overcome this contradiction. To real-
ize the benefits that these principles promise, Flaherty's ninth and tenth principles
call for the establishment of special rules and regulations governing access to and
use of personal information. They also call for the specification of appropriate
civil and criminal penalties for their abuse. As has been the historical lament,
"laws are meant to be broken," and "the exceptions make the rule"; the realization
of the goals that these principles are meant to support depends on the collective
will of the people to enforce the rules and to punish offenders if necessary. The
protection of individual privacy requires a level of vigilance and commitment to
these principles that simply does not exist. As Barrington Moore suggests, "It is
about as plain as anything can be that big bureaucracies are here to stay and that
attempts to restore privacy and individual autonomy by dismantling bureaucra-
cies as such are doomed to failure:' 171 The only alternative appears to be another
bureaucracy; yet the level of resistance in the United States to the idea of a data
protection agency has not diminished despite the continuing increase in public
concerns about privacy.
The eleventh principle, that which ensures individuals of the right to have ac-
cess to records in order to evaluate, challenge, and correct inaccuracies, places too
great a burden on individuals. There is little doubt that information about indi-
viduals exists in some detail in hundreds of files, most of which these individuals
are probably unaware of. It is only when there is a problem that they can trace to
the use of a particular list, such as a credit report or an insurance file, that they
might be led to request access to their files. It was only in the heat of publicity
about government "dirty tricks" that members of the public began to request ac-
cess to the files held on them by the FBI. Yet it shoµld certainly be clear by now
that a great many other files contain inaccuracies that serve to limit or constrain a
person's options. Perhaps it is not too far-fetched to imagine that people will come
to check up on their informational health in the same way that they make periodic
visits to their physicians and take periodic note of their own physical health status.
A list of preventive screenings of classes of files may come to be as important and
as routine as periodic checks of weight, blood pressure, and cholesterol levels.
A DATA PROTECTION REGIME 225

Flaherty's twelfth principle, the right to be forgotten, to become anonymous,

and to make a fresh start by destroying almost all personal information, is as in-
triguing as it is extreme. It should be possible to call for and to develop relation-
ships in which identification is not required and in which records are not gener-
ated. For a variety of reasons, people have left home, changed their identities, and
begun their lives again. If the purpose is nonfraudulent, is not an attempt to es-
cape legitimate debts and responsibilities, then the formation of new identities is
perfectly consistent with the notions of autonomy I have discussed. The courts
and the legislatures have developed ways to make it possible for corporations to
even reduce their obligations through the declaration of bankruptcy. Corpora-
tions, unlike individuals, can be rather easily dissolved and formed anew on ac-
tion of their boards of directors. Why should corportions as fictional persons al-
ready have rights that natural persons still long to enjoy?
 8
CONCLUSION

AND SO TO CONCLUDE
In Majid Tehranian's book, Technologies ofPow~r, we are left with two options, to-
, talitarianism or communitarian democracy.' Realist to the end, Tehranian also
suggests that the communitarian option might be opposed by two second-best al-
ternatives: limited success or co-optation. The future is indeterminant, and the
trajectory remains hidden because the past is never, ever, truly repeated. Some
form of r ehabilitated democracy is a common theme in the final pages of most
works that offer comment on contemporary society. Samuel Bowles and Herbert
Gintis suggest that our future lies in the direction of a "post-liberal democracy."2
This future is also uncertain, dependent as it is on a revolutionary expansion of
personal rights against the competing expansionary claims of property. These
more democratic futures depend on the successful conversion of liberal individu-
alism into a collective awareness of common interests, which will transform the
discourse of rights into a discourse of radical empowerment. These visions are
steadfastly idealist and resist the pessimism that flows from a more structuralist
theory of domination.
Can we agree with Tehranian that there is no telos, or essential purpose or end,
that is inherent in the technology that we have defined as the panoptic sort? Can
the technologies developed during the control revolution ii1 late capitalism be
transformed to serve a democratic purpose, or is such a system of control inher-
ently antidemocratic? 3 Does workplace democracy need a system of disciplinary
surveillance? Does a democratic public sphere need political strategists armed
with sharply focused citizen profiles? Does an efficient market need consumer re-
search? Advertising and promotion? Segmentation and targeting? Or are these ac-
tivities incompatible, mutually inconsistent, contradictory, and antagonistic to
the notion of free actmg, fully informed rational producers and consumers?
It has been and remains my view that the panoptic sort is an antidemocratic
system of control that cannot be transformed because it can serve no purpose
other than that for which it was designed-the rationalization and control of hu-
man existence. This is a different vision from that which I once held as a youth.
Then, social engineering was a good thing. Social engineers would operate the
panoptic system in the interests of the "World Community." Social engmeers

227
22S CONCLUSION

wou\d correct problems in people just as surely as civil, chemical, and electrical
engineers and aerospace technicians corrected problems in the flow of rivers, the
fertility of the soil, and the time it took to get from here to there. Yet today, envi-
ronmentalists are not alone in their assessment of the consequences of allowing
these engineers and their employers a free hand in bringing nature under the con-
trol of science. We are seriously at risk. Estimates vary widely, but the ranges be-
tween ten and one hundred years do not speak well for the changes in the quality
of life we may experience anywhere along the road to an almost certain global ca-
tastrophe.
We have little reason to rejoice about the success of social engineering either.
The insanity of the urban core reflects a hopelessness that is reproduced by the op-
eration of the panopic sort-a discriminatory technology that selects out and re-
wards self-identification as deviant and dysfunctional and increases the sharpness
of distinctions that are then reified and institutionalized. Panopticism identifies,
breeds, cultivates, and reproduces failure.
Robert En tman's book, Democracy Without Citizens, talks about the emergence
of an American democracy in decline. This political environment, which Entman
describes as a "spiral of demagoguery, diminished rationality in policymaking,
heightened tendency toward symbolic reassurance and nostalgic evasion of con-
crete choices, and ultimately misrepresentation of the public," is a joint product of
two institutions, the government and the press, independently pursuing strategic
rather than democratic goals.4 Entman's solutions for the problems of American
democracy do not give one hope: government financing of national news organi-
zations run by the major political parties is a solution that moves as close to disas-
ter as anything I could imagine.5 But if not this, by what means are we to realize
participatory economic democracy? Is the development of a movement, a U.S.
Green party, for example, of the sort which Tehranian seems to suggest, some-
thing that we will approach with the aid of specialists, professionals, or strategists?
Will we mount a direct mail campaign? Is this a process that involves leadership?
Another vanguard perhaps? Will the state stand by? How will the corporate giant,
Culture, Inc., respond to our "please, hold" while we get our new political act to-
gether so we can take it on the road?6
How is it that the emancipatory and critical project of Jurgen Haber mas is to be
realized when the the actors in the period of transition to democracy believe they
must act strategically rather than democratically?

If actors are interested solely in the success, i.e., the consequences or outcomes of their
actions, they will try to reach their objectives by influencing their opponent's defini-
tion of the situation, and thus his decisions or motives, through external means by
using weapons or goods, threats or enticements. Such actors treat each other strate-
gically. In such cases, coordination of the subjects' actions depends on the extent to
which their egocentric utility calculations mesh.7
CONCLUSION 229

Habermas argues that "agreement in the communicative practice of everyday

life rests simultaneously on intersubjectively shared propositional knowledge, on
normative accord, and on mutual trust." 8 None of these requirements obtains at
any scale that could be called substantive, and the operation of the panoptic sort,
which is a strategic rather than a communicative technology, does not require nor
support their attainment. The panoptic sort reduces the need for communicative
understanding by increasing the isolation of individuals. At its ultimate level of
development, each individual might have contact only with the network (after aJI,
what needs can we imagine that cannot eventually be met more efficiently
through an automated panoptic system?). Although trust might be valued instru-
mentally, i.n the absence of viable alternatives trust becomes a luxury, rather than a
structural requirement.
The ethical principles that must be shared before communicative action can
leave strategic communication to history cannot be infused in the same way that
stannous fluoride can be introduced into the water supply. Moral development is
a process oflearning, one that takes historical time. Habermas cites assessments of
the level of moral consciousness in the United States that place more than half of
the population below the "postconventional" target, and others have shown peo-
ple to be quite capable of regression to prior stages of moral development. 9 To
subject this process of moral development to la technique, perhaps even using the
panoptic sort to facilitate the identification of those in need of remedial work,
cannot serve the desired ends.
The privacy agenda, which includes the creation of an independent bureau-
cracy that would have the responsibility for ensuring the survival of privacy inter-
ests, is similarly problematic. David Flaherty is quite sober in his recognition of
the difficulties that privacy commissioners face.

The harsh reality is that data protectors run the risk of being only a tiny force of ir-
regulars equipped with pitchforks and hoes waging battle against large technocratic
and bureaucratic forces equipped with lasers and nuclear weapons. This is especially
true for their essential work in the public sector, where they are not simply a part of
the government, but the primary protector of citizens in their relations with the gov-
ernment itself. ·
The issue is essentially one of power.... ln terms of external conflicts over power
relations, data protection agencies are squeezed between power holders and the
powerless in trying to foster public support for their goals. 10

Flaherty also recognizes the inherent tension that keeps the state from acting
aggressively to restrict the development and use of any technology, especially new
forms of information technology that are seen to be the wave of the future, which
might carry a troubled economy into the next Kondratieff upswing. Because of
these pressures, and his self-proclaimed status as an optimist, Flaherty holds on to
the hope that a data protection agency, which will articulate and pursue privacy
interests on a continuous basis, might keep the pan optic system under control. He
230 CONCLUSION

is entitled to his dreams. But because he fails to pay sufficient attention to the im-
portance of disciplinary surveillance to the survival of corporate capital,
Flaherty's analysis ignores a critical dynamic that makes the assumption of gov-
ernment control unrealistic.
The panoptic sort does not engender trust and a sense of community. Quite
the opposite is the result. And, as we understand the notion of deviation amplify-
ing, positive feedback loops in the general theory of systems, growing mistrust
leads to expanded surveillance, and each cycle pushes us further from the demo-
cratic ideal.

WHAT IS TO BE DONE?
Jacques Ellul describes a future in which the technological system has reached its
highest level of integration. It is an image of the future that leads one to ask
"What's v.'l'ong with this picture?"

lt will not be a universal concentration camp, for it will be guilty of no atrocity. It

will not seem insane, for everything will be ordered, and the stains of human passion
will be lost amid the chromium gleam. We shall have nothing more to lose, and
nothing to win. Our deepest instincts and our most secret passions will be analyzed,
published and exploited. We shall be rewarded with everything our hearts ever de-
sired. And the supreme luxury of the society of technical necessity will be to grant
the bonus of useless revolt and of an acquiescent smile. 11

My sense is that this is not the kind of future that any of us would design. 12 It is not
the future of our dreams and fantasies. But it is the future that is promised by the
panoptic sort, and it is a future that we can see in faint outline.
We are, as Stewart Brand suggests, engaged in the work of "inventing the fu-
ture," but this future, as Marx reminds us, is never faithful to our design. This is
true in large part because each of us has our own incomplete vers.i ons of the more
complete design. Indeed, because the design process is ongoing, many of us are
working with versions that are obsolete, that have been replaced or superseded,
but somehow we missed the notice or discarded the mailing. Whether through
forgetfulness, carelessness, or childlike stubbornness, some of us refuse to join the
project and climb aboard this train as it begins to pick up speed.
It is the work of critical scholarship to raise doubts in the minds of the other
passengers, to give voice to their unspoken concerns about the competence of the
engineers, to validate their mistrust of the digitized voices that announce the next
station or the final destination. lt is the work of critical scholarship to speak to the
engineers, to wonder aloud with them about whether the tracks will carry a train
this long, tills fast, that far.
In L. Frank Baum's great story, The Wonderful Wizard ofOz, we are provided a
vantage point from which to see great trickery and illusion. When Dorothy and
CONCLUSION 231

her friends came before the Wizard, each saw a different representation. Dorothy
saw Oz as a great head, the scarecrow saw a lovely lady, the tin woodman beheld a
terrible beast, and the lion envisioned a ball of fire-all illusion. Within the
panoptic future, addressability and verifiability mean that it is much more likely
that each of us will be exposed to a different, customized, administratively tailored
image of our immediate environment, our risks, our options, and the opportuni-
ties for the realization of our dreams. In the Wizard of Oz, it was Toto, scared by
the lion's roar, that knocked over the screen and revealed the Wizard as a "little,
old man, with a bald head, and a wrinkled face," rather than an all-knowing, all-
seeing, and all-powerful granter of wishes (who always demanded something in
exchange). Perhaps because it takes more energy than one can, or perhaps should,
bring to bear to knock down the screens around the pauoptic machine, critical
scholarship should be focused on generating small holes or tears in the screen that
will allow others to see more clearly how the illusion is produced.
My project is not the lion's roar, just a tiny rent in the screen. There is much
more to be seen. Make a hole for yourself, or help me to widen the one that I have
already begun.
NOTES

CHAPTER 1
1. Henry L. Wells was the director of Credit Research, Spiegel, Incorporated. The point-
ing system W<tS described in a seminar paper entitled "New customer credit reporting sys-
tem;' pp. 4-21. In a photocopied collection entitled Numerical Pointing Plans for Evaluating
Consumer Risks. The Second Consumer Credit Symposium. University of Pennsylvania,
January 10, 1963.
2. Kevin Robins and Frank Webster, "Cybernetic capitalism: Information, technology,
everyday Life;' pp. 44- 75. In V. Mosco and J. Wasko [eds), The Political Economy ofInforma-
tion. Madison: University of Wisconsin Press, 1988.
3. Webster's New Collegiate Dictionar;i. Springfield, MA: G. & C. Merriam, 1974·
4. A similar definition from the Oxford English Dictionary 1s pr~vided in Gerald R.
Winslow's Triage and Justice, Berkeley: University of California Press,"1982, which provides
a detailed exploration of the inequities and moral dilemmas involved in medical triage.
5. Webster's Third New International Dictionary (unabridged). Springfield, MA: G. & C.
Merriam, 1976.
6. The notion of information subsidies suggests that actors with the reasons and the re-
sources necessary to produce influence over the actions of others will do so in part through
the provision of information that favors one option over another. This information is seen
as a subsidy because it reduces the costs an individual would face if she had to gather the in-
formation on her own. See Oscar H. Gandy, Jr., Beyond Agenda Setting: Informatiori Subsi-
dies and Public Policy Norwood, NJ: Ablex, 1982.
7. Jurgen Habermas, The Structllra·l Transformation of the Public Sphere: An Inquiry into
a Category ofBourgeois Society. Cambridge, MA: MIT Press, 1989.
8. Stephen K. White, The Recent Work ofJurgen Habermas: Reason, J11Stice and Moder-
nity. New York: Cambridge University Press, 1988.
9. Oscar H. Gandy, Jr., "The political economy of communications competence:' pp.
108-124. In V. Mosco and J. Wasko [eds), i988, op. cit. ·
10. Randall Bartle~t, Economics and Power: An Inquiry into Human Relatio~ships and
Markets. New York: Cambridge University Press, 1989, pp. 3-8.
n. T. B. Bottomore, Karl Marx: Selected Writings in Sociology and Social Philosophy. New
York: McGraw-Hill, 1964.
12. David J. Sholle, ''.Critical Studie.~: From the Theory of Ideology to Power/
Knowledge." Critical Studies in Mass Communication 5 (March i988):16-41.
13. Nicholas Garnham, "A political economy of mass communication," pp. 24-30. In
Capitalism and Communication: Global Culture and the Economics ofInformation. London:
Sage Publications, 1990.

233
234 NOTES

14 Robert Horwitz, The Irony of Regulatory Reform: The Deregulation ofAmerican Tele-
communications. New York: Oxford University Press, 1989.
15. Anne Branscomb, "Property rights in information," pp. 603-642. ln M. Gurevitch
and M . Levy [eds], Mass Communication Review Yearbook. Vol 6. Newbury Park, CA: Sage,
1987; Anne W. Branscomb, Who Owns Information? Occasional Paper No. 2. New York:
Gannett Center for Media Studies, May 1986.
i6. Amitai Etzionj, Genetic Fix. New York: Macmillan, 1973; Troy Duster, Backdoor to
Eugenics. New York: Routledge, 1990.
17. Karl Marx, from the preface of his Contribution to the Critique of Political Economy
[1859J excerpted in Botto more, 1964, op. cit., p. 52.
18. Dan Schiller, "How to think about information:' pp. 27-43. lo V. Mosco and J. Wasko
[eds], The Political Economy ofInformation. Madison: University of Wisconsin Press, 1988.
19. Brian Burkitt, Radicnl Political Economy: An In troduction to the Alternative Econom-
ics. New York: New York University Press, 1984; Howard J. Sherman, Foundations ofRadical
Political Economy. Armonk, NY: M. E. Sharpe, 1987; Richard D. Wolff and Stephen A.
Resnick, Economics: Marxian Versus Neoclassical..Baltimore, MD: Johns Hopkins Univer-
sity Press, 1987.
20. Joseph .A. Schumpeter, Cqpitalism, Socialism and Democracy. 4th ed., 11th impres-
sion. London: Unwin University Books, 1966, p. 28.
21. Arun Bose, "Modern Marxian political economy," pp. 90-115. In David K. Whynes
[edl, What is Political Economy? Eight Perspectives. New York: Basil Blackwell, i984. Bose
suggests the emphasis was in the original, but the reference qid not include the citation
from Marx, referring instead to his 1980 volume Marx on Exploitation and Inequality.
22. Dan Schiller, 1988, op. cit., p. 36.
23. Jacques Ellul, What I Believe (trans. Geoffrey W. Bromiley). Grand Rapids, Ml:
Eerdmans Publishing, 1989, p. t35.
24. Ibid., p. l37·
25. Ellul might have derived this from the orientation of Marx, who envisioned the real-
ization of man's potential as the essence of liberty, which could only be obtained after the
development ofproductive capacity made such personal development possible.
26. William Julius Wilson, The Truly Disadvantaged: The Inner City, The Underclass, and
Public Policy. Chicago: University of Chicago Press, 1987.
27. Ellul, 1989, op. cit., pp. 138- 139.
28. C. George Benello, "Technology and Power: Technique as a Mode of Understanding
Modernity," pp. 91-107. In C. Christians and J. Van Hook [eds], Jacques Ellul: interpretive es-
says. Champaign, IL: Unjversityoflllinois Press, 198L
29. Jacques Ellul, The Technological Society (trans. John Wilkinson). New York: Vintage
Books, 1964, p. viii.
30. Similar criticisms have been lodged against the otherwise impressive analysis of in-
formation technology published by James Beniger, The Control Revolution: Technological
and Economic Origins ofthe Information Society. Cambridge, MA: Harvard University Press,
1986. See, for example, G. ). Mulgan, Communication and Control: Networks and the New
Economics ofCommunication. New York: Guilford Press, 1991.
31. Joseph A. Schumpeter, 1966, op. cit., pp. 81-86.
32. Rogers Brubaker, The Limits ofRationality: An Essay on the Social and Moral Thought
ofMax Weber. London: Allen and Unwin, 1984.
NOTES 235

33- David Beetham, Max Weber and the Theory of Modern Poli.tics. London: Allen and
Unwin, 1974, p. 65.
34. Reinhard Bendix, Max Weber. An Intellectual Portrait. Garden City, NY: Doubleday
(Anchor), 1962, pp. 458- 459.
35. Beetham, 1974, op. cit., p. 242.
36. Frank Webster and Kevin Robins, Information Technology: A Luddite Analysis. Nor-
wood, NJ: Ablex, 1986, P-73-
37- Robert G. Meadow, "Political campaigns, new technologies and political competi-
tion," pp. 5- 16. In R. Meadow [ed], New Communication Technologies in Politics. Washing-
ton, D.C.: Washington Program of the Annenberg School of Communications, 1985.
38. Christopher Dandeker, Surveillance, Power and Modernity: Bureaucracy and Disci-
pline from 1700 to the Present Day. New York: St. Martin's Press, 1990, p. 9.
39. Robert J. Holton and Bryan S. Turner, Max Weber on Economy and Society. New
York: Routledge, 1989,-p. 66.
40. Brubaker, 1984, op. cit, p. 19.
4i. Meheroo Jussawalla and Chee-Wah Cheah, "Economic analysis of the legal and pol-
icy aspects of information privacy," pp. 75-102. In M. Jussawalla and C. Cheah [eds], The
Calculus ofInternational Communications. Littleton, CO: Libraries Unlimited, 1987.
42. Brubaker, 1984, op. cit., p. 42.
43. Randall Bartlett, Economics and Power: An Inquiry into Huma11 Relations and Mar-
kets. New York: Cambridge University Press, 1989.
44. Michel Foucault, Discipline and Punish: TI1e Birth of the Prison (trans. Alan
Sheridan ). New York: Vintage Books, 1979; especially Chapter 3, "Panopticism," pp. 195-
230.
45- Ibid., p. 201.
46. Ibid., P- 203
47. Oscar H. Gandy, Jr., and Charles Simmons, "Technology, privacy and the demo-
cratic process." Critical Studies in Mass Communication 3 (June 1986):155- 168.
48. Kevin Robins and Frank Webster, 1988, op. cit
49- Michel Foucault, "Classifying," pp. 125- 165. In The Order ofThings: An Archeology of
the Human Sciences. New York: Vintage Books, 1973.
50. Paul Rabinow, "Introduction," in P. Rabinow [ed), The Foucault Reader. New York:
Pantheon, 1984.
51. Mary Tew Douglas, How Institutions Think. New York: Syracuse University Press,
1986, p. 100.
52. lbid., p. 20.
53- Jelena Grcic-Polic and Oscar H. Gandy, Jr., "The emergence of the marketplace stan-
dard:' Media Law and Practice (1991):55-64.
54- Peter G. Moore, The Business ofRisk. Cambridge: Cambridge University Press, 1983.
55. Stanley Cohen, Visions ofSocial Control: Crime, Punishment and Classification. Cam-
bridge: Polity Press, 1985, p. 196.
56. Ibid., P- 267.
57- Rick Roderic!<, Habermas and the Foundations ofCritical Theory. New York: St. Mar-
tin's Press, 1986.
58. Stephen A. Resnick and Richard D. Wolff, Knowledge and Class: A Marxian Critique
ofPolitical Economy. Chicago: University of Chicago Press, 1987.
236 NOTES

59. Anthony Giddens, The Constitution ofSociety: Outline of the Theory ofStructuration.
Cambridge: Polity Press, t984.
60. David Held and John B. Thompson [eds], Social Theory ofModern Societies: Anthony
Giddens and His Critics. Cambridge: Cambridge University Press, 1989.
6i. Ann Showstack Sassoon (ed], Approaches to Gramsci. London: Writers and Readers,
1982.
62. George Gerbner, Larry Gross, Michael Morgan, and Nancy Signorielli, "Living with
television: The dynamics of the cultivation process;' pp. 17-40. In J. Bryant and D. Zillman
(eds], Perspectives on Media Effects. Hillsdale, NJ: Lawrence Erlbaum, 1986.
63. Anthony Giddens, The Nation State and Vialence. Berkeley: University of California
Press, 1985.
64. Ibid., p. 309.

