Spatial inequalities and regional

development

HENDRIK FOLMER and JAN OOSTERHAYEN, editors

University of Groningen/Holland

SPRINGER-SCIENCE+BUSINESS MEDIA, B. V.
Library of Congress Cataloging in PubHcadon Data
Regional Science Symposium, University of Groningen, 1977.
Spatial inequalities and regional development.
"Held at the Faculty of Economics of the University of
Groningen."
Includes index.
1. Regional planning-Congresses. 2. Regional economics-
Congresses. I. Folmer, Hendrik. II. Oosterhaven,
Jan. III. Groningen. Rijksuniversiteit. Faculteit der
Economische Wettenschappen. IV. Title.
HT391.R3S4 1977 309.2'S 79-13176
ISBN 978-90-481-5806-5 ISBN 978-94-017-3046-4 (eBook)
DOI 10.1007/978-94-017-3046-4

Copyright ©1979 by Springer Science+Business Media Dordrecht

Originally published by Martinus Nijhoff Publishing in 1979

No part of this book may be reproduced in any form by print,

photoprint, microfilm or any other means, without written
permission from the publisher.
List of contributors

Paul Drewe is Professor of Physical Planning Research at the Department of Architec-

ture at the Technical University of Delft.

Ulf Erlandson is Lecturer of Economic Geography at the Department of Geography at

the University of Lund.

Hendrik Folmer is Lecturer of Statistics and Research Methods at the Department of

Geography at the University of Groningen.

John Friedmann is Director of Urban Planning at the University of California at Los

Angeles.

Wim Hafkamp is Lecturer of Mathematics at the Department of Economics at the

University of Amsterdam.

Hans Heijke is Head of the Department of Labour Market Research of the Netherlands
Economic Institute at Rotterdam.

Stuart Holland lectures in Economics at the State University of Sussex.

Leo Klaassen is President Director of the Netherlands Economic Institute and Professor
of Regional and Social-Economic Research at the University of Rotterdam.

Peter Nijkamp is Professor of Regional and Environmental Economics at the Free

University of Amsterdam.

Jan Oosterhaven is Lecturer of Regional Economics at the Department of Economics

at the University of Groningen.

Harry W. Richardson is Professor of Economics and Urban and Regional Planning at

the School of Urban and Regional Planning of the University of Southern California.
VI LIST OF CONTRIBUTORS

Walter Stohr is Director of the Interdisciplinary Institute for Urban and Regional
Studies and Professor of Regional Policy at the University of Economics of Vienna.

Franz TOdtling is Lecturer of Regional Economics and Regional Policy at the Univer-
sity of Economics of Vienna.
Preface

In September 1977 a 'Regional Science Symposium' was held at the

Faculty of Economics of the University of Groningen in the Netherlands.
Organized because of the recent establishment at the Faculty of
Economics of a group that is engaged in teaching and research in the field
of regional science, the aim of the symposium was to make university
members more familiar with regional science and to introduce the newly-
created group to the national and international scene.
Two separate topics were selected, of potential interest to both re-
searchers and policy-makers.
The first, spatial inequalities and regional development, was chosen
because of its central place in regional science. Authors from several
disciplines were asked to approach this theme from a general, policy-
orientated point of view. This ensured the enlightenment of the various
dimensions of spatial inequality and its implications for regional policy.
The results have been collected in the volume Spatial Inequalities and
Regional Development.
The second theme focused on spatial statistical analysis. This branch
of statistics is a relatively new one which receives growing attention
among researchers in the field of applied regional science. The meeting
on this topic concentrated on new results of research on the use of appro-
priate statistical and econometric methods for analyzing spatial data.
The papers concerned have been collected into another volume, Explora-
tory and Explanatory Statistical Analysis of Spatial Data.
Both volumes contain mainly papers presented at the symposium. Some
additional papers have been included to improve the consistency of the
volumes. All contributions have been revised before final publication.
In this process critical remarks made by invited discussants at the sympos-
ium proved to be very helpful. We believe that these efforts have helped
considerably to improve the quality of both volumes.

Groningen, September 1977 The editors

Contents

LIST OF CONTRIBUTORS v
PREFACE VII

INTRODUCTION XIII

1. SPATIAL INEQUALITIES AND REGIONAL

DEVELOPMENT: A FRAMEWORK
HENDRIK FOLMER AND JAN OOSTERHAVEN

1.1 A comprehensive model and its measureableness 2

1.2 Four systems 4
1.3 Policy implications for regional development 15

Part I. Analytical studies in spatial inequalities and regional development

2. ON THE CONTRADICTIONS BETWEEN CITY AND

COUNTRYSIDE 23
JOHN FRIEDMANN

2.1 Contradictions: cosmic and historical 24

2.2 Functional and territorial bases of social transformation 28
2.3 Concepts of city and countryside 31
2.4 The pseudo-resolution of contradictions in the core countries of
capitalism 35
2.5 The globalization of the conflict 39

3. ENVIRONMENTAL PROTECTION AND SPATIAL

WELF ARE PATTERNS 47
WIM HAFKAMP AND PETER NIJKAMP
X CONTENTS

3.1 Spatial welfare profiles 50

3.2 Environmental externalities: a foonal framework 55
3.3 Redistribution effects of a tax-subsidy scheme 65
3.4 A multiregional equilibrium model for environmental quality 71
3.5 Conclusion 81
A. Extensions of the environmental externalities model 82
B. Conditions for a generalized Pareto optimum 89

4. CONTACT POTENTIALS IN THE EUROPEAN SYSTEM

OF CITIES 93
ULF ERLANDSSON

4.1 Job functions and contact activity 94

4.2 Contact systems and regional development in Sweden 95
4.3 Accessibility in the European system of cities 100
4.4 Personal contact networks and regional development 112

5. HUMAN REACTIONS TO SPATIAL DIVERSITY:

MOBILITY IN REGIONAL LABOUR MARKETS 117
J .A.M. HEYKE AND L.H. KLAASSEN

5.1 Migration and mobility in four European countries 119

5.2 Migration and mobility of foreign workers 122
5.3 Psychological distances 126
5.4 Final remarks 127

Part II. Policy-oriented studies in spatial inequalities and regional development

6. SPATIAL EQUITY: SOME ANTI-THESES TO CURRENT

REGIONAL DEVELOPMENT DOCTRINE 133
WALTER STOHR AND FRANZ TODTLING

6.1 Spatial equity and regional policy 133

6.2 Conceptual bases of current regional development practice 137
6.3 Some empirical evidence of recent spatial development trends:
the material effects of regional development policy 146
6.4 Selective spatial closure as a strategy for increasing the resilience
of spatial systems? 152
CONTENTS XI

7. AGGREGATE EFFICIENCY AND INTERREGIONAL

EQUITY 161
HARRY W. RICHARDSON

7.1 Definitional problems and related conceptual issues 162

7.2 Theoretical considerations 169
7.3 The Williamson curve 173
7.4 Other efficiency-equity compatibilities 176
7.5 Equity impacts of regional policies 177
7.6 Policy implications 180

8. CAPITAL, LABOUR AND THE REGIONS 185

STUART HOLLAND

8.1 The fallacy of self-balance theory and current regional policy 186
8.2 EXisting power structure and regional policy 201
8.3 Socialisation and change 210

9. INTEGRATED REGIONAL PLANNING: AS APPLIED

TO THE NORTHERN NETHERLANDS 219
PAULDREWE

9.1 Background information on ISP 220

9.2 A strategic choice framework 223
9.3 ISP: its making and its products 230
9.4 ISP, spatial inequalities and regional development 247

SUBJECT INDEX 255

Introduction

Spatial inequalities and regional development have been problems of out-

standing importance in regional science. It is not far from the truth to
state that differences between regions have been amongst the major causes
that have contributed to the originating and development of scientific
thought about spatial regularities and related problems.
In spite of their importance spatial inequalities and regional develop-
ment have been dealt with somewhat fragmentarily in regional science.
Studies involving just one type of spatial inequality with only a limited
number of relations between different aspects are dominant. Most theories
and models deal with, e.g., housing conditions, income distribution or
environmental quality as isolated subjects. When different types of
inequality are taken into consideration, one type is mostly considered as
exogenous and the other one as endogenous. In this way regional science
is reduced to the mere sum of traditional disciplines like (regional)
economics, geography and sociology.
However, the dynamic interdependence of the different types of
inequality ought to be the central subject in a study of spatial inequalities
and regional development as regional science is, e.g., defined l as 'the
synthetised (integrated) analysis of the political, economic, sociological,
cultural and psychological factors affecting the development of a meaning-
ful region or system or regions'.
Although fragmentary approaches are quite natural at the initial stage
of a new discipline, and although many important insights have been
gained in this way, there are severe limitations to these approaches.
In the first place important aspects of regional development have been
more or less neglected. Problems like the disruption of traditional, social
and cultural structures or the diminishing possibilities of certain types of
self-determination due to economic integration and development, have
1. W. Isard, Introduction to Regional Science. Prentice-Hall, Englewood Cliffs, 1975,
p.5.
XIV INTRODUCTION

mostly been ignored.

Second, although it is defensible from a scientific point of view to
concentrate on one type of spatial inequality only and to abstract from
other relevant types, this is not at all responsible from a policy point of
view. In real life, decision makers are confronted with the total com-
plexity of a (system of) region(s). Plan formulation and plan execution in
one field neglecting interdependencies with other fields most probably
leads to serious new problems, as has been illustrated by many examples
in real life experience with regional planning. Inasfar as regional scientists
claim to contribute to the practice of regional planning, their discipline
has to provide a framework that exposes all relevant consequences of
policies directed to, e.g., diminishing undesired spatial inequalities.
It is the intention in this volume to deal with several types of spatial
inequalities and regional development policies, thereby paying attention
to the two limitations mentioned above. The contributions are grouped in
two broad categories which, naturally, partly overlap. Part I contains
studies that are mainly directed to the analysis of spatial inequalities.
Part II contains studies that are mainly aimed at the evaluation and
formulation of regional development policies.
Chapter I describes the outlines of a comprehensive model of spatial
inequalities and regional development policy, thus providing a general
framework for the other contributions in this volume.
In Chapter 2 a typology of contradictions is presented. The various
types of contradictions, and especially the eternal one between func-
tional and territorial bases of social transformation, may be seen as the
moving force(s) behind spatial inequalities and regional development.
Friedmann applies his conceptual framework to the contradictions
between town and countryside at various levels of spatial aggregation
along the path of time.
In Chapter 3 a welfare-theoretical analysis of environmental spatial
externalities in a multiregional economic system of polluting and non-
polluting regions is made. Ha/kamp and Nijkamp analyse how systems
of charges, subsidies and pollution standards may produce several types
of interregional welfare equilibria, fulfilling, e.g., Pareto optimality
conditions. In addition, they illustrate the potential applicability of their
framework with some empirical experiments.
INTRODUCTION xv
In Chapter 4, Erlandsson derives accessibility measures for the 100
largest European cities from the structure of the European transport
network. Next contact potentials are derived from a combination of the
accessibility measures and the distribution of population over the cities
studied. Both results are analysed in the context of the strategic impor-
tance of face-to-face contacts for performing most decision-making and
important research and development activities.
Chapter 5 contains the results of two migration studies. Heijke and
Klaassen explain migration flows from mobility, i.e. the propensity to
move and the size of the migration impulses. The mobility structure of
different groups of migrants is found to be influenced by population
density and average distance within countries and by language kinships,
i.e. psychological distance between countries.
Close along the conceptual lines laid out by Friedmann, Chapter 6
devotes special attention to an aspect of spatial inequalities that has
received little attention thus far: the sociocultural dimension of regional
welfare. Stohr and Todtling describe current regional policies as actually
reinforcing the process of functional integration, the cause of at least
part of the present regional problems by means of its externalities forced
upon smaller territorial units. They end their analysis with several pro-
posals for 'selective spatial closures' as a strategy for increasing the
resilience of regions.
Chapter 7 critically reviews various traditional regional economic
concepts and theories on the compatibility or trade-off between aggregate
efficiency and interregional equity. Against this background Richardson
evaluates the socioeconomic performance of existing regional policies, and
concludes that - depending on the scope of the defmitions of equity and
efficiency and on the concrete case at hand - a policy which reconciles
both objectives is tractable.
In Chapter 8 Holland analyses the economic and political power struc-
ture as a major hindering factor in solving the regional problem at a
national level. Substantial differences in labour costs between developed
countries and Third World countries and diminishing transport costs form
a major cause of the stagnation of industrial growth in developed
countries. These differences together with monopoly market structure and
globally operating multinational corporations are also analysed as the
cause which prevents current indirectly-operating regional policies to work
in the way traditional theory falsely expects them to work. Holland ends
his analysis of the economic and political power structure with a plea for
XVI INTRODUCTION

democratised selective socialisation of both production and consumption.

The last chapter evaluates the ambitious project of integrated multi-
dimensional planning for the northern part of the Netherlands with the
strategic choice framework as checklist. Drewe reports major failures in
the 'planning of the Planning' and a lack of studies in the effectiveness
of existing regional policy instruments. The result is that the ambitious
project becomes more a part of the problems for this region than a start
to their solution. From this experience guidelines are drawn for more
decentralised planning starting in the regions concerned and for more
relevant educational programmes in the field of regional science.

Hendrik Folmer and Jan Oosterhaven

1. Spatial inequalities and regional
development: A framework
HENDRIK FOLMER and JAN OOSTERHAVEN*

The concepts of spatial inequality and regional development have been

widely dealt with in regional science and in related base disciplines like
(regional) economics, sociology and political science. In spite of this,
spatial inequality is an ill-defined term in the literature, as is shown in an
economic context by Richardson in Chapter 7. Often it is used as a
synonym for inequity. In this chapter inequality is defined as the differ-
ence in spatially-defined variables as, e.g., population density and income
per capita. This implies that spatial inequality may both refer to a desired
and to an undesired difference. For instance, variety in landscape is
generally regarded as desirable, whereas large spatial differences in living
conditions are mostly unwanted. Spatial inequality associated with ideo-
logical commitment, describing an undesired difference, will be dermed
as spatial inequity.
As regional science, being one of the social sciences, is concerned with
the overall complexity of spatially-defined social systems, an efficient
framework for measuring and evaluating spatial inequalities is to compare
the profiles (see Section 3.1) of the various subsystems which constitute
the (system of) regional social system(s). Although a much more refined
classification is possible, it will suffice for our purposes to distinguish the
following regional subsystems:
- demographic subsystem
- sociocultural subsystem
- economic subsystem (including communication networks and techno-
logy)
- ecological subsystem
Regional development can now be dermed as the dynamic aspect of
a regional system, or, more precisely, as the changes that take place
in the states of the regional system as time progresses. In this respect
it is important to note that although changes in a regional system proceed

* University of Groningen, the Netherlands.

2 A FRAMEWORK

continuously over time, the measurement of these changes takes place

predominantly at discrete time intervals.
In this chapter a general framework will be outlined, both for the
analysis of spatial inequalities and for the formulation of policies for
regional development.

1.1 A COMPREHENSIVE MODEL AND ITS MEASUREABLENESS

The state of the combined regional subsystems is influenced by autono-

mous developments and by interventions, i.e. governmental actions in the
subsystem under study, both in the region under consideration and in all
other relevant regions. In addition, lagged autonomous developments and
interventions, both in the region under consideration and in other regions,
co-determine the state of the regional subsystem.
If it is assumed that there are n relevant regions and m + 1 relevant time
periods then the above statements can be formulated mathematically as
follows: 1
m n m
Si(t) =~ AikS/t-k) + ~ ~ Bjks/t-k) +
k=0 j=1 k = 0
t:f=i
m n m
~ lik s/t - k) + ~ ~ ~jk s/t - k) + e·
-I
0.1)
k =0 j= 1 k = 0
j=l=i
where:
§/t) : state vector of regional system i at time t;
Aik : influence matrix of autonomous developments in region i, which
expresses the influence of the regional state at time t - k on the
state at time t;
B· k
-I
influence matrix of autonomous developments in regionj =1= i dur-
ing period t - k to t on the state of region i at time t;
influence matrix of interventions in region i, which expresses the
influence of interventions during period t - k to t on the state at
time t;

1. It is assumed that the elements of.~i(t) can be of a nominal, ordinal as well as of an

interval or ratio-level, whereas the elements of the matrices A, B, C and I can be
both relations with the relevant elements of .§i(t) as arguments or coefficients, so
that the model is stated in its most general form.
A COMPREHENSIVE MODEL AND ITS MEASUREABLENESS 3

influence matrix of interventions in regions;

vector of exogenous variables, e.g. the geographical situation of a
region.

When a state vector for every regional system in the set of regions under
consideration has been established, the state matrix W of the multi-
regional system (see Section 3.1) has been determined and the framework
for determining spatial inequalities has originated. Analysis of spatial
inequalities now amounts in essence to comparing the state vectors of
the various regions under study. By comparing the state vectors of the
various regions with an 'ideal' state vector the spatial inequity pattern of
the various regions under study comes into existence, which is of crucial
importance from a policy point of view.
In regional science as well as in other social sciences serious problems
have prevailed with respect to the measurableness of theoretical concepts,
like 'the state of the economic or sociocultural subsystem of a region',
because of their non-observable and multi-dimensional character. Mostly
a single measurable indicator was selected (e.g. income per capita for the
socioeconomic situation of a region, see Reiner 1973), or an index was
constructed (income, investments in schools, hospitals, etc., see Richardson,
Section 7.1) to represent the theoretical construct. More adequate models
have been developed, especially in psychometrics, which also present
interesting possibilities for regional science.
Briefly summarized, these models go as follows. Let a hypothesis
between two sets of variables be derived from theory. None, one, or both
subsets of variables are theoretical constructs or, as they are called in psycho-
logical literature, latent variables. If the vector of endogenous latent
variables is denoted by 11 and the vector of exogenous latent variables by
~, then, under certain restrictions, which often can be weakened, (Le.
linearity, measurableness at interval or ratio level, etc.), the hypothesis
can be formulated as:

B'!] = !'~ + r (1.2)

where:
!J : matrix of relationships between the endogenous latent variables;
r : matrix of relationships between the latent exogenous and endo-
genous variables;
r: vector of random disturbances.
4 A FRAMEWORK

Now either from theory or in an empirical way, observable indicators

are selected which represent the latent variables. If we denote the observ-
able endogenous variables by the vector 1 and the observable exogenous
variables by;! then we have the following measurement equations:

f = ~y'!l+!l (1.3)

4 = Ax~+§ (1.4)

where:
A: matrix of relationships between the latent variables and their
indicators;
§ and § : vectors of random measurement errors.

For more information about these models, and especially about the
estimating and testing procedures, see Goldberger and Duncan (1974) and
Joreskog (1976).
The measurement problem with these kinds of linear structural equa-
tion models is no longer how to reduce a theoretical construct to a single
indicator or a single index but, instead, how to fmd as complete a set as
possible of valid observable indicators - which certainly does better
justice to the complexity of the theoretical construct than the use of a
single indicator or index.

1.2 FOUR SUBSYSTEMS

In this section the subsystems mentioned above will be reviewed in such

a way that a framework for the analytical part of this volume comes into
being.

1.2.1 The demographic subsystem

The demographic subsystem consists of the human population regarding
its size, structure and distribution in space as well as the changes which
take place in these aspects.
The demographic subsystem is in many ways an influence on the well-
being of the people living in a region. Probably the problems of urban
regions with a very high population density (pollution, congestion, short-
age of housing, etc.) are the most familiar influences. Other features of
FOUR SUBSYSTEMS 5

the demographic subsystem which are of importance for regional well-

being, both directly, and indirectly as co-determinants of the other
regional subsystems, are the age structure of the population and migration.
In this volume the migration aspect of the demographic subsystem is
paid attention to. In order to clarify the importance of migration for
regional population development a model developed by Wilson (1974)
will be taken here as a starting point. If we assume that there are N rele-
vant regions then the model reads as follows:

gl (t,t+1) M12 (t,t+1) M13 (t,t+1)

M21 (t,t+1) M2 (t,t+1) M23 (t,t+1)
(2.1)

p!'f:t+1)

where:
pi(t) the age-specific population vector of region i;
(i(t,t+T) li(t,t+T) +Si(t,t+T) ;
Mii(t,t+T) matrix of gross immigration rates for region i;
with:

~
o. b~ (t,t+1) , b~+l (t,t+1) ...... b~(t,t+1). ...... o~
1/(t,t+1) 0... 0
0... ... 0

birth rate in age group r in region i from t to t+T ;

o o 0 ... o
S~I(t,t+1) 0 0 o
o sh(t,t+1) 0 o

o 0 ... SkR-l (t,t+1) 0

with: S~_l(t,t+1) rate of transition of age group r-l to age group r in

region i (composed of death rate and emigration rate);
6 A FRAMEWORK

o 0... 0 ... o
mYt (t,t+T) 0... 0 o
o m#k(t,t+T) 0 o , with:

o 0 ...

rate of gross immigration from population in age

group r-l in regionj at time t to age group r in
region i at time t+ T.
Thus far the cohorts have been structured by age groups only. Although
omitted here, it is a straightforward task to subdivide the population by,
e.g., sex (see Rogers and McDougall 1968).
A serious problem in regional population development is the inaccuracy
of the estimation of migration. There are so many factors that influence
migration that it is extremely difficult to construct a model that encom-
passes all the factors involved. A rather successful model for describing
and predicting migration is the entropy maximising model as developed
by Wilson (1974). This slightly modified model (omitting reference to the
time period t to t+ T) reads as follows:

(2.2)

where:
M/i gross migration from region i to region j of age group r;

o .r /
= /I
1"""/
.r pi.
r '
where:
/1/ = g,/ ~{ + €i
O.r total emigration from age group r in region i;
/1/
/
propensity of age group r to emigrate from region i;
p.r
/
population in region i in age group r;
f!:r vector of regression coefficients;
~./ vector of variables relevant for age group r which characterize
region i (push factors);
ti disturbance term;
TiI.1" = _r ".r + [).I '.
b TLI
FOUR SUBSYSTEMS 7

where:
12r vector of regression coefficients;
v~
-I
vector of variables relevant for age group r which characterize
region j (pull factor);
disturbance term;

A( = [~ T{ ['(dij)r 1 ,
I
where:
[T(dij ) : distance function between region i and j relevant for age group r.

It is now possible to write m~_ 1 (t, t+ T) from (2.2) as follows:

m ij (t t+ T) = M~.- 1 (t t+ T)IP jet) (2.3)

"- 1 ' II' r
Migration plays the central role in the contribution of Heijke and Klaassen
to this volume. In their national studies (Section 5.1) the number of
people between 15 and 65 years of age is taken as the push factor and the
number of jobs in industry and services as the pull factor. In their inter-
national studies (Section 5.2) wage differences between emigration and
immigration countries were added as an explanatory variable. The simple
specifications chosen are due to a great extent to the notorious lack of
data. From the residuals per country group used in the international study,
a systematic factor relating to psychological distance (language kinship)
was detected (Section 5.3). Hence, Chapter 5 shows some important links
from the economic and sociocultural subsystems to the demographic
sUbsystem. 2
At an analytical level important feedbacks from the demographic sub-
system to the economic subsystem are discussed. Both Richardson
(Section 7.2) and Holland (Section 8.1) have severe doubts about the
working of migration of labour towards an equalization of regional
incomes per capita and of regional unemployment levels. Drewe (Section
9.4) reports on more general doubts about the effects of a population
redistribution policy on a broad spectrum of objectives.