CHAPTER 2
1. By this I am.referring to the prescrecning requirements, or "front-end verification:'
that is required if an individual applies for federally supported social welfare benefits. It
also includes the growing number of biological and genetic screens that are now required of
newborns.
2. Reference is made here to Foucault's notion of power as being based in relations and
where power does not truly exist unless the agent is free to act in more than one way. Power
is demonstrated when the choice is limited according to the wishes of power. Power is also
demonstrated when choices freely made limit future choices. See Michel Foucault, "After-
ward. The Subject and Power:' pp. 208-226. In Hubert Dreyfus and Paul Rabinow, Michel
Foucault: Beyond Structuralism and Hermaneutics. 2d ed. Chicago: University of Chicago
Press, 1983.
3. Michel Foucault, The Order of Things: An Archeology of tlie Human Sciences. New
York: Vintage Books, 1973, p. 136.
4. Troy Duster, Back Door to Eugenics. New York: Routledge, 1990, p. 97.
5. Daniel Hahneman and Amos Tversky provide a detailed and elegant discussion of the
nature of risk aversion and the absence in linearity that characteri2es the valuation of losses
and gains. "Choices, values and frames:· pp. 153-172. In N. Smelser and D. Gerstein [eds],
Behavioral and Social Science: Fifty Years ofDiscovery. Washington, D.C.: National Academy
Press, 1986.
6. Randall Bartlett, Economics and Power: An inquiry into Human Relations and Markets.
New York: Cambridge University Press, i989, p. 30.
7. Ibid., p. 46.
8. Amitai.Etzioni, The Moral Dimension. New York: Free Press, 1988, pp. 246- 247.
9. C. Edwin Baker, "Posner's privacy mystery." Georgia Law Review 12 (1978):476. Baker
argues that because existing preferences are subject to influence in that laws not only affect
the realization of preferences but also affect what preferences we have, and furtl1er, because
the influence of existing preferences is likely to be conservative rather than progressive, the
existing preferences ought not to be given as justifications for restrictions on liberty.
10. Bartlett, 1989, p. 101.
11. Samuel Bowles and Herbert Gintis, Democracy and Capitalism. New York: Basic
Books, 1987.

i1
NOTES 237

12. lbid., p. 77.

i3. This is a point made quite forcefully by Mark Hepworth and Michael Waterson, "In-
formation technology and the spatial dynamics of capital." information Economics and Pol-
icy, 3 (1988):143- 163.
14. Bartlett, 1989, pp. 127- 128.
15. Several of these approaches to policy formation are explored in Oscar H. Gandy, Jr.,
" Public relations and public policy: The structuration of dominance in the information
age:• pp. 131-163. In E. Toth and R. Heath [eds], Rhetorical and Critical Approaches to Public
Relations. Hillsdale, NJ: Lawrence Erlbaum, 1992.
t6. G. William Domhoff, The Power Elite and the State: How Policy is Made in America.
New York: Walter de Gruytcr, 1990.
17. Michael Schwartz [ed], The Structure of Power in America. New York: Holmes and
Meier, t987, p. x.
18. Edward Laumann and David Knoke, The Organization State: Social Choice in Na-
tional Policy Domains. Madison: University of Wisconsin Press, 1987.
19. Michel Foucault, Discipline and Punish: The Birth of the Prison. (trans. Alan
Sheridan). New York: Vintage Books, 1979. Thls is Foucault's core work in thls regard, al-
though later comments and expansions arc noted in Hubert Dreyfus and Paul Rabinow,
Michel Foucault. Beyond Structuralism and Hermaneutics. 2d ed. Chicago: University of
Chicago Press, 1983; and Paul Rabinow [ed] , The Fo11cault Reader. New York: Pantheon,
1984; Graham Burchell, Colin Gordon, and Peter MilJer [eds], The Foucault Effect: Studies in
Governmentality. Chicago: University of Chicago Press, 199i.
2.0. Michel Foucault, "Questions of method;' p. 84. In Graham Burchell, Colin Gordon,
and Peter Miller [eds], The Foucault Effect: Studies in Govern mentality. Chicago: University
of Chicago Press, 1991.
2.1. Foucault identifies other common dichotomies: mad/sane, dangerous/harmless,
normal/abnormal, which are common to systems of coordination and control.
22. Michel Foucault, 1979, op. cit., p. 197·
2.3. We wish to avoid the error of some critics of Foucault who seem to have focused on
the Panopticon as a structure and lost sight of panopticism as an administrative technique.
2.4. John Dinwiddy, Bentham. Oxford.: Oxford University Press, 1989, p. 8.
25. Foucault, t979, op. cit., p. 2.02..
26. Tbid., p. 205.
2.7. [bid.
28. Frank Webster and Kevin Robins, Information Technology: A Luddite Analysis. Nor-
wood, NJ: Ablex, 1986; Kevin Robins and Frank Webster, "Cybernetic capitalism: Informa-
tion, technology, everyday life;' pp. 44- 75. fn V. Mosco and J. Wasko [eds], The Political
Econdmy of Information. Madison: University of Wisconsin Press, i988; Kevin Robins and
Frank Webster, "Information as capital: A critique of Daniel Bell," pp. 95-117. In). Slack and
F. Fejes [eds], The Ideology ofthe Information Age. Norwood, NJ: Ablex, 1987.
29. Dreyfus and Rabinow, 1983, op. cit., p. t6o.
30. Foucault describes a system of classificatory codes that distinguished the very good
from the mediocre, the bad, and the "shameful" classes. FoucauJt, Discipline and Punish,
pp. 181-182..
31. Michael Lewis, "Leave home without it." New Republic (September 4, i989):i9-22.
32. fbid., p. 183.
238 NOTES

33. Foucault, Discipline and Punish, p. 189.

34. Giddens's critici.s m of functionalism in general and of creeping functionalism in
Marxist theory is discussed in his The Constitution ofSociety. Cambridge: Polity Press, i986,
pp. 15- 19, and in his A Contemporary Critique of Historical Materialism. Berkeley: Univer-
sity of California Press, 1981, pp. 293-297.
35. Giddens, 1981, p. 26.
36. Giddens, 1981, p. 46.
37. Giddens, 1981, p. 47.
38. John B. Thompson, "The theory of structuration," pp. 56-76. ln D. Held and J.
Thompson [eds] , Social Theory ofModern Societies: Anthony Giddens and His Critics. Cam-
bridge: Cambridge University Press, i989.
39. Thompson, 1989, p. 63.
40. Giddens, i986, op. cit., p. xxx.
41. Giddens, 1981, op. cit., p. 36.
42. Bartlett, 1989, op. cit., pp. 43- 44.
43. The concept of hegemony, associated with Antonio Gramsci, and the critical theory
of the Frankfort School and its revisionist emanations in the work of Jtirgen Habermas,
stand as the dominant exemplars of a Marxian response to this problem.
44. Erik Olin Wright, "Rethinking, once again, the concept of class structure:' pp. 269-
348. In Erik Olin Wright [ed], The Debate on Classes. London: Verso, 1989; Erik Olin
Wright, "Class boundaries and contradictory class locations," pp. 112-129. In A. Giddens
and D. Held [eds), Classes, Power and Conflict. Berkeley: University of California Press,
1982.
45. Wright, i989, p. 276.
46. Erik Olin Wr ight, "Models of historical trajectory: An assessment of Giddens's cri-
tique of Marxism;' pp. 77- 102. In Held and Thompson, i989, op. cit
47. Wright, i989, p. 89.
48. Selections from Max Weber, "Economy and Society:' pp. 60-n In A. Giddens and
D. Held [eds], Classes, Power and Conflict. Berkeley: University of California Press, i 982;
"Classes, Status Groups and Parties;' pp. 43-61. In W. G. Runciman, Weber: Selections in
Translation (trans. Eric Matthews). Cambridge: Cambridge University Press, i978.
49. This was a style of life which Thorsten Veblen had subjected to such unreniitting
criticism in his Theory of the Leisure Class. New Yo rk: New American Library, 1953·
50. Giddens and Held, p. 67.
51. Arthur L. Stinchcombe, Information and Organizations. Berkeley: University of Cali-
fornia Press, 1990, p. 274.
52. Ibid., p. 292.
53. £bid., p. 294.
54. The nature of the dual labor market and the problems of the urban ghetto have been
explored in some detail by William Julius Wilson in The Truly Disadvantaged. Chicago:
University of Chicago Press, i987; see also: Richard M. Cyert and David Mowery [eds],
Technology and Employment: Innovation and Growth in the U.S. Economy. Washington,
D.C.: National Academy of Science Press, 1987; and John D. Kasarda, "Urban industrial
transition and the underclass:' Annals: AAPSS 501 (January i989):26-47.
55. Stinchcombe, 1990, p. 303.
NOTES 239

56. The mainstreaming hypothesis, as a specification of the cultivation hypothesis, sug-

gests that people whose concrete social experience would generally produce quite different
social perceptions tend to have more similar perceptions ifthey share the common cultivat-
ing experience of televiewing. The movement of those in the upper classes is the opposite of
that of those in the lower classes, but both groups move toward a common mainstream. A
discussion of mainstreaming and other aspects of the cultivation hypothesis can be found
in George Gerbner, Larry Gross, Michael Morgan, "Living with television: The dynamics of
the cultivation process;' pp. 17-40. In J. Bryant and D. Zillman [eds], Perspectives on Media
Effects. Hillsdale, NJ: Lawrence Erlbaum Associates, 1986; and Nancy Signorielli and Mi-
chael Morgan (eds], Cultivation Analysis. Newbury Park, CA: Sage, 1990.
57. Stinchcombe, op. cit., p. 303.
58. Geoff Mulgan, Communication and Control: Networks and the New Economics of
Communication. New York: Guilford, 1991, p. 53.
59. Giddens, 1986, op. cit., p. q1.
60. Ibid., p. 261.
61. Oscar H. Gandy, Jr., "The political economy of communications competence," .pp.
108-124. In V. Mosco and J. Wasko [eds], The Political Economy of Information. Madison:
UniversityofWisconsin Press, i988.
62. Derek Gregory, "Presences and absences: Time-space relations and structuration
theory;' pp. 185- 214. In Held and Thompson, 1989, op. cit.
63. Gregory, 1989, p. 208.
64. These perspectives on technology are discussed in a recent book by Majid Tehranian,
Technologies of Power: Information Machines and Democratic Prospects. Norwood, NJ:
Ablex, 1990. Tehranian offers what he calls a "contextualist" perspective, by which the use of
technologies depends primarily on the existing structures of power into which they are in-
troduced.
65. Jacques Ellul, The Technological Society (trans. John \Alilkinson). New York: Vmtage
Books,1964
66. Jacques Ellul, Propaganda: The Formation of Men's Attitudes (trans. Konrad Lellen
and Jean Lerner). New York: Vintage Books, i973.
67. Ellul, 1973, p. x.
68. Ibid., p. xvii.
69. Clifford G. Christians, "Hegemony and the technological system: Gramsd versus
Ellul." Conference paper. Annual meeting of the Association for Education in Journalism
and Mass Communication, Boston, MA, August 1991.
70. See especiaUy, Antonio Gramsci, Selections from the Prison Notebooks. London: Law-
rence and Wishart, 1973; Anne S. Sassoon [ ed], Approaches to Gramsci. London: Writers and
Readers, 1982; and selections in A. Mattelart and S. Siegelaub [eds], Communication and
Class Struggle. Vol. 1, Capitalis.m, Imperialism. New York: International General, 1979·
71. Jurgen Habermas's emphasis on the influence of commercialization in the transfor-
mation of the role of the press as an aspect of the traditional public sphere and the site of
opinion formation does not reach Ellul's level of attention to questions of technique. See
his The Structural Transformation of the Public Sphere: An Inquiry into a Category of Bour-
geois Society (trans. Thomas Burger). Cambridge, MA: MIT Press, 1989.
72. John Wilkinson, "Translator's Introduction" to Ellul, The Technological Society, 1973,
p. xv.
 NOTES

73. Ellul, 1973, p. 134.

74. C. George Benello, "Tecbnology and power: Technique as a mode of understanding
modernity;' pp. 91- 107. In C. Christians and T. Van Hook [eds], Jacques Ellul: interpretive es-
says. Champaign, IL: University of Illinois Press, 1981.
75. Ellul, 1973, p. 135.
76. David Lovekin, Technique, Discourse and Consciousness: An Introduction to the Phi-
losophy ofJacques ElluL Bethlehem, PA: Lehigh University Press, 1991, p. 143.
77. Ellul, 1973, op. cit.
78. Ellul, 1964, p. 43.
79. Ibid., p. 47.
80. Ibid., p. 54.
81. Ellul, 1964, p. 171. This development ofa planning role for economics and economists
is at the heart of John Kenneth Galbraith's enfolding of technjque and propaganda in his
work on the management of demand. See his The New lndustrial State. New York: Signet,
1968; and especially his Economics and the Public Purpose. Boston: Houghton Mifflin Co.,
1973. Not even the subversive economics of E. F. Schumacher, as expressed in his Small Is
Beautiful: Economics As IfPeople Mattered (New York: Harper and Row, 1973) managed to
avo id the pursuit of technology and planning, as though calling it "appropriate" somehow
hid the fact that the concern was with efficiency and productivity.
82. I recall, and not without some pleasure, taking courses in a department at Stanford
called Engineering-Economic Systems, which was involved with assessing the teclinological
efficiency of engineering systems designed to solve social as well as material problems.
There were also courses in the Food Research Institute concerned with economics, devel-
opment, and national planning clearly not limited to agriculture; another school offered
courses in the economics of education, which mused about "investments in human capi-
tal" and "educational production functions." In association with the medical school, Victor
Fuchs offered a popular course on the economics of health, which focused on the cboices
made by individuals following Gary Becker's notions of household prodµction, which con-
ceptualized inviduals as "producing" their own health or illness. The introduction of eco-
nomic concepts and efficiency rationales into the logic of the courts has been seen by sev-
eral commentators to have pushed aside the values of fairness, justice, Cind equity. The
spread of economics beyond its disciplinary borders shows no signs of abating.
83. Ellul, 1964, p. 411.
84. Ibid.
85. James Beniger, The Control Revolution: Technological and Economic Origins of the In -
formation Society. Cambridge: Harvard University Press, i986.
86. Dennis Wrong, "Introduction." In D. Wrong (ed], Max Weber. Englewood Cliffs, NJ:
Prentice-Hall, 1970.
87. Rogers Brubaker, The Limits ofRationality: An Essay on the Social and Moral Thought
ofMax Weber. London: Allen and Unwin, 1984, p. 2 .
88. Brubaker, 1984, pp. 3-4.
89. Christopher Dandeker, Surveillance, Power and Modernity: Bureaucracy and Disci-
pline from 1700 to the Present Da)~ New York: St. Martin's Press, 1990.
90. Robert J. Holton and Bryan S. Turner, Max Weber on Economy and Society. New
York: Routledge, 1989, p. 23.
91. David Beetham, Max Weber and the Theory of Modern Politics. London: Allen and
Unwin, 1974, p. 65.
NOTES 241

92. Holton and Turner, 1989, p. 44.

93. John Elster and John Roemer [eds]. Interpersonal Comparisons of Well-Being. New
York: Cambridge University Press, 199i.
94. Holton and Turner, p. 53.
95. See the discussion of the "emergence oflegal rationality" in Weber's political sociol-
ogy in Reinhard Bendix, Max Weber: An Intellectual Portrait. Garden City: Doubleday, i962,
pp. 391-496.
96. Bendix, 1962, p. 414.
97. Harry Braverman, Labor and Monopoly Capital. New York: Monthly Review Press,
1974.
98. Frarik Webster and Kevin Robins, Information Technology: A Luddite Analysis. Nor-
wood, NJ: Ablex, 1986.
99. This process of automation is discussed from the perspective of personal experience
by Harley Shaiken in his Work Transformed: Automation and Labor in the Computer Age.
New York: Holt, Rinehart & Winston, i984.
100. Webster and Robins, 1986, pp. 328-347.
101. Beniger, i986.
102. Kevin Robins and Frank Webster, "The revolution of the fixed wheel: Information,
technology and Social Taylorism:' pp. 36-91. ln P. Drummond and R. Paterson [eds], Tele-
vision in Transition. London: British Film Institute, 1986.
103. Beniger, 1986, p. i5.
104. Ibid., j>. 16.
io5. Spiros Simi tis, "Reviewing privacy in an information society:' University ofPennsyl-
vania Law Review 135 (1987):733.
106. This is to say nothing about the debates in science about the impossibility of mea-
suring anything that does not change that which is being measured. To this we add the in-
sight that "identity" is a construct, rather than a fact. Mario Bunge argues that not only are
repeated measures necessarily at variance with each o ther, but increasing the precision or
quality of measuremen.t only increases the unrepeatability of any measure. Mario Bunge,
Causality and Modern Science. 3d rev. ed. New York: Dover Publications, 1979, p. 267.
107. Oscar H. Gandy, Jr., "A critical challenge: The approach of political economy to the
study of communication and information." Journal of Media Economics. io7 (Summer
1992): 23- 42.
108. Mario Bunge, op. cit., p. 271.
io9. Anthony Giddens, "A reply to my critics;• pp. 249-301. Jn D. Held and). Thompson,
1989, op. cit., pp. 288- 289.
110. The Heisenberg principle that denies the possibility of measuring anything without
changing that which is being measured, however, is certainly an issue worthy of consider-
ation.
m. Eugene Webb, Donald Campbell, Richard Schwartz, and Lee Sechrest, Unobtrusive
Measures: Nonreactive Resea,rch in the Social Sciences. New York: Rand McNally, 1971.
112. Giddens in Held and Thompson, p. 289.
113. Bunge, 1979, pp. 14-19.
u4. Bunge, 1979, p. 353.
115. Bunge, 1979, pp. 330- 331.
 NOTES

n6. Thomas L. Haskell, The Emergence ofProfessional Social Science: The American Social
Science Association and the Nineteenth-Century Crisis ofAuthority. Urbana: Un iversity of Il-
linois Press, 1977, p. 16.
u7. Haskell, i977, p. 251. .
u8. Max Weber from Wirtschaft und Gesellschaft. In W. G. Runiciman [ed], Weber: Se-
lections in Translation. i978, p. 350.
u9. Christopher Dandeker, Surveillance, Power and Modernity: Bureaucracy and Disci-
pline from 1700 to the Present Day. New Yor.lk: St. Marti11's Press, 1990, p. 9.
120. The surveillance of telecommunications workers was explored in great detail in the
study of workplace surveillance by the U.S. Office of Technology Assessment, The Electronic
Supervisor. Washington, D.C.: U.S. Government Printing Office, 1987. See also: Andrew
Clement, "Office automation and the technical control of information workers," pp. 217-
244. In V. Mosco and J. Wasko feds], The Political Economy of Information. Madison: Uni-
versity of Wisconsin Press, 1988.
121. Peter T. Kilborn, "Vvorkers using computers find a supervisor inside." New York
Times. December 23, i990, sec. A.
122. Richard Barnet and Ronald Miiller; Global Reach. The Power of Multinational Cor-
porations. New York: Simon and Schuster, 1974.
123. Dandeker, 1990, p. 154· See also: Mark Hepworth, Geography of the Information
Economy. New York: Guilford Press, 1990.
124. Dandeker, 1990, p. 210.
125. Reinhard Bendix, Max Weber: An Intellectual Portrait. Garden City: Doubleday,
1962, p. 482.
126. Julien Freund, The Sociology ofMax Weber. New York: Vintage Books, 1969, p. 229.
127. Bendix, 1962, pp. 424-425.
128. Haskell, 1977, p. 89.
129. Andrew Gamble, The Free Economy and the Strong State: The Politics ofThatcherism.
Durham, N.C.: Duke University Press, 1988.
i30. Gamble, 1988, p. 184.
131. The Reagan administration's corporatist evolution began with the inyolvement of a
privately managed and funded commission: "The President's Private Sector Survey on Cost
Control:' referred to as the Grace Commission, that made extensive recommendations on
the transformation of the government. See: U.S. Congress. Senate Committee on Govern-
mental Affairs. Oversight of the Grace Commission Report. Hearing, May 9, i985. Washing-
ton, D.C.: U.S. Government Printing Office, 1985. The involvement of the conservative
Heritage Fonndation and its Mandate for Leadership represents another means through
which an attack on the bureaucracy was designed and implemented.
132. Gamble, 1988, p. 50.
133· Weber's discussion of these problems focused on a kind of double government that
might characterize an authoritarian regime, but the operation of executive power through
the Office of Management and Budget was not unlike that of an army or secret police force
charged with whipping the government bureaucracy into alignment with the new party
line. See the discussion ofthis point in Bendi.x, 1962, p. 467.
134. Wilson Dizard, The Coming Information Age. 3d ed. New York: Longman, i989,
originally published in i982.
135. Peter Hall and Paschal Preston, The Carrier Wave: New Information Technology and
the Geography of lflnovation, 1846-2003. London: Unwin Hyman, 1988. Hall and Preston
NOTES 243

suggest that the next global economic expansion might be seen as the fifth in a series of
such cyclical waves as described by Vasily Kondratieff and as widely discussed in terms of
the long waves of the business cycles explored by Joseph Schumpeter.
136. Jorge Schement and Leah Lievrouw, "Capitalism and the industrial origins of the
information society:' pp. 33- 45. In Schement and Lievrouw [eds), Competing Visions, Com-
plex Realities: Social Aspects ofthe Jnformation Society. Norwood, NJ: Ablex, 1987.
137· Marc Uri Porat, The Information EconomyVois.1- 2. Washington, D.C.: U.S. Depart-
ment of Commerce, 1976.
138. See for example, Meheroo Jussawalla, Donald Lamberton, and Neil Karunaratne
[eds), The Cost of Thinking: Information Economies of Ten Pacific Countries. Norwood, NJ:
Ablex, 1988; Raul Katz, "Measurement and cross-national comparisons of the information
work force." The Information Society 4 (4) (1986): 231-277.
139. It sh~uld be noted that a portion of Porat's work did in fact utilize an input-output
table of the United States to estimate what the employment consequences might be if there
were a reduction in defense expenditures.
i40. Meheroo Jussawalla, "lnformation economies and the deve.l opment of Pacific
countries:' pp. 15- 43. In M. Jussawalla, D. Lamberton, and N. Karunaratne, 1988, op. cit.
{41. Dan Schiller discusses what he sees as a counterproductive debate about a relatively
useless distinction that remains as a vestige of Marx's emphasis on surplus value rather than
the wage relation. See: Dan Schiller, "How to think about information:' pp. 27-44. In V.
Mosco and J. Wasko, 1988, op. cit. A classical Marxian analysis of the information economy
that explores the implications of a growing pool.of nonproductive workers is to be found in
Patricia Arriaga, "Toward a critique of the information economy." Media, Culture and Soci-
etn (1985): 271-296.
142. Gales of creative destruction and swarms of innovations are constructs popularized
by Joseph Schum peter through his emphasis on the importance of the entrepreneur to the
periodic crises of capitalism that we know as the business cycle. Joseph Schumpeter, Capi-
talism, Socialism and Democracy. London: Unwin University Books, 1966.
143· Peter Hall and.Paschal Preston, The Carrier Wave: New Information Technology and
the Geography of In nova tum, i846-2003. London: Unwin Hyman, 1988.
i44. Stinchcombe, i990, p. 362.
145· G. J. Mulgan, Communication and Control. New York: Guilford Press, 1991.
146. Murray Laver, Information Tech110logy: Agent of Change. Cambridge: Cambridge
University Press, 1989.
147. Theodore .Rosz.ak, The Cult of Information. New York: Pantheon, 1986. Roszak is es-
pecially critical of the dangers that flow from treating models, especially simulations, as
though they were reality, rather than as a set of hypothetical assumptions about the nature
of reality. The problem with simulations is that they tend to make reality much more neat
and tidy than it actually is; pp. 68-70.
148. Arno Penzias, Ideas and Information: Managing in a High-Tech World. New York: W.
W. Norton, 1989.
i49. Steven Yearley, Science, Technology and Social Change. London: Unwin Hyman,
1988.
150. Charles Jonscher, "Information resources and economic productivity." Information
Economics and Po/icy1(1983):13-35.
151. Webster and Robins, 1986, op. cit.
244 NOTES

152. Marcel J. Schoppers, "A perspective on artificial intelligence in society." Communi-

cation 9 (1986):195- 227.
153. Katz, 1986, op. cit.
154. Arriaga, 1985, op. cit.
155. Klaus Lenk, "Information technology and society;• pp. 273-310. In G. Friedrichs and
A. Schaff [eds], Microelectronics and Society: For Better or Worse. Oxford: Pergamon Press,
1982.

CHAPTER 3
1. Klaus Krippendorff, Content Analysis. Beverly Hills, CA: Sage, 1980.
2. Krippendorff, 1980, p. 21.
3. Krippendorff, 1980, p. 27.
4. U.S. Congress. Office of Technology Assessment, Informing the Nation: Federal Infor-
mation Dissemination in an Electronic Age. OTA-CIT-396. Washington, D.C.: U.S. Govern-
ment Printing Office, October 1988, pp. 64- 67.
5. Michael J. Weiss, The Clustering ofAmerica. New York: Harper & Row, ~988.
6. Gary D. Bass, Executive Director, OMB Watch, Washington, D.C.: "Testimony Before
the Legislation and National Security Subcommittee of the House Government Operations
Committee Regarding Reauthorization of the Paperwork Reduction Act:' July 25, 1989.
7. Priscilla M. Regan, "Privacy, government information, and technology." Public Ad-
ministration Review 46 (6) (November/December 1986):629- 634.
8. Robert Mitgang, Dangerous Dossiers. New York: Donald Fine, Inc., 1988.
9. Ibid., p. 250.
10. Frank J. Donner, TheAge ofSurveillance: The Aims and Methods ofAmerica's Political
Intelligence System. New York: Alfred A. Knopf, 1980.
11. Herbert N. Foerstel, Surveillance in the Stacks: The FBI's Library Awareness Program.
New York: Greenwood Press, 1991.
12. Gary T. Marx, Undercover: Police Surveillance in America. Berkeley: University of Cal-
ifornia Press, 1988; see also, Gerald Dworkin, "The serpent beguiled me and I did eat: En-
trapment and the creation of crime." Law and Philosophy 4 (1985):17-39.
13. Donner, 1980, op. cit., p. 73.
14. U.S. Congress. Office of Technology Assessment, "Update on computerized criminal
history record systems:' Appendix A, pp. 129- 134. In Federal Government Information Tech-
nology: Electronic Record Systems and Individual Privacy. OTA-CIT-296. Washington, D.C.:
U.S. Government Printing Office, June 1986.
15. U.S. Congress. Office of Technology Assessment, Automated Record Checks of Fire-
arm Purchasers: Issues and Options. OTA-TCT-497. Washington, D. C.: U.S. Government
Printing Office, July 1991. .
16. Donner, 1980, op. cit., p. 321.
17. Don H. Zimmerman, "Record-keeping and the intake process in a public welfare
agency;' pp. 319-344· In S. Wheeler [ed], On Record: Files and Dossiers in American Life.
New York: Russell Sage, 1969.
18. U.S. Congress. Office of Technology Assessment, June 1986, op. cit.
19. The development of mandatory screening and subsequent registry of persons with
such genetic characteristics is described by Troy Duster in Back Door to Eugenics. New York:
Routledge, 1990, p. 53.
NOTES 245

20. David H. Flaherty, Privacy and Government Data Banks. London: Mansell, 1979, p.
249.
21. Paul Starr, "The sociologyofofficial statistics." pp. 7- 57. In W. and P. Starr [eds], The
Politics. ofNumbers. New York: Russell Sage Foundation, 1987.
22. Ibid., p. n .
23. David H. Flaherty, Protecting Privacy in Surveillance Societies. Chapel Hill: University
of North Carolina Press, 1989, p. 81.
24. Ibid., p. 83.
25. A body of sophisticated research has been developed that focuses on the strategies
that might be used to reduce the risk of disclosure of information about a person's identity
and his or her personal attributes. See, for example, George T. Duncan and Diane Lambert,
"Disclosure-limited data dissemination." Journal ofthe American Statistical Asssociation 81
(393) (March 1986):10- 18; and Duncan and Lambert, "The risk of disclosure for microdata."
Paper presented at the Third Annual Research Conference of the Bureau of the Census,
March 29-April 1, 1987.
26. Dawn Nelson, "Record linkage: Confidentiality from the perspective of the U.S. Bu-
reau nf the Census, pp. 325- 333. In European Communities Commission, Eurostat News:
Protection of Privacy, Automatic Data Processing and Progress in Statistical Documentation.
Luxembourg: Office for Official Publications of the European Communities, 1986.
27. Joseph W. Duncan and William C. Shelton, Revolution in United States Government
Statistics 1926-1976. Washington, D.C.: U.S. Department of Commerce. Office of Federal
Statistical Policy and Standards, October 1978.
28. Testimony of John A. McLain, Joint Hearing on "1990 Census Planning-
Questionnaire Subjects:' Subcommittee on Federal Services, Post Office and Civil Service
of the Committee on Governmental Affairs, Senate; and the Subcommittee on Census and
Population of the Committee on Post Office and Civil Service. House. May 14, 1987. Wash-
ington, D.C.: U.S. Government Printing Office, 1987.
29. Duncan and Shelton, i978. Unless otherwise noted, the broad history of other gov-
ernment statistics has been derived from this source.
30. Max Eveleth, Jr., "How the private sector uses data and technology," pp. 24-45. In
Data Uses in the Private Sector. Washington, D.C.: U.S. Department of Commerce, March
i974.
31. Jerome M. Clubb, "Computer technology and the source of materials of social sci-
ence history." Social Science History10 (2) (Summer 1986):97-114.
32. U.S. Congress. Office of Technology Assessment, Informing the Nation: Federal Infor-
mation Dissemination in an Electronic Age. OTA-CIT-396. Washington, D.C.: U.S. Govern-
ment Printing Office, October i988.
33. Herbert I. Schiller and Anita R. Schiller, "Libraries, public access to information, and
commerce:' pp. 146-166. In V. Mosco and J. Wasko [eds], The Political Economy ofInforma-
tion. Madison: University of Wisconsin Press, i988; see also Paul Starr and Ross Corson,
"Who will have the numbers? The rise of the statistical services industry and the politics of
public data:' pp. 415- 447. ln Alonso and Starr, 1987, op. cit.
34. Kenneth MacCrimmon, Donald A. Wehrung, with W. T. Stanbury, Taking Risks: The
Management of Uncertainty. New York: Free Press, 1986.
35. Arthur L. Stindtcombe, Information and Organizations. Berkeley: University of Cali-
fornia Press, 1990.
 NOTES

36. Stinchcombe, 1990, p. 245.

37. Robert H. Coombs and Louis J. West [eds], Drug Testing: issues and Options. New
York: Oxford University Press, 1991.
38. Steven J. Wisotsky, "The ideology of drug testing." NOVA Law Review n (2) (Winter
1987):763-778.
39. J. Michael Walsh and Jeanne G. Trumble, "The politics of drug testing;' pp. 22- 49. In
Coombs and West, 1991, op. cit.
40. Neil Karunaratne, "An input-output approach to the measurement of the informa-
tion economy," pp. 61-70. In M. Jussawalla, D. Lamberton, and N. Karunaratne [eds], The
Cost ofThinking: Information Economies ofTen Pacific Countries. Norwood, NJ: Ablex., 1988.
41. H. Laurence Ross, "Personal information in insurance files;' pp. 203-222. In S.
Wheeler [ed], On Record: Files and Dossiers in American Life. New York: Russell Sage, 1969.
42. Ibid., p. 219.
43. Stephen Schleifer, "Trends in attitudes toward and participation in survey research."
Public Opinion Quarterly50 (1986): 17- 26.
44. Randall Rothenberg, "Surveys proliferete but answen dwindle:' New York Times.
October 5, 1990, p. A1.
45. Hugh M . .Beville, Jr., Audience Ratings: Radio, Television, Cable. Hillsdale, NJ: Law-
rence Erlbaum, 1988, pp. 226- 227.
46. Schleifer, 1986, op. cit., p. 22.
47. Beville, 1988, op. cit.
48. Meg Cox, "Nielsen announces new technology for People Meters." Wall Street Jour-
nal. June i, 1989. p. B6.
49. Frederick Bleakley, "Citicorp's Folly? How a terrific idea for grocery marketing
missed the targets." Wall Street Journal. April 3, 1991, p. Al.
50. "Anlerican Express plans to publish 20 magazines." Friday Report. April 20, 1990, p.1.
51. Timothy Noah, "Welfare recipients collect benefits with ATM cards." Wall Street
Journal. April 10, 1991, p. B1.
52. This issue has been raised in the early days of the experimental program. Jeffrey
Kutler, "U.S. Report raises privacy concerns in electronic benefit payments." American
Banker. May18, 1988, p. 32.
53. Thomas E. McManus, Telephone Transaction-Generated Information: Rights and Re-
strictions. Cambridge, MA: Harvard University, Program on lnfonnation Resources Policy,
May1990.
54. Patrick Barwise and Andrew Ehrenberg, Television and its Audience. London: Sage
Publications, 1988.
55. Arbitron's "ScanAnlerica" system was designed to increase the reach of a market test.
Because the program would utilize a "people meter:' with regular prompts to increase the
reliability of estimates of viewing by persons in conjunction with a scanning unit in the
home with which products with the universal products code purchases from any vendor
could be scanned into the database at home, the service was not dependent upon nor lim-
ited to participating stores. ScanAmerican plans were described in Beville, 1988, op. cit., pp.
292- 293.
56. Summary Scan! Marketing report. Chicago, IL, March 3, 1989.
57. Joyce Kole and Roseann Neilsen, "The coupon revolution." Direct'. April 20, 1991, p.
33.
NOTES 247

58. Cauzin Systems, Inc., of Waterbury, Connecticut, described the character and use of
its "Softstrip" technology in 1989.
59. David Paulin, «supermarket uses data strip to build customer profile." Direct. Feb-
ruary20, i989, p. 87.
60. Cauzin Systems, Inc., "Frequently asked questions abo ut Cauzin's direct response
system." Promotional sheet, December 1989.
61. Roger A. Clarke, "Information technology and dataveillance." Communications of
the ACM31 (5) (Mayi988):49$-512.
62. Ibid., p. 499.
63. Ibid., p. 502.
64. Ibid.
65. John Markoff, "American Express to buy 2 top supercomputers." New York Times.
October 30, 1991.
66. Stewart Brand, Inventing the Future at MIT. New York: Viking Penguin, i988, pp.
181- 200.
67. Peg Kay and Patricia Powers [eds], Future Information Processing Technology- 1983.
NB$-:500-103. Institute for Computer Sciences and Technology. National Bureau of Stan-
dards. Washington, D.C.: U.S. Department of Commerce, August i983. NBS is now the Na-
tional Institute for Standards and Technology (NIST).
68. Ibid., p . 75.
69. Ibid., p. 77.
70. Barbara Elazari, "Amex designs an OLTP system." Datamation. November 15,
1986:96-108.
7i. Ibid., p. 96.
72. Dimitris N. Chofras and Henrich Steinmann, Expert Systems in Banking: A Guide for
Senior Managers. New York: New York University Press, 1990.
73- lbid., pp. 39-40.
74. lbid., pp. 227- 230.
75. Ibid., p. 234.
76. Ibid., p. 270.
77. U.S. Senate. Committee on Governmental Affairs, "Report: The computer matching
and privacy protection act of 1987:· Washington, D.C.: U. S. Government Printing Office.
1988.
78. Anne M. Barry, "Defamation in the workplace: The impact of increasing employer
liability." Marquette Law Review 72 (1989 ): 264-303.
79. Gilbert Fuchsberg,"Employers' use of accident records raises specter of blacklisted
workers." Wall Street Journal. July i6, i990, p. B1.
80. Troy Duster, Backdoor to Eugenics. New York: Routledge, i990, p. 3. Philip Bereano, a
professor of engineering and public policy at the University of Washington, convinced me
of the importance of genetic testing and the technology of genetic engineering as compo-
nents of the panoptic sort during a visit to the campus in Seattle in Novernben99i. Bereano
had testified several times before congressional committees and had explored the civil lib-
erties issues in several conference papers. See his "Testimony on DNA Identification Sys-
tems: Social Policy and Civil Liberties Concerns" before the Subcomittee on Civil and Con-
stitutional Rights. Committee on the Judiciary. U.S. House of Representatives. March 22,
1989.
 NOTES

81. Ronald Frank, William Massey, and Yoram Wind, Market Segmentation. Englewood
Cliffs, NJ: Prentice-Hall, 1972.
82. Ronald Frank and Marshall Greenberg, "Zooming in on TV Audiences." Psychology
Today. October 1979, pp. 92-103; see also Ronald Frank and Marshall Greenberg, Audiences
for Public Broadcasting. Beverly Hills, CA: Sage, 1982.
83. Patrick Barwise and Andrew Eh renberg, Television and its Audience. London: Sage
Publications, i988. Simmons was influenced primarily by their early studies for the Inde-
pendent Broadcasting Authority that focused on tbe choice of television programs as a re-
flection of a decision to invest mental effort in the processing of information.
84. Ronald I. Simmons, "The relationship between cognitive style and television prefer-
ences among African-American college students." (unpublished dissertation) Howard
University, Washington, D.C., June i987.
85. Most of these multivariate statistical techniques are described in Peter Monge and
Joseph Cappella [eds], Multivariate Techniques in Human Communication Research. New
York: Academic Press, 1980.
86. Jonathan Gutman, "Techniques for audience segmentation," pp. 123-137. ln J. Domi-
nick and J. Fletcher [eds], Broadcasting Research Methods. Newton, MA: Allyn and Bacon,
i985.
87. Jae-On Kim and Charles W. Mueller, Introduction to Factor Analysis. Beverly Hills,
CA: Sage, 1978.
88. Mark Aldenderfer and Roger Blashfield, Cluster Analysis. Beverly Hills, CA: Sage,
1984.
89. Marija Norusis, "Stacking Beers: Cluster Analysis;' pp. 165-191. ln SPSSx Advanced
Statistics Guide. Chicago: SPSS Inc., i985.
90. William R. Klecka, Discriminant Analysis. Beverly Hills, CA: Sage, i980.
91. Paul E. Green and Vithala Rao, Applied Multidimensional Scaling. New York: Holt,
Rinehart & Winston, i972. See also Joseph Kruskal and Myron Wish, Multidimensional
Scaling. Beverly Hills, CA: Sage, i978.
92. Frank, Massey, and Wind, i972, op. cit., pp. 165-175.
93. Joseph Cappella, "Structural equation modeling: An introduction;' pp. 57-110. In P.
Monge and J. Cappella, 1980, op. cit.
94. James B. Rule, Douglas McAdam, Linda Steams, and David Uglow, "Documentary
identification and mass surveillance in the United States." Social Problems 31 (2) (1983):222-
234.
95. Ibid., p. 214.
96. Jerome Svigals, "Smart Cards-A critical decision point." Journal of Retail Banking
i9 (1) (Spring 1987):43-55.
97. lbid., p. 47.
98. Joseph Eaton, Card Carrying Americans. Totowa, NJ: Rowman & Littlefield, i986, p.
140.
99. Michel Foucault, The Order of Things: An Archaeology of the Human Sciences. New
York: Vintage Books, 1973.
100. Edward Eugene Gallahue, "Some factors in the development of market standards,
with special reference to foods, drugs, and certain other household wares." (dissertation)
Washington, D.C.: Catholic University Press, 1942..
101. J. Beniger, 1986, op. cit., pp. 344-346.
NOTES 249