2. More information about migration, especially in relation to economic conditions,

can be found in Wilson (1970), Andersson (1977), Karlquist and Suickars (I 977),
Andersson and Plant (1977), Gustavsson Harsman and Suickars (1977); concerning
international migration see Wadensjo (1977).
8 A FRAMEWORK

1.2.2 The sociocultural subsystem

The sociocultural subsystem of a region is, strictly speaking, identical to
the overall regional system which is the concern of this chapter. This fact
causes problems in defining the various subsystems because one of the
fundamental prerequisites of defining differences, namely exclusiveness,
will be violated. However, for reasons of analysis the abovementioned
four subsystems have been distinguished. With this in mind we will define
the sociocultural SUbsystem as man in his behaviour of self-realisation.
One aspect of human self-realisation is production of goods which are
scarce, but this aspect will be excluded from the sociocultural subsystem.
We will define it as the constitutive element of the economic subsystem.
In order to get an understanding of what is meant by self-realisation we
refer to the model developed by Maslov for individual self-realisation.
Maslov conceives of the following five levels of needs, arranged in an order
from 'lower needs' to 'higher needs':
- physiological needs (hunger, thirst, shelter, etc.);
- safety needs (security, order, stability, etc.);
- belongingness and love needs (affection, identification, etc.);
- esteem needs (prestige, success, self-respect, etc.);
- need for self-actualisation (independency, creativeness, etc.).
According to Maslov a 'lower' need must be adequately satisfied before
the next 'higher' need can fully emerge in a person's development (see
Krech and Crutchfield 1965). The degree to which these needs are satis-
fied is generally denoted by the term 'welfare'.
An important aspect of welfare is the distinction between individual
welfare and collective welfare. In order to understand the concept of
collective welfare, it has to be noted that individual welfare is in essence
socially determined, partly by the social activities of the individual and
partly by factors beyond his control. In the same sense collective welfare
is not the aggregate of individual welfare, but the welfare of a group, the
members of which are related to each other. Collective welfare, then, is
the degree to which the group is able to realise its goals or, stated in
another way, the measure by which the group is able to satisfy its needs.
This process of self-realisation of groups is also dependent on the societal
conditions the group is enmeshed in. The goals of groups differ at different
levels, i.e. small primary groups, like the family, and larger groups like
territorial or functional groups.
With regard to the welfare of individuals and groups there does not
exist a theory as well-developed as, inter alia, that for the field of popula-
FOUR SUBSYSTEMS 9

tion development. Neither have many comprehensive empirical studies

been conducted. An important contribution both to theory-building and
measurement in the field of individual welfare is the study by Norlen
(1977), who applies a linear structural relation model as described in
(1.2)-( 1.4) to a data set obtained from a survey of Swedish Level of Living
Conditions conducted in 1968. In this survey living conditions for the
adult Swedish population between the ages of 15 and 75 years were
studied on the basis of a sample of some 6,000 individuals.
Norlen does not start from the Maslov model, but instead uses the
areas health, work, economy and consumption in the widest possible sense
of the word as a framework for a set of Level of Living components. The
components used by Norh~n are an elaboration of the following com-
ponents used in the 1968 Swedish Level of Living Conditions survey:
- health and medical care;
- employment and working conditions;
- schooling;
- economic resources;
- childhood and family conditions;
- housing conditions;
- nutrition;
- leisure time and recreation.
In the sense of equation (1.1) the abovementioned areas of health, work,
economy and consumption can be viewed as parts of the vector !1 and the
Level of Living Components as parts of the vector ~.
In this volume, Stohr and Todtling treat the above-defined problem of
welfare of individuals and small groups in relation to economic develop-
ment programmes. They document a growing dissatisfaction in regional
communities about the increasing externalities of large-scale functional,
economic and political developments. These developments include current
regional policy; they threaten the right of individuals and small groups to
determine their immediate natural and human environment and diminish
the possibilities to exert adequate control over those external influences
which affect their way of living.
The threatening of human well-being, as analysed by Stohr and
Todtling, is in essence also the central problem Friedmann puts forward,
although in the different context of the contradiction between conditions
in town and countryside. From Friedmann's analysis it can be derived that
conditions in modem metropolitan areas are unfavourable for man in
satisfying his higher needs of belongingness, esteem and self-realisation.
10 A FRAMEWORK

Even the lower needs of health and shelter are threatened. All three
authors refer to the same causes: fun·ctional specialisation and integration.
In Stohr and Todtling's view, this means a one-sided attention in
regional development for economic efficiency. As a remedy for this
problem they suggest, inter alia, that economic objectives have to be
integrated in a broader set of policy goals which do justice to the total
complexity of human well-being. To Friedmann, functional integration
means that man and his primary social structures based on territorial
integration have come under the domination of growing anonymous
corporate and bureaucratic power. In Friedmann's definitions functional
integration implies territorial integration and the converse, so that the
total destruction of neither is possible. Nevertheless there exists at the
moment a serious lack of balance. In order to restore the natural situation
of equilibrium a recovery and strengthening of territorial values and
organisations seems necessary.
Of the many relations of the sociocultural subsystem to the other sub-
systems, we additionally want to mention the changing attitudes towards
modern production and consumption methods, which may induce changes
in production technology and environmental protection. Of the influences
the sociocultural subsystem exerts on the demographic subsystems most
known are the attitudes towards birth rate and family structure, but
migration also is influenced, as noted by Heijke and Klaassen.

1.2.3 The economic subsystem

In this section we aim at a broad review of the factors that affect a
region's economic activity and how externalities of economic activities
can be incorporated in the economic subsystem. As to the economic
activity of a region, we begin with the income-generating factors, which
can be divided into five categories. For each of these categories the factors
that influence location will be globally inventorized.
The first income-generating factor refers to social security, transfer
payments, pension plans and property incomes. This complex category
mainly includes derived forms of income which are not directly depen-
dent on the location of economic activities but on the persons who receive
income in these categories. As most have relative freedom of movement,
their living places will be primarily determined by conditions favourable for
their well-being: climate, environmental quality, services of all kinds, etc.
However, both preferences as well as possibilities to act in accordance
with preferences will differ for, inter alia, different income, age, and
FOUR SUBSYSTEMS 11

educational categories and the precise kind of income type.

The second 'major' income-generating factor is the public sector. The
distribution of governmental spending is partly determined by historical
and institutional variables. Governmental spending has become an impor-
tant instrument in regional development, and this aspect of it will be dealt
with in Section 1.3, referring to policy implications.
A third primary income-generating factor is agriculture. Factors which
determine the distribution of agriculture are primarily the necessary
natural preconditions such as soil quality, climate, etc. Secondly, even
if the natural conditions are favourable, agriculture has to compete with
other economic activities like industry, traffic, etc.
A fourth, and much more complex, primary income-generating factor
is industrial activity. Lots of theories and models have been developed in
this field, ranging from partial location analysis and general location
theory to game theory. However, most of these theories suffer from too
many abstractions from reality, e.g. assumptions of homogeneous space
and rational behaviour (see Paelinck and Nijkamp 1975).
In order to determine which factors influence the regional differences
in industrial location, some understanding of the behaviour of business
firms is needed. We assume that business firms behave in such a way as to
promote their continuity and growth. This purpose is, inter alia, achieved
by maximizing their profits. In a heterogeneous space the maximum profit
will vary by location because both revenue and cost function vary by
location. Besides, revenue and cost functions are uncertain and depend on
time, because of technical and socioeconomic changes (Smith 1971).
Because of spatial cost and revenue differences incentives come into
existence for firms to change location (Harris and Hopkins 1972).
In the highly-developed Western countries real wage costs have increased
substantially, while communication costs have decreased considerably.
This has led relatively labour intensive industries to relocate to developing
countries like Taiwan, Singapore, etc., where labour costs are only a frac-
tion of the cost in Europe and North America. This process of relocation
has qontributed to what is presently viewed as one of the most serious
problems in already developed industrial countries: growing unemploy-
ment. Another major change is the emergence of multinational corpora-
tions and monopoloid market structures for many (industrial) products.
This combination of changing factors forms the central theme of
Holland's analysis. He argues the inadequacy of all indirectly operating
regional policy at the national level. The scale of multinationals is so
12 A FRAMEWORK

large that they can internalise most classical economies of localisation and
agglomeration at almost any place, thereby marginalising the smaller
enterprises in the less-developed regions of more developed countries.
On the other hand these same factors contribute to the expansion of
industrial growth poles in some tax and labour union havens in Third
World countries. However, this expansion is reached at the costs of
growing intra- and interregional inequalities (see Sections 6.3 and 7.5),
particularly with regard to non-income components of welfare as Richard-
son argues in Chapter 7.
The fifth important primary income-generating factor is denoted by
the general tenn 'service sector'. It covers a conglomerate of economic
activities like banking, insurance, medical and social care, education,
etc. The location of these activities is in general dependent on the loca-
tion of the activities it provides services for. Lots of models have been
developed to describe and predict the size and location of these activities,
of which the economic base and Lowry models probably are most well-
known.
So far this section has been devoted to the traditional income-generat-
ing factors. These factors, however, are not sufficient to explain the
concentration of decision-making functions in, e.g., industry, banking,
government, and the concentration there of such activities as research
and development in major urban regions. The location of these kinds of
activities is mainly detennined by the opportunity they provide for
immediate, regular and frequent face-to-face contacts between the holders
of the associated functions (see Section 4.1). This aspect of spatial in-
equality at the Western European level is the main theme of Erlandsson's
contribution. He derives the total number of contact possibilities for
inhabitants of the major European urban regions as a function of their
numbers and the existing high-speed transportation network, which
provides the face-to-face communication facilities between these regions.
Decision-making and research functions in tum attract a great number
of other functions which supply them with infonnation processing,
accounting and other kinds of services. Therefore their concentration also
explains the concentration of other kinds of administrative functions. In a
way this process is comparable to the same kind of processes described
by the Lowry model for basic industrial activities and dependent service
sectors.
As decision-making and most administrative functions generally are
better paid than, inter alia, manufacturing and primary functions, this
FOUR SUBSYSTEMS 13

also helps in understanding regional income differences. Finally, when

decision-making and research and development functions are relocated,
e.g., to less-developed areas, accessibility in the form of high-speed trans-
portation is of crucial importance (see also Section 4.2).

1.2.4 The environmental subsystem

The environmental subsystem consists of the organic and inorganic world
which man lives in and which he uses to satisfy his needs. In traditional
economics the ecological subsystem was treated as a datum, because of
the abundance of natural resources. Especially since the early 1960s,
this situation has changed dramatically. As a consequence of, inter alia,
the enormous expansion of world population, the economic expansion
of the highly-developed nations and squandering production techniques,
the exhaustibility of natural resources was recognized. Moreover, with
environment used as the dumping-grounds for the waste of production
processes in such an accelerating way, it was seen that the limits of the
self-cleaning capability of the environment in many respects had been
reached or even exceeded. The environmental subsystem therefore can no
longer be treated as exogenous. As a consequence it has received growing
attention in, e.g., regional science.
The interrelationship between man and his environment has been
analysed and modelled by, inter alia, Daly (1968), Isard (1969), Victor
(1971) and Nijkamp (1977). In this section we will use an extension of
the Isard-Daly model to explicate the importance of the ecological sub-
system. This extension corresponds in essence to (1.1) with one region
considered at one point of time:

~
Economic Sociocultural Demographic Environmental
From subsystem subsystem SUbsystem subsystem

Economic A +
subsystem
Sociocul tural A +
subsystem (2.4.)
Demographic A +
subsystem
Environmental X X X A
subsystem

We can view (2.4) as an aggregate input-output flows matrix which includes

14 A FRAMEWORK

both economic, social, demographic and ecological processes. The

diagonal matrices (A) contain the coefficients representing the internal
structures, whereas the last row (x) and column (+) matrices represent,
respectively, the flows from the environmental subsystem to the other
subsystems and vice versa.
As outputs from the ecological subsystem one can mention:
- inputs into the economic subsystem: forestry, wild life, domesticated
plants and animals, minerals, etc.;
- inputs into the sociocultural subsystem: clean water and air as impor-
tant health conditions, outdoor recreation facilities, etc.;
- inputs into the demographic subsystem: destroyed natural environ-
ments and pollution which cannot be absorbed - as important factors
influencing migration and maybe birth and death rates at an unknown
scale.
The outputs from the other subsystems into the ecological subsystem
can broadly be defined into three categories: pollution, pollution abate-
ment and intended changes in the natural environment for its own sake.
Pollution, being an unintended by-product of human activity, interferes
in the biophysical relationship among organisms, including man. Pollution
has far-reaching consequences both for the environmental subsystem and,
through it, for the other subsystems. In the first place pollution interferes
with the various biochemical cycling processes whereby nitrogen, carbon,
phosphorus, etc., are absorbed by organisms. As a consequence of this,
certain organisms are poisoned and disappear whereas the conditions for
others become extremely favourable so that they become dominating.
Pollutants may also be directly toxic to people as is well-known from
many examples. Secondly, pollution generates a vicious circle. As
pollution produces disturbances in the environment, the environmental
ability to withstand further pollution is reduced.
Although pollution from the economic sUbsystem probably is the most
discussed, the other subsystems also influence the environment in many
ways, both directly and indirectly. These influences can in essence be
related to the density of population as far as the demographic subsystem
is regarded, and to the degree of material welfare and the desired con-
sumption style as far as the sociocultural sUbsystem is concerned.
From the standpoint of spatial inequities the essential feature of pollu-
tion is its reduction of the quality and quantity of environmental outputs
into the other subsystems, which generally can be compensated for only
at high costs, if at all.
A POLICY FRAMEWORK 15

Intended reductions of pollution, either by means of technical measures

after the pollution has been produced or by means of changes in the
production process itself, is called abatement and is the second above-
mentioned kind of interference in the environmental subsystem. Abate-
ment can be organized at different levels by different means. It is the cen-
tral theme of the contribution by Hafkamp and Nijkamp. They analyse
theoretically the conditions under which a multi-regional equilibrium may
be obtained in the context of a set of regionally-conflicting priorities regard-
ing maximization of production and pollution abatement. In addition, in
implementing an empirical application of their model, they reveal the
major difficulties we still have to confront in building theoretically-based
empirical models of more than one subsystem.
Attention to environment is also paid in Chapters 6 and 9. Stohr and
Todtling, e.g., point at several negative externalities resulting from the
dominance of functional private- and state-owned organisations, with
their predominantly one-sided economic goals, over smaller territorial
communities.
The last kind of output into the ecological subsystem concerns the
conscious attempts of man to actively restore or even create new natural
environments, compared to which abatement is defined as conscious
attempts to only minimise man's (negative) outputs into nature. This last,
positive, type of interference may take different forms and is bound to
become more important as the overall deterioration of environmental
quality proceeds.

1.3 POLICY IMPLICATIONS FOR REGIONAL DEVELOPMENT

To outline a policy framework for (partially) planned regional development,

attention has to be paid to the decision-making process, because the
quality of any policy is largely dependent on this process. Our second task
is to deal with conflicting interests within and between the regions under
consideration.

1.3.1 The decision-making process

Regional development is defined above as the change which takes place in
the state matrix !II of a set of regions. This change may be both positive
and negative in two senses. First, one or more regions may achieve a
higher or lower level of welfare, and second, the inequity in regional
16 A FRAMEWORK

welfare levels may be diminished or increased. Both aspects of regional

development are intermingled because they depend on the level of spatial
aggregation, as is shown in the contribution of Richardson.
In the process of regional development a crucial role is played by the
regional policies pursued by governments, both at a regional and at a
national level. They continuously intervene in demographic, economic,
sociocultural and ecological processes. These interventions can be defined
as a (regional) policy if they form a set of actions which aim at con-
scious, consistent and comprehensive guidance or control of society. It
should be noted that the word 'comprehensive' does not intend to express
the notion that a policy should cover all activities of people. It merely
implies that in preparing a policy all thinkable effects should be taken into
consideration.
As to the process of decision-making the following components can be
distinguished (see, inter alia, Dror 1968; McLoughlin 1969; Churchman
1970):
I. Establishment of goals and of the utility function relating to these
goals.
2. Establishment of interrelations and resources.
3. Preparation of a complete set of available alternative policies.
4. Preparation of a complete set of valid predictions of the costs and
benefits of each alternative and calculation of the net benefit of each
alternative.
5. Identification of the best alternative.
These components can be combined and arranged in different ways so
that different kinds of decision-making models come into existence. If it is
assumed that the goals are given and that the components can be arranged
as phases (2) to (5) without feedbacks, then the pure rationality model
appears. This model has been severely criticised (see, inter alia, Dror 1968;
Ozbekham 1969) in the sense that this model can never be completely
fulfilled in reality. Boguslav (1965), among others, points out that this
model is ineffective and time-consuming. The failure of the Integrated
Structure Plan (lSP) for the northern part of the Netherlands is partly due to
tp.ese characteristics, as is argued by Drewe in Section 9.3.
Many alternatives to the pure rationality model have been suggested:
strategic and dynamic planning, optimal policy, etc. (see Kassouf 1970).
These kinds of models have as common characteristics that goals are not
viewed as given, but instead as changing in the course of time. Uncertain-
ties regarding the other components of the decision-making process are
A POLICY FRAMEWORK 17

also introduced, which causes the necessity of feedbacks between the

different components. In this way a cyclical conception of the planning
process comes into being in which the policy goals are established in
phases. This cyclical conception of planning constitutes the basis character-
istic of the strategic-planning framework which Drewe uses as a checklist
in his evaluation of the ISP.
Finally, the influence interest groups can exercise in the decision
making process is of great importance. This aspect is described by Holland
in Section 8.2. From his analysis it becomes clear that the question of the
power of different interest groups to pursue their own goals at the cost of
the achievements of the goals of other groups is crucial. Therefore the role
of interest groups should be an integral part of regional analyses aimed at
a scientific underpinning of the (implicit) regional policies of all state
authority.

1.3.2 The treatment of conflicting goals

The establishment of goals poses one of the most difficult problems in any
planning process, since frequently goals inherently conflict with one
another. Not only in the recurring phase of their establishment, but also
in other phases, goals and the conflicts they generate play explicitly or
implicitly a major role in the decision-making process. Conflicts are
present within regions between the various components of regional
welfare, and between regions as a result of spatial externalities and the
limited availability of resources wanted by all regions in the system.
Conflicts within regions should be solvable in principle by regional
authorities, whereas conflicts between regions, again in principle, should
be solvable at a national or international level. In practice much regional
policy is centralised by national authorities. But as shown by Hafkamp
and Nijkamp a hierarchical de centralised planning could do the job as
well or even better (for a more general treatment of conflicting goals,
see Salih 1975). From their analysis it becomes clear that there may be
quite some room for autonomous regional decision-making as long as a
framework of indirect controls, e.g. prices and standards, is established.
These indirect measures should integrate diverging regional interests at a
higher level. In this view regional and national authorities have distinct
responsibilities of their own which should not be intermingled. Their
deliberate unison proved to be an unsatisfactory and less fruitful element
in the preparation of the ISP for the northern Netherlands. Hence, Drewe
also calls for a separation of regional and national responsibilities in
Section 9.4.
18 A FRAMEWORK

An important category of conflicting goals at the national level is

between aggregate efficiency and interregional equity as analysed by
Richardson. He argues in Section 7.6 that a complex and balanced set of
actions may be capable of reconciling these goals.
Finally, it is important to note that regional goals and regional policies
are not only pursued by territorial units and their organisations. With the
dominance of functional organisation in present-day society (see Sections
2.4, 2.5, 6.2 and 7.1) it seems clear that conflicts between functional and
territorial organisation may be even more important than those between
territorial units at the same and at different levels of spatial aggregation.
In Chapter 2 conflict between territorial and functional organisation are
analysed in depth. The policy proposals put forward in Chapters 6 and 8
deal especially with this major contradiction.
Different approaches are possible, depending partly on the goals given
priority and partly on the analytical diagnoses made. When goals regarding
the immaterial aspects of regional welfare are more prominent, a low level
of spatial aggregation may be chosen to combat spatial externalities of
functional development (Section 6.4). When the more material aspects,
e.g. employment and income, are put forward a higher national level
seems more likely to be able to deal with this conflict (Section 8.3).
The national state also can be viewed as a strongly functionally-oriented
organisation because of its predominantly departmental structure of
decision-making. When the analyses stress this functional character of the
national state as another cause of local problems (Section 6.2) this may be
a reason for the choice of a lower spatial scale for territorialising func-
tional power.

1.3.3 Concluding remarks

In outlining a framework for analysing spatial inequalities and formulating
policies towards regional development, it has become clear that theoreti-
cally an umbrella can be constructed. The next important step, however,
is the very troublesome and difficult task of filling the relatively empty
theoretical boxes and of providing more relevant regional science know-
ledge for policy-making purposes. Especially knowledge about power
distributions in society and about the working of large-scale organisations
(see also Section 9.4), which hitherto received relatively little attention,
should be incorporated in regional science to make it of more relevance
for the actual underpinning of regional policies.
 19

REFERENCES

Andersson, A., 'Economic analysis of demographic problems', in Andersson and Holmberg (1977).
Andersson, A., and P. Holmberg (eds.), Demt>graphic, economic and social interaction. Ballinger,
Cambridge, Mass., 1977.
Andersson, A., and Th. Plant, 'Economic analysis of the supply and demand for labour in Regions',
in Andersson and Holmberg (1977).
Boguslav, R., The new Utopians: a study of systems design and social change. Prentice Hall, Engle-
wood Cliffs, 1965.
Churchman, C. and W. Kant, 'A decision theorist', Theory and Decision 1,1970,107-116.
Daly, H.E., 'On economics as a life science',Journal of Political Economy, vol. 76,1968,392-406.
Dror, Y., Public policy making reexamined. Chandler, San Francisco, 1968.
Goldberger, A.S. and O.D. Duncan (eds.) ,Structural equation mt>dels in the social sciences. Seminar
Press, New York, 1974.
Gustavsson, J .R., B. Harsman and F. Snickars, 'Intraregional migration and the housing market', in
Andersson and Holmberg (1977).
Harris, C.C. and F.E. Hopkins, Locational analysis. Lexington Books, London, 1972.
Isard, W., 'Some notes on the linkage of the ecological and economic systems', Papers of the
Regional Association , vol. 22, 1969, 107-118.
Joreskog, K.G., 'Structural equation models in the social sciences: specification, estimation and
testing', Invited paper for the symposium on Applications of Statistics, Dayton, Ohio, 1976.
Karlq uist, A. and F. Snickars, 'Regional migration and labor markets', in Andersson and Holmberg
(1977).
Kassouf, S.,Normative decision making. Prentice Hall, Englewood Cliffs, 1970.
Krech, D. and R.S. Crutchfield,Elementsofpsychology. Knopf, New York, 1965.
McLoughlin, B., Urban and region planning: a systems approach. Faber and Faber, London, 1969.
Norlen, U., 'Development of a statistical model for social indicators', in Andersson and Holmberg
(1977).
Nijkamp, P., Theory and application of environmental economics. North Holland, Amsterdam,
1977.
Ozbekhan, H., 'Toward a general theory of planning', in Jautsch (ed.), Perspectives of planning.
OECD,Paris, 1969.
Paelinck, JR. and P. Nijkamp, Operational theory and methods in regional economics. Saxon
House, Farnborough, 1975.
Reiner, T.A. 'Welfare differences within a nation' ,Papers of the Regional Science Association, vol.
32,1973.
Rogers, A. and S. McDougall, 'An analysis of population growth and change in Slovenia and the
rest of Yugoslavia', Working Paper 81, Center for Planning and Development Research, Univers-
ity of California, Berkeley, 1968.
Salih, K., 'Goal conflicts in pluralistic multi-level planning for development' ,International Regional
Science Review, vol. 1, no. 1,1975,49-72.
Smith, D.M., Industrial location: an economic geographical analysis. Wiley, London/New York,
1971.
Victor, P A., Input-output analysis and the study of economic and environmental interaction.
Ph.D. thesis, University of Columbia, 1971.
Wadensjo, E., 'Some factors determining international migration', in Andersson and Holmberg
(1977).
Wilson, A.G., Entropy in urban and regional mt>delling. Pion, London, 1970.
Wilson, A.G., Urban and regional mt>dels in geography and planning. Wlley, London, 1974.
Part I: Analytical studies in spatial
inequalities and regional development
2. On the contradictions between city
and countryside
JOHN FRIEDMANN*

The bourgeoisie has subjected the country to the rule of towns.