102. Gerald R. Winslow, Triage and Justice. Berkeley: University of California Press, 1982.
103. Starr, 1987, op. cit, pp. 43-44.
104. Mary Tew Douglas, How Institutions Think. New York: Syracuse University Press,
1986, p. 58.
105. Ibid., p. io1.
106. Foucault, 1973, p. 131.
107. John E. Hunter, "Factor Analysis;' pp. 229-257. In P. Monge and J. Cappella [eds],
Multivariate Techniques in Human Communication Research. New York: Academic Press,
1980. Hunter suggests that one approach to factor analysis, trait theory, asks "What are the
fundamental traits which underlie a given set of observed variables?" p. 228.
108. James A. Anderson, "Some preliminary thoughts on the elaboration of audiences:'
Conference paper. Association for Education in Journalism and Mass Communication,
Washington, D.C.: August, 1989.
109. David Margolick, "Finding jury is first test in celebrated rape trial" New York
Times. October 31, 1991, p. BS.
110. Newton N. Minow and Fred H. Cate, "Who is an impartial juror in an age of mass
media?" American University Law Review40 (2) (Winter 1991):660.
111. Nancy Reichman, "Managing crime risks: Toward an insurance based model of so-
cial control." Research in Law, Deviance and Social Control 8 (1986):151- 172.
112. Ibid., p. 166.
113. Peter W. Huber, Liability: The Legal Revolution and Its Consequences. New York: Ba-
sic Books, 1988.
114. James K. Stewart, "From the d irector." National Institute ofJustice Research in Brief
March 1987, p.1. .
115. Norval Morris and Marc Miller, "Predictions of dangerousness in the criminal law:·
National Institute ofJustice Research In Brief March 1987, p. 3.
116. Ibid., p. 4.
117. This is a point made elegantly by Troy Duster, 1990, op. cit., pp. 97-101.
118. Reichman, 1986, p. i57.
119. James Rule, Private Lives and Public Surveillance (Social Control in the Computer
Age). New York: Schocken Books, 1974·
120. Ibid., p. 178.
121. Ibid., p. 179·
122. Peter McAllister, "Early warning on delinquencies." Retail Control. March 1986, pp.
16-27.
123. Svigals, 1987, op. cit., p. 47.
124. Michael J. Weiss, The Clustering ofAmerica. New York: Harper & Row, 1988.
125. Eveleth, 1973, op. cit., p. 34.
126. Starr and Cors~n, 1987, op. cit., p. 426.
127. Philip Porado, "Tiger pause." Campaigns and Elections. April-May 1990, pp. 49-:;1.
128. TIGER: The Coast-to-Coast Digital Map Data Base. Washington, D.C.: U.S. Depart-
ment of Commerce. Bureau of the Census, November 1990.
129. Ibid., p. 4.
i30. Weiss, 1988, op. cit., p. u.
13i. Ibid., p. 13.
250 NOTES

i32. Oscar H. Gandy, Jr., and Charles E. Simmons, "Technology, privacy and the demo-
cratic process:' Critical Studies in Mass Communication 3 (2) (June 1986):155-168. See also
the "i.nside view" by political consultant Matt Reese. "From telephone to telelobby," p. 104.
In R. Meadow, New Communication Technologies in Politics. Washington, D.C.: Washington
Program. Annenberg School of Communications, 1985.
133. David Beiler, "Precision politics:' Campaigns and Elections. February/March 1990,
p.33.
134. Kevin Kramer and Edward Schneider, "innovations in campaign research: Finding
the voters in the 1980s;' pp. 19-s2. In Meadow, 1985, op. cit.
135. Ibid., p. 23.
136. Beiler, 1990, op. cit., p. 35.
137. Campaigns and Elections. February/March 1990, p. 35.
138. Joel E. Book (Book and Company Marketing Services), "Target market publishing:
The revolutionary marketing communications potential of database-driven magazine pub-
lishing systems." Executive briefing, Orlando, Florida, December 1989.
l39· Ibid.
140. Chorafas and Steinmann, 1990, p. 232.
141. Conquest/Direct promotion, Octo.ber 1989.
142. Personalized copy of 1989 Epsilon Data Management letter to shareholders, Febru-
ary 20, 1990, p. 2.
143. Charting Your Course: For Fundraising and Membership Development. Burlington,
MA: Epsilon, n.d., p. 4.
144. Ibid., p. 10.
145. Gannett Co. news release, December 7, 1989.
1.46. Information included in a package from Gannett Telemarketing in response to a
query about the business, March 12, 1990.
147. Eleanor Novek, Nikhil Sinha, and Oscar Gandy, "The value of your name:' Media,
Culture and Society 12 (1990 ):525-s43.
148. Personal correspondence, March 9, 1990.
149. "Equifax reverses 'target marketing' policy:' Privacy journal. August 1991, p. 1.
150. "Privacy-risk assessment program mounted by TRW; some criticism." DM News.
November u, 1991, p. 238.
151. Raymond Wacks, Personallnformation: Privacy and the Law. New York: Oxford Uni-
versity Press, i989.
152. "The DBA consumer file vs. the competition- A competitive analysis." Photocopy
of a broadsheet. Englewood, NJ: Database Anlerica Companies, 1990.
153. Product announcement in the direct marketing newsletter, Friday Report. April 13,
1990, p.1.
154. Tom Koch, journalism for the 21st Century: Online Information, Electronic Databases,
and the News. New York: Praeger, 199L That newspapers themselves have also entered the
business of selling access to their files and the data they generate about their customers bas
been noted by Jean Ward and Kathleen Hansen, "Journalist and librarian roles, informa-
tion technologies and newsmaking." Journalism Quarterly 68 (3) (Fall 1991):491498.
155· Starr and Corson, 1987, op. cit.
156. "Meredith Online with 'largest' database:' DM News. October 28, 1991, p. i.
l57· Starr and Corson, op. cit., 1988, p. 420.
NOTES

158. U.S. Department of Commerce. National Telecommunications and Information

Administration, NTIA Information Services Report. 88-235. Washington, D.C.: U.S. Depart-
ment of Commerce, August 1988, p. 29.
159. The RBOCs, created after the consent decree that broke up AT&T, involved restric-
tions on the lines of business that the still large and powerful common carriers could enter.
Pressure from other information service providers who feared competition from the
RBOCs was exercised through the oversight of District Judge Harold Greene. After staying
his own order to allow the RBOCs into the information business, the U.S. Court of Appeals
vacated the stay on October 7, i991. The American Newspaper Publishers Association inl-
mediately began a campaign to put the giant companies back into their cages. See Telecom-
munications Reports. Octoben4, 1991, pp. 1- 7.

CHAPTER 4
1. Meredith Mendes, "Privacy and computer-based information systems:' pp. 193-264.
ln B. Compaine [ed), Issues in New Information Technology. Norwood, NJ: Ablex, 1988.
2. American Express Company, 1984 Annual Report. New York: American Express, 1985,
p.3.
3. American Express Company, 1985 Annual Report. New York: American Express, 1986,
p.4.
4. American Express Company, i986 Annual Report. New York: American Express, 1987,
pp.4-5.
5. Ibid., p. 31.
6. American Express Company, 1987 Annual Report. New York: American Express, 1988,
pp.8-15.
7. American Express Company, 1988 Annual Report. New York: American Express, i989,
pp.14-15.
8. "Privacy task force being formed:' Friday Report. January 6, 1989, p. 1.
9. There is no intention ofleaving the impression that the inlage of American Express is
one without any warts. In a case in the U.S. Court of Appeals, D.C. circuit, Amex was
chided for its pursuit of its interests through its "extraordinary use of interrogatories . . . the
various sets of interrogatories and their answers are in the hundreds of pages. They run as
far afield as inquiring the name of every Jaw firm the plaintiff bad been affiliated with since
1951." In the court's view "the length, scope and detail of tbe interrogatories propounded by
American Express suggest a strategy of attrition rather than a legitimate discovery of the
facts needed to resolve a dispute over the account." Oscar S. Gray, v. American Express Com-
pany. No. 83-1475. U.S. Court of Appeals. 743F. 2d 10 (1984) at 20.
io. David Enscoe, "Privacy debate goes public." Target Marketing. January i989, p. 35.
11. Ibid., p. 39.
u. American Express Company, Form 10-K. Annual Report to the Securities and Ex-
change Commission. New York American Express, March 1989.
13. For a useful discussion of the changes in the United States and in advanced industrial
economies, which have been labeled by some as information economies, the following rep-
resent good sources: Vincent Mosco and Janet Wasko [eds], The Political Economy of Infor-
mation. Norwood, NJ: Ablex, 1988; Jennifer Slack and Fred Fej~s [eds], The Ideology of the
Information Age. Norwood, NJ: Ablex, 1987.
252 NOTES

14. TRW, Toe., 1984 Annual Report. Cleveland, OH: TRW, Inc., 1985.
15. Ibid., pp. 16-17.
16. "TRW credit data flexes its muscles." Privacy Journal April 1989, pp. i, 4, 6.
17. TRW, Inc., 1985 Annual Report. Cleveland, OH: TRW, Inc., i986, p. 4.
18. Ibid., p. 7.
19. Ibid.
20. Ibid., p. ll.
21. Michael Van Buskirk, cited as a source in William Ecenbarger, "They know who you
are." Philadelphia Inquirer Magazine. February 4, 1990, p.15.
22. "TRW credit data flexes its muscles:' Privacy Journal. April 1989, p. 6.
23. Michael Miller, "Six states sue TRW over credit-reporting practices." Wall Street Jour-
nal. July 10, 1991, p. Bi.
24. Equifax, Inc., 1984 Annual Report. Atlanta, GA: Equifax, Inc., 1985, p. 2.
25. Ibid., p. 9.
26. Ibid., p. 15.
27. Ibid.
28. Equifax, Inc., 1985 Annual Report. Atlanta, GA: Equifax, Inc., 1986, p. 2.
29. Equifax, Inc., 2986 Annual Report. Atlanta, GA: Equifax, Inc., i987, p. 3.
30. Ibid., p. u.
31. Ibid.
32. Ibid., pp.18-19.
33. Equifax, Inc., 1987 Annual Report. Atlanta, GA: Equifax, Inc., i988.
34. Hovator v. Eqiiifax, Inc. U.S. Court of Appeals for the Eleventh Circuit, July 30, i987,
823 F. 2d 413.
35. Equifax, Inc., i988 Annual Report. Atlanta, GA: Equifax, Inc., 1989, p. 14.
36. Ibid., p. 22.
37. David Linowes, Privacy in America. Urbana, IL.: University of illinois Press, 1989;
David Flaherty, Protecting Privacy in Surveillance Societies. Chapel H ill, NC: University of
Nor th Carolina Press, 1989.
38. Mark Nadel, "Rings of privacy: Unsolicited telephone calls and the right of privacy."
Yale Journal on Regulation (1986):99-128.
39. Congress passed the Telephone Consumer Protection Act in 1991 and charged the
FCC with its implementation. Tn its efforts to balance what it sees as legitimate privacy in-
terests witll the "continued viability of beneficial and useful business services:' which gen-
erated hundreds of millions of dollars from unsolicited calls, the commission clearly fa-
vored the industry position. See: "FCC stands up for telemarketing; hits do-not-call
database option." DM News. April 27, 1992, p. 1.
40. James Katz, "U.S. Telecommunications privacy policy." Telecommunications Policy
(December 1988):353-368.
41. Brian Lane, quoted in: Robert M. Entman, State Telecommunications Regulation: To-
ward Policy for an Intelligent Telecommunications Infrastructure. Report of an Aspen Insti-
tute Conference, Aspen, Colorado. July 9-13, 1989. Truro, MA: Aspen Institute, p. 2i.
42. James B. Ginty, vice president AT&T, Philadelphia, quoted in A. Gnoffo: "They've
got your number." Philadelphia Inquirer. February 4, 1990, sec. C1, p. 8.
43. R. Fannin,"The last great hope?" Marketing and Media Decisions. February 1988, pp.
24-30.
NOTES 253

44. The question of the sale of the company's member lists has been addressed by Ari
Solomon, but there is no firm distinction between general membership lists and the identi-
fication of persons who repond to client ads. See: Ari Solomon, "The Consequences of
Prodigy:• Privacy Journal. April 1990, pp. 4-5·
45. "PRODIGY Service Member Agreement." White Plains, NY: Prodigy Services Com-
pany, 1989, p. 8.
46. Seymour Lusterman, Managing Federal Government Relations. Research Report No.
905. New York: Conference Board, 1988.
47. Direct Marketing Association Guidelines for Ethical Business Practices. Direct Market-
ing Association, mailed to the author, January 1990.
48. "Automated number identification systems growing." Friday Report. March 3, 1989,
pp.1-2.
49. Ed Burnett, " How to cope with list problems and abuses:' Directions. March/April
1988, pp. 1-6.
50. R. Baker, R. Dickinson, and S. Hollander, "Big brother t994: Marketing data and the
IRS:' Journal ofPublic Policy and Marketing 5 (J988):227-241.
51. Alexander Hoffman, "A Statement to the House Committee on the Judiciary." U.S.
Congress. House. Subcommittee on Courts, Civil Liberties, and the Administration of Jus-
tice. Hearings: "1984 Civil liberties and the National Security State." Washington, D.C.:
Government Printing Office, i984, pp. 320-321.
52. Ray Schultz, "Big compilers say no to the FBI:' DM News. May 4, 1992, p. i.
53. "Can direct marketers police themselves?" Target Marketing. April 1989, p. 16.
54. Ibid.
55. "TRW credit data flexes its muscles." PrivtUy Journal. April 1989.
56. Q uoted by David Enscoe in "Privacy debate goes public." Target Marketing. January
1989, p. 34.
57. Robert f. Posch, "How the law(s) of'privacy' impact your business." Direct Market-
ing. October 1987, p. 78.
58. Ibid., p. 80.
59. Ibid., p. 102.
60. Robert Posch, "Nuisance/privacy infractions-part two." Direct Marketing. January
1988, p. 103.
61. Roy Schwedelson, February 23, 1989. Audio tape 39-0088 LI, Hoke Communica-
tions, Inc., Garden City, NY; abstracted in: "Privacy vs. Free Speech:' (May 1989). Direct
Marketing. May 1989, pp. 41-46.
62. Ibid, p. 42.
63. Ibid., pp. 45-46.
64. David Linowes, 1989, op. cit.
65. My research throughout this entire project was aided by the fine work of a number of
graduate research assistants. However, one student, Elizabeth Van Hom, was especially
helpful in the design and implementation of this survey of business leaders.
66. Rose Harper, Mailing List Strategies: A Guide to Direct Mail Success. New York:
McGraw-Hill, 1986.
67. Analysts have suggested that cross-marketing was pretty much in its infancy because
the majority of firms, including those in financial services, did not have their customer data
254 NOTES

in formats that would facilitate matching to support cross-marketing. See Dan Bencivenga,
"Pinpointing Investors." Target Marketing. September 1988, p. 32.
68. Alan Westin and Lance Hoffinan, "Privacy and security issues in the use of personal
information about clients and customers on micro and personal computers used in office
automation." Contractor Report prepared for the U.S. Office of Technology Assessment.
February 1985.
69. Equifax, Inc., The Equifax Report on Consumers in the Information Age. Atlanta, GA:
Equifax, Inc., 1990, p. 77.
70. David Linowes, 1989, op. cit., p. 40.
71. Arthur L. Stinchcombe, Information in Organizations. Berkeley: University of Cali-
fornia Press, 1990.

CHAPTER 5
I. T. R. Young makes a distinction between public opinion, mass opinion, and social
opinion. For Young, public opinion represents the primary source of democratic social
change, which is the authentic product of communicative interaction within the public
sphere. Young, along with Habermas and others, suggests that the public sphere has been
subverted and transformed thxough its penetration by state and corporate interests and
that manufactured mass opinion is the result. Social opinion represents the middle dis-
tance in which the emancipatory potential of public opinion is limited by the influence of
history, tradition, and established cultural beliefs and serves primarily to reproduce thjngs
as they are: "Social opinion reproduces existing social forms while public opinion is
pointed toward new and different forms of social life." He continues, "News is the stuff of
public opinion; propaganda is tl1e heart of social opinion. It is a mystification to conflate
them and does much mischlef to the public opinion policy process:' See T. R. Young, "Pub-
lic opinion, mass opinion, and social opinion: The constitution of political culture in the
capitalist state;' pp. 264-266. In V. Mosco and J. Wasko [eds], The Critical Communications
Review. Vol. 3. Norwood, NJ: Ablex, 1985.
2. Anthony Giddens, The Constitution .of Society. Cambridge: Polity Press, 1984, pp. 49-
104.
3. Amos Tverskyand Daniel Kahneman, "Rational choice and the framing of decisions:'
pp. 60-89. In K. Cook and M. Levi [eds J, The Limits of Rationality. Chicago: University of
Chicago Press, t990.
4. A classic in the field of social research is the work by Eugene Webb, Donald Campbell,
Richard Schwartz, and Lee Sechrest, Unobtrusive Measures: Nonreactive Research in the So-
cial Sciences. Chicago: Rand McNally, 197L For an updated assessment of approaches to the
problems of measurement see Delbert C. Miller, Handbook of Research Design and Social
Measurement. 5th ed. Newbury Park, CA: Sage, 199i.
5. Joseph Lepkowski, "Telephone sampling methods in the United States," pp. 73-98. In
R. Groves et al. [eds], Telephone Survey Methodology. New York: John Wiley and Sons, i988.
6. Owen Thornberry and James Massey, "Trends in United States telephone coverage
across time and subgroups," pp. 25-69. In Groves, et al., 1988, op. cit
7. My research assistants, Jerry Baber and Catherine Preston, administered each of the
groups, with Baber leading the discussions through a "nominal group technique" that uti-
lized a weU -structured set of questions and exercises.
NOTES 255

8. The term panoptic sort was never used. The letters of invitation and the question-
naire each participant completed prior to beginning the group interview spoke only about
privacy and practices related to information privacy.
9. Kristen Eddy, "Lit majors are not lepers:' Washington Post. November 22, 1987, p. C5.
io. Kurt Dubowski, "Drug use testing: Scientific perspectives." Nova Uiw Review u (2)
(1987): 416-552.
n. Raymond Wacks, Personal information: Privacy and the Uiw. Oxford: Oxford Univer-
sity Press, 1989.

CHAPTER 6
2. The 1978 survey was administered by Louis Harris and Associates (Harris) for Sentry
Insurance, the 1983 survey was administered by Harris for Southern New England Tele-
phone, and the i990 survey was administered by Harris for Equifax, Inc. All secondary
analyses of the Harris data were performed without weighting. Although the Harris organi-
zation has developed a high level of precision in the construction of its samples, even their
procedures generate samples that require weighting if they are to be projectable to the U.S.
population. The published reports of these Harris surveys use data weighted to more
closely approximate proportions in key demographic groups. My secondary analyses are
focused primarily on uncovering relationships between variables, rather than emphasizing
distributions in the population. This approach will, on occasion, produce estimates of dis-
tributions that are at variance with those published by Harris, Westin, or the sponsors of
the surveys.
2. Oscar H. Gandy, Jr., "Telecommunication and Privacy," a study funded by a grant
from AT&T through the Center for Communication and Information Science and Policy at
the University of Pennsylvania, 1988-1990.
3. The limitation of the number of call-backs to three, in addition to dominance of the
quota over the representativeness of the sample produced a sample that grossly
overincluded female respondents. The bulk of the analyses of this dataset were performed
with a weighting factor for gender that reduced the effective presence of females to 52 per-
cent and the effective sample size to 1,008. Concerns about concentrating an oversample
through weighting are quite different from those that expand categories that are underrep-
resented. Such is dearly not the case in this analysis.
4. One broad search was conducted through the Roper Center for Public Opinion Re-
search in November of 1987 that collected summaries of responses to questions about pri-
vacy from 1975 to 1987 in order to compare levels of concern with attention to privacy issues
in the press.
5. Alan Westin, Privacy and Freedom. New York: Atheneum, i967.
6. Raymond Wacks, Personal Information, Privacy and the Law. Oxford: Oxford Univer-
sity Press, w89.
7. "Pentagon considers data access controls:' Transnational Data and Communications
Report. August 1986, p. 5; Philip Doty, "Federal research and development (R&D) as intel-
lectual property;' pp. 139-171. In C. McClure and P. Hernon [eds], U.S. Scientific and Tech-
nical lllformation (STI) Policies: Views and Perspectives. Norwood, NJ: Ablex, 1989.
8. James Katz and Annette Tassone, "Public opinion trends: Privacy and information
technology." Public Opinion Quarterly 54 (1990 ): 125- 143.
 NOTES

9. Ibid., p. 138.
10. For the scale with eleven variables (direct marketers excluded) Chronbach's alpha=
.87.
n. Comments made during c.onferences, seminars, and conversations about her work
by Mary J. Culnan, an associate professor at the School of Business Administration,
Georgetown University.
i2. Hearings were held in April 1980 before the Senate Committee on Banking, Housing,
and Urban Affairs. The Subcommittee on Consumer Affairs included testimony from Eu-
gene T. Merrigan, senior vice president of marketing, Equifax Services, Inc.
13. Twenty-six variables were measured through Likert-type scales, ranging from 1 (dis-
agree very strongly) to s (agree very strongly). SPSSx, principle components analysis, gen-
erated ten factors explaining 79.2 percent of the total variance.
14. John S. Detweiler, "Monitoring 'entitlement' attitudes in the 'sos." Public Relations
Review 12 (3) (Fall 1986):28-40.
15. Herbert McClosky and Alida Brill's exploration of public opinion regarding civil lib-
ertie.s, including those they identified with regard to personal privacy, indentified a wide-
spread willingpess of people to limit the rights of criminals and even those who might de-
part from a community's standards. See their Dimensions of Tolerance: \Vhat Americans
Believe About Civil Liberties. New York: Russell Sage Foundation, i983, especially Chapter 5,
"The rights of privacy and lifestyle:' pp. 171-231.
i6. The factor analysis method was principal components with varirnax rotation as gen-
erated by SPSSx. The four factors explained 74. 8 percent of the measured variance. For the
identification of the factors, those variables with factor loadings of .5 or higher were used.
Additionally, those variables that were highly correlated with more than one factor were
eliminated after the factors bad been rotated.
i7. Katz and Tassone, op. cit., p. i39.
i8. Respondents' rankings were determined empirically, after-the-fact, rather than by a
direct question.
i9. \<Vhose Business is it Anyway? Washington, D.C.: National Business League, January
1990.
20. Reported in the Equifax Report, p. 14.
21. Katz and Tassone, 1990, op. cit., p. i42.
22. The factor analysis method was principal axis factoring with varimax rotation. It
should be noted that the original instrument generally only allowed for an affirmative, a
negative, and a don't know/not sure response. Two modifications were made that have im-
plications for the interpretation of the factors and correlation-based analyses. For the bulk
of the items used in the factor analysis, the items were transformed into dichotomous
dummy variables, taking the values of zero and one, whereby the value of one was assigned
to those that were the privacy affirming, or most fearful, responses. All other responses, in-
cluding the nonresponses, were coded as zero. This allowed the full data set to be used in all
correlations. At the same time, such a coding undoubtedly served to underestimate the ex-
tent of privacy concern in that it treated all qualified (unsure) responses as not concerned.
The second option, taken less frequently, was to assign the not sure response to the middle
category. Because the number was always small, the resultant distribution was grossly bi-
modal. It was for this reason that rather conservative estimates of significance were.used to
guide my interpretation. The likely result is a type II error, missing relationships that might
have appeared if the distributions were more appropriate to the tests applied.
NOTES 257

23. Chronbach's alpha was utilized as a measure of reliability.

24. This might be seen as a fairly extreme view in the context of the current popularity of
a cognitive science approach. Political ideology might be seen as an intervening variable
that provides an indication of an information processing screen that influences how indi-
viduals understand their changing environment. Jn this way ideology, like personality and
values, can be seen as causal or determinative of other more dynamic opinions and
behaviors. That is not the position that is taken at this stage in the analysis.
25. The correlations were r =. 15 for African-Americans and r = . 09 for Hispanics.
26. Ronald Ioglehart, The Silent Revolution: Changing Values and Political Styles Among
Western Publics. Princeton: Princeton University Press, 19n; Ronald 1nglehart, Culture Shift
in Advanced Industrial Society. Princeton: Princeton University Press, 1990.
27. Ioglehart, i977, p. 7.
28. Ibid., p. 293.
29. Milton Rokeach, The Nature ofHuman Values. New York: Free Press, 1973.
30. Oscar H. Gandy, Jr., "ls that all there is to love? Values and program preference:· pp.
207- 219. Io S. Thomas [ed], Studies in Mass Communication and Technology. Norwood, NJ:
Ablex, 198i. This study utilized a modified version of the Rokeach value sort in combina-
tion with discriminant analysis to predict the program preferences of college students. At-
tributes of the programs were seen to be related to the values that were dominant in student
rankings.
31. Rokeach, 1973, p. 62.
32. Jbid., p. 73-
33. For an examination of changes in ideological self-identification see John P. Robinson
and John A. Fleishman, "Ideological identification trends and interpretations of the
Liberal-Conservative balance." Public Opinion Quarterly 52 (1988): 134-145.
34. Analysis was performed with unweighted data, n = 1,250, using gender as a control
variable, primarily to assess the extent of its interaction with age.
35. Westin's "concern over privacy index" was described in his testimony before the Sub-
committee on Government Information, Justice, and Agriculture. House Committee on
Government Operations, Washington, D.C.• April 10, 1991. The additivity of the measures
is questionable (alpha = .29), although its use as a basis for an ordinal classification of re-
spondents has been shown to have considerable predictive ability with regard to other mea-
sures of privacy concern.
36. Nancy Signorielli and Michael Morgan !eds], Cultivation Analysis: New Directions in
Media Effects Research. Newbury Park, CA: Sage, 1990.
37. The standard deviation= 2.34 hours.
38. "When you read the papers, how much attention do you pay to stories about politics,
government or public affairs?"
39. Theoretical positions that suggest that television and the media system in general
contain contradictory tendencies may be reintegrated from a theoretical stance that argues
that different media forms play different symbolic functions. The more "selective" media
serve the interests of the more elite members of the population because they deal with ideo-
logical questions that are as yet unsettled.
40. This was measured by response to the question "How concerned would you guess
the average American is about threats to their privacy?" On a scale from 1- 10, l =of no con-
cern and 10 =of very great concern.
 NOTES

41. Alleo Barton and R. Wayne Parsons, ''Measuring belief system structure:· Public
Opinio11Quarterly41 (1977h59-18o; George Bishop, "The effect of education on ideologi-
cal consistency." Public Opinion Quarterly 40 (1976): 337-348; Pamela Conover and Stanley
Feldman, "'Tbe origins and meaning of Liberal/Conservative self-identifications:' Ameri·
can journal of Political Science 25 (4) (1981):617-645; John Fleishman, "Types of political at-
titudestructure: Results of a cluster analysis." Public Opinion Quarterly50 (1986):371-386.
42. Louis Harris and Associates, Despite gains conservatives still mino rity?' The Harris
0

Survey. February 27, 1986, p. 1..

43. fohn Fleishman, "Trends in self-identified ideology from 1972 to i982: No support
for the salience hypothesis." American journal ofPolitical Science30 (J) (1986):517-s41.
44. Milton Rokeach, The Nature ofHuman Va foes. New York: Free Press, 1973.
45. Geo.rge Gerbner, Larry Gross, Michael Morgan, and Nancy Signorie!Ji, "Living with
television: The dynamics of the cultivation proce.ss," pp. 17-40. In J. Bryant and D. Zillman
[eds], Perspectives 011 Media Effects; and George Gerbner, "Epilogue: Advancing on the path
of righteousness (Maybe)," pp. 249-262. ln N. SignorielH and M. Morgan [eds). Cultivation
Analysis: New Directions in Media Effec1sResearcf1. Newbury Park, CA: Sage, 1990.

CHAPTER 7
1. See Roger Noll, "The political and institutional context of communications policy,"
pp. 42- 65. ln M. Snow [ed]. Marketplace for Telecomm1mic11t'io11s. New York; Longman,
1986; Vincent Mosco, "Perspectives on the state and telecommunications system;• pp. 85-
108. ln The Pay-Per Society. Norwood, Nf: Ablex, (989; Afan P. Hamlin, "Public choice, mar-
kets and utilitarianism," pp. u6-138. In D. Whynes [ed ], What is Political Economy? Oxford:
Basil Blackwell. 1984.
2. Richard F. Hixon, Privacy in a Public Society: Human Rights in Conjlic1, New York: Ox-
ford University Press, 1987, p. 46.
3. AJaa R Westin, Privacy and Freedom. New York: Atheneum, i967.
4. David Flaherty, Protecting Privacy in Surveillance Societies. Chapel Hill: University of
North Carolina Press, 1989.
5. Richard F. Hixon, Privacy in aPublic Society: Human Rights in Conflict. New York: Ox-
ford University Press, 1987.
6. .Barrington Moore, Jr., Privacy: Studies in Social and C11lt11ral History. Armonk, NY:
M. E. Sharpe, 1984, pp. 268, 285. The historical basis in levels ofsocial development for dif-
ferent meanings of privacy is discussed in several other key sources, including Richard ~
Posner, "Privacy, secrecy and reputation." Buffalo Law Review 28 (1979).
7. Leon Pestinger, A Theory of Cognitive Dissonance. Evanston, IL: Row, Peterson and
Company, 1957.
8. EUsabeth Noelle-Neumann, "Spiral ofsilence: A theory of public opinion." Jciumal ~
Commwrication 24 (1974): 43- 5i.
9. David L. Bazelon, "Probing privacy." Gonzaga Law Review u (.4) (Summer i977):,5fr,..-
619.
10. Gary L. Bostwick, "A taxonomy of privacy: Repose, sanctuary, and intimate deci-
sion." California Law Review64 (1976):1, 451.
u. Roe v. Wade, 410 U.S. ll3 (1973).
t2. Bostwick reflects this unfortunate confusion by specifying Uthe privacy of ia~
decision" in his suggestion that "this privacy is less 'freedom from' and more 'freedom t.o.-
NOTES 259

and cites Griswold v. Connecticut, 38t U.S. 479 (1965) as the case that recognized the zone of
intimate decision and generated this migration from the core interests in controlling ac-
cess.
13. Samuel D. Warren and Louis D. Brandeis, "The right to privacy." Harvard Law Re-
view 14 (s) (December 15, 1890 ): 213, note i.
14. Ibid., p. 196.
15. Ibid., p. 198.
16. Ibid., p. 204.
17. Ibid., p. 209. note 1.
18. Ibid., p. 2t5.
19. James H. Barron, "Warren and Brandeis. The Right to Privacy, Harvard Law Review
(1890): Demystifying a landmark citation:' Suffolk University Law Review 13 (4)4 (Summer
1979):874-922.
20. This is Barron's claim, ibid., pp. 910-911.
2i. Edward J. Bloustein, "Privacy as an aspect of human dignity: An answer to Dean
Prosser:' New York University Law Review39 (December 1964):962- 1,007.
22. Ibid., p. 969.
23. William Patton and Randall Bartlett, "Corporate 'persons' and freedom of speech:
The political impact of legal mythology." Wisconsin Law Review (1981):494-5u; Herbert
Hovenkamp, Enterprise and American Law, 1836-1937. Cambridge, MA: Harvard University
Press, 1991; Morton J. Horwitz," 'Santa Clara' revisited: The development of corporate the-
ory:• pp. 13-63. ln W. Samuels and A. Miller [eds), Corporations and Society: Power and Re-
sponsibility. New York: Greenwood Press, 1987; Kent Middleton, "Commercial Speech and
the First Amendment." (unpublished dissertation) University of Minnesota, 1977.
24. Bloustein, op. cit., p. 971.
25. Ibid.
26. Ibid., p. 974.
27. Ibid., p. 979.
28. Ibid., p. 987.
29. Shilbey v. Time, .Inc., 45 Ohio App. 2d 69, 341 N.E. 2d 337 (1975) in Jonathan P. Gra-
ham, "Privacy, computers and the commercial dissemination of personal information."
Texas Law Review65 (7) (June 1987), pp, t,413-1,415.
30. Lamont v. Commissioner ofMotor Vehicles, 269 P. Supp. 880 (S.D.N.Y.), aff'd, 386 F. 2d
449 (zd Cir. i967) cited in Graham, p. 1, 415.
31. Blaustein, op. cit., p. 989.
32. A great many things are bought and sold in the marketplace, and much of that in-
volves a loss of dignity. Prostitution is only the most obvious example, and the efforts to
deny the trnde the protections of the law may be taken as evidence of the salience of the
threat to dignity as well as the concern that its consequence is not limited to the primary
parties in the transaction but is threatening to the moral fabric of the society.
33. National Public Radio story, "Weekend Edition;• December 29, i991, included a
comment by a satisfied customer who uses his device faithfully, and "if he's over the limit,
he doesn't drive." Counterarguments were heard from representatives of Mothers Against
Drunk Driving who argued that the device encourages driving while under the influence of
alcohol and that targeting those who had already demonstrated a tendency toward such
dangerous behavior was irresponsible.
260 NOTES

34. Blaustein, op. cit., p. 999.

35. Blaustein, op. cit., p. 1, 003.
36. Spiros Simi tis, "Reviewing privacy in an information society." University ofPennsyl-
vania Law Review135 (1987)707-746.
37. Ibid., p. 731.
38. Anne W. Branscomb, "Property rights in information," pp. 603-642. In M. Gurevitch
and M. Levy [eds], Mass Communication Review Yearbook, Vol. 6. Newbury Park, CA: Sage,
1987; previously published in B. Guile [ed], Information Technologies and Social Transfor-
mation. Washington, D.C.: National Academy Press, 1985.
39. Ibid., p. 603.
40. This refers to an advantage such as that derived from the monopoly claims on the
electronic frequency spectrum provided with an exclusive broadcast license. Tax subsidies,
import restrictions, patents, and copyrights all represent state guaranteed advantages that
carry with them some expectation of public benefit.
4i. Herbert Schiller, Culture, Inc. : The C-Orporate Takeover of Public Expression. New
York: Oxford University Press, i989.
42. The conflicting positions taken by activists and others within the homosexual com-
munity over the practice of "outing" has been discussed in Alida Brill, Nobody's Busi11ess:
The Paradoxes of Privacy. Reading, MA: Addison-Wesley, i990, pp. i28-142. Gay activists
justify outing as a means of normalizing homosexuality by increasing public recognition of
its prevalence, especially as it regards persons of importance who have avoided labeling or
even as it refers to suspicion of their difference in the past.
43. As par t of Title III of the federal Superfund Amendments Reauthorization Act
(SARA), the Emergency Planning and Community Right-to-Know Act as passed in 1986 re-
quired manufacturers to provide public estimates of the amounts of their "routine" emis-
sions of toxic chemicals into the environment. Six years later, the environmental movement
was actively discussing the "right to know more!' See for example, "The next steps for right
to know." Environmental Action 23 (3) (November/December 1991): 19-24·
44. U.S. v. Miller, 425 U.S. 435 (1976).
45. Quoted in Bazelon, 1977, op. cit., from California Bankers Association 11. Shultz, 416
U.S. 21, 78 (1974), p. 609.
46. Branscomb, op. cit., p. 615.
47. The Supreme Court in 1991 struck down a New York law, referred to as the "Son of
Sam" law, which sought to limit the realiiation of profit by criminals who commercialized
their exploits in the mass media. Part of the rationale in the New York courts, and the
countless state courts that used the rule as a model, was that victims rather than criminals
ought to be compensated from books about criminal activity. Although recognizing a legit-
imate state interest in compensating victims, the restriction on particular forms of speech
was not considered to be justifiable under the First Amendment. See Paul M. Barrett,
"High Court rejects 'Son of Sam' law, citing criminals' rights of free speech." Wall Street
Journal. December n, 1991, p. A4. The prior restrain t o n publication by governmen t em-
ployees who might be working with "sensitive" information still remains as a restriction on
the realization of profit.
48. Edward Pio man and L. C lark Hamilton, Copyright: Iritellectual Property in the Infor-
mation Age. London: Routledge and Kegan Paul, i980; Sandra Braman, "Trade and infor-
mation policy." Media, Culture and Society u (1990 ): 361-385.
NOTES

49. Robert J. Posch, "How the law(s) of 'privacy' impact your business." Direct Market-
ing (October 1987):78. That free speech rights find constitutional expression in the First
Amendment and that privacy rights are found only in a penumbra of emanations is the
product of a narrow and self-interested interpretation. Fourth Amendment limitations of
state access seem equally explicit, even if the word privacy is not used.
50. bM News, a trade paper for the direct marketing industry, reported the availability
of a list of recent callers to 900 numbers, including callers to adult, dating, and psychic
lines. The same issue offered listings of the 50,000 "active members" of the Penthouse Book
and Video Society, which "offers 'hard-to-find' adult reading and viewing material." De-
cember 2, i991. Offering of the lists by the publication would not be seen as publication in
the same way that posting of the lists might.
51. C. Edwin Baker, Human Liberty and Freedom ofSpeech. New York: Oxford University
Press, i989, especially Chapter 9.
52. Ibid., pp. 201- 202.
53. Ibid., pp. 205-206.
54. Ibid., p. 211.
· 55. Kenneth C. L~udon, Dossier Society: Value Choices in the Design ofNational Informa-
tion Systems. New York: Columbia University Press, 1986, p. 367.
56. Priscilla M. Regan, "Public use ofprivate information: A comparison of personal in-
formation policies in the United States and Britain." (unpublished dissertation) Cornell
University, 1981, pp. 361- 368.
57. John C. 13ennett, "Regulating the computer: A comparative study of personal data
protection policy." (unpublished dissertation) University offllinois at Urbana-Champaign,
1986, p. 298.
58. Flaherty, 1989, p. 309.
59. David P. Linowes, Privacy in America. Urbana, IL: University of Illinois Press, 1989, p.
82.
60. "Select Committee to Study Governmental Operations with Respect to lnteUigence
Activities." U.S. Senate. 94th Congress. First Session, i975.
6t. Flaherty, op. cit., p. 309.
62. George B. Trubow, "Watching the watcheirs: The coordination offederal privacy pol-
icy." Software Law ]ournal3 (3) (Summer t989) :391- 4u.
63. David H. Flaherty, "The need for an American Privacy Protection Commission."
Government Information Q11arterly1 (3) (1984):246.
64. Kenneth James Langan, "Computer maLching programs: A threat to privacy?" Co-
l1~mbia Journal of Law and Social Problems 15 (1979 ):142-180.
65. Testimony of Janlori Goldman, ACLU, on S. 496, the Computer Matching and Pri-
vacy Protection Act of i987, before the Subcomm ittee on Government Information, Justice
and Agriculture. House Government Operations Committee, June 23, 1987, p. 8.
66. Flaberty, 1989, p. 357. .
67. Herbert Hovenkamp, Enterprise and American Law, 2836-1937. Cambridge, MA:
Harvard University Press, 1991.
68. Morton J. Horwitz, "The doctrine of objective causation," pp. 201-213, in David
Kairys [ed], The Politics ofLaw: A Progressive Critique. New York: Pantheon, i982.
69. Peter W. Huber, Liability: The Legal Revo.lution and its Consequences. New York: Ba-
sic Books, 1988.
262 NOTES