It has created enormous cities, has greatly increased the urban
population as compared with the rural, and has thus rescued a
considerable part of the population from the idiocy of rural life.
Just as it has made the country dependent on the towns, so it has
made barbarian and semi-barbarian countries dependent on the
civilized ones, nations of peasants on nations of bourgeois, the
East on the West (Karl Marx and Friedrich Engels, Manifesto of
the Communist Party).

The locus classicus on the question of rural-urban relations is The German

Ideology (Marx and Engels). Because it sets forth the essential themes to
be developed here, it will be useful to quote the pertinent passage in
extenso:

The existence of the town implies, at the same time, the necessity of administration,
police, taxes, etc.; in short, of the municipality, and thus of politics in general. Here
first became manifest the division of the population into two great classes, which is
directly based on the division of labour and on the instruments of production. The
town already is in actual fact the conception of the population, of the instruments of
production, of capital, of pleasures, of needs, while the country demonstrates just the
opposite fact, isolation and separation. The antagonism between town and country
can only exist within the framework of private property. It is the most crass expres-
sion of the subjection of the individual under the division of labour, under a definite
activity forced upon him - a subjection which makes one man into a restricted town-
animal, the other into a restricted country-animal, and daily creates anew the conflict
between their interests. Labour is here again the chief thing, power over individuals,
and as long as the latter exists, private property must exist. The abolition of the antag-
onism between town and country is one of the first conditions of communal life, a

* University of California, Los Angeles. The research for this essay was done while
the author was on a Guggenheim Foundation fellowship at the Centre for Environ-
mental Studies in London, England. Earlier drafts received very helpful comments
from Peter Marris, Edward Soja, Clyde Weaver, and the editors ofthe present volume.
24 CITY AND COUNTRYSIDE: CONTRADICTIONS

condition which again depends on a mass of material premises and which cannot be
fulfilled by the mere will, as anyone can see at first glance ... The separation of town
and country can also be understood as the separation of capital and landed property,
as the beginning of the existence and development of capital independent of landed
property - the beginning of property having its basis only in labour and exchange
(Marx and Engels 1970, p. 69).
In this famous passage, Marx and Engels refer to the 'antagonism' between
town and countryside, tracing the conflict to the spatial division of labor.
But only a year or two later, in the Communist Manifesto, they wrote
about the 'gradual abolition of the distinction between town and country'.
And to confuse matters still further, a footnote in the English translation
of this work reminds us that 'in the German edition of 1848, the word
"distinction" reads "antithesis" '(Marx and Engels 1975, p. 60).
Now antagonism, distinction, antithesis, and contradiction - this last
term being currently in vogue - are clearly not synonymous; each conveys
a different theoretical meaning. Because clarification of this central
concept is essential to my task, 'contradiction' is made the subject of an
extended theoretical inquiry in Section 2.1. This is followed, in Section
2.2, by an examination of the nature of the contradiction between the
functional and territorial bases of social integration. A final component
of the theoretical framework is added in Section 2.3, where the related
concepts of city and countryside are discussed in terms that will be useful
to the further development of my topic. The remaining two sections are
devoted to an extended analysis of the clash between rural and urban-
centered interests. Section 2.4 concerns the pseudo-resolution of this
clash within the core countries of western capitalism, while section 2.5
projects the argument into a global context. The question is: if rural-
urban contradictions were 'successfully' resolved within the advanced
capitalist nations, could not this process be replicated with a view to the
ultimate integration of the world economy into a single, homogeneous
system of functional relationships?
2.1 CONTRADICTIONS: COSMIC AND HISTORICAL

Contradictions are everywhere in the world. Without contradic-

tions there would be no world (Mao Tse-tung, On the Ten Great
Relationships).

In ordinary usage, the notion of contradiction suffers from a certain

imprecision of meaning. Some of this, no doubt, must be attributed to
history. Originally a term in logic ('statement A "contradicts" statement
CONTRADICTIONS: COSMIC AND HISTORICAL 25

B' in the sense that A and B cannot be logically maintained at the same
time), it came to be applied, in the hands of German idealist philosophers
(Kant, Hegel), to relations in reality (Williams 1976,91-93). Subsequently,
it played a key role in the formalization of dialectical materialism. 1 In
Marxist thought, contradiction acquired specific, if sometimes elusive,
historical and existential meanings; it became simultaneously a condition
and principal method of revolutionary practice.
In this paper, I shall use contradiction to refer to a standing in mutual
opposition of two social forces which, though interpenetrating and
clashing, are complementary to one another, comprising a unity, a
whole. 2 The broad scope of this definition can be rendered more precise
by distinguishing its several moments. Contradictions that are constitutive
of the human condition, I propose to call cosmic contradictions. An
example would be the contradictions that exist between the individual
self and the matrix of social relations that sustains us and through which
we become human. A cosmic contradiction also exists between the two
opposing bases of social integration: functional and territorial. I shall
return to this contradiction in the next section. For the present, let it
suffice to note that cosmic contradictions can change in only two

1. It was Engels, rather than Marx, who formalized the dialectic as a way of viewing
the world and so transformed the loose notion of contradictions-in-reality into a
more rigorous concept (Engels 1976; OHman 1971, Chapter 5).
2. The unity (the whole, the entity) arises from the contradictory relation; it is not
pre-existing to it. Thus there are no ultimate entities in the universe, only forces-
in-relation. According to Colletti (1975), this is the only true dialectical form of
opposition, but it exists exclusively in logic. Relations of opposition in reality, on
the other hand, are instances of contrariety (Realrepugnanz), and are without
logical contradiction.
This distinction is made in order to save Marxism for science, or possibly the
other way round. The argument is that, in a unity of opposites, neither of the two
component forces has ontological reality by itself; it is a mere shadow, whose
existence requires the constant presence of the other. Adopting Colletti's position,
both proletarians and bourgeois would, in themselves alone, be merely fictions
( epiphenomena).
Colletti's argument is not convincing. To reach it, he has to dip into idealist
philosophy. From an historical materialist point of view, the proletariat exists not
for itself alone, but only as the dialectical opposite of capitalist social forces.
Eliminate capitalism, and the proletariat, in its specific historical appearance and
mission, is eliminated as well. Within capitalism, however, the proletariat is 'real'
enough and not, as Colletti seems to think, an instance of the metaphysical imagina-
tion. For a non-dialectical 'systems' view of this same problem, see Edgar S. Dunn,
J r. (1971, Chapter VI).
26 CITY AND COUNTRYSIDE: CONTRADICTIONS

respects: in their external appearance (or fonn) and in the balance of

opposing forces. They can never be resolved in any final way.
All other (non-cosmic) contradictions, I call historical. Historical contra-
dictions are non-pennanent. When their resolution can be accomplished
without destroying the unity they comprise, they are non-antagonistic.
But the resolution of antagonistic contradictions will constantly bring
forth new unities. Antagonistic contradictions are the major generative
force in history. 3
Non-antagonistic contradictions arise within the context of antagonistic
struggle. Capitalist competition for market dominance, for example, is
non-antagonistic: the market areas in contention may be divided by
mutual agreement, leaving each finn in monopolistic control over a
portion of the whole. Similarly, in the lexicon of Maoist ideology, some
contradictions are said to arise 'among the people'. But again, their resolu-
tion - through communal education, for example - will not violate the
essential unity of the people which is dermed by their continuing antagon-
istic struggle against the hostile forces of capitalism. What detennines the
nature of a contradiction is the correct method of its resolution. Non-
antagonistic contradictions will yield to peaceful methods; antagonistic
contradictions will not.
This taxonomy is helpful in clarifying the theoretical use of the con-
cept. Equally important, however, are the fonns contradictions take in
the general movement of history. According to the well-known Hegelian
schema, contradictions are resolved through specific actions that 'surpass'

3. The distinction between antagonistic and non-antagonistic contradictions is taken

from Mao Tse-tung (1968,70-71):
'Contradiction and struggle are universal and absolute, but the methods of resolving
contradictions, that is, the forms of the struggle, differ according to the differences
in the nature of the contradiction. Some contradictions are characterized by open
antagonism, others are not. In accordance with the concrete development of things,
some contradictions which were originally non-antagonistic develop into antagon-
istic ones, while others which were originally antagonistic develop into non-antagon-
istic ones . . . Economically, the contradiction between town. and country is an
extremely antagonistic one both in capitalist society ... and in the Kuomintang
areas in China.... But in a socialist country and in our revolutionary base areas,
this antagonistic contradiction has changed into one that is non-antagonistic; and
when communist society is reached, it will be abolished.
Lenin said, 'Antagonism and contradiction are not at all one and the same.
Under socialism the first will disappear, the second will remain.' That is to say,
antagonism is one form, but not the only form of the struggle of opposites; the
formula of antagonism cannot be arbitrarily applied everywhere.
CONTRADICTIONS: COSMIC AND HISTORICAL 27

them, leading to a new and higher unity, a movement which suggests the
image of a forward-surging spiral. 4 In the process, although the formerly
opposing forces are destroyed and vanish from earth, some of the 'positive'
and usually secondary elements are rescued into the next ('higher') stage
of the historical process, where they are retained. 5 But every new unity,
in turn, will generate new forces in opposition to itself, and so the process
continues without let-up, destroying, surpassing, retaining, in a continuing
upward spiral of historical progress. 6
In this ontology, there are no 'cosmic' contradictions. Instead, the
outward appearance of historical phenomena in Hegel veils the unity of
spirit that pervades all things. All contradictions strain towards an expres-
sion of this unity. And only when the 'final' contradiction has been
overcome will humankind embark upon the true course of its history.
This mystical vision animates such fundamentally divergent thinkers as
Karl Marx (the 'higher phase of communist society') and Pierre Teilhard
de Chard in ('Omega').7 Hegel 'stood on his head', one must conclude, is
still indubitably Hegel (Althusser 1969).
Just as Hegel's philosophy, together with its materialist inversion, is,
in the final analysis, monistic (Le. reducible to a single, indivisible unity),
so traditional Chinese, including present-day Maoist, thought is funda-
mentally dualistic (Le. ultimate reality is two, not one). The idea of an

4. According to Friedrich Engels, the most important dialectical laws are the
'transformation of quantity to quality - mutual penetration of polar opposites and
the transfonnation of each other when carried to extremes - development through
contradiction or negation - spiral form of development' (cited in Ollman 1971,
p.55).
5. The German equivalent of 'surpassing' - au/heben - means both a 'raising up' and
a 'preserving' and thus contains the notions of progressive change as well as
continuity.
6. Hegelian dialectics, with its dual moment of destruction and preservation in histori-
cal processes, appeals to a romantic streak in the German imagination. The Valhallian
rhetoric of destruction (which also appears in Joseph Schumpeter's celebrated
phrase, 'gales of creative destruction', referring to economic innovations under
capitalism) accounts for much of the romantic appeal of Marxist writings.
7. With the achievement of a classless society from which all coercive (political) power
is absent, antagonistic contradictions will also have been overcome. At this point,
pre-history comes to an end, and humankind's true history begins to unfold.
Marx's profound intuition of this history is perhaps best expressed in a short
paragraph of the Critique of the Gotha Program (Marx 1972, p. 17). In all fairness,
it should be noted that what is historical transformation in Marx becomes trans-
figuration with Teilhard de Chardin. All the same, de Chardin's 'hyper-personal'
at the Omega-point of History is still this side of Heaven (de Chardin 1965, Book
IV, Chapter 2).
28 CITY AND COUNTRYSIDE: CONTRADICTIONS

ontogenetic unity ('world spirit') is utterly alien to this thought; all

existing unities are made up of the opposing forces of yin and yang.
Thus, in the earliest of Chinese classics, the Book of Changes, we may
read: 'The two principles in opposition are united in a relation based on
homogeneity; they do not combat but complement each other. The
difference in level creates a potential, as it were, by virtue of which
movement and living expression of energy became possible' (Wilhelm
1968, 281-282). Here the emphasis is on the complementarity of
opposites, to the point where the ancient commentator even questions
whether or not the two opposing forces express a dualistic point of view
('A relation based on homogeneity'). Several thousand years later, Mao
Tse-tung would emphasize the aspect of struggle between them, as each
elemental force presses against its opposite (and opposing) force, attempt-
ing to maintain, to increase, to reduce, or to invert the relative imbalance
between them, changing one thing into another in an infinite sequence of
historical transformations. For Mao: 'the unity of opposites is the most
basic law . . . There is no such thing as the negation of the negation.
Affirmation, negation, affirmation, negation . . . in the development of
things, every link in the chain of events is both affirmation and negation
... The life of dialectics is the continuous movement towards opposites.
. . . Engels spoke of moving from the realm of necessity to the realm of
freedom, and said that freedom is the understanding of necessity. This
sentence is not complete ... Does merely understanding make you free?
Freedom is the understanding of necessity and the transformation of
necessity - one has some work to do, too . . . Thus it is that only by
transformation can freedom be obtained' (Schramm 1974,226-230).
In this dualistic view, contradictions may change their specific, historical
form, but inherently they remain the same, merely alternating in their
position of relative strength or weakness. It follows that the flux of
history is cyclical rather than progressive. All apparent unities are consti-
tutive of forces in contradiction and struggle.
Throughout the remainder of this essay, I intend to follow this (non-
Hegelian) dualistic interpretation of the concept of contradiction.

2.2 FUNCTIONAL AND TERRITORIAL BASES OF SOCIAL TRANSFORMATION

Every social formation, by definition, must be conformed or 'integrated'.

That is the essential condition of its continued stability and growth. The
BASES OF SOCIAL TRANSFORMATION 29
meaning of integration, however, is an elusive one (Seymour 1976,3-10).
It is not identical with homogenization, alignment, equilibrium. These
terms signify an absence of life, the punto muerto of maximum entropy.
The integration of a social formation is rather the form of its struggle to
become. Integration is, therefore, akin to differentiation, opposition, and
tension; it is the result of forces locked in perpetual struggle.
In any social formation, there exist two principal bases of social integra-
tion, functional and territorial. Together, they constitute a 'cosmic' unity
of opposites - apart, yet together. The first term refers to linkages among
entities organized into hierarchical networks on a basis of self-interest.
Where a relationship between two entities is functional, one is using the
other as an instrument to accomplish a purpose of its own. With respect
to power, such relationships are symmetrical only in the limit. By contrast,
territorial integration refers to those ties of history and sentiment that
bind the members of a geographically bounded community to one
another. 8 Territorial communities are communities informed by a deep
attachment to their territorial base. Their consciousness curves back upon
itself and comes to rest in its history. Transcending interests of class - a
functional concept - such a community defines itself collectively in
struggle against the outside world. 9

8. This description of the two bases of social integration bears a certain family resem-
blance to such well-worn sociological distinctions as organic and mechanical
solidarity (Durkheim) and society and community (Tonnies). These earlier
concepts, however, were modelled, not as a 'unity of opposites' but as 'ideal types'
of two contrasting (and mutually exclusive) forms of social organization which
might subsequently be applied to the study of social history (from the primary
relations of Gemeinschaft to the secondary, associative relations of Gesellschaft,
from 'traditional' society to 'modern'). By contrast, functional and territorial bases
of social integration are complementary forces that stand in contradiction to each
other; they are complex social forms which arise from this double interaction.
9. All social integration above the level of the small, face-to-face group occurs on the
basis of either territory or function. This is true of even nomadic tribes whose
migrations are fixed within a given territorial range or of Jews in the diaspora whose
common history is comprehensible only with reference to their expUlsion from
their ancestral home in Judea. (Cf. Porteous 1976).
In territorially integrated societies, history and place are one. This is beautifully
illustrated in the following description, based on the Chinese Book of Rites (she-
king), on the founding of a new city:
'There is a settled order in which one must proceed. The ramparts are first raised:
they are the most sacred part of the town. Afterwards the temple of the ancestors
[Miao 1 is built. Care is taken to plant trees meanwhile (hazels and chestnuts) whose
fruits and berries will be offered to the ancestors, and those trees which are used
 30 CITY AND COUNTRYSIDE: CONTRADICTIONS

As territorial entities expand through either conquest or peaceful

annexation, they must increasingly resort to corporate and bureaucratic
power for the control and management of their domain. Concurrently,
their internal or normative power is weakened, until it is no longer able to
make an adequate response to the iinpacts - mostly disintegrative - of
increased functional power. But territorial power is similarly threatened by
an expansion of corporate power (resulting, for example, from the
efficient extraction of monopoly profits). Corporate power will seek to
impose its own principle of social organization, the pursuit of private gain,
on territorial forms of organization.
Some of the major distinctions between functional and territorial
organizations are shown schematically below. 1o

to make coffins and sonorous drums. "In ancient times when the plan of the capital
was traced, they did not fail to choose the most important rising ground in the
kingdom and constitute the ancestral temple, nor to select trees of the finest
growth to make the sacred forest." When the walls, the altars and the plantations
which are to give sanctity to the town are finished, the palace and houses are built'
(Granet 1958).
The symbolic quality of this historical action reveals the reverence with which
human beings approach the task of staking out a territorial core for their lives.
Yet territory is merely the ground for human history. It is collective experience
which creates the essential bond between human beings. Thus Duchacek (1970,
p. 20): 'It should be recognized that men identify not only with a territory as such
but also (if not primarily) with its political and economic system, its methods and
goals, and with its history and its destiny' (emphasis added).
10. This distinction is also recognized by Miguel Morales (1976, p. 83):
'A third tendency, finally, has been shown to exist in the formal and a-historical use
of the term region. This has contributed to an asymmetrical relation: on one hand,
the evident descriptive power of the term, and on the other, its inability to explain.
In these cases, two aspects have dominated the scene: in the first, space is con-
ceived mathematically, in abstract form, which allows one to know areas based on
topological considerations; in the second, which is identified with empirical systems
analysis, the historical meaning of the process ofregionalization as a specific social
process has been practically dissolved in its phenomenological appearance. Regions,
in that case, are converted into mere study areas or areas for action without signifi-
cant historical contents; eventually, they become formalized planning areas which
in their practical aspects are manifestly efficient' (translated from the Spanish;
emphasis added).
CONCEPTS OF CITY AND COUNTRYSIDE 31

FUNCTIONAL ORGANIZATION TERRITORIAL ORGANIZATION

TIME rate of change (dy/dt) history (concrete succession of

events)
SPACE cost of overcoming distance regional geography
(e.g. 'gravity models')
SYSTEM unbounded, but global in the physically bounded (partial closure)
limit ('spaceship earth ') at neighborhood, local, regional,
national, and imperial levels
SOCIAL RELATIONS hierarchical, instrumental hierarchical, ascriptive
FORM OF contractual, associational reciprocal, communal
ORGANIZATION
POWER BASE utilitarian and coercive normative and coercive
INTEGERS individual, folk,
interest group, tribe,
social class, nation,
corporation people

2.3 CONCEPTS OF CITY AND COUNTRYSIDE

The theoretical framework for an analysis of rural-urban relations will be

completed with the distillation of analytically useful concepts of city and
countryside. In line with the distinction between functional and territor-
ial organization, two very different conceps of the city, together with
their corresponding concepts of countryside, may be distinguished.
To begin with the functional concept, cities appear as spatially organ-
ized subsystems of society that are characterized by a high relative density
of popUlation and a predominance of non-primary (specifically non-
agricultural) activities. In this perspective, the designation of 'city' serves
as a metaphor for a functional, primarily economic clustering of human
activities which creates its own topological (abstract) space. Such 'cities'
have obviously no history; their temporal dimension is best expressed as
a rate of change, and their performance is assessed from primarily an
economic point of view. At issue is their efficiency as centers of produc-
tion, distribution, and capital accumulation, as formulated by Lucien
Romier:

... large cities are formed by national and international connections and movements;
their future is bound up in a network of vaster activities; and their destiny crosses
administrative and even territorial borders to follow the general trade routes (in
Marilitegui 1971,p. 174).
 32 CITY AND COUNTRYSIDE: CONTRADICTIONS

Jose Carlos Mariategui gave succinct expression to this functional concept.

'A great city,' he wrote in 1931, 'is basically a market and a factory'
(Mariategui 1971, p. 178). The functional mentality is also at work in
such doubtlessly useful but vapid concepts as Standard Metropolitan
Statistical Area (SMSA) and Functional Economic Area (FEA) (Berry
1969). These 'areas' - no longer even nominally cities - are linked into a
system, the so-called space-economy, whose structure is mediated through
a measure of generalized access (lsard 1956). Although the economic
space of a nation can be mapped, showing a differentiated profile of
appropriate index values (which themselves are chosen to express an
urban bias), only that space will be drawn which is to some extent
already integrated with the market economy (Berry 1975). In this
perspective, rural areas will show up as economic areas whose low 'profile'
is different in quantity but not in quality from so-called cities, or they
remain virtually blank, much as uncharted map areas of the Antarctic used
to be portrayed on ancient maps.ll
This perception is reflected in the practice of corporate interests that
reside at the principal nodes of the space economy. Rural areas are
scanned by them for their potential instrumental value in expanding
corporate production and for accumulating wealth. To the extent that
they may have such value, rural 'peripheries' will be organized to supply a
steady stream of resources to a growing market economy whose ultimate
extent is global.
In the contrasting language of territorial organization, the city appears
as a bounded, territorial unit, part of a specific social formation that is
characterized by the relatively high density of its population, a predomin-
ance of non-primary (non-agricultural) activities, and a degree ofpolitical
autonomy whose theoretical value may range from Virtually zero to 100
percent. 12 A territorial city is therefore an historical and political entity,

11. Such a map is produced by Edward W. Soja (1968b, fig. 47) for post-independence
Kenya. This map (together with the accompanying text) distinguishes between the
political (national) boundaries of Kenya and the 'effective national territory' which
occupied no more than a fifth of the former. The only data shown for the 'empty'
four-fifths are 'Shifta attacks and ethnic conflicts'.
12. Political autonomy was virtually absent in the Islamic city of the Middle ages
(Lapid~s 1967). The Greek polis, on the other hand, enjoyed complete autonomy
in principle (Finley 1973). In practice, of course, 'many poleis were joined in
"leagues" with their neighbors; others were in a permanent condition of subjection
to the more powerful' (Pounds 1969, p. 136).
CONCEPTS OF CITY AND COUNTRYSIDE 33

that is interconnected with other, similarly constituted units in a variety

of ways. It is a place, with a physiognomy of its own, and it always has a
nameP Walls are symbolic of territorial cities, but they are not a defining
characteristic. 14
Against the background of such cities, rural areas - the countryside -
are also territorially defined, and rural populations are seen as being
grouped and bound together by a remembered past, a certain pattern of
relations between man and land, a common dialect or language, a set
of religious beliefs, familiar customs. Under territorial rule, rural regions
are typically subdued and politically dominated by the city. Their
appearance in history as politically autonomous areas is as brief as it is
infrequent.
Howard Spodek provides us with one of the best images of the territo-
rial city and its reciprocally dependent countryside. His objects are the
headquarters towns in Saurashtra, India, up to the advent of British rule
in 1863:

Two elements in the structure of headquarters towns impeded regional and market
integration. First, within the headquarters town, the military ruler and the professional
and business classes - particularly those with wide-ranging contacts - often battled
with one another. The goal of the ruler, which was usually to maximize political
control over his subjects, conflicted with the goals of the merchants and professionals,
which were to make profits and to exercise skills and talents. The ruler usually did not
want the town market and court to be independent entities. He wanted them subject
to his will ...
Towns in India impeded political and economic integration in a second way. The
towns, which may have served as hinges vertically linking higher and lower levels of the
polity, nevertheless served as military headquarters 'for ·antagonistic relations with

13.A magnificent <territorial' portrayal of a city is found in Kollek and Pearlman

(1968). Territorial thinking is beginning as well to enter contemporary planning
practice (see Lynch 1976).
14.Max Weber (1958) made its protective walls one of the defining characteristics of
the western city. The symbolic value of walls in defining a place leans on the
following paragraph from Lewis Mumford's The City in History (1961, p. 52):
'Now all organic phenomena have limits of growth and extension, which are set
by their very need to remain self-sustaining and self-directing: they can grow at the
expense of their neighbors only by losing the very facilities that their neighbors'
activities contribute to their own life. Small primitive communities accept these
limitations and thus dynamic balance, just as natural ecological communities
register them.'
The walls are drawn 'protectively' around a particular uniqueness; they define the
'placeness' of a place (see also note 9).
34 CITY AND COUNTRYSIDE: CONTRADICTIONS

neighboring towns, so to speak, horizontally. Frequently, towns were militantly hos-

tile to one another. This hostility could have led to creative competition, but at least
as often it resulted in warfare. Persistent warfare and the threat of war inhibited
political and market integration.
These two tensions frequently marked the urban system in pre-Independence India:
tension between the land-holding town rulers and the non-landed merchant and pro-
fessional classes; and tensions between neighboring towns which served as headquarters
for local, often militarily antagonistic rajas ...
When the British began to dominate the Saurashtra region politically, they found
162 fortified settlements in its 25,000 square miles ... Trade between kingdoms was
minimal . . . The politically and culturally fragmented peninsula had no unifying
urban hierarchy.
Struggles within the town walls between military rulers, usually Rajputs, and the
merchants and professionals are copiously documented both in Indian and British
sources ... The rulers were tied to the land and capital city, for these were the bases
of their power. The merchants and professionals, however, saw geographical mobility
as their main lever of power. They bargained with the prince, threatening to desert and
take their capital and skills with them if their demands were not met. Often they did
leave, either by choice or forced out by coercion. Loyalty to the state did not often
develop (Spodek 1973,254-256).