70. Tureen v. Equifax, 571 F 2d 411 8th Cir. i978, cited in Jonathan P. Graham, "Privacy,
computers and the commercial dissemination of personal information." Texas Law Review
65 (7) (June i987):I,416.
71. Susan Gardner, "Wiretapping the mind: A call to regulate truth verification in em-
ployment." San Diego Law Review 21 (1984): 295- 323.
72. A distinction is made between reliability and validity. Presumably, a core attribute
such as honesty should not be highly variable within individuals. An unreliable instrument
might produce a low honesty score one day and a high score another day, which could indi-
cate a change in honesty that had not taken place. The reliability of the measure might be
affected by the humidity in the room, the amount of coffee consumed by the person being
tested, the time of the day when the test is taken, or any one of a hundred other factors that
might cause the scores to vary. Although an unreliable measure can be valid only by chance,
in the case of predictive validity, the concern with validity turns on the relationship be-
tween the test score and some unmeasured, but genuine, attribute of the individual's per-
sonality, which, if known , could predict a variety of behaviors. Predictive validity, then, is
concerned with the accuracy of the test in the identification of persons who would, if given
the opportunity, act dishonestly.
73. Ann M . Barry, "Defamation in the workplace: The impact of increasing employer li-
ability." Marquette Law Review72 (1989): 264-303. Barry discusses the emergence of a "neg-
ligent reference theory" by which employers seek actions against those who write refer-
ences for not providing relevant information that they held and by which the failure to
disclose information is defined as negligent, p. 301. This liability is in conflict with suits
charging negligent hiring, defined as a "breach of the employer's duty to make an adequate
investigation of an employee's fitness before hiring him," p. 302. A comprehensive examina-
tion of employer liability that emphasizes privacy dimensions is found in Richard M.
Howe, "Minding your business: Employer liability for invasion of privacy." Paper presented
to the Association of Life Insurance Counsel, White Sulphur Springs, WV, May 21, 1990.
74. Raymond Wacks, Personal Information, Privacy and the Law. Oxford: Oxford Uni-
versity Press, 1989.
75. Ibid., p. i81.
76. Ibid., p. 238.
77. ibid .• p. 230.
78. Richard A. Posner, "Privacy, secrecy and reputation." Buffalo Law Review 28 (1979),
p.5.
79. Amitai Etziooi takes detailed exception to this claim and provides nllll1erous exam-
ples that indicate the influence of moral, nonselfish motivations. Amitai Etzioni, The Moral
Dimension: Toward a New Economics. New York: Free Press, i988.
80. Posner, op. cit., p. 9.
81. The fact that these estimates are themselves subject to race, class, and gender bias is
ignored in Posner's and other neoclassical responses to critiques of discrimination.
82. This has been rather inadequately measured in terms of the proportion of African -
American and Hispanic residents of the state.
83. Robert Posch, "Why are we returning to redlining?" Direct Marketing (March
1990):77.
84. Posner, 1979, op. cit., pp. 44-45.
85. Ibid., p. 50.
NOTES

86. Andrew Cassel, "Loan-rejection study puts banks on defense." Philadelphia Inquirer.
October 28, 1991, p. D1; Anthony R. Wood, "Brokers having qualms over loan criteria." Phil-
adelphia Inquirer. October 28, 1991, p. D1.
87. Simitis, 1987, op. cit., p. 735.
88. Laurence A. Benner, "Diminishing expectations of privacy in the Rehnquist Court:'
John Marshall Law Review22 (4) (Summer 1989):525-876.
89. Benner notes that the force of the warrant clause was weakened by the Court in a de-
cision regarding the right of housing inspectors to gain access to private property without
the necessity of a warrant demonstrating probable cause. Thus a distinction for administra-
tive, rather than criminal searches was established. See Camara v. Municipal Court, 387 U.S.
523, 538-539, 1967, cited p. 831.
90. Ibid., p. 835.
91. The case of California v. Greenwood, 108 S. Ct. at 1,627 cited in Benner, p. 856.
92. See Skinner v. Railway Executives' Association, 109 S. Ct. 1,402 (1989) cited in Benner,
p. 843.
93. Ibid., p. 845.
94. See Katz v. United States, 389 U.S. 347 (1967) cited in Benner, p. 852.
95. See California v. Ciraolo, 476 U.S. 207 (1986) cited in Benner, p. 859.
96. Ibid.
97. James J. Tomkovicz, " Beyond secrecy for secrecy's sake: Toward an expanded vision
of the Fourth Amendment privacy province." Hastings Law Journal 36 (5) ( March 1985):736.
98. Loretta E. Murdock, "The use and abuse of computerized information: Striking a
balance between personal privacy interests and organizational information needs." Albany
LawReview44 (3) (April 1980):610.
99. See for example, Robert Ellis Smith, Compilation ofState and Federal Privacy Laws,
i988. Washington, D.C.: Privacy Journal, t988.
100. Colin J. Benn~tt, "Computers, personal data, and theories of technology: Compar-
ative approaches to privacy protection in the 1990s." Science, Technology and Human Values
16 (1) (Winter 1991):51-69; Colin J. Bennett, "Regulating the computer: Comparing policy
instruments in Europe and the United States." European Journal of Political Research 16
(1988):437- 466.
101. David Chaum, "Security without identification: Transaction systems to make big
brother obsolete:' Communications of the ACM 28 (10) (October i985):1,030-1,044.
102. Bennett, 1991, p. 58.
103. Flaherty, 1989, op. cit.. p. 382.
104. Ibid.. p. 365.
105. " How the American public views consumer privacy issues in the early '90s and
why:' Testimony of Alan F. Westin before the Subcommittee on Government Information,
Justice and Agriculture. House Committee on ·Government Operations, Washington, D.C.,
April 10, 1991.
106. Baker, 1989, op. cit., p. 213.
107. Ibid., p. 212.
108. Robert Babe angered economists with his return to the fundamentals in the warn-
ing he delivered to them and other policymakers at a telecommunications p.o licy confer-
ence. See his ulnformation industries and economic analysis: Policymakers beware," pp.
 NOTES

123-135. Tn 0. Gandy, P. Espinosa, and J. Ordover [eds], Proceedings from the Tenth Annual
Telecommunications Policy Research Conference. Norwood, NJ: Ablex, 1983.
109. Meheroo Jussawalla and Chee-Wah Cheah, Chapter 4, "Economic analysis of the
legal and policy aspects of information privacy;' pp. 75- 102 in their book The Calculus ofIn-
ternational Communications. Littleton, CO: Libraries Unlimited, i987.
uo. Rohan Samarajiva and Roopali Mukherjee, "Regulation of 976 services and dial-a-
porn: Privacy and policy implications for the intelligence network." Conference paper. Sev-
enteenth Congress of the International Association for Mass Communication Research,
Bled, Yugoslavia, August 1990.
111. Jussawalla and Cheah, op. cit., p. 89.
u2. Ruth Faden and Tom Beauchamp with Nancy King, A History and Theory of In-
formed Consent. New York: Oxford Unversity Press, 1986, p. 8.
u3. Ibid., p. 28.
u 4. l am quoting Cardoza in the 1914 case of Schloendorff v. Society of New York Hospi-
tals, p. 123.
u5. Martin R. Gardner, "Consent as a b<!r to fourth amendment scope--A critique of a
common theory." Journal ofCriminal Law and Crimi11ology71 (4) (1980):443- 465.
u6. Gary T. Marx, Undercover: Police Surveillance in America. Berkeley: University of
California Press, 1988.
117. fbid., p. 107.
u8. Edward Laumann and David Knoke, The Organizational State: Social Choice in Na-
tional Policy Domains. Madison: UniversityofWisconsin Press, i987.
119. Janlori Goldman responded to my question while she was serving as a panelist at a
conference organized by Telecommunications Reports, "Caller-ID, ANI, and Privacy:' Wash-
ington, D.C., October 15-16, i990. We continued our discussion brjefly following a panel on
privacy at the Nineteentli Annual Telecommunications Policy Research Conference, Solo-
mons, MD, September 28- 30, 1991.
l20. Robert Kastenmeir, opening statement (August 3, i988), Joint Hearing on Video
and Library Privacy Protection Act of i988. Subcommittee on Courts, Civil Liberties, and
the Administration of Justice of the House Committee on the Judiciary and the
Subcomittee on Technology and tlie Law of the Senate Committee of the Judiciary. Wash-
ington, D.C.: U.S. Government Printing Office, 1989.
121. Testimony of Vans Stevenson in hearings cited above, pp. 75- 78.
122. Testimony of Richard Barton in hearings cited above, p. 88.
123. "DMA comes out against video privacy legislation as written." Friday Report. Au-
gust 5, i988, p. i.
124. This commission, headed by Dav.id Linowes, investigated issues of personal privacy
between 1975 and 1977 and made recommendations for private as well as public sector pol-
icy regarding personal information.
125. Video Privacy Protection Act of i988. PL 100- 618, 102 Stat. 3,196; 2d, November 5,
1988.
126. Steve Bates, "A blockbuster debate about privacy." Washington Post. January 1, i991,
p. El.
127. As has been noted, transaction processing is a multibillion dollar industry with high
growth potential because telecommunications networks provide interconnections to
countless point-of-purchase terminals that have become central to the operation of bank-
NOTES

ing, retail, and transportation activities. The corporations that provide the telecommuni-
cations and computer services that define this market have made controversial daims re-
garding their rights to utilize the information that is generated through their delivery of
these services.
128. Paul Shultz, Caller ID, ANI and Privacy: A Review of the Major Issues Affecting Num-
ber Jdentification Technologies. Washington, D.C.: Telecommunication Reports, i990, p. 4.
129. State of New York Public Service Commission. "Proceeding on motion of tbe com-
mission to review issues concerning privacy in telecommunications." 9o-C--0075. January
31, 1990. The primary resource in this communication was the "Memorandum and Ques-
tions" provided by Commissioner Noam.
130. Part of the argument in Smith v. Maryland, 442 U.S. 735 (1979) was that persons
should know that the telephone company could and would use available technology to
identify the numbers from which illegal or harassing calls would be made. Thus all that
would be necessary would be widespread publicity about the ayailability of Caller-ID or
other technology to eliminate any claims based on reasonable expectations.
131. The attorney general of North Carolina called for an investigation of Southern Bell
because corporate executives had written a memorandum encouraging the generation of
letters to utility commissioners. The memo included copies of ten sample letters that were
to be used as "guidelines" in generating this flood of public response. In addition, execu-
tives were war~ed to use personal rather than company stationery in writing versions of
these letters. Details of this charge were presented in "N.C. attorney general seeks probe of
Southern Bell Caller-TD campaign." Privacy Times. Marchi, 1990, p. 3.
i32. Police in several jurisdictions have indicated that Caller-ID has threatened several
undercover operations and the lives of police and persons cooperating with police. See for
example, "Florida law enforcers oppose Caller-ID, Southern Bell plan." Privacy Times. Au-
gust 14, i990, p. 2.
i33. Thomas E. McMailus, Telephone Transaction-Generated Information: Rights and Re-
strictions. Report P- 90- 5. Cambridge, MA: Harvard University, Program on Information
Resources and Policy, May i990. .
134· U.S. Congress. Qffice ofTechnology Assessment. Critical Connections: Communica-
tion for the Future. OTA-CIT-407. Washington, D.C.: U.S. Government Printing Office,
January 1990.
135. Signalling System No. 7 is the current standard for control signaling on digital tele-
communication networks.
i36. Testimony of Rohan Samarajiva, Call Management Services Trial, The Public Utili-
ties Board of Manitoba, i991.
137. Although Congress passed the Telephone Consumer Protection Act in 1991 to con-
trol these systems and the associated abuses, it appears that the Federal Communications
Commission refused to honor the intention of the legislation. See "FCC stands up for tele-
marketing; Hits do-not-call database option." DM News. April 27, 1992, p. I.
138. Susan Burnett Luten, "Give me a home where no salesmen phone: Telephone solici-
tation and the First Amendment." Hastings Constitutional Law Quarterly7(Fall1979):i29-
164; Mark S. Nadel, "Rings of privacy: Unsolicited telephone calls and the right of privacy."
Yale Journal ofRegulation 4 (December i986):99-128.
139· A regulatory rather than a voluntary option along these lines has been considered
with regard to the placement of an asterisk, or other identifying mark, next to a name in a
266 NOTES

directory that would inform callers that unsolicited calls were unwelcome. This option fa.
cilitates the mounting of a "no trespassing" sign on the telephone. Yet blanket bars against a
category of calls defined as "unsolicited" appear to represent an unmanageable and ineffi-
cient response for all but a very small number of persons.
140. The problem of vaguenes.s in the definition of those classes of callers that would be
seen as intolerably harassing by a reasonable person is raised by M. Sean Royall, "Constitu-
tionally regulating telephone harassment: An exercise in statutory precision." University of
Chicago Law Review 56 (4) (1989):i.403- tAJ2.
141. We note that having an unlisted or unpublished telephone number provides no
guarantee that individuals can even pay to limit access to their homes or their persons by
phone. Any call made to another party, especially a commercial party that has some form of
CNI, will introduce those numbers into the pool.
142. A citation is frequently made to the case of Smith v. Maryland, 422 U.S. 735 (1979) in
which a pen register, a device that records the numbers called from a particular telephone,
was used to gather evidence of obscene and harassing telephone calls. The court argued
that since the device did not record the content of the communications, only the fact that
they were made, it was sufficiently unlike a wiretap to make the requirement of a prior war-
rant unnecessary. The court argued further that because it is necessary to indicate the num-
ber one wishes to call, even if mechanically, to the telephone company, there could be no
reasonable expectations of privacy in those numbers. Glenn Chatmas Smith, "We've got
your number! (Is it Constitutional to Give it Out?): Caller identification technology and
the right to informational privacy." UCLA Law Review 37 (1989):145-223. This provides a
detailed critique of the Smith court's view and its limited applicability to the technology of
Caller-ID.
143· "Caller-ID declared legal by South Carolina County Judge." Privacy Times. Decem-
ber 4, 1990, pp. 2- 3.
144· David B. Hack, "Caller I.D. and Automatic Telephone Nwnber Identification. CRS
Issue Brief, IB9.0085. Washington, D.C.: Congressional Reference Service. Library of Con-
gress, February u, t991.
145· Amy S. Rosenberg, "Rising call for unlisted numbers." Philadelphia Inquirer. May 6,
1990,p. Al.
146. Women frequently request that they be listed in telephone directories only by ini-
tials rather than by given name so as not to provide information to potential harassers
about homes with single women. Professional women and feminists, however, have pres-
sured the telephone companies to include individual listings of their names as well as list-
ings for their husbands.
147. Remarks by Will Dwyer, CEO of Private Lines, Inc., of Beverly Hills, CA, as pre-
sented to the Telecommunications Reports conference on Caller-ID, ANI, and Privacy,
Washington, D.C., October 15, 1990.
148. Gary Marx included such claims, as well as others, in his influential testimony be-
fore the Pennsylvania Public Utility Commission. Docket No. R-891200, Mayi989. ·
149. Julie Johnson, "ANJ leads the way." Telephony. April 10, 1989, p. 34-
150. American Express reportedly changed its inst ructions to its service operators after
cardmembers reported surprise and some irritation with the knowledgeable greetings they
received because operators knew who they were before they actually answered the call.
151. Ibid., p. 35.
NOTES 267

i52. This is from a presentation by Steve Mulcahy, Manager, corporate marketing, 800
Service for MCI Communications, to the Telecommunication Reports conference on
Caller-ID, ANI & Privacy, Washington, DC, October 15-16, i990.
i53. James Rule and Paul Attewell. "What do computers do? Social Problems 36 (3)
(1989):225-241, p. 237.
154· These are the comments of Brian R. Lane, Managing Director for Marketing and
Regulatory Planning, NYNEX Service Company, at a conference on "State Telecommuni-
cations Regulation;• Aspen, Institute, Aspen, Colorado, July 9- 13, 1989.
155. Hack, 1991, op. cit., p. 7.
156. Marx, testimony, op. cit., pp. 10- 11.
157. These comments are in regard to the Proceeding on the Motion of the Commission
to Review Issues Concerning Privacy in Telecommunications, Case 90-C-0075, March 15,
1990.
158. These comments by Cathleen Black were delivered at an address to the National
Press Club on October 7, i991, as quoted in Telecommunications Reports, October, 14, 1991,
p.5.
159· We are reminded of the example of question framing, which opposes the Roper
form of the question of economic justification: "Don't you agree that watching commer-
cials is a reasonable price to pay for receiving quality television entertainment?"- which
finds nearly the same proportion of the population agreeing that "television commercials
represent an annoying interruption."
160. Evan Hendricks, Trudy Hayden, and Jack Novik, Your Right to Privacy. 2d Edition.
Carbondale, ll: Southern Illinois University Press, i990.
161. Meredith W. Mendes, "Privacy and Computer-based Information Systems," pp.
193-264. In B. Compaine [ed ), Issues in New Information Technology. Norwood, NJ: Ablex,
i988.
i62. Laumann and Knoke, op. cit., 1987.
163. The assumption of an absence of an individual's legitimate interest in records held
by others, added to the already weakened standards of reasonableness following Katz,
started a decline in privacy fortunes that has not yet begun to slow.
164. By distribu ting readily available information about the demographic makeup and
consumer orientation of the more than 120 million households on CD-ROM discs, Lotus
and Equifax expected their product to revolutionize marketing among the small businesses
that had been less active users of consumer list services. After the Wall Sreet journal pub-
lished an article describing this proposed product, computer activists distributed copies
and comments through their electronic bulletin boards and electronic mail networks. The
resultant swarm ofleners and threats led Lotus to withdraw the product, at a reported loss
exceeding $8 million.
· 165. Michael W. Miller, "Coming to your local video store: Big Brother." Wall Street Jour-
nal. December 26, 1990, p. 9.
i66. Jerry Berman and Janlori Goldman, A Federal Right of Informational Privacy: The
Need for Reform. Washington, D.C.: Benton Foundation, i989.
167. David F. Linowes, "The U.S. Privacy Protection Commission. A restrospective view
from the chair." American Behavioral Scientist 26 (5) (May/June 1983):577-590.
168. Council ofEurope, Convention for the Protection ofIndividuals With Regard to Auto-
matic Processing ofPersonal Data; OECD. Guidelines Governing the Protection ofPrivacy and
Transborder Data Flows of Personal Data.
268 NOTES

169. Flaherty, 1989, p. 380.

170. Simitis, 1987, pp. 740-74i.
171. Moore, i984, op. cit., p. 288.

CHAPTER 8
i. MajidTehranian, Technologies of Power. Norwood, NJ: Ablex, 1990.
2.Samuel Bowles and Herbert Gintis, Democracy and Capitalism. New York: Basic
Books, i987.
3. James Beniger, The Control Revolution. Cambridge, MA: Harvard University Press,
1986.
4. Robert M. Entman, Democracy Without Citizens: Media and the Decay of American
Politics. New York: Oxford University Press, 1989, p. 128.
5. Ibid., p. 136.
6. Herbert I. Schiller, Culture, Inc. New York: Oxford University Press, 1989.
7. Jiirgen Habermas, Moral Consciousness and Communicative Action (trans. C.
Lenhardt and S. Nicholsen). Cambridge, MA: MIT Press, 1990, p. 133·
8. Ibid., p. 136.
9. Ibid., p. 175.
io. David Flaherty, Protecting Privacy in Surveillance Societies. Chapel Hill: University of
North Carolina Press, 1989, pp. 393-394.
u. Jacques Ellul, The Technological Society. Alfred A. Knopf, 1964, p. 427.
12. Stewart Brand, The Media Lab. Inventing tf1e Future ai M.T.T. New York: Penguin
Books, 1987.
I
ABOUT THE BOOK AND AUTHOR

The consensus is clear. Personal privacy will become the dominant issue of the 1990s. Yet a
focus on privacy, as we have come to .understand it so far, all but guarantees that we ignore
the implications of the privacy debate at the more fundamental levels of individual auton -
omy, collective agency, and bureaucratic control. The Panoptic Sort helps us to understand
just what is at stake when the bureaucracies of government and commerce gather, share,
and make use of an almost unlimited amount of personal information to manage the social
and economic systems within their spheres.
Unlike Foucault's panoptic prison, which involved continual, all -encompassing surveil-
lance, the current panoptic system depends upon the ability of operators to classify and
then separate disciplinary subjects into groups in a way that increases the efficiency with
which the techniques of correct training or rehabilitation may be applied to each individ-
ual. This book describes in full detail the design and use ofthe pan optic operation, with ex-
amples from marketing, employment, insurance, credit management, and the provision of
governmental social services.

Oscar H. Gandy, Jr., is professor of communication at the Annenberg School for Com-
munication at the University of Pennsylvania.
INDEX

Access to information, 60, 68, 85, 90, 94, 138, Application process, 62, 69, 74-76, 100, 132,
175, 181, 189- 192, 194,199,203,208, 147
215-216,219,224 Appropriation, 185, 187
remote, 93 Arriaga, Patricia, 52
ACLU. See American Civil Liberties Union Artificial intelligence, 96
A. C. Nielsen Company, 65, 93 Assessment, 15-18, 24, 35, 182
Activism, 166, 168- 169, 209 Association, 217
Actuarial model, 85 AT&T. See American Telep hone and
Addressability, 231 Telegraph, Inc.
Administration, bureaucratic, 40, 46 Attewell, Paul, 218
Administrative purpose, 57, 74 Attitudes, 149, 155- 156, 175
Advertising, 70, 104, 193, 227 Attributes, personal, 17
African-Americans, 65, 77, 82, 150, 152, 168, Audiences, 62, 65, 77, 84
200-201,218 analysis of, 65
Age, 152, 155- 157, 160, 163, 169-170 Auditability, 6 1
cohort, 155, 160-161 Audit trail, 2 I7
group, 156 Authentication, 71
Authority, 47-48
Agency, 7, 12, 26, 29, 46
Authori7.ation, 16, 24, 27, 73, 96-97, 122,
AIDS, 107
207,217
American Civil Liberties Union (ACLU), 56,
third party, 97
J96, 209, 210, 221
Automatic number identification (ANJ), 91,
American Express (Amex), 24, 66, 71, 73, 90,
109, 164, 212, 217
95- 97, 102
Automation, 51, 61, 65, 73
American Express card, 73
Automobile, 35-36
American Library Association, 210 insurance, 64, 74, 163
American Newspaper Publishers Association, registration, LOS, 187
219 .
Autonomy, 3, 135, 180-181, 189, 194, 205,
American Student List Company, 91 221- 225
American Telephone and Telegraph, Inc.
(AT&T), 25, 33, 67, 103, 137, 141, 152, Background checks, 76
212 Baker, C. Edwin, 193-194, 205-206
Amex. See American Express Bank records, 191
Anderson, James, 83 Bank Secrecy Act, 191
AN!. .See Automatic nwnber identification !3ar codes, J 29
Annoyance, 215 Barron, James, 184-185
Answering machine (telephone), 103 Bartlett, Randa ll, 3, 9, 18-19, 29
Apple Computers. 72, 93 Barton, Richard, 210-21 l

271
272 INDEX

Barwise, Patrick, 69, 77 surveillance, 4·6, 189

Bawn,L.Frank,230 systems, 54, 178
Bazelon, David, 18 1 Bureau of Labor Statistics, U.S., 59
Beetham, David, 8, 40 Bureau of the Census, U.S., 58-59, 61, 88,
Behavior modeling, 72 141, 161
Behaviorscan, 69 Burge, W. L., 101- 102
Belcore. See Bell Communications Research Burnett, P,,d, 93, J05
Bell, Daniel, 50-51 Business establishments, 110
Bell Atlantic Corporation, 2 l 4
Bell Communications Research (Belcore), Ca.lifomia, 216
214 Caller-ID, 68, 103, 141. 148-149, 161, 164,
Bell Operating Companies, 103, 219 203,212- 217
Bendix, Richard, 7 blocking, 219, 221
Benello, C. George, 37 Calling number identification (CN I), 68, 105,
Beniger, fames, 39, 41-42, 49, 5 l 211-214, 217-218
Benner, Dennis, 106 Calling party identification, 217
Benner, Laurence, 201-203 Cambridge Reports, 145
Bennett, John, 195, 203-204 Capitalism, 18, 25, 29, 35, 40-41, 49- 50, 227
Bentham, Jeremy, 22, 23 Cardoza, Benjamin, 208
Better Business Bureau, 100 Cate, Fred, 84
Bias, 16, 18, 124
Causation, 44-45, 197
in assessment, 54
CD-ROM, 93, 142
Bibliographic databases, 93
Census,57-58,82, 130,134
Big Brother, 140
data, 87, 105
Biometriddentifier, 81
Central Intelligence Agency (CIA), 56
Birth certificate, 80
Change-of-address service, 106
Black, Cathleen, 219
Chawn, David, 203
Blacks. See African-Americans
Cheah, Chee-Wah, 206-207
Blockbuster Enter tainment, 211, 222
Bla ustein, Edward, 185- 189 Chilton Corporation, 99
Christians, Clifford, 36
Blue Cross/Blue Shield, 81
Bork, Robert, 138, 210 Church, Frank, 195
Bose, Arun, 5 CIA. See Central Intelligence Agency
Bostwick, Gary, 181 Citicorp, 66, 87
Bowles, Samuel, 19, 227 Civil rights, 203- 204
Brand, Stewart, 72, 230 Claritas Corporation, 88, 91
Brandeis, Lo uis, 182- 186 Clarke, Roger, 71
Brand names, 82 Class, 169
Bransco01b,Anne,4, 190-193 analysis, 30-32
Braverman, Harry, 41 consciousness, 29-33, 48
Brubaker, Rogers, 9, 39, 40 ·divisions, 174
Bullock, Richard, 100 positions, 175
Bunge, Mario, 44, 45 social, 29, 150, 153, 174
Bureaucracy, 7-8, 42, 46-48, 55, 95, 194-195, Classification of indjviduals, 10, 15, 16, 17,
203-204,207- 208,224 24, 35, 43, 81-86, 92, 150, 161, 180, 191
corporate, 62, l 95 Cluster analysis, 78
Bureaucratic CNI. See Calling number identification
imperative, 61 Coercion, 186, 190
organizations, 18 Cohen, Stanley, 11
INDEX 273

Cohort lists, 91, 104, 108, 217

aging, 160 privacy, 96
measurement, 157 profiles, 130, 218
Collective bargaining, 31 Content analysis, 53-55, 80
Collectivity, 178 Contracts, 9, 184, 190, 205, 207
Commercialization, 183, 186, 192 implied, 183, 190. 220
Commercials, 69 Control, I, 6, 9- 12, 21, 24, 33, 42, 47, 75, 82,
Commoditization, 34 85- 86, 138, 143, 164, 179, 181, 189, 213,
of information, 2, 52 215,227,230
Common law, 183-184, 188, 221 loss of, 129, 164, 181, 207
Communication, 3, 33, 42, 229 Control Revolution, 41, 49, 227
strategic, 177, 228-229 Coordination, 33, 66
technology, 37 Copyright, 182-183
Communicative competence, 3, 34 Corporate database, 66, 107
Company Corporate person, 197
size, 114, 121- 122 Corson, Ross, 94
type, H9 Cost-effectiveness, 75
Competition, 62 Council of Europe, 223
Compliance, 175 Coupons, 70, 221
Computerization, 24, 46, 47, 64, 66, 218 CPNI. See Customer, proprietary network
Computer Matching and Privacy Protection · information
Act, 75 Credit, 98, l 08, 222
Computers, 38, 51, 55-56, 64, 74, 83, 98, 104, application, 133
130, 137' 143, 204 authorization, 80, 131, 211
crime and, 100 bureau ,64,92-93, 141, 163,201
errors with, 130 cards, 66, 73, 80, 89, 93, 126, 131, 133, 146,
files, 100, 127 206
security, 10 l data,98-99, 101,108-l09
Confidentiality, 59, 141, 144, 179, 188 history, 44, 98, 129, 200
Conformity, 24 . rating, 20 ··
Congress, U.S .. 122, 195 records, 67, 93, 129, l36, 145
Consciousness, 28, 30, 33, 39, 123-124 report, 20, 98, 132, 201, 224
discursive, 28, 124 reporting firms, 108
practical, 28, 124 risk, 100, 130
Consent, 132, 186, 188, 207-208, 210, 211, scoring, 87
214,224 transactions, 81, 192
implied, 210 Criminal justice system, 86
See also Informed consent Criminals, 145, 149
Consequentiality, 17, 59, 191 privacy rights of, 145
Conservatism, 155, 165 records of, 192
Constitution, U.S., 138, 208 Critical Theory, 2, 230-231
Constraint, 181 Cross-marketing, 109
Consumerism, 8, 102 Cultivation analysis, 12, 164
Consumers, 42, 6H52, 64, 66-70, 85, 93- 95, Customer, 47, 62, 69, 83, 209
99, 102- 104, 142, 163, 193, 2 10, 216, 227 data, 96, 109, 222
behavior of, 70, 77, 79, 85, 124, 150 files, 66
credit, 86, 92, 10 l information, 97, 109-112, 120
databases, 66, 92 lists, 113, 193
information, 65, 93, 98, 108 profiles, 94
274 JN DEX

proprietary network information (CPNI), Disclosure, 185- 187, 189, 190--191, 193-194,
219 206, 211, 216
Discriminant analysis, 78-79
Dandeker, Christopher, 8, 40, 46-47 Discrimination, 8~1, 107, 200-201
Data, 54, 57, 102- 106 racial, 20 1
census,58 Dishonesty, 198
enhancement services, 93 Distanciation, 12, 33-34
gathering,58, 132- 133, 142, 145 Dizard, Wilson, 49- 50
linkage, 204 DMA. See Direct Marketing Association
protection. 99, 204, 224, 229 Documentation, 24, 80- 8 1
raw, 56 Domhoff, G. William, 20
reduction, 55 Domination, 19, 227
sharing, 96 process of, 27, 33, 48
Database, 70, 76, 94, 106, 197, 204 systems of, 25
electronic, 98 Donnelly Marketing Information Services, 90
management, 101 Donner, Frank, 56
ma.r keting, 62, 90, 95, 107 Double hennaneutic, 44
searching software, 139 Doubleday & Company, 106
vendors, 212 Douglas, Mary, 10, 83
Database America, 93 Douglas, William 0 ., 192
Dataveillance, 71 Driver's license, 80
Debit cards, 81 Drugs
Deception, 208 abuse of, 61
Decision support system, 72 screening, 100
Defamation, 188 testing for, 61 , 86, 127, 132, 202
Deficit Reduction Act, 57 use of, 64, 138
Dehumanization, 40 Due process, 182, 196
Democracy, 133, 179, 189, 227- 228 Dun and Bradstreet, 90, 110
Demographic Duster, Troy, 17, 76
classification, 120
identifiers, 77 Economic
information, 69, 86-87, 104, 113 b ase, 30, 39
Department of Conunerce, U.S., 61 rationality, 40
Department ofEducation, U.S., 75 Economics, 4-5, 8, 39
Deskilling, 41, 46 neoclassical, 19, 50
Determination, process of, 4445 Economists, 39, 52, 78
Detweiler, John, 144 Education,152-154, 168- 170, 190
DIALOG/Knowledge Index, 94 Efficiency, 7-11, 22, 23, 36- 37, 41-42, 48-49,
Dignity, 185-186, 188-189, 198 59,67, 71, 74,148,194-196,202,207
Direct mail, 107, 126, 141, 228 Ehrenberg, Andrew, 69, 77
Direct marketing, 95, 97- 98, 104, 106-108 Elazari, Barbara, 73
industry, 91, 96, !07-108, 163, 193, 210, Elig ibility, 62, 76
222 Ellul, Jacques. 5- 7, 36-39, 42, 230
Direct Marketing Association (DMA), 97, Embarrassment, 181
104-108,209-211,215,219-220,223 Employees, 60, 75, 100, i08
Direct Marketing Association of Long Island, Employers, 20, 75, 135, 144-145, 147, 154,
107 200
Disciplinary surveilJance, 54, 230 Employment, 76, 128, 139, 144, 148, 198
Discipline, 24 Entman, Robert, 228
INDEX 275

Entrapment, 56 Flaherty, David, 57-58, 179, 195- 196, 204,

Entrepreneurs, 38 222- 225,229-230
Epsilon Data Managem ent, 90 Fleish m an, John, 165
Equality, 40, 49, 166, 168-169, 173, 177 Focus group interview, 125
Equifax, 92-93, 95, 99, 100- 105, 108, 122, Focus groups, 136, 143, 145, l 47
140, 142 Fordism, 41
Ethnicity, 152 Fortune 500, I 08
Etzioni, Amitai, 19 Foucault, Michel, 9-11, 16, 21-24, 34, 39,
Europe,204 82-83, 190
European Community, 97, 223 Foundations, 20
Evaluation, 24 Fourteenth Amendment, 182
Examination, 24 Fourth Amendmen t, 20 1-203
Executives, 11 l, 122 France, 22
Expectations, role of, 26, 32 Frank, Ronald, 77, 79
Fraud,100, 145,200
Experiments, 68-69
detection, 74, 10 I
Expertise, 27, 48
Fraudulent applications, 76
Expert systems, 73
Freedom, 2, 9, 40, 42, 49, 166, 169, 173, 180,
Exploitation, 186-188, 192
185- 186, 188-189, 193,203,205,208
Exposure to information, 164
Freedom of Information Act. 127
Externality, I 8, 206
Free speech, 107
· Front-end verification, 71, 76
Factor analysis, 78, 83, l 7S
Functionalism, 25, 37
Fair Credit Reporting Act (FCRA), 99, I 01,
Futurists, 51- 52
107
Fair Financial information Practices Act, 142
Galbraith, John K., 56
Fair information practices, 204, 223
Gallahue, Eugene, 82
Fairness, 149 Gamble, Andrew, 49
False light, 185, 188 Gannett Telemarketing, 9 l
FBI. See Federal Bureau of Investigation Gardner, Susan, 198, 208
FCC. See Federal Communications Gateways, 94
Comniission Gender, 152, 160, 168-170
FCRA. See Fair Credit Reporting Act General Motors Corporation, 42
Federal Bureau of Investigation (FBI), 56, Genetic
106, 128, 133, 224 analysis, 132
Federal Communications Comffiission characteristics, 76
(FCC), 103, 219, 222 manipulation, 6
Federal Register, 196 screening, 77, 124
Federal Reserve Board, 20 I Geodemographic
Files,54-57,66,80,98, 110, 120, 127, 133, analysis, 55
139, 141, 143-144, 192,195,223-224 clustering, 87
analysis of, 71 Georgetown University, 99
creation of, 62 Gerbner, George, 12, 32, 164, 174
rnatching·of, 57 Germany, 58
paper, 56 Giddens, Anthony, 11-13, 25-34, 44-45, 48,
sharing of, 133 123-124, 177
system of, 57 Gintis, Herbert., 19, 227
First Arnenctmen t, 215 Goldman, /anlori, 196, 209
rights under, 193 Gossip, 183-184
 INDEX

Government, 130, 133, 135-136, 175, 197, process of, 15-17, 24

204,228-229 racial, 43
bureaucracy, 48-49, 210 Ideology, 12, 49, 150, J 69
commercial data use, 145 self-identification, 165
data gathering, 56-58, 130, 133 Itnage, 184,186-187
information, 190 Image processing, 72
regulation, 75 Incentives, 197
surveillance, 56 Income, 153
Graham, Jonathan, 187 Individuality, 179, 186, 189, 194, 204-205,
Gramsci, Antonio, 12, 36 227
Great Britain, 198-199 industrialization, 31
Greenberg, Marshall, 77 inequality, 3, 6, 9, 19, 34, 177, 206
Gregory, Derek, 34 Influence, production of, 21
Group membership, 78, 82, 157 Information, 2, 8, 13, 16, 34-35, 39, 41-47,
Groups, 125, 144, 149-150, 156, 175 50-53,55,60,64,66,69-70, 76, 79, 89,
Gutman, Jonathan, 78 100-103, 107, 138, 143, 149, 179, 183-
184, 188, 189--193, 197-198,200,206,
Habermas, )llrgen, 3, 228, 229 213,220
Hair analysis, 132 business, 100-101, 103, 219
Hall, Peter, 49 demand for, 197
Hamilton, Alexander, 197 demographic, 57
exposure to, 164
Harassment, 215
false, 148
Harris, Louis and Associates, 102, 140, 144,
gathering, 46, 55, 128, 132, 138, 196
150, 153
individually identifiable, I, 3, 5, 9, 58, 94,
Harris Data Center, 137
104, 210
Haskell, Thomas, 45, 48
-intensive organizations, 141
Headquarters, corporate, 110
market, 62, 206, 220
Health insurance medical, 199
See Insurance, medical negative, 133-134
Hegemony, 12, 36 policy, 95, 99
Hispanic, 83, 152-153 privacy, 138-139, 189--190, 193, 203
Hixon, Richard, 178-179 processing, 42, 71-72
Hoffman, Alexander, 105, 109 property rights in, 103, 190, 192-193
Holton, Robert, 40 relevance of, 133
Homeowner's insurance, 64 as a resource, 2 I
Homosexuality, 191 sensitivity of, 134, J39, 153, 198, 200
Horwitz. Robert, 197 services, 68, J03, 105
Households, 57, 59, 93, 97, 102, 125 sharing, 110, 134, 138, 141- 142, 145, 149,
income of, 70, 153 170, 191
Hovencarnp, Herbert, 197 strategic, 54
Hughston, Thomas, 216 system, 36, 98
technology, 8, 10, 36, 39, 41, 49, 50-51,
IBM. See International Busi ness Machines, 229
Inc. transaction-generated, 104, 146
Identification workers, 50-51
personal, 29, 33, 35, 43, 65, 78, 80-82, 86, Information Age, 4, 49, 102
89, 93, l06, 199, 211, 213-214, 216-218, lnformed co nsent, 68, I03, 189, 208, 210, 223
223,225,229 Jnglehart; Ronald, L55-156
INDEX 277

Inj ury, 184, 186, 188, 208 Justice, 40-41

Inspection, 22, 63
Institutions, l 9, 140, 175 Kahneman, Daniel, I24
influence of, 25- 27, 37, 38 Kastenmeir, Robert, 210--211
lnstrumentalism, 38, 45 Katz, James, 139-140, 145
Instrumentalists, 20 Katz, Raul, 52
Insurance, 45, 62, 64, 85-86, 107, 131, 133, Katz v United States, 202
147 Kennedy family, 84
companies, 154, 184 Knoke, David, 21, 209, 222
industry, I45. 173 Knowledge, 19, 21, 28, 34, 35, 38, 40-4I, 45-
medical, 132, 147-148, 207 47, 50, 53, 73,82, 124, 174, I 79, 184,
Integrated Services Digital Network (ISDN), 190, 199
212 from experience, 174
Intellectual property, 182- 184, 192- 193, 207 Knowledgeability, 12, 26, 28-29, 36, 52, 123
Intelligence, 25, 41, SJ, 53, 55, 60, 74, 81, 98, Knowledge engineering, 51, 73
139, 143-144, 161, 199, 223 Koch, Tom, 93
Intelligent network, 103 Kondratieff, Nikolai, 49
Kramer, Kevin, 89
Intelligent systems, 51
In teract ions, da.ta generating, 26-29, 33, 34, Krippendoff, Klaus, 53- 54
173,208
Labo~ 19, 26,30, 41-42,50
Interactive media, 42
cont ract, 19-20, 31, 33, 60
Interests, 178, 180, 189, 198, 209, 213, 222,
division of, 41, 46
227
market, 30, 32
1nternalization, 18 I
nonproductive, 5, 50
Internal Revenue Service (IRS), 56, 59, 74-
surveillance of, 46
76, 87, 105- 106, 130, 133, 141, 144, 145
Lamont v Commissioner of Motor Vehicles, 187
International Business Machines, Inc. (IBM),
Las Vegas, 216
94, 104
Laudon, Kenneth, 194
In terpretation, 34, 44
Laumann, Edward, 2 1, 209, 222
Interviews, 125
Law, system of, 8-9, 40, 43, 101, 136, 138,
Intimacy, I82, 186
148, 177-178, I82-183, 191, 197, 209,
Int rusions, 213
2IS,222
Invasiveness, 134, 138, 202
Legitimacy, 35, 47-48, 132, 144, 186, 194
Investigations, 69, 71, I63, I90, 197, 207
Legitimate business purpose, 101, I03, 132,
IRS. See Internal Revenue Service 154, 193, I96-l97, 211
ISDN. See Integrated Services Digital Legitimati.o n, 27, 48, I 75, 196
Network crisis. 49
Lenk, Klaus, 52
Jackson, Andrew, 197 Liability, 76, 85, 107, 197- 198, 208
Jones, T homas, 90 Liberal ideals, 155
Jonscher, Charles, 51 Liberals, 152
Journalism, 183 Liberty, 186, I89-190, 193-194, 205
Journalists, 93, I91 theory, 193, 205-206
Junk mail, 130 Lie detector tests, 132
Jurors, 84. 135 Lievrouw, Leah, 49
Jury d uty, 127 Life cycle, 26, 157, I 75
jury selection, 84 Life insurauce, 64
Jussawalla, Meheroo, 50, 206-207 Lifestyle, 63, 69, 77, 89, 92, 109, I31, 147
 INDEX