To describe the territorial city, Spodek had to set it off against the func-
tional city. And, indeed, both fonns of description are jointly necessary
for a complete account: the city is a node in a global network of economic
relations; it is also a place, possessing a political life of its own, a history,
a unique appearance, a mode of life, a sense of being. 15 After four
centuries of steady attrition, this second, psycho-historical dimension is
only vestigially present. 16 Both corporate and bureaucratic power have
become consolidated at the national level, extending cities into an
intricate network of functional relations. Symbolized in the literature as
'megalopolis' (e.g. Gottman 1961; Soja 1968a), this network is more

15. Historically, there has been a gradual evolution of the city from place to node. This
evolution is beautifully traced and recorded in Mumford (1938). A more dialectical
urban history in which the fate of the city as a place becomes intertwined with the
fate of the national economy is found in Conot (1974). The 'great city' of Conot's
saga is Detroit.
16. According to Charles Tilly (1976, p. 374), ' ... the two centuries after 1700 pro-
duced an enormous concentration of resources and means of coercion under the
control of national states, to the virtual exclusion of other levels of government .
. . . A whole series of organizational changes closely linked to urbanization, indus-
trialization and the expansion of capitalism greatly reduced the role of the communal
group as a setting for mobilization and as a repository for power; the association
of one kind or another came to be the characteristic vehicle for collective action.
The rise of the joint-stock company, the political party, the labor union, the club
all belong to the same general trend.'
THE PSEUDO-RESOLUTION OF CONTRADICTIONS 35

accurately described as a fabric of invisible relations (financial, com-

munications, freight and people movements). As a result, the .city has
become a staging ground of conflicting national and international forces.
Whether it will re-emerge as an organic element in a revitalized cultural
region, as in Lewis Mumford's vision of a 'regional city', remains to be
seen. It may be noted, however, that regionalism is beginning to re-emerge
as a potent political force in many countries, especially in Europe. 17
The cosmic contradiction between city and countryside is therefore
between the functional organization of bureaucratic and corporate power,
on the one hand, and territorially organized power - at all the relevant
scales - on the other. It is between two contradictory yet complementary
modes of social existence. Other, historical contradictions are similarly
present, for instance, in Howard Spodek's account of territorial warfare
in Saurashtra, or in the competition among corporate powers for resources
and markets in the periphery of major economic regions. Whereas these
historical contradictions are amenable to some form of resolution, the
cosmic contradiction is, ultimately, not.

2.4 THE PSEUDO-RESOLUTION OF CONTRADICTIONS IN THE CORE

COUNTRIES OF CAPITALISM

Coincident with the spectacular rise of capitalist forms of organization,

the industrial revolution unleashed a series of violent forces that, in the
end, brought on the nearly total collapse of local forms of territorial
governance. Among these forces were relentless economic growth, the
continuous technological transformation of both commodities and their
methods of production, explosive demographic expansion, vast inter-
regional migration, the settlement of new continents, and the consequent
de-population of the older rural areas.
As cities expanded, they also burst their traditional 'containers', spilling
out into the surrounding countryside. Where they were still surrounded by
a wall, the wall, in a fateful and symbolic gesture, was torn down. The old
unity of territory and governance was beginning to crack. In the more
densely populated regions, cities actually fused as physical entities, form-
ing vast sheets of urbanism, Patrick Geddes' 'conurbations'.
17. Duchacek (1970) is undoubtedly the best general study on the subject. For Spain,
see the extremely provocative Asalto al Centralismo (1976). For Great Britain:
Hechter 1975. For France: Dulong 1975 and n.d.
36 CITY AND COUNTRYSIDE: CONTRADICTIONS

Cities were becoming unbounded spatial structures. It was therefore

only a matter of time before a new 'science' would spring up to measure
the dimensions of this new formation, trace its topography, and discover
the sources of its 'lawful' behavior. The Regional Science Association was
founded in 1954. The flavor of its dominant interest is captured in the
following introduction to a major paper delivered at the Association's
first annual meeting. According to the author:

This article suggests a method for describing the magnitude of a city's influence on
its surrounding hinterland. And a method is proposed for partitioning an entire terrain
into metropolitan regions or even urban trade areas (Carroll 1955, p. 0-2).

The functional reductionism of this approach is unmistakeable. About the

same time, Don Martindale and Gertrud Neuwirth wrote their laconic
epitaph commemorating the passing of the territorial city:

The modern city is losing its external and formal structure. Internally it is in a state of
decay while the new community represented by the nation everywhere grows at its
expense. The age of the city seems to be at an end (Weber 1958, p. 62).

The new urban scale, coincident with the appearance of a national hier-
archy of cities and a nationally integrated economy, was ultimately
justified by referring to the immense gains in economic efficiency that
were supposed to be realized (Borts and Stein 1964; Mera 1975). The
immediate evidence for this was the seemingly irresistible growth of
national production and the resulting affluence of the population at large.
These gains were dearly bought, however. Bourgeois complacency was
profoundly shaken by the accumulation of misery and squalor in cities,
by cyclical instability, by massive structural unemployment, by life-
destroying pollution, by the rapid exhaustion of natural resources, by
increasingly devastating and senseless wars, by widespread social aliena-
tion. All the same, many 'costs' were thought to be the unavoidable
byproducts of a modern industrial economy:

. . . rapid economic progress is impossible without painful readjustments. Ancient

philosophies have to be scrapped; old social institutions have to disintegrate; bonds of
caste, creed, and race have to be burst; and large numbers of persons who cannot keep
up with progress have to have their expectations of a comfortable life frustrated. Very
few communities are willing to pay the full price of economic progress (United
Nations 1951, p. 15).
THE PSEUDO-RESOLUTION OF CONTRADICTIONS 37

That, as it happened, was the collective wisdom of a group of international

experts which had been asked to advise the United Nations on measures
of full employment in the underdeveloped countries of the world. It
brings to mind Joseph Schumpeter's exultation in the process of 'creative
destruction', whose break with the past he understood to be radical and
complete (Schumpeter 1947, Chapter 7). This conscious turning away
from the past, this virulent anti-historicism, found its ultimate expression
in the celebrated phrase of Henry Ford: History is bunk! It might have
been more accurate to say that history had been reduced to a bundle of
growth rates.
This dismissal of history as a source of relevant knowledge pointed to
a more general phenomenon: the loss of territoriality and the correspond-
ing loss of social bonds that were not based exclusively on calculation.
Gains had become privatized, and this made it easier to ignore the social
costs of corporate expansion which accrued primarily to territorial
entities. IS
As local territorial organizations disintegrated under the impact of
fundamental rationalization, class conflict replaced the rivalry of towns
and regions. This new conflict - appropriate to the newly dominant mode
of functional organization - was chiefly an urban phenomenon, but it was
present as well in those rural areas where agricultural production was
organized along corporate lines, as in California. For all practical purposes,
such areas were rural in only external appearance. In every other respect,
they had become a part of the unbounded city.
Territorial power had not altogether vanished, however; it had merely
been displaced to the national level. Although its reintegration around the
nucleus of the state occurred slowly and unevenly - it had begun much
earlier in France, for instance, than in Germany or Italy - it was chiefly
during the 19th and 20th centuries that local power was rendered all but
impotent (Holborn 1954).
Just as territorial power re-emerged at the national level, so corporate
power was re-organized on a predominantly national basis. Colonial
empires were not yet part of an internationalized network of relations,
they merely extended the national economy into the backyard of distant
regions overseas. On the other hand, the hierarchy of corporate command
IS. Kenneth Arrow (1963) was awarded the Nobel Prize for demonstrating that a
community welfare function, which is essentially a territorial concept, cannot be
logically derived from the assumptions of the reigning economic paradigm which
ascribed fundamental reality to individual utilities.
38 CITY AND COUNTRYSIDE: CONTRADICTIONS

was by no means identical with that of territory. Economic locations, for

instance, were almost completely independent of the power of the state.
Although raison d'etat might suggest a greater degree of regional balance,
policy and incentive measures were largely impotent to guide entre-
preneurial investments into the more backward regions of the country.
(Stohr and Todtling, see Chapter 6). Not only were the instruments too
weak to affect the private calculus significantly, but the political will
was often lacking for subordinating corporate to territorial power. So long
as it was possible to privatize gains while socializing losses, one did not
have to try too hard to equalize the benefits of economic growth across
all regions. As economists were eager to point out, 'people prosperity'
should always be preferred to 'place prosperity'. If areas were bypassed
by economic progress (which is to say that they were unattractive as
potential locations for private businesses), it would be irrational, they
argued, to direct social resources from their most efficient employments
to economically unattractive regions. The obvious solution is to facilitate
the movement of people from the backward areas to major nodes of
opportunity - the growth poles of the national (and international)
economy. Alternatively, backward areas could be made accessible to
metropolitan consumers for second homes and outdoor recreation. They
would thus be brought integrally into the urban field (Friedmann and
Miller 1965). The argument was more than theoretical; it was based on
observable trends. As urban infrastructure spread across the continent
in Western Europe and North America, homogeneous economic spaces
were created that made it possible to locate production facilities almost
anywhere and still be within overnight trucking distance from major
markets.
Within the core countries of western capitalism, functional patterns of
social integration were winning out over divided territorial powers almost
everywhere. Local powers were dissolved and reintegrated at the national
level. And as the contradictions between city and countryside were slowly
overcome, the class struggle was left as the major and unchallenged
conflict within the national domain.
With the substantial victory of corporate power, older social distinctions
between rural and urban areas all but vanished. Left was a set of spatially
continuous indicators which failed to reveal any fundamental differences
in the spatial structure of activities (Berry 1975).19
19.As a result, of this evening out of rural-urban differences, amenities are becoming
increasingly decisive for industrial location. This shift in locational criteria from
GLOBALIZATION OF THE CONFLICT 39

The annihilation of significant territorial units at sub national levels did

not, however, succeed in destroying their reality in the consciousness of
people. Reacting to the threat of social disintegration before the relentless
advance of corporate power, the advocates of the new regionalism in
western Europe are beginning to demand greater political autonomy, the
subordination of corporate to territorial interests, a reduction in the
spatial division of labor, and the creation of economically more self-
reliant regions (Gaviria 1976; Friedmann and Weaver 1978, Chapter 8).

2.5 THE GLOBALIZATION OF THE CONFLICT

The parallel and, indeed interdependent growth of corporate and national

power, extending over a period of two centuries, made it seem natural
to assume a virtual identity of interests. Evidently, there was more than a
grain of truth in Marx's and Engel's sarcastic reference to existing govern-
ments in Western Europe as the Executive Committee of the Bourgeoisie.
This presumptive identity of interests occasioned a theory of imperialism
in which the wealth of metropolitan countries in the North was gained by
subduing and exploiting colonial peoples in the southern hemisphere.
Neo-imperialism, it was argued, continued this policy, but chiefly in the
guise of unequal exchange relations (Lenin 1970; Frank 1967;Amin 1974).
What may have been true of the 19th and early 20th century, however,
when colonies were little more than distant national peripheries, is
certainly no longer true in the era of communication satellites and trans-
national corporations. A partial interdependence of interests obscures the
fundamental difference between the functional, non-territorial nature of
the corporation and the historical sense of unity which constitutes the
basis of national territorial power. 20
Until fairly recently, corporate enterprise was typically restricted to
operate within national boundaries. The costs of overcoming distance had
laid down a graded pattern of spatial efficiencies, yielding the hierarchical
resources and transport costs to social amenities (Klaassen 1968) means that
settlement patterns are becoming independent of economic location. Increasingly
in the future, jobs will follow people, and not the other way around.
20. As is well known, when faced with the task of expelling foreign invaders of their
own countries, both Stalin and Mao relied on nationalist sentiments to rally the
population in defense of their soil. It may be that the belly is the ultimate task
master; all the same, ideas and sentiments seem real enough when millions are
prepared to lay down their lives in defense of them.
40 CITY AND COUNTRYSIDE: CONTRADICTIONS

system of nodal areas through which economic space is organized. But the
pattern of spatial efficiencies which in the course of two centuries had
expanded to the size of national boundaries, drastically changed as
economic growth came to be based on science, and as transport and
communication technologies improved, shrinking first national and then
global distances to the size of a village.
This change, manifested in a radical reduction of transport and com-
munication costs, was reinforced by an equally dramatic shift in relative
emphasis from fixed plant and equipment to finance capital. This de-
materialization of the forces of production facilitated the merging of
financial resources from different points of origin ('money doesn't smell')
and, in a quick progression, led to the de-nationalization of corporate
power.
The real power of finance capital resides in its global mobility. When
profits on investment are as high as 40 percent a year, as in many Third
World countries, investments can be written off in less than half a decade
without loss, and the recovered capital transferred to any part of the globe
that happens to be temporarily more attractive. 21 Cheap, almost instan-
taneous communication, and now even supersonic jet travel, has made
finance capital virtually independent of location. As a result, very large
cities - Tokyo, Los Angeles, Mexico City, Sao Paulo, New York, London
- have become primary centers within a global interlocking system of
banking and communication. But as financial centers, their world-wide
dominance has been 'etheralized'; it is no longer territorial.
The occasional alliance between finance capital and territorial power is
not a necessary but an historical relation. To a very large extent, finance
capital has become independent of territorial power. 22 This is true for
even leading capitalist countries, a development which is only now begin-
21. In a recent year, for example, employment in Mexico's border industries (i.e.
American firms operating in Mexico under a special arrangement which permits
them to take advantage of low-cost labor), fell from 80,000 to 30,000. The wages
of Mexican workers had risen, and Haiti had once more become politically attrac-
tive. Thus, in the early 1970s, Haiti, where women work for 40 cents a day, experi-
enced an enormous industrial boom whose counterpart was economic depression
in northern Mexico! (See Dillman 1976; Biderman 1974.)
22.According to Holland (1975, p. 368): 'The new imperialism is different. The parent
country has given way to the parent company, which rarely has any interest in
developing the infrastructure or strengthening the administration of the country
in which its subsidiaries operate. It leaves the social costs of running some of the
poorest countries in the world to local governments while maximizing the private
henefits of the multi-national corporation which it heads.'
GLOBALIZATION OF THE CONFLICT 41

ning to be noted. A functionally integrated world economy no longer

requires a well-defined geographic center as an element of its structure. 23
Or, as Melvin Webber (1964) put it in a celebrated phrase, it has become a
'non-place urban realm'. The lodestone of this realm, its constant North,
is corporate profits, a criterion that includes as well the safeguarding of
essential raw materials and the expansion and protection of markets. It is
not geared to any national or local interest. 24
The global division of labor is organized through a system of point
locations and connecting flows of commodities, information, and financial
instruments. This spatial dimension of corporate expansion is accomplished
with the assistance of quasi-territorial institutions, such as the European
Economic Community and non-territorial organizations, including the
23. The following quotation from the Chairman of Citicorp demonstrates fairly clearly
the thinking which stands behind my assertion of the loss of a geographic center for
the emerging world economy:
'As a last resort, all the multinational company can do in its relations with a sov-
ereign state is to make an appeal to reason. If this fails, capital, both human and
material, will leave for countries where it is more welcome. Whether or not there is
a shortage of capital is the subject of debate, but no one asserts there is a surplus.
Since men and money will in the long run go where they are wanted and stay where
they are well treated, capital can be attracted but not driven.
In the long run, it all comes down to this: the figure of the global company in
anyone area will be determined by the degree to which a particular government is
willing and able to sacrifice the material well-being of its citizens to non-economic
factors. Everything we've discussed thus far will be resolved almost automatically
when our nation-states make up their minds concerning this one basic question.
The reality of a global market place has been the driving force pushing us along
the path of developing a rational world economy. Progress that has been made owes
almost nothing to the political imagination .... Far too many of the world's people
have now seen what the global shopping center holds in store for them. They will
not easily accept having the doors slammed shut by nationalism' (Wriston 1976).
These remarks were meant for British ears, and they were pointed at British pol-
icy! Now Citicorp is based in New York City, and it is possible to argue that what is
true for the rest of the world is not true for the United States, where territorial and
corporate interests are largely one and the same. I grant that there exists a certain
historical coincidence between them, but it is not a structural phenomenon. Citi-
corp's location in New York is a matter of relative convenience. If New York
should ever become incapable of being governed, Citicorp would pack its tapes
and books and shift its headquarters to London or some other hospitable city
where essential banking and communication services are available.
24. This statement does not preclude the possibility of a temporary conjunction of
corporate and national (territorial) interests. It is these conjunctions that lend
credence to expressions such as 'American imperialism' (see, e.g., Joint Economic
Committee 1977). I wish to insist, however, that such conjunctions are not struc-
tural; they are ephemeral as the wind.
42 CITY AND COUNTRYSIDE: CONTRADICTIONS

International Monetary Fund, the World Bank, and the Organization for
Economic Cooperation and Development. Major world capitals play a
coordinative role, but their choice as a location for corporate staff is largely
a matter of historical accident and convenience (Heenan 1977). London is
a major financial market; it also happens to have the cheapest secretarial
labor force in Western Europe. A step or two further down the financial
hierarchy, is Beirut. But when Beirut was engulfed in civil war, much of
the financial capital moved to Amman. And if Amman should capsize,
then there is Kuwait or some other principality along the Arab Gulf. Let
it be noted: the first route of British supersonic jets was to Bahrain!
What is happening today is clearly the attempt by corporations to
replicate their earlier national phase on the scale of the whole globe. As a
consequence, national territories in the world periphery are penetrated at
suitable points, and these points are gradually incorporated and absorbed
into the whole global network of economic relations (Michalet 1975). The
earlier effort had successfully wiped out most local territorial entities as
effective political actors. It had reconcentrated power at the national
level, and it had realigned people's lives according to functional principles.
Today, the denationalized corporation seeks to achieve the same purpose
by weakening the integrity of nation states, undermining their power and
authority to act. Its long-term project is to totalize the principle of
functional integration and to resolve all social questions by the calculus
of profit. As Citicorp's Walter B. Wriston said: 'Far too many of the world's
people have now seen what the global shopping center holds in store for
them. They will not easily accept having the doors slammed shut by
nationalism. 25
Or will they? That is indeed the fateful question of our time. Will
territorial power, after all, succumb to corporate and bureaucratic power,
history to economic growth? Should we imagine the globalization of the
economy to the exclusion of the territorial principle? Or, more precisely,

25, See Note 23. Methods of conducting business such as those revealed by the Lock-
heed scandals of recent memory and involving pay-offs to heads of state and other
top-ranking government officials are only a tiny glimmering of what must be
extremely widespread practices. The Vice-President of the United States had to
resign for accepting bribes, and former Prime Minister Tanaka is standing trial in
Japan on similar charges. As Quisling of Norway sold out to Nazi Germany, so
Agnew and Tanaka have sold out to corporate interests. Is there a difference?
(Quisling was shot after the war as a collaborator. Agnew has become a successful
author-businessman. Tanaka went to jail.)
GLOBALIZATION OF THE CONFLICT 43

should we imagine a world in which the principle of territorial integration

has become the pliant tool of corporate power?
The latter course is conceivable, even though it may not be desirable,
but the former, involving the totalization of the principle of functional
organization, is not. In a unity of opposites, one aspect cannot be
'totalized' without destroying the very unity which allows it to exist:
to be integrated implies to be perpetually engaged in struggle.
The real question is rather the terms of that struggle. Which force is to
be dominant? Up to now, functional forces have been nearly everywhere
victorious. Yet, the territorial principle is far from being down and out:
its maintenance has proved to be essential to the hegemonic rise of
corporate power. It was needed for the general maintenance of economic
and political stability; for the provision of economic services not profit-
able on private account; for the mobilization of investment capital over
and above the capacity of the private sector; for the management of the
'social costs' of private enterprise, including pollution, resource degrada-
tion, unemployment, inflation, and income inequality; for the reproduction
of the labor force through the provision of minimum essential housing,
health, and education; and for the penetration and control of peripheral
economies.
The managers of the global system are quite aware of their continued
need for some degree of territorial integration at the level of the nation-
state (Peccei, 1969). But in accepting this need, they are also obliged to
accept the basic contradictions that accompany territorial integration: its
potential conflict with all functionally integrated structures.
The danger is real and ever-present: that functional forces will under-
mine their own stability and power by dissolving territorial integrity. With
rapidly mounting social costs, and tensions within national social forma-
tions rising to nearly intolerable levels of intensity, this danger has never
been so great as now. It is the threatened breakdown of the world economy
that foreshadows a possible reconstruction of the social order along
communalist lines and a concomitant recovery and strengthening of
territorial values.
44 REFERENCES