Linkage,24,66,69, 199,223- 224 Massey, William, 79

services, 92 Mass media, 131, 166, 175, 184-185
Linowes, David , 108, 122, 195 exposure, 174
List influence, 164
brokers, 141 use, 168
industry, 105 Matching, 74-76, 195-196, 204-205
management, I 08 MCI, Inc., 218
products, 108 Measurement, 7, 43-44, 61, 68, 77, 79, 83,
vendors, 105, 113, 210 123, 125, 198
Lists, 74-76, 91, 93, 105-107, 113, 129, 193- Mechanization, 38
194, 209- 210, 217, 220 Media Lab, The, 72
mailing, 105, 122, 136, 210 Medical
Lobbyists, 209, 222 costs, 147
Lockwood, David, 31 examinations, 76
Loss prevention, 17, 83 hjstory, 63
Lotus Development Corporation, 93 infonnation, 199
Lotus Marketplace, 142, 222 insurance. See Insurance, medical
Lovekin, David, 38 Medical Information Bureau (MIB), 64
Lying, 148 Medicare, 76
Men, 150
McAllister, Peter, 87 Mendes, Meredith, 221
Machlup, Fritz, 50 Meredith Corporation, 94
McManus, Thomas, 67-68, 2J3 Merton, Robert, 6
Magnetic strip technology, 81 Metromail, 106
Mailing lists, 90-91, 146, 210 MIB. See Medical Information Bureau
companies, 142, 153 Microprocessors, 35, 51
Mainstreaming, 32, 174 Miller, Mac, 86
Management, 41-42, 46, 71 Miller, Michael, 222
Maritz Mar keting, 137 Minow, Newton, 84
Market, 19, 31, 42, 48-49, 66, 123, 177, 188, Misrepresentation, 148
190, 194,200,205, 209,227 MIT. See Massachusetts Institute of
failure, 206 Technology
power, 187 Mitgang, Herbert, 56
research, 23, 42, 65, 68-70, 150 Mitsubishi, 73
researchers, 103 Modems, 94
segment, 77, 92 Monitoring, 29, 46, 61, 71, 124, 126, 130
segmentation, 65, 79 reflexive, 28, 43
solution, 206 remote, 47
Marketing, 10, 17, 24, 42, 47, 96-97, 101, Monopoly supply, 207
104, 108-1 09, l ll-113, 129, 131. 135- Moore, Barrington, 179, 224
136, 142, 211, 213 Morris, Norvall, 86
database, 62, 217 Motor vehicle records, 101
personal information, 142 Mukherjee, Roopali, 206
surveys, 64 Muldoon, Katie, 106
Marx, Gary, 56, 208, 219 Mulgan, Geoff, 33
Marx, Karl, 4-5, 8, 25, 29-30, 35, 39, 230 Multidimensional scaling, 78-79
Marxists, 25, 29, 50, 52 Murdock, Loretta, 203
Massachusetts Institute of Technology (MIT) ,
72 Nadel, Mark, 102
INDEX 279

Naming, 83 Patent, 182

Narrowcasting, 90 Pattern recognition, 72
National Bureau of Standards, U.S., 72 Pennsylvania, Commonwealth of, 148
National Consumers League, 147 Pennsylvania Public Utility Commission, 219
National Institute of}ustice, U.S., 85 People for the American Way, 21 1
National Instimte on Drug Abuse, U.S., 61 Performance assessment, 47, 60, 62
National Multiple Sclerosis Society, 90 Personae, 180, 203
National Rifle Association, 90 Personal
Natural sciences, 197 computers, 92, 94
Negative option, 21 0-211, 219-220, 224 experience, 175
New England Telephone, 102 identification, 66, 80-81, 155
News media, 100, 126 Personal identification number (PIN), 81,
Newspapers, 164, J 69 149
New York Public Service Commission, 219 Personal information, 35, 42, 54, 58, 62, 64,
New York State, 187, 212 80,90,95-96,99, 100-103, 111, 122-
Nielsen. See A. C. Nielsen Company 123, 125, 134, 138-140, 143, 147, 152,
Nixon, Richard, 195 161, 163- 164, 178, 183, 187-188, 190-
Noam, Eli, 212-213 194, 198,201,203, 205-206,209,212,
Noelle-Neumann, Elizabeth, 181 223,225
Normalization, 10-11, 17, 24, 42 collection of, 111, 131
Norms, 39, 47-48 market for, 93, 183
profiles, 92
OECD. See Organization for Economic sale of, 153
Cooperation and Development sensitivity of, 92, 139
Office of Management and Budget (OMB), sharing of, 64, 129, 134, 207
U.S., 55, I96 systems, 22 3- 224
Office of Technology Assessment (OTA), Personality, 180, 183, 185-188
U.S., 55, 213 Personalization, 70, 138
OMB. See Office of Management and Budget Personhood, 80, 179, 186, 198-199
On-line data, 93- 94 corporate, 185
Organization for Economic Cooperation and Persons, 193
Development (OECD), 223 fictional, 193, 225
Organizations, 17, 19, 21, 25, 31-33, 43-44, Photographers, 182-183
46,60,64, 76, 80-8 1, 135, 139-140, 164, PIN. See Personal identification number
173, 190, 194, 197, 209, 213, 215, 220 Plesser, Ronald, 211
Orwell, George, 140 Point-of-purchase systems, 56, 66, 96
OTA. See Office of Technology Assessment Policy, 40, 49, 203- 204
Ownership of information, 52 orientation, 155
Political
Pacific Bell, 219 climate, 165
Paine, Thomas, 31 consultants, 210
Panopticism, 9- 11, 21-22, 181,224, 228 cynicism, l 57
Panopticon, 9, 22- 23, 34, BI interest, 166, 169
Panoptic system, 5, 41, 123, 148, 173, 181, involvement, 166, 173
228-229 orientation, 155, 165-166
orientations toward, 166 surveillance, 195
Panoptic technology, 122 Political economy, 19, 43, 197
Parallel processing, 71- 72 Political targeting, 89
Participation, political, 179, 189 Polygraph, 198
280 INDEX

Porat, Marc, 50 threats to.• 163, 170, 196

Posch, Robert, 106-107, 193, 200 torts, 185
Positive op tion, 221 zones, J81, 211
Posner, Richard, 200- 201 Privacy Act, 55, 195-1.96, 204
Postal Service, U.S., 106 Privacy Protection Study Commission, 21 1,
Poulantzas, Nicos, 20 222
Power, 3, 8, 11- 12, 15, 18- 23, 60, 80, 82-83, Privatization, 48, 52, 60
133, 135, 149, 173, 175, 177, 181 , 186, PRIZ.M, 55, 91
190, 195, 197, 203, 205, 209, 229 Probable cause, 202
nature of, 3, 27, 30, 33- 34, 37-38, 43, 45- PRODIGY, 94, 104
47, 52 Product differentiation, 82
technology of, 3, 21 · Productivity, 51, 61, 200
Power/knowledge, 8, lO Professionalization, 41, 49
Powerlessness, 173 Profi les, 63, 71- 74, 79, 101-102, 109, Jl3,
Power structure research, 20 120-121, 126, 129-130, 139, 143-144,
Prediction , 45, 84, 86, 101, 144, 150, 161 146, 180, 186, 195, 212, 217
Preferences, individual, 19, 78, 124, 194 criticism of, t31
Preprocessing Propaganda,36,38-39,53
as control technology, 16, 42-43, 55 Property rights, 183, 205
Prescreening, 71, 75-76, 201 Prosser, William, 185-186, 188
Press, 175, 184, 222, 228 Psychographics, 47
Press coverage, 141, 148 Psychological testing, 144
Preston, Paschal, 49 Public affairs, 165, 168-169
Prevention, 56 Public figures, 128, 186-187
Privacy, 52, 58, 68, 92- 93, 95-97, 102, 106- Public interest, I I , 100, 179, 184, 193-194,
108, 124-127, 132, 135, 137- 138, 140, 197,205
147, 161, 165, 178-179, 182-224,229 Publicity, 184-187, 222
activists, 103, 164, 212 right of, 183, 187
advocates, l 0 l, 204, 222 Public opinion, 37, 61, 64, 102, 123- 125, 137,
claims, 107 143, 150, 156, 160
concerns, 138, 140, 143, 160, 163, 165, 169, social origins of, 149
199 surveys,65, 89, 126, 147
crisis, 99 Public policy, 107, 110, 209
debate, 106 Public relations, 58
definitions of, 127 Public sphere, 2- 3, 37, 182, 227
enhancing technologies, 103 Punishment, 24.
expectations of, 139, 192, 201- 203, 213
fundamen talists, 135 Questionnaires, 132
interests, 102, 114, 179, 181, 185- 186, 198,
202-204, 213- 217 Racial
invasion of, 102- 103, 105- 107, 126, 128, discrimination, 20 I
135-136, 147,175, 188,206,210 identification, 43, 152
issue, 105, 108, 111, 120 Radio Shack, 66
laws, 170, 188, 198, 200 Ratings, 84
loss of, 138, 140, 187, 196, 206- 207 Rationalism, 40-41, 48, 50, 80, 107, 227
policy, 95, 204 Rationalization, 7- 9, 15, 39-42, 47, 49, 51-
preferences, 166 52, 136
rights, 145, 179, 183, 187, 190, 196, 200, Reagan, Ronald, 211
209 Reagan administration, 61, J 39, 144, 195, 205
INDEX 281

Reaganism, 48 R. R. Donnelley, 89
Reasonableness, 199, 20 1-202, 208, 213, 219 Rule, James, 74, 218
Reciprocity, I 49 Rules, 27, 29, 39, 43, 48, 218
Records, 17, 20, 22, 24, 44, 46, 56-58, 68, 73,
75, 142, 180, 184, 187, 191-192, 196, Samarajiva, Rohan, 206, 219
199,217,225 Sample, 54, 65, 69, 78, 124-125, 156
linkage of, 58 bias, 65, 157
medical, 64 probability, 59
public, 57 Scanning, 69
systems, 55 technology, 65-66, 69
Redlinu1g,58,88, 130,200,218 Schement, Jorge, 49
Reflexive monitoring, 28, 43 Schiller, Dan, 5
Refusal rates, 65 Schneider, Edward, 89
Regan, Priscilla, 194 Schumpeter, Joseph, 5, 7, 35
Registration, 22, 24 Schwedelson, Roy, 107-108
Regulation,40,97, 103, 145-146, 170, 177- Scoring system, 92
178, 194,205,207, 224 Screening, 71, 76, 86, 132, 214, 216, 224
Rehnquist, William, 202 Sears, Roebuck and Company, 66, 94, 104
Reichman, Nancy, 85-86 SEC. See Securities and Exchange
Relational database, 96, 129 Commission, U.S.
Relationship, 62, 175, 191, 220, 225 Securities and Exchange Commission, U.S.
Relevance of information, 133 (SEC), 97
Reliability, 54, 69, 71, 81, 198 Security, 109, 113, 122
Remote sensing, 59 Segmentation, 3, 8, 11, 23, 77- 79, 90, 220,
Reproduction, wcial, 11- 12, 26 227
Resistance, 147 Self-definition, 43, 8F-84, 228
Resources, individual, 27, 29, 33-34, 47-48, Self-determination, 194, 208, 213
95, 106, 122 Sensitivity of information, 92, 134, 199
Responsibility, 189, 197 Sentry Insurance, 102
Reverse directories, 215- 216 Service
Rights, individual, 19, 31, 68, 74, 107, 123, relationship, 62
147, 179, 182,184, 190--195,203-204, sector, 32
209,225,227 Services. 62, 75
t~ be let alone, 182, 189 Sexual activity, 199
to know, 190, 192 Sexual identity, 43
regime of, 200, 218 Sexual orientation, 134, 139, 199
Risk, 16, 17, 20, 63, 76, 85-87, 97, 100--102, Shi/bey v Time, Inc., 187
142, 145, 175, 181, 191, 228, 231 SIC. See Standard Industrial Classification
assessment, 23, 85-86, 92, I 97 Simitis, Spires, 42, 189, 201, 223
avoidance of, 17, 85-86, 101 Simmons, Ronald, 77
management, 85 Singapore, 50
Robbin, Jonathan, 88 Skin11er v Railway Labor Executives
Robins, Kevin, i, 8, 10, 23, 41-42, 51 Associatiorr, 202
Roe v Wade, I82 Sloan, Alfred P., 42
Rokeach, Milton, 156, 166, 169, 173 Sloanism, 42
Roper Center for Public Opinion Research, Smart card, 8 1, 87
137, 140 Smith, Adam, 35
Ross, H. Laurence, 62-64 Smithsonian Institution, 93
Routinization, 12, 42 Social classification, 82-83
282 INDEX

Social control, 1, 9, 12, 86 Surveillance, 8-10, 15, 21, 22, 23, 24, 34, 40,
Social costs, 20l 46, 47, s4-ss, 67, 71, 80-81, 126, 138,
Social Democrats, 58 143, 163, 181, 195-196,202,227,230
Socializatio11, 26 disciplinary, 21, 23, 58
Social marketing, 90 government, 49, 57
Social policy, 43 market, 66
Social science, 21, 2:>--24, 43-44, 84 police, 56
Social scientists, 44-45, 59, 124 political, 56-57, 195
Social Security Administration, U.S., 57 Surveys. 58-59, 64, 82, 102, 108, 125, 132,
Social security card, 80 137, 141
Social security number, 74, 76, 80, 106, 127, refusal to participate in, 65
143, 149.204 telephone, 64
Social systems, 25, 33, 37, 45, 49, 177, 189, Systems, 17, 25, 37, 177, 230
197,200 control, 23
Social Taylorism, 41 needs of, 25, 26
Soft.mip technology, 70 social, 25
Software, computer, 36, 72-73, 77 technical, 36
Sorting, 1- 2, 82, 91
Speech, 185, 193-194,215,219 Target
commercial, 107, 193-194 market, 89-90
cor porate, 185, 193 Targeting, 10, 87, 89-90, 220, 227
Spiegel Corporation, 1 political, 89
Stan dard Industrial Classification (SIC), 110, system, 92
116 Tassone, Annette, 139-140, 145
Standardization, 42-43, 46 Taste, individual, 38
Starr, Paul, 57- 58, 82- 83, 94 Taylor, Frederick W., 41, 51
State, 25, 48, 107, 178, 182, 186, 189-197, Taylorism, Social, 41
203,205,223,228-229 Techn ique, 36-39
State bureaucracy, 48, 49 Technological imperative, 36
Statistical modelin g, 10 I, 120 Technology, 2- 8, 17, 24, 3S-39, 42, 83, 123,
Statistical p rediction, 86 160, 164,203-204,214,220,227,229
Statistical services, 94 autonomous, 36-38
Statistics, 8, 10, 55, 59-60, 131 commun ication, 10, 33
government,57- 58 discriminatory, 53, 80, 228
Status, 74, 147 predictive, 84
groups, 31, 40 sophistication in, 109, 116, 121-122
Stewart, James K., 85 Technophobia, 160
Stinchcombe, Arthur, 31- 33, 60 Tehranian, Maj id, 227-228
Strategic planning, 60 Telecommunications, 24, 46-47, 67, 96, 109,
Structuralists, 20 213.2 17
Structuration, l l-13, 25- 27, 32-34, 45, 48, industry, 103
177 networks, 33, 51, 68, 71
Subscriptions, 192- 193, 199 Telemarketing, 18, 33, 47, 96, 102-103, 107,
lists from, 2 J l 109, 112, I 14, 141, 165, 215, 220-221
Su percomputers, 71 Telephone, 102-103, 109, 125, 148, 200, 215
Su perstru cture, 30 answering machine, 217
Supreme Court, U.S., 182, 191, 197, 201, 210, call detail records, 67, 146, 192
222 companies, 141, 163, 213- 214, 219
Surplus value, 5, 41, 50 di rectory, 216
INDEX

numbers, 66, 76, 93, 103, 105- 106, 109, University of Pennsylvania, 75
120, 148. 199, 206, 216 UPC. See Universal Product Code
numbers, unpublished, 105, 216 USA Today, 9 l
number services (800, 900, etc.), 13, 67- 68, Usefulness, 193
91, 97, 103, 164, 212 Utility, 38
solicitation, 130, 221
subscribers, 67 Validity, 54, 69
survey, 125 Value, 84, 183, 189
t ransaction-generated information change, 155-156
(ITGI), 67-68, 94, 164 Values, 19, 33, 40, 124, 156, 166, 177, 185,
Telephone Preference Service, I 04 189,208, 222
Telesphere Conununications, 91 Verification, 57, 76, 196, 231
Television viewing, 164-165, 168-170, 173 Victimization, 17, 154
Thatcher ism, 48 Video and Library Protection Act, 2 10
TIGER mapping technology, 88 Video Pr ivacy Protection Act, 209
Thompson, E. P., 27, 31, 32 Videotape rentals, 192
Time and space relationship, 26 Virtual reality, 34
Toffler, Alvin, 47, 51 Vulnerability, 22 l
Tomkovicz, James, 203
Wacks, Raymond, 92, 139, 198-200
Trade secret, 182, 191
Warner/Amex, 96
Training, 24
Watergate, l 64, 195
Transaction-generated information, l 05, 126,
Webe~t.fax,7-9,30,35,39,40,46--48
129, 132, 146
Webster, Frank, I, 8, 10, 23, 41-42, 51
Transaction processing, 13, 96-97, 213
Weighting, 65
Transactions, 66, 68, 71-73. 80-83. 100, 102-
Welfare, social, 18, 19, 20, 40, 43, 144, 147,
103, 113, 141- 142, 184, 200, 202, 206,
184
2 13, 22 1
recipients, 67
Triage, L, 82, 223
system, 57
Trust, 32, 58, 135, 140, 157, 161, 163, 173-
Westin, Alan, 102, 109, 137-138, 164, 179,
175, 229-230
181, 183,205-206,213
TRW, Inc., 92-93, 95, 98-99, 102, 105-106,
Whistle blowers, 67
108, 132-133, 21 l
Wilkinson, John, 37
TTGI. See Telephone, transaction-generated Wilson, William }., 6
information Win d, Yoram, 79
Tureen v Equifax, 197 Wiretapping, 148
Turner, Bryan, 40 Withdrawal, 147
Wizard ofOz, The, 231
Uncertainty, 45 Women, 150, 152
Understanding, 29, 32, 36 Workers, 31-32, 39, 42,46--47,60, 73, 76, 144
Uneeda Biscuit, 82 compensation claims, 76
Unemployed, 25 Workforce, 61
Unemployment, 20, 22 Workplace democracy, 227
Unions, 31, 32 Wright, Erik Olin, 30-31
Un°ited States, 49, 50, 51, 56, 58-59, 88, 92,
107, 139, 155, 182, 195, 198, 203, 223, Yellow pages, I 05
224,229
United States v Miller, 191- 192, 222 Zip code, 87-88, 92- 93, 107, 1 13, 120
Universal Product Code (UPC), 69 Zone of intimate decision, 182
Praise tor The Panoptic Sort

·'With clarity, grace, and erudition Oscar H.Gandy critically analyzes the ever-growing information
technology octopus. Comprehensive, integrative, and provocative, the book is must reading for
those seeking to understand the multiple challenges of contemporary panoptic systems.''

- Gary Marx
University of Colorado- Boulder

''This book will take its place on a shelf of the seminal works of the decade and alongside
1984 and The Power Elite. Its theme is that access to and instrumental use of data about our
lives and activities confer power of unprecedented scope and magnitude. Information technology
and data banks exacerbate the inequalities that are tearing us apart. Ranging from issues of
privacy and marketing to survey research and public policy, technically informed for the infor-
mation specialist and economist and accessible to readers from the social and political sciences
and the humanities, Gandy's work is required reading for scholars and researchers who wish
to gain acritical-and sobering-understanding of the newly emerging dynamics of the Infor-
mation Age.' '
- George Gerbner
The Annenberg School for Communication
University of Pennsylvania

" ...thoughtful and thought-provoking. Gandy combines incisive analysis of social theory with
innovative empirical studies. The result is adeeply troubling portrait of our endangered privacy.
There is no better book on the subject.''
-James E. Katz
Bellcore

For order and other information, please write to:

Westvlew Press
5500 Central Avenue • Boulder,
36 Lonsdale Road • SummertOI

CoVlf design: Pally Cbrlsle1111n