REFERENCES
Althusser, L.,For Marx. Pantheon, New York, 1969.
Amin, S., Accumulation on a world scale: a critique of the theory of underdevelopment. 2 vols.
Monthly Review Press, New York, 1974.
Arrow, K.,Social choice and individual values. 2nd ed. Wiley, New York, 1963.
Asalto al centralismo: la reivindicacion de la autonomw en las nacionalidades y regiones del estado
Espaiiol. Avabce/Intervenci6n, Madrid, 1976.
Berry, B.J.L., Metropolitan area definition; a re-evaluation of concept and statistical practice.
Rev. ed. Bureau of the Census Working Paper no. 28, Washington, D.C., 1969.
Berry, B.J.L., 'The geography of the United States in the year 2000', in J. Friedmann and W.
Alonso (eds.), Regional policy: readings in theory and applications. MIT Press, Cambridge,
Mass., 1975.
Biderman, J., Enclave development: the case of multinational assembly industries on Mexico's
northern border. Master's thesis, Department of City and Regional Planning, University of
California, Berkeley, 1974.
Borts, GR. and J.1. Stein, Economic growth in a free market. Columbia University Press, New
York,1964.
Carroll, J.D., Jr., 'Spatial interaction and the urban-metropolitan regional description'. Papers
and Proceedings, Regional Science Association, vol. 1, DI-14, 1955.
Colletti, 1., 'Marxism and the dialectic', New Left Review, no. 93,3-30,1975.
Conot, R., American odyssey: a unique history of America told through the life of a great city.
William Morrow, New York, 1974.
de Chardin, P.T., The phenomenon of man. Harper and Row, New York, 1965.
Dillman, C.D., 'Maquiladoras in Mexico's northern border communities and the border industrali-
zation program', Tijdschrift voor Econ. en Soc. Geografie, vol. 67,138-50,1976.
Duchacek, I.D., Comparative federalism: the territorial dimension of politics. Holt, Rinehart and
Winston, New York, 1970.
Dulong, R., La question Bretonne. Presses de la Fondation Nationale des Sciences Politiques, Paris,
1975.
Dulong, R., La question regionale en France. 1 Monographies, Ecole des Hautes Etudes en Sciences
Sociales, Centre d'Etudes des Mouvement Sociaux, Paris, n.d.
Dunn, E.S., Jr., Economic and social development: a process of social learning. The Johns Hop-
kins Press, Baltimore, 1971.
Engels, F., Anti-Diihring. Foreign Languages Press, Peking, 1976.
Finley, M.I.,Democracy ancient and modern. Rutgers University Press, New Brunswick, N.J., 1973.
Frank, A.G., Capitalism and underdevelopment in Latin America: historical studies of Chile and
Brazil. Monthly Review Press, New York, 1967.
Friedmann, J. and J. Miller, 'The urban fields', Journal of the American Institute of Planners, vol.
31,312-320,1965.
Friedmann, J. and C. Weaver, Temtory and function: the evolution of regional planning doctrine.
Edward Arnold, London, 1978.
Gottman, J., Megapolis: the urbanized northeastern seaboard of the United States. Twentieth
Century Fund, New York, 1961.
Gaviria, M.,Ecologismo y Ordenacwn del Temtorio en Espana. Cuadernos para el Diagolo, Madrid,
1976.
Granet, M., Chinese civilization. Meridian Books, New York, 1958.
Hechter, M., Internal colonialism: the Celtic fringe in British national development, 1536-1966.
Routledge and Kegan Paul, London, 1975.
Heenan, D.A., 'Global cities of tomorrow', Harvard Business Review, vol. 5, 79-92, 1977.
Holborn, H., The political collapse of Europe. Knopf, New York, 1954.
Holland, S., The socialist challenge. Quartet Books, London, 1975.
Isard, W., Location and space economy: a general theory relating to industrial areas, land use,
trade, and urban structure. Wiley, New York, 1956.
Joint Economic Committee, Congress of the United States, The United States response to the new
REFERENCES 45
international order: the economic implications for Latin America and the United States. U.S.
Government Printing Office, Washington, D.C., 1977.
Klaassen, 1.H., Social amenities in aret1 economic growth: an analysis of methods for defining
needs. OECD,Paris, 1968.
Kollek, T. and M. Pearlman, Jerusalem: sacred city of mankind: a history of forty centuries.
Weidenfeld and Nicholson, London, 1968.
Lapidus, I.M., Muslim cities in the later middle ages. Harvard University Press, Cambridge, Mass.,
1967.
Lenin, V.I.,Imperialism: the highest stage of capitalism. Foreign Languages Press, Peking, 1970.
Lynch, K.,Managing the sense of a region. MIT Press, Cambridge, Mass., 1976.
Mao Tse-tung, Four essays on philosophy. Foreign Languages Press, Peking, 1968.
Mariategui, J.C., Seven interpretive essays on Peruvian Reality. University of Texas Press, Austin,
1971.
Marx, K., Critique of the Gotha Program. Foreign Languages Press, Peking, 1972.
Marx, K. and F. Engels, Manifesto of the communist parry. Foreign Languages Press, Peking, 1975.
Marx, K. and F. Engels, The German IdeokJgy. Part One.C.J. Arthur (ed.},International Publishers,
New York, 1970.
Mera, K., Income distribution and regional development. University of Tokyo Press, Tokyo, 1975.
Michalet, CA., Les firmes multinationales et la nouvelle division internationale du travail. Inter-
national Labour Office, Geneva, 1975.
Morales,AM., 'Consideracionesgenerales sobre la planificaci6n urbano-regional en America Latina',
Revista Interamericana de Planificaci6n, vol. 10,78-91,1976.
Mumford, L., The city in history: its origins, its transformations, and its prospects. Harcourt, Brace
and World, New York, 1961.
Mumford, 1., The culture of cities. Harcourt, Brance, New York, 1938.
OIlman, B., Alienation: Marx's concept of man in capitalist society. University Press, Cambridge,
1971.
Peccei, A., The chasm ahead. Collier-Macmillan, London, 1969.
Porteous, J.D., 'Home: the territorial core', The Geographical Review, vol. 66, 383-390, 1976.
Pounds, N.J.G., 'The urbanization of the classical world', Annals of the Association of American
Geographers, vol. 59, no. 1,1969.
Schramm (ed.},Mao Tse·tung unrehet1rsed: talks and letters, 1956-71. Penguin, London, 1974.
Schumpeter, J.A., Capitalism, socialism and democracy. 2nd ed. Harper and Bros., New York,
1947.
Seymour, J.D., China: the politics of revolutionary reintegration. Crowell, New York, 1976.
Soja, E.W., 'Communication and territorial integration in East Africa: an introduction to transac-
tion flow analysis', East Lake Geographer, vol. 4, 39-57, 1968a.
Soja, E.W., The geography of modernization in Kenya. Syracuse University Press, Syracuse, N.Y.,
1968b.
Spodek, H., 'Urban politics in the local kingdoms of India: a view from the princely capitals of
Saurashtra under British rule', Modem Asian Studies, vol. 7, 253-275, 1973.
Tilly, C., 'Major forms of collective action in Western Europe, 1500-1975', Theory and Society,
vol. 3, 365-376, 1976.
United Nations, Measures for the economic devekJpment of underdevekJped countries. Report by
a group of experts. Department of Economic Affairs, New York, 1951.
Weber, M., The city. Free Press, Glencoe, 111.,1958.
Webber, MM., 'The urban place and the nonplace urban realm', in M.M. Webber et al.,Explorations
into urban structure. University of Pennsylvania Press, Philadelphia, 1964.
Wilhelm, R. (ed.), The I Ching, or book of changes. 3rd ed. Routledge and Kegan Paul, London,
1968.
Williams, R.,Keywords: a vocabulary of culture and society. Fontana, London, 1976.
Wriston, W.B., People, politics and productivity: the world corporation in the 1980s. Citicorp,
London, 1976.
3. Environmental protection and spatial
welfare patterns
WIM HAFKAMP and PETER NIJKAMP*

The problem of an efficient allocation of scarce resources has been studied

quite extensively in economics. One of the basic and often unsolved ques-
tions in theoretical analyses, however, is the 'efficiency-equity' dilemma.
This dilemma implies that an optimal use of resources does not necessarily
guarantee a social equity. The existence of significant structural welfare
discrepancies indicates that this dilemma is also hard to solve in reality.
These discrepancies are also reflected in spatial inequalities.
Now the question arises whether a relationship exists between spatial
welfare inequalities and the quality of the physical environment in which
people live. The extensive literature in this field does not reflect a
common view on this problem. Instead, two main streams based on
opposing views may be distinguished.
A first approach takes for granted the higher efficiency of capital and
labour in more productive regions (for example, due to favourable loca-
tional conditions). As a result the income of people in these areas will be
higher, while on the other hand the quality of their physical environment
will be lower. Consequently, the conclusion can be drawn that wealthier
people live in more polluted areas.
The alternative view states that poor people are not capable of choosing
a residential location with a satisfactory environmental quality due to the
higher price of these locations (see among others Enzensberger 1973;
Harvey 1970). Consequently, the conclusion is drawn that poorer people
live in areas with a relatively low environmental quality.
It should be noted that these opposing views do not exclude each other.
Any statement on the relationship between regional welfare, spatial
inequality and environmental quality is co-determined by the spatial scale
of analysis. The first view expressed above may hold true for a global
regional subdivision, whereas the latter view may be relevant for a detailed

* University of Amsterdam and Free University, Amsterdam, respectively.

48 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

intra-urban subdivision. Hence, a priori no unifonnly valid statement on

this relationship may be drawn. Instead, it may be worthwhile to
construct a more general framework for studying these relationships,
viz. the existence of externalities.
Our economic world is a world of externalities. There are numerous
human activities which affect (directly or indirectly) the welfare position
of other human beings without charge (in the case of costs) or without
reward (in the case of benefits). Since the time of Marshall externalities
have been mainly considered as abnonnal elements in an otherwise
rational market system. During the post-war period, however, techno-
logical progress and innovation as well as the socioeconomic developments
involved have led to a technical and social structuring of our society
characterized by a high degree of interdependence among all members.
This interwoven pattern of all social, economic and technical decisions
and activities has constituted the fonnal background for the occurrence
of many external effects.
In a spatial setting the existence of external effects is also quite evident,
because a spatial system is nonnally an open system with a wide variety
of spatial spill-over effects (cf. also Paelinck and Nijkamp 1976). It is not
far from reality to state that space is the medium through which externali-
ties are transferred from one economic agent to another, as is apparent
from environmental pollution, noise nuisance, congestion and the like.
Spatial externalities can in general terms be described as unpriced impacts
of an activity at a certain place upon individuals or groups, at a different
place, who are not directly involved in that activity (Nijkamp and Rietveld
1977). These externalities arise from the fact that not all socially relevant
spatial consequences of an activity are reflected in market transactions
(cf. Dear 1977). Consequently, in a situation of externalities a competitive
private enterprise economy does not guarantee an efficient allocation of
resources (see the traditional analysis of Pigou 1930; or the more modern
analyses of Baumol and Oates 1975; Maler 1974).
Although it has to be admitted that the inadequate working of the price
system in the case of externalities may preclude a socially optimum alloca-
tion of resources, it should be added that the inadequate working of this
price system is mainly due to the fact that the economic decisions of
individuals and groups do not take into account the unpriced (negative or
positive) impacts involved: economics is the science of human decisions,
and uncompensated impacts are a result of human decisions. More explicit
attention to these impacts (reflected among others in the parameters of a
ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE 49

welfare function) would certainly lead to more balanced decisions and to

a lesser occurrence of uncompensated impacts.
In general, spatial externalities show a declining intensity from a certain
source in space onward. This distance-decay pattern of externalities bears
a close resemblance to diffusion phenomena (cf. Hordijk and Nijkamp
1977) and leads to the notion of spatial externality fields (cf. Harvey
1973). In a sense these fields represent the spatial distributive impacts of
unpriced spill-overs.
The neglect of uncompensated spatial impacts has led to unbalanced
spatial developments during the dynamic post-war period. The period of
uncontrolled rapid economic growth has influenced the spatial welfare
patterns of nations. For example, many densely populated and indus-
trialized areas have shown a relatively high rise in production and income,
whereas environmental quality has seriously suffered from this growth
process. An inverse situation can be observed in many rural areas. Further-
more, even within rapidly growing regions the inequality of income has
undergone a rise (cf. Bartels and Nijkamp 1976; Stohr and TOdtling
Chapter 6). Uncontrolled growth processes appear to be clearly reflected
in spatial welfare patterns, not only at the regional level but also at the
urban level (cf. Pahl 1971 ; Vipond 1974).
The neo-classical view takes for granted that the mobility of goods and
production factors will ultimately lead to a tendency for a spatial
equalization of factor prices and commodity prices, so that efficiency
criteria and equity criteria can be reconciled (cf. Richardson 1973 and
Section 7.2). The neglect of natural or physicahegion-al advantages, the
neglect of power elements and of institutional bottlenecks, the over-
assessment of mobility and of the spread effects of economic growth, and
the neglect of externalities have, however, led to a theory with a rather
narrow scope.
Until recently, spatial externalities were often regarded as abnormal
phenomena in economic analyses. Fortunately, this myopic view of
economics has been increasingly abandoned: externalities are phenomena
inherent in the spatial economic structure of our society and they should
therefore play a crucial role in socially relevant economic analyses, espe-
cially in the field of environmental problems.
In this contribution, the policy control of externalities accruing from
environmental decay will be analyzed in more detail, especially the spatial
impacts of environmental externalities. A central role is played by
so-called spatial welfare profiles in a multi-regional system. These profiles
50 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

constitute the basis for an integrated welfare and policy framework

between conflicting priorities, in particular economic growth and environ-
mental protection. After an analysis of discrepancies between spatial
patterns in an economic-ecological sense, attention will be focused on a
spatial equilibrium analysis by means of multiple diverging welfare criteria.
In this respect, this contribution proposes a theoretical Paretian frame-
work as the basis for an integrated socioeconomic and environmental
analysis of a multi-regional system. This equilibrium analysis serves to
describe the conditions under which a multi-regional equilibrium may be
obtained, given a set of conflicting priorities within this system (such as
maximization of consumption and pollution abatement). Particular atten-
tion will be paid to public controls (such as charges and subsidies, either
centralized or decentralized) in this multi-regional system in order to
achieve a satisfactory level of environmental quality. The formulation of
optimal multi-regional policies will be facilitated by using the notion of
Pareto-optimality and some of its modem variants.
The last part of this chapter concentrates on an empirical illustration by
means of a two-region model. The first aim of the model is to demonstrate
how the notion of a regional welfare profile can be used in a multi-
regional optimization framework. The second aim is to investigate how far
the present state of regional economic and environmental model-building
is capable of dealing with the theoretical notions of multi-regional equili-
brium and Pareto-optimality. This chapter concludes with an outline of
further research to operationalize the abovementioned equilibrium
notions in a dynamic setting.

3.1 SPATIAL WELFARE PROFILES

The evolution of a system of regions can be described in various ways. A
traditional measure for the economic state of a spatial system is income
per capita. This measure may be a reasonable criterion for welfare in the
case of a perfect competitive system characterized by full information and
a fully operating price system, but it still neglects many essential elements
of human life such as residential living conditions and the quality of
working life (see also Hueting 1974). The increased attention to the
quantitative and qualitative aspects of human well-being and the call for a
greater responsibility of man for his surroundings (e.g. environmental
management and resource management), have led to a need for a broader
view of welfare or well-being (including also non-monetary or qualitative
elements).
SPATIAL WELFARE PROFILES 51

This need runs parallel to the tendency to describe social well-being by

means of a set of social indicators (cf. Bauer 1966; Perloff 1969; and
Smith 1973). Many spatial impact analyses (for example, the well-known
Environmental Impact Statements) are to a certain extent based on the
'indicator movement'. A basic problem in the use of these indicators is
formed by the lack of systematics and coherence among the various
indicators. A significant advantage of this approach, however, is the fact
that many aspects of human welfare falling outside the realm of the price
mechanism (such as externalities and collective goods) can be taken into
account.
In our opinion, the state of a spatial system can be represented in an
adequate manner by means of a multidimensional profile vector M' encom-
passing the values of all relevant (socio)economic, spatial and environ-
mental variables as elements. In many traditional analyses M' was assumed
to include only (socio)economic variables such as average income,
employment, investments, etc. Sin~e the insight has grown that many
aspects of human welfare are to be related to a broader set of state
variables (such as availability of public facilities, urbanisation rate, access-
ibility, environmental quality, congestion, etc.), the attention of welfare
economists, regional scientists and planners has increasingly focused on a
more integrated view of regional and urban policy-making (cf. Van Delft
and Nijkamp 1977; Klaassen and Botterweg 1976; Nijkamp 1977).
For a systematic analysis of the wide variety of welfare components it
may be meaningful to subdivide M' into a series of sub-profIles:

WE the economic profIle encompassing income, growth, investments,

differentiation of economic structure, consumption, etc.;
Ws the social profIle including the quantitative and qualitative state of
the labour market, degree of unionization , degree of social stability,
social public facilities, etc.;
Wp: the physical-spatial profile related to location patterns, residential
land use, industrial land use, volume of transport networks, etc.;
WM: the environmental profIle including ecological quality characteris-
tics, pollution, quantity of natural areas, congestion, etc..

Given a spatial system composed of R regions r (r = 1, ... ,R), the state

of such a multi-regional system can be represented. by means of a state
matrix If (see (1.1)).
52 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

1 .................. . R

!Ys
( 1.1)
W=

Each row of (1.1) represents a multi-regional cross-section of one welfare

component, while each column of (1.1) represents a regional welfare
profile. The elements of W may be measured in cardinal, ordinal or
qualitative units, but here the assumption is made that W encompasses
cardinal values.
The dynamics of a spatial system can be represented by an extension
of (1.1) with the time dimension t (t = 0, . . . 1) leading to a spatio-
temporal block for a welfare profile (see Figure 1).

I
I
I
I

,, .-L-
!Yp , .... "
.... "
!!1M -,-,,'
1 ...... r .. .... . R
Figure 1. A spatiotemporal welfare proflle.

Matrix (1.1) and Figure 1 constitute the foundation stones for measuring
discrepancies in welfare profiles between regions as well as the inter-
temporal evolution in these discrepancies. Two aspects of measuring
SPATIAL WELFARE PROFILES 53

spatial discrepancies deserve broader attention, viz. the coherence between

the elements of the multi-regional welfare profile and the choice of an
appropriate distance metric for spatial (or spatiotemporal) discrepancies.
Suppose a system with 2 regions characterized by the welfare profiles
11'1 and 11'2. Then a simple (linear) model describing the spatial inter-
dependence may be:

0.2)

where ~ is a vector with exogenous variables. The semi-reduced form

structure of region I (assuming that region I does not influence region 2) is:

( 1.3)

or:

(1.4)

The structure of C is worth further attention. If C is block-diagonal

(according to the main division of welfare profiles 11'E' 11's' 1:l'p and 1:l'M)'
there is mutual independence between the 4 welfare profiles at the inter-
regional scale and mutual interdependence at the interregional scale
within each welfare profile. Matrix C is block-diagonal when both A and
11 are block-diagonal matrices subdivided according to the same blocks
associated with the abovementioned welfare profiles.
When" is a block-triangular matrix according to the following structure:

(1.5)
C=

then the economic profile of region 2 dominates that of region I with

respect to all other profiles, etc. In a sense such a spatial dominance bears
54 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

much resemblance to the Von Neumann-Morgenstern game-theoretic

concept of dominance (see Davis and Whinston 1962).
In many cases, however, there is a complete spatial interdependence
among the welfare proflles, so that these cases are characterized by mutual
spill-over effects and mutual externalities.
In addition to the intraregional and interregional coherence some atten-
tion should be paid to the measurement of spatial discrepancies. The
regional welfare components are measured in different dimensions
(money, quantities, indices, etc.). A prerequisite for determining spatial
discrepancies is a standardization of the spatial welfare proflle matrix W.
When the typical elements of Ware denoted by Wir (i = 1, ... ,I and r =
1, ... R), a meaningful standardization may be:

, if i is a benefit criterion,
0.6)
W ir
1-- , if i is a cost criterion,
w:t'
1

where wi * is an 'ideal' value or achievement value, which may be cal-

culated as:

w:t'
1
= max
r
W
ir
(1.7)

Relationship (1.6) guarantees that 0 .;;; Vir';;; 1, so that this performance

index is higher as w ir approaches its ideal or achievement value.
Next, the average spatial inequality of the welfare components of two
regions rand r' can be calculated by means of the following (unweighted)
Minkowski p-metric:
I
srr , = (~ (Vir - vir ''f)l/P ,p;;;' 1 ( 1.8)
1= 1

It is clear that this Euclidean distance and the rectangular distance are
special cases of (1.8), viz. if p = 2 and p = 1, respectively. Further details
on the use of the Minkowski p-metric are contained, among other sources,
in Bartels and Nijkamp (1976), Beckenbach and Bellman (1961), Green
and Cannone (1970), Nijkamp and Rietveld (1977) and Shephard (1966).
For a discussion on a weighted welfare metric, see Nijkamp and Rietveld
(1977).
ENVIRONMENTAL EXTERNALITIES FRAMEWORK 55

The Minkowski p-metric can also be used to measure intertemporal

changes in the welfare components of one individual region. In a similar
way one may measure the welfare distance of each region separately
with respect to the overall ideal value of the spatial system.
The foregoing distance analysis can also be used to measure spatial
discrepancies for each welfare profIle separately, so that the discrepancies
between every two regions can be represented by means of four distance
indices fjr' , sfr' , I,'r' , and ~, . Finally, the unweighed total welfare
index vr of each region can also be represented by means of a distance
metric, viz.:
I
= (l: v. p )l/p (1.9)
i= 1 IT

Clearly, the latter index can again by subdivided into the four above-
mentioned welfare profIles, so that the economic welfare index of region
r can be denoted by vEr' the environmental welfare index by vMr ' etc. In
the following sections the use of these welfare indices in an interregional
equilibrium system will be exposed in more detail.

3.2 ENVIRONMENTAL EXTERNALITIES: A FORMAL FRAMEWORK

Environmental externalities are non-market effects (positive or negative)

which result as a side-effect of economic activities of producers and
consumers (including governmental agencies) and which affect welfare
or profit conditions of other economic agents through uncompensated
spill-overs for the physical surroundings of these agents (cf. Nijkamp
1977). Traditionally, the welfare-theoretic Pigouvian solution to the
externalities problem is to measure in pecuniary units the net social
benefits (i.e. benefits minus costs accruing from the externalities) and to
impose a subsidy or charge of an equal amount on the source of the
spill-over effects. Alternatively, the solution of the 'new welfare econom-
ists' is to use the Hicks-Kaldor compensation scheme so as to compensate
the victim for the decline in his welfare or utility.
Clearly, these approaches require much information, especially as far
as the monetary assessment of externalities is concerned: the translation
of a wide variety of welfare components (including social and environ-
mental variables) into a one-dimensional 'measuring rod of money' is
fraught with difficulties. A system of standards (regulations, prohibitions,
56 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

etc.) is at first glance somewhat easier, but such a system suffers basically
from the same problem, because here also the welfare-theoretic implica-
tions of fixing a set of standards are to be assessed (Nijkamp 1977).
There is a wide variety of instruments to control environmental extern-
alities and environmental damage. Examples are changes in moral attitudes
(e.g. a call for a responsible stewardship (Nijkamp 1976) or moral coer-
cion), standards (e.g. maximum emission norms, prohibitions, rationing of
pollutants, constructioI). of public or private abatement installations),
charges (per unit of pollutant, per unit of input or per unit of output),
and subsidies (e.g. compensation for decline in welfare, production or
property values, stimulation of clean activities, stimulation of adjusted
environmental research, creation of equivalent environmental facilities).
The problem of spatial discrepancies in welfare profiles of diverse
regions will now be discussed from the point of view of environmental
externalities. Special attention will be devoted to the construction of an
interregional equilibrium model based on earlier studies of d'Aspremont
and Gerard-Varet (1976), Dr'eze and de la Vallee Poussin (1971), Malin-
vaud (1972) and Tulkens and Schoumaker (1975). For the sake of
presentation a set of simple hypothesis for the analysis of environmental
externalities in a spatial setting will be made (see also Tulkens and Schou-
maker 1975):
- the spatial system is composed of a set R =1rl r = I, ... ,R } of regions.
- R is subdiyided into a set U =1u I u = I, .. ( , U } of po1!utir:.g repons
and a set N = 1n I n = I, ... ,N } of poilu ted regions (U u N =R ,or
U+N=R).
- each region u has a production consumption set Gt = 1(xu. Pu)1 Xu > 0,
Pu > o} ,where Xu and Pu represent, respectively, the private production
and the emission of pollution of region u. This production set is a
feasible set which may be restricted by various side-conditions arising
from the economic system itself (e.g. capacity constraints, input-output
linkages, etc.).
- each region u has a (strictly quasi-concave, continuous and differen-
tiable) welfare function Wu (xu' pu) with the following properties:
w'Xu = awu(xu ,pu)/axu > ° (2.1)

and, 't:I Xu > 0,3 Pu such that:

> 0, ifpu < Pu
w~u = awu(xu ' pu)/apu I (xu' pu) = 0, ifpu = Pu (2.2)
< 0, ifpu > Pu
ENVIRONMENTAL EXTERNALITIES FRAMEWORK 57

Conditions (2.1) and (2.2) require somewhat more explanation. First, it

should be noted that these conditions are based on the most general
feasible relationships between Pu and xu' viz. Pu ~ 0 and Xu ~ O. This
implies in fact an unspecified general technological relationship between
pollution and production (a so-called multi-function) 1
In the second place, it is clear that the elements of the regional welfare
profile, Xu and Pu' are the arguments of the welfare function. Conditions
(2.1) and (2.2) imply, inter alia, a preference mapping of the production
consumption set. This means that the assumption is implicitly made that
the production Xn can be divided among consumption and abatement
investments. The consumption part of Xn exerts a positive influence on
welfare. Consequently, condition (2.1) presupposes non-satiation with
regard to private commodities: the more the better.
Condition (2.2) presupposes the existence of a critical threshold level
Pu for environmental pollution. Such a threshold level for pollution is
based on the assumption that for each production level a feasible pollu-
tion level can be identified that maximizes welfare. It should be noted
that the satiation level Pu corresponds to a situation of free pollution
For ease of exposition, a negative impact of pollution on the welfare of
the polluter is excluded by assumption, except for the case that a pollu-
tion abatement program has to be carried out as a result of public policy.
In that case (Pu < Pu) the welfare of the polluting regions will undergo a
decline due to the higher abatement costs, paid for by lower consumption.
However, after an abatement program the welfare of polluted regions
will increase (see (2.4) ). On the other hand, it is plausible to assume that
given a level xu, any rise in pollution beyond the threshold level Pu
implies inefficient production activities because there is no reason to
pollute more than would correspond to the normal pollution level. This
would also lead to a decline in welfare of the polluting regions.
Furthermore:
- each region u has a choice set Su (i.e. a set of regional decision variables)
which is equal to the production consumption set Cu ' i.e. Su = Cu'
- each region n has a production consumption set Cn = { (xn , zn)1 xn > 0,
zn « O}, where zn represents the immission (i.e. input) of pollutants in
region n.

1. Clearly, the specification of a poilu tion emission model is a special case of such a
multi-function.
58 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

- each region n has a (strictly quasi-concave, continuous and differentiable)

welfare function wn (xn • zn) with the following properties:

w'
xn
= aWn (xn • zn)/axn >0 (2.3)

and

w' = aWn (xn • zn )/azn~ 0 (2.4)

zn
The assumption is made that the welfare functions of polluting regions do
not include environmental quality Zn of surrounding regions n as argu-
ments; in other words, the set -0 of polluting regions does not care about
the environmental decay of the set of polluted regions N.
- each region n has the following choice set Sn = {xn I xn > O}.
- the spatial system" has a fixed total production capacity x* (> 0), so
that any spatial allocation of commodities should satisfy the additivity
condition:
U N
~ X +~ x = x* (2.5)
u=l u n=l n

- the spatial system has no initial stocks of pollutants, so that initially

the following conditions are valid:
pu = 0, V u
(2.6)
Zn = O,'Vn
- environmental quality Z is uniformly dispersed or at least perceived as
such through all polluted regions and made up of minus the sum total
of all emitted pollutants from all polluting regions;
U
Z = - ~ pu (2.7)
u=l

where the assumption is made that environmental quality is the same in

all regions, i.e.:

(2.8)

The latter assumption can easily be relaxed on the basis of more informa-
tion about spatial diffusion patterns and environmental damages (see
Section 3.4).
ENVIRONMENTAL EXTERNALITIES FRAMEWORK 59

In a sense, elements Xu and xn can be regarded as items representing an

economic profile index of regions u and n. Similarly, Pu and zn can be
regarded as an environmental welfare index related to the environmental
profile of regions u and n. The choice set of regions u E jj and n EN can
be extended in a straightforward manner with multiple economic and
environmental indicators, but for ease of presentation this extension will
not be carried out here for the moment.
Commodities Xu and xn are considered as necessities (hence Xu > 0
and xn > 0); Z (~ 0) is essentially an index for environmental quality (the
higher the better) and, given condition (2.8), it can be regarded as a public
good (cf. Milleron 1972; Samuelson 1954) because the 'use' of environ-
mental quality by the one region does not exclude its availability for the
other region. Since Z may be equal to zero, condition (2.4) may be an
equality (see Baumol and Oates 1975).
The foregoing remarks on preference mappings can be illustrated by
means of the Figures 2, 3 and 4 which show indifference curves for the
sets of regions 0', Nand [; u N.

Figure 2. Indifference CUlves of Wu (xu' pu)' \;f u.

Figure 2 represents a set of indifference curves between Xu and Pu of a
polluting region u; the shape of this curve represents the satiation pheno-
menon of Pu ' so that a certain non-extreme level of utility can be achieved
with two different values of pu. Clearly, an optimum utility level and a
related pollution level Pu corresponds to each value of Xu (cf. point Pu
corresponding to xu; see also condition (2.2) ).
60 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

Figure 3 depicts a set of indifference curves between xn and zn for

polluted regions. Condition (2.4) which includes a situation of a maximum
environmental quality (i.e. zn = 0) implies that the tangent to each
indifference curve in all points (xn = 0) runs parallel to the zn -axis.

Figure 3. Indifference curves of wn (x n ' zn)' '\;f n.

Finally Figure 4 integrates Figures 2 and 3 into an open Edgeworth

box. If there are only two regions u and n, then the length of the axis
between the two origins should be equal to x*; it is obvious that also Pu =
- zn = - z.
Z

Figure 4. Edgeworth box for (x ,p ) and w (x'

W
U U u n n' Z n ) .
ENVIRONMENTAL EXTERNALITIES FRAMEWORK 61

It is clear that the abovementioned hypotheses are rather restrictive.

On the other hand, many hypotheses can be generalized straightforwardly,
although this will complicate the analysis. A situation, however, of reci-
procal environmental externalities, of multiple commodities, sectors and
pollutants, and of explicitly defined production functions will be taken
into consideration later. For the moment, attention will be focused on a
tax-charge system to arrive at an interregional equilibrium.
Before· deriving the conditions for interregional equilibrium in the
following section some states of the interregional system should be
defined.
A feasible state of the abovementioned spatial system is a (U+N)-tuple
of the following vectors {(xu' pu) , (xn , zn)}' u E U, n EN, which fulfJ1
conditions (2.5), (2.7) and (2.8). In other words, all points in the open
Edgeworth box of Figure 4 are feasible solutions.
A Pareto state is a feasible state for which no other feasible state
exists that is better than the original state for at least one region without
being worse for the other regions. Therefore, a Pareto state is any feasible
state {(x~, p~) , (x~, z~)} , for which no other feasible state {(xu', Pu'),
(x,:, z':)} exists, such that:

Vu (2.9)

and

Vn (2.10)

with a strict inequality for at least one u or n.

Clearly, line AA' ftom Figure 4 is the set of all tangency points between
the indifference curves of region u and n, and satisfies these Pareto condi-
tions. It can l?,.e proved that the maximization of welfare of- a certain
region u (u E U) affecting welfare of all other regions n (n EN) through
environmental spill-overs leads to a Pareto optimum state if and only if
this state is feasible and satisfies the following conditions (see Baumol and
Oates 1975; Malinvaud 1972):
N
w' (w'
PI4 Xu
= n=l
~ w' /w'
Z xn
Vu (2.11 )

This condition states that the marginal rate of substitution au between

pollutants and commodities in the polluting region should be equal to
62 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

the aggregated marginal rates of substitution an between environmental

quality and commodities in the polluted regions. Condition (2.11) can
also be written as:
N
au = 1: an , 't;fu (2.12)
n =1
A disagreement equilibrium of a spatial system is a (U+ N)-tup1e of the
following vectors {(xu' pu) , (xn , 2)}, u E lJ, n E N, with the following
properties:

subject to: Xu ~ Xu (2.13)

Pu ~ 0

and

(2.l4)
A U A
Z = -1: P
u= 1 u

This situation is illustrated by point B in Figure 4 for a given Xu and xn '

which might be the result of a system of pollution standards.
A disagreement state may occur in the case of free pollution, where the
polluting regions are free to choose their waste discharges independently
and where no negotiation takes place, even when the abatement costs are
lower than the benefits of a higher environmental quality in the polluted
regions. It is a typical case of unidirectional spatial externalities. Such a
dominance of the polluting regions (i.e. an absence of environmental
policy) may lead to a less optimal interregional situation. In this case, the
polluting regions are free to choose an optima11eve1 of Xu and Pu ' so that
their marginal rate of substitution will be equal to zero, i.e. au = O. The
polluted regions, however, can only adjust to the actions of regions u
(u E 0), so that their marginal rate of substitution will in general be
unequal to zero, i.e. an ~ O. Hence, normally, a disagreement equilibrium
is not a Pareto solution (unless a.. = 0); see also Figure 4.
ENVIRONMENTAL EXTERNALITIES FRAMEWORK 63

Finally, the notion of a Pigou-Baumo! equilibrium (PBE) will be dis-

cussed. Given a disagreement equilibrium, the question arises of how to
move the spatial system to a Pareto optimum. A first way of doing so is
to introduce a system of standards (for example, Pu .;;;; pu). The transition
path to the Pareto optimum and the efficient interregional allocation of
commodities is then, however, not a straightforward or unambiguous
result. Therefore, a more flexible strategy, viz. a system of charges, may
be considered (see Pigou 1930; Baumol 1972). A system of charges,
however, requires a price system as well as a redistribution mechanism.
The following new variables now have to be introduced: 7r> 0 (unit price
of commodities x), T;;;' 0 (unit effluent charge on emission of pollution
p), Y u (lump-sum transfer related to region u), and Yn (lump-sum transfer
related to region n).2
It is clear that environmental protection by means of a spatially
differentiated charge system will affect the economic welfare positions of
all regions and hence their development paths. The existence of one price
7r is based on the assumption of a real exchange model in which a perfect
market requires one price for the commodity sector. A fully operating
price system would require the introduction of a monetary sector.
Anyhow, it is clear that all charges and subsidies are to be redistributed
within the total spatial system.
A PBE is now defined as a quadruple Q = [{(xu' Pu), (xn ' z)} ; (7r , T) ;
(Yu' Yn ); (xu' xn)] with the following properties:

(2.15)

and

(2.16)
subject to: 7rXn + Yn = 7rXn

2. The income transfers have a positive sign in case of payments by region u and n,
and a negative sign in case of receipts by region u and n. The sign of these transfers
depends on the type of taxation and redistribution (e.g. charges or compensations).
64 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

Furthermore the trajectory from a disagreement equilibrium towards a

PBE should satisfy the following additivity conditions:

(2.17)

and
u P'"
1: = -z'" (2.18)
u=1 u

Evidently, the set of PBEs may also include Pareto and disagreement
equilibria.
It is clear that (2.15) through (2.18) constitute an optimization
program for U+N regions subject to constraints on regional budgets and
environmental quality. The introduction of a tax may lead to a decline in
Pu and hence to a rise in z. The redistribution of the tax proceeds will
affect the regional production distribution. However, the precise direction
of the income transfers Yu and Yn and its consequences for the commodity
sector are unknown so far. This is co-determined by the tax regime at
hand.
The way in which the redistribution takes place will be discussed later.
Suffice it to say for the moment that Yu and Yn lead to a spatial redis-
tribution of the tax proceeds, as is easily seen by adding up and combin-
ing (2.15) and (2.16):

or by using (2.17)
u '" U N
71:
u=1 u
P + 1: ~ + 1: Y
u=1 u n=1 n
=0 (2.20)

The latter result indicates that the total tax proceeds are redistributed
among all regions. Thus the charge-subsidy schemes presuppose a (hypo-
thetical) central agency.
The next section will be devoted to an analysis of the redistribution
effects on regional welfare profIles as a consequence of a policy of
environmental protection by means of a tax-subsidy scheme.
REDISTRIBUTION EFFECTS OF TAX-SUBSIDY 65

3.3 REDISTRIBUTION EFFECTS OF A TAX-SUBSIDY SCHEME

A disagreement equilibrium can be transformed into a Pareto equilibrium

along the trajectory of the series of PBEs. It is reasonable to assume that
in the case of a PBE an effective charge on pollution has to be fixed such
that the marginal rate of substitution between the production of
pollutants and commodities is equal to their respective price ratios:

au = Tlrr (3.1)
or when rr is taken as a numeraire (Le. rr = I):
(3.2)

A description of a transition process to a Pareto optimum requires a

dynamic model in order to determine an equilibrium value of T and the
consequences for the spatial allocation of other welfare elements. The
effects of alternative spatial redistribution policies may be studied as well
(see also Tulkens and Schoumaker 1975).
Substitution of (3.2) into the Pareto conditions (2.12) gives:
N
T = L a (3.3)
n=l n

which is the optimality condition for an optimal effluent charge. As long

as (3.3) is not fulfilled, the difference between T and L an can be used as
a measure of closeness with respect to the Pareto optimum.
If all preyious variables are time-varying variables depending on t, it
is plausible to fix the effluent charge in each time period such that:
• N
T = L a - T (3.4)
n=l n

where T is defined as dr/dt (notice that T = 0 at the beginning). In other

words, the change in the tax rate is equal to the total marginal damage
(measured in utility terms) over all regions minus the marginal benefits
(measured in utility terms) of the polluting regions (cf. (3.2) ). It is also
clear that during the whole transition phase:

1r = 0 (3.5)
66 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

because the private good is chosen as the numeraire. Two aspects of an

environmental protection policy may now be inv~stigated.
First the pollution charge in all regions u E U will lead to a change in
pollution emission, Pu' so that the additivity conditipn for pollution
dispersion and environmental quality becomes:
u • •
~ p =-z (3.6)
u=1 u
~learly, the change in pollution tax, fJu ' to be paid by each region u E
U is equal to d(rPu)/dt, so that we may define:

(3.7)

Second, the spatial redistribution effects of pollution taxes may be

analyzed. Two cases will be distinguished here, viz. the 'polluter pays'
principle (a) and the 'pollutee pays' principle (b).
a) The effluent charges are redistributed over all regions u E ij (as a
subsidy) and n E N (as a compensation) according to the respective
fixed allocation parameters a and f3n established by the environmental
protection agency;

(3.8)

and
• U
Y n + f3n ~ fJu
u=1
=0 (3.9)

It is clear from (2.20) that a necessary condition for a PBE solution is:
N
a+ ~
n=1
f3n = 1, 0~.Bn~1 (3.10)

b) The polluted regions n E N pay a certain amount of money in order to

stimulate polluting regions u E [j to diminish their pollution. In this
case the polluting regions receive the compensating payments from the
polluttees. This situation implies income transfers from regions n, so
that then Yn > O. If it is assumed that these payments per unit of
3 If a = 1, then y + 8u = 0, but Yn = 0, so that all tax receipts are redistributed
among polluters. Yr a =0, all effluent charges are transferred to the poilutees.
REDISTRIBUTION EFFECTS OF TAX-SUBSIDY 67

environmental quality correspond to the marginal rates of substitution

between environmental quality and commodities, the (change in)
compensating payments by region n E N is:
(3.11 )

where'Y is a control parameter indicating the proportion of compensat-

ing payments transferred by region n. If, in addition, effluent charges
are levied in all regions U E {; and if they are entirely redistributed
within {; itself, (Le. Q = I), then the income transfers related to regions
U E tJ are:

(3.12)

The dynamic process described by (3.4), (3.5), (3.6), (3.8) and (3.9) (or
(3.11) and (3.12) ) implies a continuous adjustment from a disagreement
z
equilibrium to a new equilibrium, where again T = Pu = =Yu =Yn = O.
The reason for a continuous shift is the fact that the 'budget' constraint
passing through point B in Figure 4 will rotate to the left in the case of
effluent charges (see Figure 5). In this case a shift from B to C has taken
place, which implies a transition from a disagreement equilibrium B to a
Pareto equilibrium C (at the line AA') via a series of PBEs. The location of
the Pareto optimum that is the outcome of this process depends on the
redistribution parameters Q, f3n and 'Y. Figure 5 also illustrates that Pu
declines as T increases.

Figure 5.
An Edgeworth box with a disagreement
equilibrium B and a Pareto optimum C.
68 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

It can easily be proved that the transition phase proceeds along a series
of continuous PBEs. The initial point B itself already reflects a PBE, since
it satisfies conditions (2.15) to (2.20). It is clear that the transition
process is continuous, because the utility functions are continuous
differentiable with respect to their (time-varying) arguments. Next, the
maximization of (2.15) and (2.16) subject to (2.17) and (2.18) also
guarantees that (2.15) and (2.16) are fulfIlled during the whole process, so
that the process satisfies this PBE-condition.

In addition, on the basis of (2.15), (2.16), (3.5), (3.7), (3.8) and (3.9)
one may derive:

(3.13)

and
• U
xn = f3.n ~
u= 1
()
u
(3.14)

so that by means of (3.10) the following result is obtained:

u. N.
~x+~x =0 (3.15)
u=1 u n=1 n

which proves the fulfIlment of (2.17). Finally, condition (3.6) guarantees

that condition (2.18) is satisfied, so that the whole process satisfies all
conditions of a PBE.4
The final equilibrium point of this transition process can be proved to
be a Pareto optimum satisfying (2.12) (see for a general proof Milleron
1971; Tulkens and Schoumaker 1975).
This Pareto optimum leads to a lower emission of pollution generated
by regions U E 0. Depending on the redistribution policy (Le. the values
of a, I3n and 'Y), the polluting and the polluted regions arrive at a lower or
higher utility level, a lower or higher production and so forth. Scince sev-
eral options (for example, the 'polluter pays' principle) are open for such
a combined environmental protection and spatial redistribution policy,
one may construct a table encompassing the various results, as in Table 1.

4. In the case of (3.11) and (3.12) a similar result can be obtained.

REDISTRIBUTION EFFECTS OF TAX-SUBSIDY 69

Table 1. Results of a combined environmental protection policy and spatial redistribu-

tion policy on regional welfare arguments

N
a ~ (3 'Y Wu Xu Pu wn xn z
n =1 n

0 1 0 <0 <0 <0 >0 >0 >0

The first (illustrative) row of Table 1 represents a 'polluters pay' prin-

ciple in which all tax proceeds are entirely redistributed to polluted
regions. It is clear that the distance metric discussed in (1.8) can now
be applied to measure the overall welfare discrepancy between the regions
of system fJ and N. Clearly, the welfare profiles implicit in Table 1 are
based on a rather restrictive view of the economic process, but several
generalizations will be proposed in Appendix A.
The conclusion can be drawn that under various tax regimes a Pareto
optimum can be attained. The redistribution effects of the tax proceeds
upon polluting and polluted regions depend on institutional factors (such
as the 'polluter pays' principle and the 'pollutee pays' principle), which
can be analyzed by means of the allocation parameters.
The foregoing analysis demonstrated an externalities confJict between
two regional systems fJ and N. Each region u E fj and n E iii has its own
welfare function that is to be maximized. However, an independent
maximization cannot be carried out, because system fj affects welfare
of system N through environmental externalities, as represented in
Table 2.
Table 2. A conflict matrix between welfare functions of different regions

(X~ ,p~) (x~ ,zo)

wu(x~,p~)
wu(AU ,pu)
*
u 0
wn(xn ,z) wn(x,:, - ~ p )u wn(~'zo)
u=1

The diagonal elements represent an independent welfare maximization of

wu(xu ,pu) and wn (xn ,z) by regions u E fj and n EN, respectively. These
values represent the maximum maximorum of w Uo (xu ,pu) and wn (xn ,z)
70 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

with corresponding argument values X~, p~, x~ and!o. The right upper
element of Table 2 is empty, because regional system N does not influence
system 6. Inversely, the left lower element represents welfare of regions
n E iii, when polluting regions act as dominating decision-makers. Conse-
quently, the welfare loss Lu of a region n ENdue to environmental
externalities is:

(3.16)

By means of pollution taxes and a spatial redistribution policy these

welfare losses can be influenced as exposed before.
Instead of one comprehensive regional welfare function one may also
consider its arguments as separate welfare indicators (see also Section 3.1).
In this case the objective functions of regions U E fj become:

max Xu
maxpu' ~fpu < ~} (3.17)
max -Pu' lfpu > Pu or: min I Pu - Pu I
Similarly, the objective functions of regions n EN become:

maxxn l (3.18)
maxz ~

In this case Table 2 can be transformed as in Table 3.

Table 3. A conflict matrix between different welfare elements

xO pO xO zO
u u n

xO xu(p~) * *
u
pO *
pu(x~) u *

* * *

Each element of Table 3 indicates the value of an argument on the left

side determined by an optimum value of other arguments on the top of
the table. An asterisk indicates no influence. The right-upper block shows
MODEL FOR ENVIRONMENTAL QUALITY 71

that the present problem is characterized by non-reciprocity. Equilibrium

solutions of the abovementioned conflicting priorities can be found
inter alia by means of pay-off strategies (cf. the Von Neumann dominance
concept) and compromise strategies (see Nijkamp 1977). Public policy
(by means of charges) can again be used to arrive at a more harmonious
interregional equilibrium. The effectiveness of public policy is co-deter-
mined by the power of the dominant regions.
In conclusion, regional welfare profiles and multi-regional conflict
matrices are useful tools to represent the welfare consequences of a set of
competitive regions in an environmental framework.
The foregoing analysis is based on perfect information, assuming that
a polluter will react directly on an effluent charge and a public decision-
maker will adjust directly the marginal taxes. Whether or not the effluent
charges correspond to the marginal pollution damage is hard to calculate,
so that in reality a sensitivity analysis of effluent charges should be
carried out in order to test the effectiveness of these charges. The effects
of charges on price structures of private commodities are also ignored
here, because there is only one private good. In Appendix A an attempt
has been made to make the foregoing approach more realistic by intro-
ducing multiple sectors, multiple pollutants, multiple products, produc-
tion factors, side-conditions arising from capacity constraints, etc. It
appears that several environmental principles (e.g. the 'polluter pays')
can be directly integrated in such a more general model. Obviously, the
applicability of such a model is determined by the availability of reliable
data. In the following section an attempt will be made to illustrate the
difficulties inherent in the application of the foregoing ideas via an
empirical model. The aim is to investigate how far such an equilibrium
analysis can be applied in an empirical context.

3.4 A MULTIREGIONAL EQUILIBRIUM MODEL FOR ENVIRONMENTAL

QUALITY

In this section a multiregional pollution model will be discussed. The

starting-point is an empirical model recently developed in the Netherlands.
Given the lack of reliable data, however, it is more reasonable to consider
this model as a learning model on the basis of which a set of simulation
experiments can be carried out. The results of these experiments may
clarify the mechanism and effectiveness of abatement policies; the notions
72 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

of Pareto optimality and allied concepts discussed above may also be

considered. So the theoretical framework presented in Sections 3.1-3.3
constitutes the background for the various experiments (A through E).
Experiments A through D serve merely to illustrate the use of a multi-
regional pollution model for calculating regional welfare profIles; experi-
ment E introduces a tax and redistribution system in order to relate the
multiregional model to the concepts of Sections 3.2-3.3.
The reason why this empirical model is being introduced is that it is
worthwhile examining whether the foregoing formal models may be
linked to practical pollution abatement policies. Clearly, there is a dis-
crepancy between the welfare-theoretic analysis of the previous sections
and a simple empirical optimization model. But, by placing such a model
in the framework of a Paretian welfare analysis, the analysis may shed
light on the future steps to be undertaken in further research. Particularly
experiment E of this section is an attempt to integrate the elements of
the foregoing Paretian analysis, viz. a pollution charge is established on the
basis of differences in the redistribution of tax proceeds via well-known
principles such as 'the polluter pays' and 'compensation of victims'.
The empirical model used in our analysis is a multiregional input-output
model with spatial externalities accruing from diffusion of pollutants. The
model itself is an adjusted and linearized version of a pollution model
recently developed by Coupe (1976; see also Nijkamp 1977). Two Dutch
regions are distinguished in Coupe's study, denoted here by region 1 and
region 2. These two regions are assumed to be linked to each other by
means of production relationships and environmental externalities. Two
sectors will be assumed, viz. a non-polluting sector (A) and a polluting
sector (B) (aggregated from the regional input-output tables for the
Netherlands). Five pollutants are taken into account, viz. particulates (1),
S02 (2), CO(3), NO x (4) and HF(5).
Clearly, the general welfare functions from the previous sections have
to be specified now. The regional welfare functions are assumed to encom-
pass value added and environmental pollution as arguments. Pollution
taxes per region will be introduced as a means to arrive at a multiregional
equilibrium for the spatial externalities arising from environmental
pollution. The model will be presented in a comparative static version,
while some adjustments for a dynamic version will be suggested.
In theory, the regional objective functions of such a model should be
strictly quasi-concave, continuous and differentiable welfare functions.
The welfare functions employed here are simplified and only related
MODEL FOR ENVIRONMENTAL QUALITY 73

to the maximization of value added (net of environmental damage

measured against as yet unknown costs). Given the linear character of
these welfare functions, the Pareto solution discussed in the previous
sections is not directly applicable in its original sense. However, by means
of a tax and redistribution system the essential ingredients of a Pareto
approach are present (see experiment E), except that the mathematical
specification of the (linear) objective functions should be adjusted. When,
for example, instead of (4.1) a quadratic penalty function with respect to
a priori fixed achievement levels would have been assumed, the original
Paretian analysis might be carried out directly. Due to lack of information
about these achievement levels and for the sake of simplicity, however,
the abovementioned linear objective function will be maintained in this
chapter. Consequently, the numerical applications are merely to be seen as
simulation experiments in the field of abatement strategies, regional
welfare profiles and tax-redistribution schemes.
The multiregional objective function is:
_ T T "\T "\T
maxw -.PI ~I +.P2~2-~1~1 -~2~2 (4.1)

where 1'i (i = 1, 2) is a (2 xl) vector with value added coefficients for sector
A and B, ~i a (2 x 1) vector with sectoral production volumes (measured
in Df!.), 1i a (5 x 1) vector with cost coefficients for 5 pollutants and ~i a
(5 xl) vector with concentration levels of 5 pollutants (measured in Jlg/m 3 ).
The constraints of the model are input-ouput constraints, employment
constraints and pollution constraints.
The input-output constraints are:

(I-A.)x·;;;;'f·*
- - I -I _I
(4.2)

where Ai is the input-output matrix of region i and It a prespecified

vector of final demand elements. In the model at hand, It is set equal to
the final demand vector from the year 1960. -
The employment constraints for each region are:

(4.3)

1. (i = 1, 2) is a (2 x 1) vector with sectoral employment coefficients

where -I
74 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

of region i, and Li the available labour stock (measured in man-years) in

region i. The elements .1 Li are included to make the model more flexible
by allowing job commuting. Clearly the following total multiregional
employment condition should also be satisfied:

(4.4)

The relationships for the generation of pollution are:

(4.5)

where Pi is a (5 x 1) vector with pollution emission of region i (measured

in tons per year), and lli a (5 x 2) matrix with emission coefficients. The
vector of waste discharges Pi can be transfonned into concentrations by
means of a spatial diffusion matrix:

(4.6)

where gi (i = 1, 2) is a (5 xl) vector with pollution concentrations in

region i, and H12 a (diagonal) diffusion matrix transfonning the emissions
of pollutants in region 2 into ambient concentrations in region 1, etc. The
following constraints may be imposed on the pollution concentrations:

(4.7)

where g* is assumed to be 4 percent higher than the pollution concentra-

tions in the least polluted region 1 in 1966 ( based on a free pollution).
The model described by (4.1)-(4.7) was used as the basis for a series of
experiments with alternative specifications of the constraints; next the
policy aspects based on pollution taxes were introduced. The model was
calculated for the year 1966. The data of the model are included in
Table 4.
MODEL FOR ENVIRONMENTAL QUALITY 75

Table 4. Data of the multiregional pollution model.

1'1 if
.73 .75 .10 .20 .09 .12 .07xl0-3 .06xl0- 3 700xl0 6 468xl0 6
.51 .60 .17 .29 .15 .28 .04xl0~ .03xl0- 3 1080xl0E7198xl0 6

!b !l11 !l12
.00 1.44 .39 .21 17,500
.00 8.80 .45 .18 97,500
.00 5.05 .30 .00 14,000
.00 5.28 .27 .11 13,500
.00 .01 .45 .18 200

~2 !!.21 !!.Z2
L 1=163xl0 3
.00 1.79 .21 .39 L z =941xl0 3
.00 9.91 .18 .45
.00 8.13 .00 .30
.00 4.90 .11 .27
.00 .02 .18 .45

Given the data of Table 4, a series of experiments A through E was carried

out. The first four experiments, A-D, were only based on employment
and/or pollution constraints. The last experiment was based on a system
of pollution charges and subsidies (in line with the exposition presented
in Section 3.3). In this way the effectiveness of a system of charges can
be compared with a system of standards (see also Nijkamp 1977, for a
discussion on this subject). These experiments are based on aggregated
regional welfare functions. The solutions reflect a Pareto solution, because
any other allocation of resources and production factors leads to a worse
result for at least one region.

Experiment A:
This experiment was simply based on a maximization of value added
without any additional pollution constraints or pollution costs (i.e. ~1 =
~z :;:: Q). The results ofthis 'free pollution' model are presented in Table 5
as a set ofregional welfare proflles.
76 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

Table Sa. The optimal regional Table Sb. The standardized regional
welfare profiles of experiment A. * welfare profiles of experiment A. **

Region 2 Region 2
Variable Variable

Income per capita 12,600 16,400 Income per capita .77 1.00
Pollution concentration 59,100 154,800 Environmental quality .38 1.00
Percentage employment 100 100 Percentage employment 1.00 1.00

* The pollution concentrations of the S pollutants are added here for the sake of
simplicity. All figures are rounded off on 10 2 units.
** Environmental quality (standardized) is defined here as 1 minus the (standardized)
pollution concentration.

It is clear that in the case of absence of pollution constraints a maximiza-

tion of regional values added leads to high values of average incomes,
pollution and employment, as is demonstrated by the outcomes of Table
5. The results of this experiment can be used to calculate standards as a
frame of reference for evaluating the welfare profiles of other experi-
ments. With regard to the income variables, the maximum average value
added (16,400 Dfl.) has been chosen as a comparative standard. With
regard to the remaining variables, a zero level of pollution and a 100
percent employment have been assumed as standards.
The degree of discrepancy (or inequality) between region 1 and 2 can
be calculated by means of the Euclidean distance S12 between the
elements of the standardized regional welfare profiles (see also (1.8) ).
The interregional inequality index S12 appears to be equal to .66. This
inequality index has only a limited meaning as such, but it can be used to
compare alternative states of the economic system associated with
alternative assumptions concerning the constraints of the model or
with alternative public policies.

Experiment B:
This experiment is based on a maximization of value added taking into
account the pollution constraints (4.7). The pollution costs and pollution
taxes are here again neglected (i.e. ~1 ::;:: ~2 ::;:: Q).The results are contained
in Table 6.
MODEL FOR ENVIRONMENTAL QUALITY 77

Table 6a. The optimal regional Table 6b. The standardized regional
welfare profiles of experiment B. welfare profiles of experiment B.

Region 1 2 Region 1 2
Variable Variable

Income per capita 12,400 13,800 Income per capita .76 .84
Pollution concentration 45,000 61,800 Environmental quality .29 .40
Percentage employment 100 100 Percentage employment 1.00 1.00

The pollution constraints appear to lead to a decline in average income of

region 2 (apart from a decline in pollution itself). The degree of inter-
regional inequality Sl2 is now equal to .14, so that the introduction of
pollution standards appears to lead to a decline in regional inequality
(compared with experiment A). This result corresponds to a priori
expectations, because a truncation of the feasible area will diminish extra-
ordinarily high values of the variables.

Experiment C:
In experiment C the assumption of a maximum value added is replaced by
a minimization of environmental damage costs (Le. PI = P2 = Q). The
pollution constraints are then abandoned. The results, based on the rigid
assumption that the unit environmental damage costs ~ (i = 1, 2) are
equal for all 5 types of pollutants, are contained in Table 7.

Table 7a. The optimal regional Table 7b. The standardized regional
welfare proflles of experiment C. welfare profJ.1es of experiment C.

Region 1 2 Region 1 2
Variable Variable

Income per capita 10,000 5,750 Income per capita .61 .35
Pollution concentration 17,600 23,000 Envirpnmental quality .11 .15
Percentage employment 82 42 Percentage employment .82 .42

The maximization of environmental quality appears to lead to low pollu-

tion levels, but also to rather low income and employment levels, especially
78 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

in the originally most productive region 2. The regional discrepancy index

SI2 appears to be equal now to .58, indicating that the interregional
inequality has undergone a decline with respect to a free maximization of
value added, but a rise with respect to a pollution standards program.
These results are again quite reasonable, because a minimization of
damage costs wi111ead to a significant decline in the income levels of both
regions (especially of the most polluting region).

Experiment D:
Experiment D is a combination of experiments A and C, except that the
unit environmental damage cost of region 1 and 2 are assumed to differ.
~1 and ~2 are calculated as follows. In a situation of free pollution (exper-
iment A) the pollution concentrations of region 1 and 2 are equal to
60,000 Ilg/m 3 and 160,000 Ilg/m3, respectively. The resulting regional
incomes for the same situation are approximately 1.6 X 109 Dfl. and
15.6 X 109 Dfl., respectively. Now the assumption is made that region 2
is prepared to pay 25 percent of its income in exchange for a clean
environment, while the less polluted region 1 is prepared to pay 12.5
percent of its income. Given this assumption, the marginal cost ('shadow-
price') per unit concentration in region 1 and 2 is approximately equal to
3,300 Dfl. and 24,000 Dfl., respectively. Hence, the vectors ~1 and ~2 are
equal to (3300, 3300, 3300, 3300, 3300)T and (24000, 24000, 24000,
24000,24000)T respectively.
The resulting welfare profiles can be found in Table 8. The income and
employment results of experiment D bear a close resemblance to those
from experiment B, but the environmental quality results show a rather
significant difference. The interregional inequality index S12 is equal to
.13, which falls also in the vicinity of the results of experiment B.

Table 8a. The optimal regional Table 8b. The standardized regional
welfare profiles of experiment D. welfare profiles of experiment D.

Region 1 2 Region 1 2
Variable Variable

Income per capita 12,000 13,500 Income per capita .73 .82
Pollution concentration 22,600 38,800 Environmental quality .15 .25
Percentage employment 100 100 Percentage employment 1.00 1.00
MODEL FOR ENVIRONMENTAL QUALITY 79

Experiment E:
Experiment E is the most integrated approach, since here pollution taxes
and subsidies are also taken into account. The pollution taxes ()i (i = 1, 2)
are levied upon the emitted pollutants, so that:

().1 = -r.Tp.
,_,
(4.8)

where Ii is a vector of marginal tax rates on pollution. The interregional

subsidies si accruing from the charge system are divided according to an
exogenous (public) allocation coefficient a:

(4.9)

Now the new objective function is assumed to be:

(4.10)

This welfare function of region 1 has to be maximized subject to a political

constraint concerning the distribution of welfare between region 1 and 2.
Here the assumption is made that after the distribution of subsidies the
average welfare of region 1 should be equal to that of region 2. Hence, the
distributional constraint is included as:

(4.11)

Clearly, now the distribution parameter a has to be chosen such that

condition (4.11) is satisfied. This can be done by deleting (4.9) from the
optimization program and by replacing it by the condition:

(4.12)

Next, Sl and S2 can be calculated, given (4.10), (4.11) and (4.12). Then
a can be derived ex post from (4.9).
The vectors of environmental cost coefficients ~i are calculated in
experiment D. The vectors of marginal pollution taxes '!i are now calcu-
lated accordingly. Given the assumed propensity of region 1 and 2 to pay
12.5 percent and 25 percent of its regional income, the total emission
80 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

in these regions in the case of a free pollution can be related to this total
pecuniary propensity for a cleaner environment in order to calculate the
corresponding pollution tax. Hence, the pollution tax rates are: II =
(5000, 5000, 5000, 5000, 5000)T and I2 = (9000, 9000, 9000, 9000,
9000).T
The results are presented in Table 9 (a appeared to be equal to .52).
These results show a high degree of correspondence with those from
e~eriment D. The interregional inequality index Sl2 is now approximately
equal to .10.

Table 9a. The optimal regional Table 9b. The standardized regional
welfare profIles of experiment E. welfare profIles of experiment E.

Region 1 2 Region 1 2
Variable Variable

Income per capita 12,900 13,400 Income per capita .79 .82
Pollution concentration 22,600 38,800 Environmental quality .15 .25
Percentage employment 100 100 Percentage employment 1.00 1.00

The foregoing model is essentially a comparative static mode. Due to

its linearity, the marginal rates of substitution are constant, so that a
continuous adjustment according to the lines exposed in previous sections
is hardly possible. This would require a non-linear model. Therefore, the
theoretical Paretian analysis of Sections 3.2 and 3.3 is only partially
reflected in experiment E.
Another possible way to dynamize the foregoing model is to carry out
an iterative procedure for the redistribution parameter a in order to arrive
at a desired interregional distribution of welfare after a series of successive
steps. This would imply a transition from a disagreement equilibrium after
a series of disequilibrium points towards a new Pareto solution. An
alternative approach would be to assume that the regional technology
can be controlled by an appropriate choice of pollution taxes so that the
input coefficients may be a continuous function of tax rates.
Although the theoretical part of the present study aims to be an equili-
brium analysis, the empirical part can be considered more as a policy
model. The adjustments are left aside, and properties of the equilibrium
(cf. a Nash equilibrium) are not investigated any further. Clearly, such
CONCLUSION 81

a policy model is only meaningful for a practical policy framework when

the empirical data base is improved.
The last step of the research agenda for the abovementioned welfare
analyses in an operational setting concerns the collection of more reliable
data, not only for the input-output and diffusion coefficients but also for
the elements of the tax-subsidy and redistribution mechanism.
In conclusion, experiment E contains several ingredients for an opera-
tionalization of a Paretian analysis in an environmental framework but the
final steps require an adjustment of the welfare function, a dynamization
of several model parameters and more accurate data. These steps will be
the focus of further research of the authors.

3.5 CONCLUSION

The notions of regional welfare proflles and of interregional welfare

distances appear to be quite useful in studying interregional inequalities
in the framework of environmental policies and regional developments.
These notions can be directly linked to a Paretian welfare analysis.
The model presented in this study is a simplified model which may be
extended in various ways. Due to lack of data, however, such an extension
is rather difficult and at the least very time-consuming.
The model includes political control parameters in several ways (via
distribution and taxation). In this contribution the relevance of equili-
brium models in the framework of public policy was exposed. It is clear
that the gap between abstract theory and operational results is still large,
although our study shows that it may be abridg~d to a certain extent.
Clearly, much research should still be done.
82 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

APPENDIX A. Extensions of the environmental externalities model

In this appendix, the analysis will be extended toward a multi-sector

pollution model. Suppose (without loss of generality) that the economy
concerned is composed of two sectors producing private commodities of
type A and B, respectively. The production of each of these commodities
gives rise to a specific type of pollutant: one pollutant, associated with
production process A, ends up in all regions of the spatial system and the
other one, associated with production process B, remains within the
borders of its own region. Commodity category B is produced in all
regions while commodity category A is only produced in some regions.
Then the following compact hypotheses which are a straightforward
extension of Sections 3.2 and 3.3 may be made:

- The spatial system contains a set R = {rl r = 1, ... ,R } of regions.

- R is subdivided into a set 0 ={u 1 u = 1, ... , U} of regions producing
commodity category A and B (denoted by x~ and x:' respectively)
and a corresponding_ pollution category (denoted by pA u and pBu ,respec-
tively), and a set N = {n] n = 1, ... ,N} producing only commodity
category B (denoted by x:) and a corresponding pollution category
(denoted by p:).
- Each region u has a production-consumption set Cu = {(x~, x:' p~ ,
p: ' zu)] x~ > 0, x: > 0, p~ > 0,
the environmental quality of region u.
p:
> 0, Zu .;;:;; O}, where Zu represents
- Each region u has a well-behaved welfare function Wu (x~ , x:'
zu) with the same properties as in (2.1) and (2.2):
p: ' p: '

W~A
u
> 0, W~B
u
> 0, w~ .~
u
°
'V x~ > 0, 3 P~ , such that
(A.l)
> 0, ifp~ < Pu
-A

= °
, ifpA
u = P~
< 0, ifp~ > Pu
-A
(idem for B)
A. EXTENSIONS OF THE ENVIRONMENTAL MODEL 83

u
- Each region has a choice set Su = {x: ' x: 'p: 'p: }.
x:
- Each region n has a production-consumption set en = {(x:' P:, zn ll
> O,p: > O,zn .;;;; Of.
- Each region n has a well-behaved welfare function wn (x:' zn) P: '
with the following properties (see (2.3) and (2.4) ):

W~B > 0, w;n ~ 0

x: ' fi: '
n

'V E such that

(A.2)
> 0, ifP:
= 0 , ifpB
n
< 0, ifP:
- Each region n has a choice set Sn = [x:' P: ]
- The spatial system has a fixed sectoral production capacity, so that:

(A.3)

(A.4)

P: P:
- The environmental quality Zu in regions u is a weighted average of
pollution type and with weights 711 and 712 (711 + 712 = 1):

(A.S)

where ~u is a diffusion parameter indicating the proportion of total

pollution of type A ending up in region u.
Note that only pollutant p~ exceeds the regional borders, so that
~ ~u + ~ ~n = 1. For an introduction to diffusion models see Nijkamp
(1977).

pollution type and P: P:

- The environmental quality zn in regions n is a weighted average of
with weights €1 and €2 (€1 + €2 = 1):
_ U A B
Zn - - €1 ~n ~ Pu - € 2Pll (A.6)
u=l
84 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

Given these hypotheses, the concepts described in Section 3.2 can again
be applied.

{(x: ' x!, P: '

P%, Zu)' (x:' P:,
A feasible state of the spatial system is composed of the set of elements
zn)} which satisfy conditions (A.3)-(A.6).
A Pareto optimum represents again a feasible situation where an increase
in welfare of one region leads to a decrease in welfare of at least one other
region. In other words, a feasible state

is a Pareto state if there is no other feasible state

such that:

and

(A.8)

with strict inequality sign for at least one condition.

It can be proved that necessary and' sufficient conditions for feasible
solutions to be a Pareto optimum are (see Appendix B):

(A.9)

(A.IO)

and

(A.II)

where the latter condition is related to the spatial diffusion of pollutants

from production process A in regions u E U.
A disagreement equilibrium is any feasible (U+M-tuple of the following
vector {(x: ' x!, P:, p!, zu)' (x:' P:, zn)} ,u E 0, n EN, which satisfies
the following properties (since the price of x: is considered as a numer-
A. EXTENSIONS OF THE ENVIRONMENTAL MODEL 8S

aire, a new price 1T for commodity category B has to be introduced):

subject to: ~ + 1T~ :eo;; ~ + 1T~

(A.12)

zu =

and

subject to: x: x::eo;;

(A.13)

In the case of free pollution, regions with production processes A will fix
their pollution levels independently, so that regions n E N suffer from uni-
directional environmental externalities from regions u E ii, while all
regions u E ii have reciprocal externalities. In this respect, only a partial
Pareto optimum may be possible, viz. with respect to production process
B which does not pollute other regions. Then only (A.9) and (A.10) are
fulfilled and not (A.1l).
Finally, aPBE is a feasible state dermed by the quadruple

(yuA ' YB.A B) (¥ =B =B)]

u ' Y n 'Yn ; u' Xu ' XII

(TA and TB being an effluent charge on pollutants from production A and

B) with the following properties (see also (2.1 S) and (2.16) ):
86 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

sub;ect
J
to: xA
u + 1TxB
u + -rA
uu pA + rBpB + yA + yB =
uu u u xA
u + 1TxB
u

and

(A.lS)

subject to: 1TxB

n
+ rBpB
n n
+ yA
n
+ yB
n
= 1TxB
n

while also the side-conditions from (A.12) and (A.13) are to be satisfied.
By adding up the side-conditions from (A.14) and (A.IS) and by combin-
ing them with those from (A.l2) and (A.13), it can easily be proved that:

= 0 (A.16)

which states that all effluent charges are redistributed over all regions.
The redistribution effects of a tax policy can be analysed in a way
analogous to the method in Section 3.3. Assuming again that the tax rate
is set equal to the marginal rate of substitution between pollution and
production, one finds:

7uA = auA (A.l7)

7uB = auB (A.18)

7nB = anB (A.19)

This implies the following dynamic pollution tax structure:

'A = U A NAB B A
7u 0u auz + € ~ 0 a a / a - r
"11 u~= l (A.20)
i n = 1 n uz nz uz u
A. EXTENSIONS OF THE ENVIRONMENTAL MODEL 87

.;.B
'n
= € 2 aB
nz - -.-B
'n (A.21)

'u = .....
.;.B a B rB
,,2 uz - U (A.22)

If the assumption is made that the relative competitive position of sectors

A and B remains equal, it is clear that ir = O. Given (A.S) and (A.6), it is
easily seen that:

Z
. = -f/luu L
~ u·P A -f/2Pu
·B
(A.23)
u
u u=l

and

Z
. =- €1
u·
5nu=l A
L P - € PnB (A.24)
n u 2

The change in the pollution taxes ()u and ()n to be paid by region u and n
is by definition:

(A.2S)

()B = fBpB + rBp· B (A.26)

u u u u u

and

()B
n
= i:BpB
n n
+ rBp·
n n
B (A.26)

When only the effluent charges related to pollutants from production

process A are redistributed over all regions by means of fixed allocation
parameters a and {J, the following conditions hold true:

Y• A
u
+ afJuA =0
'
0 ..;; a ..;; I (A.28)

yB
u + aOuB = 0 (A.29)

yB
n
+ aOnB = 0 (A.30)

and
U
Y• A + R L ()A = 0, 0..;; ~n ..;; I (A.31)
u "'n u = 1 u
88 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

Here the assumption is made that all pollution taxes associated with
process B are redistributed within the pollution-generating region itself,
so that the following conditions should be fulfilled:
N
ex + ~ {3, = 1 (A.32)
n=l n

in order to fulfil the conditions (A.16) for a PBE. _

The situation of compensating payments by regions n E N can be
analyzed in the same way. It should be noted that these compensating
payments may, in principle, be either positive or negative, dependent on
the institutional aspects of environmental management. The transition
process from an initial disagreement equilibrium (without any effluent
charge) toward a new equilibrium proceeds along a continuous set of
PBEs. Given the assumption of well-behaved welfare functions and of
continuous welfare arguments, the continuity of the process is quite clear.
The conditions for the existence of a PBE are included in the side-condi-
tions of (A.12)-(A.15) and can be proved as follows. The side-conditions
from (A.14) and (A.l5) are obviously satisfied because this is inherent in
the constrained maximization procedure. Next, by combining these side-
conditions with (A.25)-(A.31), the following result is obtained:

(A.33)

(A.34)

so that by means of (A.3 2) the following condition is fulfilled:

u 0A U 0B N 0B
=0
~ X +7l'~ X +7l'~ X (A.35)
u=l u u=l u n=l n

which proves the fulfilment of the budget constraints from (A.l2) and
(A.l3). The same holds true for the pollution conditions from (A.l2)
and (A.l3). Consequently, the transition process can be characterized as a
set of continuous PBEs.
The fmal question is whether again a Pareto solution will be attained.
It can be derived by means of the proofs provided by Malinvaud (1972)
and Tulkens and Schoumaker (1975) that this indeed is the case. The
various redistribution options (depending on ex, i3n , etc.) can again be
represented by means of a table analogous to Table 1 including the
regional welfare profiles.
B. CONDITIONS FOR A GENERALIZED PARETO OPTIMUM 89

APPENDIX B. Conditions for a generalized Pareto optimum

Assume the existence of (U+N) welfare functions w (x A , x B , PuA , puB, zu)

B B -.JL. u u
and wn (x n ' Pn ' zn) for all regions u E U and n EN. Define the following
marginal rates of substitution:

0uA -- WpA
, / WxA
' (B.la)
u u
, /w'xB
0uB -- WpB (B.lb)
u u
A -w' /W' A
°uz - zu xu (B.lc)

B -- W , /w'x B
0uz zu (B.ld)
u

OnB -w'
- p'B /w'xB (B.le)
n n
B -W' /W' B
°nz - zn xn (B.IO

A Pareto optimum for a certain region 1 can be found by maximizing its

welfare function wl(xf, x{l, pf, p{l, zd subject to conditions (A.3)-
(A.8). The corresponding Lagrange function is (for a discussion on
Lagrange functions see Paelinck and Nijkamp 1976):

L = WI - ¥ ) (wu -
u=2"u
WO) -
u
~
n=1 ''11
) (w -
n
WO) -
n
III (¥ xA
u=1 u
- 0 *)

U U A B
- ~ "u(zu + 111 ~u ~ Pu + 112Pu )
u=1 u=1
N U A B
- ~ I.In (zn +€l~n ~ Pu +€2P n ) (B.2)
n=1 u=1
The first-order conditions for a maximum are:
aL
-A =- \ W~A , Al =1 , V u (B.2a)
axu u
90 ENVIRONMENTAL PROTECTION AND SPATIAL WELFARE

aL
- = - A w' B - J.l.2 \;fu (Bo2b)
axB u Xu
u

aL U N
ap;;
= - A w' A - 111 ~ V 8u - €1 ~ vn 8n = 0, \;f u (Bo2c)
u Pu u=1 u n=1

0, \;fu (Bo2d)

\;fu (Bo2e)

aL
- = - A w' B - J.l.2 = 0, \;f u (Bo2t)
axB n xn
n

aL
. - - - A w' B - V = 0 \;fn (Bo2g)
ap: n Pn n
€2
'

aL
- A w' - v = 0 \;f n (Bo2h)
n zn n '

By combining respectively (Bo2a) and (Bo2c), (Bo2b) and (Bo2d), (Bo2a)

and (Bo2e), (Bo2b) and (Bo2e), (Bo2t) and (Bo2g), and (Bo2t) and (Bo2h),
the following results are obtained:
U N
OA = 111 ~ 8u Vu /J.l.1 + €1 ~ 8n Vn /J.l.1 (B.3a)
u u=l n=l

oB
u = 112 V)J.l.2 (B.3b)

A
°uz = Vu /J.l.1 (B.3c)

°uz
B
= Vu /J.l.2 (B.3d)

iJB
n = €2 Vn /J.l.2 (B.3e)

°nz
B
= Vn /J.l.2 (Bo3f)

Combination of (B.3e) and (Bo3t) gives the first condition for a Pareto
optimum:
B. CONDITIONS FOR A GENERALIZED PARETO OPTIMUM 91

(BA)

which states that the marginal rate of substitution between pollution and
production is equal to the weighted marginal rate of substitution between
environmental quality and production.
Combination of(B.3b) and (B.3d) gives the second Pareto condition:

(B.S)

which is similar to (BA).

By combining (B.3c), (B.3d) and (B.3f) it can be derived that:

(B.6)

Substitution of (B.3c) and (B.6) into (B.3a) gives the last Pareto condition:
u u
auA = 1/1 L 0 aA
u = 1 u uz
+ t1 nL= 1 0n auz
A aB / aB
nz uz
(B.7)

The complexity of the latter condition arises from the diffusion mechan-
ism for environmental externalities assumed with regard to pollutants
from production process A.

REFERENCES
d'Aspremont, C. and L.A. Gerard-Varet, 'Un modele de negociation internationale pour certains
problemes de pollution'. Revue d'Economie Politique, 597-620, August, 1976.
Bartels, C.P.A. and P. Nijkamp, 'An empirical welfare approach to regional income distributions'.
Socio-Economic Planning Sciences, vol. 10,117-128,1976.
Bauer, R.A. (ed.), Social Indicators. MIT Press, Cambridge, Mass., 1966.
Baumol, WJ., 'On taxation and the control of externalities'. American Economic Review, vol. 62,
307-322,1972.
Baumol, W.J. and W.E. Oates, The theory of environmental policy. Prentice-Hall, Englewood
Cliffs, 1975.
Beckenbach, E. and R. Bellman, An introduction to inequalities. Random House, New York,
1961.
Coupe, B.E.M .G., Regional economic structure and environmental pollution. Martinus Nijhoff, The
Hague, 1976.
Davis, O.A. and A. Whinston, 'Externalities, welfare and the theory of games' ,Journal of Political
Economy, vol. 70, 240-262,1962.
Dear, M., 'Spatial externalities and locational conflict', in D.B. Massey and P.W.J. Batey (eds.),
Alternative frameworks of analy sis. Pion, London, 152-167, 1977.
Delft, A. van and P. Nijkamp, Multi-criteria analysis and regional decision-making. Martinus
Nijhoff, The Hague, 1977.
92 REFERENCES

Dreze, J 11. and D. de la Vallee Poussin, 'A tatonnement process for public goods'. Review of
Economic Studies, vol. 38,133-150,1971.
Enzensberger, H.M., 'Zur Kritik der politischen Okologie'. Kursbuch 33, Rotbuch Verlag, Berlin,
142,1973.
Green, P.E. and F.J. Carmone, MultidimensioTUlI scaling and related techniques in marketing
aTUllysis. Allyn and Bacon, Boston, 1970.
Harvey, D., Social Justice and the city. Arnold, London, 1970.
Hordijk, 1. and P. Nijkamp, 'Dynamic models of spatial autocorrelation'. Environment and
Planning, vol. 9, 505 -519, 1977 .
Hueting, R.,Nieuwe Schaarste en Economische Groei. Agon Elsevier, Amsterdam, 1974.
Klaassen, 1.H. and T.H. Botterweg, 'Project evaluation and intangible effects: a shadow project
approach', in P. Nijkamp (ed.), Environmental economics, vol. 1. Martinus Nijhoff, The Hague,
3349,1976.
Mliler, K.G.,Environmental economics. The Johns Hopkins Press, Baltimore,1974.
Malinvaud, E., Lectures on microeconomic theory. North-Holland Publishing Co., Amsterdam,
1972.
Milleron, J.C., 'Recherche des extremums de fonctions de plusieurs variables avec ou sans con-
trainte sur les variables', in E. Malinvaud (ed.) , Lec;ons de theorie microeconomique. Dunod,
Paris, 257-271, 1971.
Milleron, J .C., 'Theory of value with public goods'. JOurTUllofEconomic Theory, vol. 5,419472,
1972.
Nijkamp, P., 'Toward a society without a perspective?', Institute for Christian Studies, Toronto,
1976 (mimeographed).
Nijkamp, P., Theory and application of environmental economics. North-Holland Publishing Co.,
Amsterdam,1977.
Nijkamp, P. and R. Rietveld, 'Multi-objective programming models'. Regional Science and Urban
Economics, vol. 6,253-274,1976.
Nijkamp, P. and P. Rietveld, 'Impact analyses, spatial externalities and policy choices'. Paper for
First Canadian Regional Science Association Meeting, Halifax, 1977 (to appear in Northeast
Regional Science Review).
Paelinck, J.H.P. and P. Nijkamp, Operational theory and method in regional economics. Saxon
House, Farnborough (jointly with D.C. Heath, Lexington), 1976.
Pahl, R., 'Poverty and the urban system', in M. Chisholm and G. Manners (eds.) , Spatial policy of
the British economy. Cambridge University Press, Cambridge, 126-146, 1971.
Perloff, M.S. (ed.), The quality of the urban environment. Resources for the Future, Washington
D.C., 1969.
Pigou, A.C., The economics of welfare. MacMillan, London, 1930.
Richardson, H.W., RegioTUllgrowth theory. MacMillan, London, 1973.
Samuelson, PA., 'The pure theory of public expenditure'. Review of Economics and Statistics,
vol. 36, 387-389, 1954.
Shephard, R.N:, 'Matrix structures in ordinal data'. JourTUll of Mathematical Psychology, vol. 3,
125-139,1966.
Smith, D.M., The geography of social well·being. McGraw-Hill, New York, 1973.
Tulkens, H. and F. Schoumaker, 'Stability analysis of an effluent charge and the "polluter pays"
principle'. JourTUll ofPublic Economics, vol. 4, 245-269, 1975.
Vipond, J., 'City size and unemployment'. Urban Studies, vol. 11,39-47,1974.
4. Contact potentials in the European
system of cities
ULF ERLANDSSON*

It has been argued in regional science, geography, economics, and socio-

logy that the development and/or growth potential of a region is closely
related to its interaction with other regions. In fact the more 'advanced'
a society is, the more emphasis is placed on communication possibilities,
not only in regard to the volume of interaction but also to the quality and
duration of direct face-to-face contacts. This chapter presents an approach
to what appears as a key dimension of the problem of spatial inequalities
and regional development. It focuses on the horizons of contact potentials
surrounding particular nodes within the European communication system.
The framework used to examine the contact potentials of a range of
European cities was initially developed for applications within Sweden.
The background to these studies, described in Section 4.1, is the struc-
tural changes in occupational composition that have taken place during
the last two decades in Sweden and the consequences of these changes on
regional distribution of job opportunities and interregional travel facilities.
On the basis of the Swedish experiences it seems appropriate to examine
the interdependence between changes in employment and travel facilities
in a wider European context, as one can assume that general trends in
employment and travel behaviour among emerging administrative elites
within Europe (e.g within the EEe) are similar to those in Sweden.
Section 4.2 will therefore give a broad summary of the results of the
studies dealing with contact systems and regional development in Sweden.
This summary is mainly based on investigations by Tornqvist and his
associates in Lund (Tornqvist 1968, 1970, 1973; Engstrom and Sahlberg
1973, 1974). In Section 4.3, two different investigations dealing with
opportunities for direct personal contacts (face-to-face) within the
European system of cities are reported. The methods used to examine
these opportunities are fairly simple since lack of data preclude possibili-

* University of Lund, Sweden.

94 CONTACT POTENTIALS IN EUROPEAN CITIES

ties of applying exactly the same approach as in the Swedish case. Finally,
in Section 4.4, some questions are raised concerning the prospects and
limitations of this approach and its possible implications for a regional
policy at a European scale.

4.1 JOB FUNCTIONS AND CONTACT ACTIVITY

For an understanding of the structural changes that have taken place in

occupational composition within post-industrial societies, it is necessary
to depart from the accepted division of employment into industrial, trade
and service sectors. In the Swedish studies previously mentioned, occupa-
tional activity has been regarded in terms of job functions. These are
to be found in varying numbers and combinations within all kinds of
workplaces which together constitute the private and the public sectors
of the economy.
Job functions are the basic elements in an activity system, where the
links (relationships) between these functions are different kinds of trans-
portation and communication flows. Figure 1 shows different groups of
job functions included in this activity system. The ftgure also gives
examples of duties which are characteristic of various functions as well as
the links and flows between job functions. The main links between
primary functions and manufacturing are flows of materials. Typical
linkages for service functions are buying and service trips. The administra-
tive functions, which are to be found in varying degrees in all companies
and other organizations, are linked together by flows of information.
Transmission of this information occurs by direct personal contacts,
correspondence or telephone calls.
The activity systems in countries which can today be described as post-
industrial have undergone a radical transformation. These changes can be
described as pronounced shifts in occupational structure from the primary
functions typical of an agrarian society to the manufacturing functions
which account for a considerable proportion of gainful employment in the
post-industrial society.
Until now the structural changes described have resulted in a large
concentration of job opportunities and residential facilities in major urban
regions throughout the world. Especially in developed countries, the
growth of major cities is largely the result of the expansion of service and
administrative functions. There we ftnd the administrative and manage-
CONTACT SYSTEMS IN SWEDISH REGIONAL DEVELOPMENT 95

ment personnel of trade unions and other interest organizations, financial

institutions, a large proportion of the growing public administrative sector
and a large number of service activities which are linked to these adminis-
trative functions (Hall 1966; Kristensson 1967).
JOB FUNCTIONS LINKS and FLOWS

i
ADMINISTRATIVE FUNCTIONS
Decision -making, planning, Direct personal contacts
negotiations. search. product

~ :"o~cow,
development etc.

Control. direction of
production. information-
processing. services to I ,"ro••
Al etc.

Accounting. routine
office work. services to
Al and A2 etc. Contacts via telephone
and correspondence
SERVICE FUNCTIONS

e=:U->
Business services Buying and service trips
Household services

MANUFACTURING FUNCTIONS
Processing of mot.rials. MATERIAL FLOWS
handling of goods. con- D

~---'>
struction. etc.

PRIMARY FUNCTIONS
Agriculture. mining,
energy production, etc. Goods tranportation

Figure 1. Diagrammatic representation of an activity system consisting of job

functions connected by links and flows (adapted from Tornqvist 1973).

Development in the post-industrial society has led to a far-reaching

division of labour and specialization. As a result of this, the interdepen-
dence of various work functions has increased considerably. Society
cannot function without the constant transit of goods, people and
messages.

4.2 CONTACT SYSTEMS AND REGIONAL DEVELOPMENT IN SWEDEN

4.2.1 Information transmission and employment

The following observations concerning the connections between informa-
tion transmission and employment in Sweden during the 1960s and 1970s
96 CONT ACT POTENTIALS IN EUROPEAN CITIES

appear particularly interesting from the viewpoint of regional develop-

ment and spatial inequality:
1. The most important information exchanged in society still requires
face-to-face meetings in spite of the rapid advance of telecommunica-
tions. Information flows via telecommunication and correspondence
can hardly affect the location of various activities because physical
distance is no longer an obstacle for these types of information trans-
mission.
2. Groups of administrative employees in such organizations as manu-
facturing firms, service companies and public administration, who are
involved in extensive contact activity may be identified. These
employees (AI in Figure 1) spend an average of between 30 and 50
hours a week on direct personal contacts, some of them more than 60
hours (these figures include travelling time but not pre-travel prepara-
tion time). Other employees have fewer face-to-face meetings (A 2 ).
There are also categories of administrative personnel with very few
personal contacts outside their own organization (A3 ). Contacts made
in the course of duty by blue-collar workers in the operational units of
organizations (D) outside their place of work are barely measurable
(Hedberg 1970; Sahlberg 1970).
3. Examination of interregional travel in Sweden, i.e. trips between the
centers of the country's seventy A-regions, shows that employees with
contact-,intensive functions account for 80 percent of all trips by air
and about the same proportion of first-class train travel.
4. The interregional contact network is dominated by the major urban
regions both as the origins and destinations of trips. The dominance of
Stockholm as a central node in the network is particularly striking. In
southern and central Sweden there is mutual contact activity between
the smaller regions, but in the north such contacts are remarkably few.
Nearly all contacts are directed towards Stockholm. This pattern of
regional contacts is to a great extent a result of the Swedish transport
network.
s. Studies of the changing regional distribution of employment during the
1960s and the beginning of the 1970s show that there is a defmite
connection between the need for contacts and regional concentration
(Engstrom 1970; Engstrom and Sah1berg 1973). The expansion of
employment has shifted from the operational functions (D in Figure 1)
to the administrative functions (A). The number of employees in
functions responsible for the exchange and processing of information,
CONTACT SYSTEMS IN SWEDISH REGIONAL DEVELOPMENT 97

particularly the most contact-intensive (AI)' has increased very rapidly.

As in the case of most technologically-developed societies in Western
Europe there is a tendency toward a concentration of these functions in
those regions containing the largest cities. This means that the func-
tional differentiation between the different regions of the country has
becOme increasingly accentuated.
Although the number of contact-intensive employees in Sweden may not
constitute a particularly large group of the gainfully employed, it is a very
important group from the point of view of regional policy for several
reasons. At least half of these employees are now located in the three
major urban regions of the country and there they exert a strong multiplier
effect on employment.

4.2.2 Contact requirements and travel facilities

On the basis of the empirical findings mentioned above the next step for
Tornqvist and his associates was to investigate regional differences in
Sweden in the availability of facilities for maintaining direct personal
contacts, i.e. to study the regional balance between travel facility demand
and supply (Tornqvist 1973; Engstrom and Sahlberg 1973). This was
done by calculating what Tornqvist called contact potentials, where the
accessibility of various places to one another was related to their contact
requirements. As a criterion for measuring the accessibility of various
places Tornqvist used a length of stay of single-day visits at each locality,
travel time and travel cost. The contact requirements were measured on
the basis of the actual number of contact-intensive employees within each
of the seventy A-regions in Sweden. The calculated potential value for
each of the regions was mapped, and the term contact landscape was used
for the cartographic description of the country's regional differences in
contact possibilities (Tornqvist 1973; Engstrom and Sahlberg 1973). The
calculations were made for 1960, 1965 and 1970. At all three points of
time, Stockholm had the highest contact potential, i.e. according to the
calculations, employees in the Stockholm region had the greatest possibili-
ties in Sweden for maintaining direct personal contacts. The contact
possibilities for most other regions increased, however, during the 1960s.
The changes were partly due to changes in travel possibilities brought
about by changes in transport service systems, and partly to a changed
distribution of contact-intensive employees (tending to concentrate in the
biggest cities). Figure 2 shows the Swedish contact landscape 1970 accord-
ing to Tornqvist.
98 CONTACT POTENTIALS IN EUROPEAN CITIES

1970

Figure 2. The Swedish contact landscape 1970. Numbers indicate contact potentials
for urban regions. Stockholm = 100. (Adapted from Tornqvist 1973.)
CONTACT SYSTEMS IN SWEDISH REGIONAL DEVELOPMENT 99

4.2.3 The contact landscape of the future

As a final step in investigating the Swedish contact landscape attention
was focused on the future. A series of experiments with operational
'models of Sweden' illustrating alternative future systems of cities revealed
the following speculations concerning relationships between regional
development and travel possibilities (Tornqvist 1973):
1. A continuous regional concentration of contact-intensive activities
with today's transport system results in increased efficiency of the
national contact system. This concentration of the contact-intensive
employees, primarily around Stockholm, leads to increased contact
possibilities not only for those working in the capital but also for
those working in other regions throughout the country.
2. The decentralization of contact-intensive activities, e.g. the relocation
of central government agencies, cannot be carried out without drastic-
ally reducing the ability of employees to maintain direct personal
contacts if the present interregional transport system is not changed
simultaneously.
3. Should the authorities, for whatever reason, wish to arrest a trend
toward increased inequality between various regions in the country,
it would seem that regional policy measures would be most effective if
directed at the transport sector.

4.2.4 Conclusions
The investigations previously mentioned show that contact activities occur
in a national system of cities. The embryo of the Swedish national system
of cities evolved long ago. Since this early development, a growing con-
centration of decision-making and administrative functions within the
largest cities has occurred. At the same time the accessibility of these large
cities has increased simultaneously with the extension of the transport
system and higher transport technology. On the national level, this has
resulted in the growth of a few urban centers at the expense of the
remaining national territory.
Additionally, the investigations showed that direct personal contacts
were not just maintained within a national system of cities, as foreign
contacts average about 10 percent of the number of contacts originating
in Sweden (Hedberg 1970; Sah1berg 1970). In the contact pattern formed
by decision-makers and other administrative personnel, it is thus possible
to see the outlines of an international 'business world'. This consists of a
set of large urban places which are connected with one another by rapid
100 CONT ACT POTENTIALS IN EUROPEAN CITIES

means of transport. Within these links the contact-intensive employees

are constantly mobile, attending meetings, conferences and other
activities. On the basis of these findings the studies dealing with contact
systems were expanded to all of Europe in order to investigate the travel
and contact possibilities in an international system of cities.

4.3 ACCESSIBILITY IN THE EUROPEAN SYSTEM OF CITIES l

Given the conditions within a single country (e.g. Sweden) the Tornqvist
approach concerning contact potentials seems to work. It is evident,
however, that the change from a national to an international context gives
rise to additional methodological problems. For instance, in applying the
Tornqvist model on a wider spatial scale it could in principle be ques-
tioned if the measure of accessibility of various places should be expressed
by a length of stay time as short as a single-day visit at each locality.
Moreover, should the contact requirements on an international level be
measured on the basis of the actual number of contact-intensive em-
ployees within each of these places, when a great part of the contact
requirement of a certain urban place is likely to be intra-national rather
than international? There may also be additional constraints on the
travel behaviour of administrative employees when international bounda-
ries have to be crossed. However, th~ investigations below have not really
taken these problems into consideration and they will thus have to be left
aside for further research.
The foregoing discussion of contact patterns in Sweden has emphasized
the role of the principal urban nodes in the national information web. An
extension of the study to the international level also poses the immediate
problem of identifying the appropriate European system of cities for the
analysis of international information webs. It is, of course, impossible to
regard any such system as closed for there will be many flows across its
boundaries. Most large European cities, for instance, are in frequent direct
contact with American cities. Beyond the immediate periphery of Europe,
e.g. in North Africa, are cities where daily contact is as much with the
urban nodes of Europe as with themselves.

1. The investigations reported in this paper are derived from a study of the possible
consequences of a proposed bridge across the Oresund, part of which is concerned
with the general question of accessibility in Europe.
ACCESSIBILITY IN EUROPEAN CITIES 101

Many studies have indicated the tendency for specialized occupational

groups and decision-makers to congregate in the upper echelons of the
labour market hierarchy, i.e. in the big cities and particularly in their
central areas. Here are to be found the principal and regional offices of
central and local government, of public and private industry, and of
'pressure groups' such as employers and trade union organizations. Hence
these cities should represent the most important elements in the contact
network. Therefore all European capitals and cities down to the size of
500,000 have been included in this stUdy.
Accessibility to these nodes depends, however, on their proximity to
airports and the distribution of these means some modification to the choice
of nodes. Bonn and The Hague, for example, have been excluded as separate
nodes since they do not have their own international airports but share their
air traffic with Cologne and Amsterdam, respectively. Additionally, seven
other urban centers2 are included due to their functions as important
international and regional centers.
In the study of accessibility in Europe, different attempts have been
made to investigate the opportunities for face-to-face contacts. Attention
is focused on two such attempts dealing with opportunities for direct
personal contacts between ninety-eight European urban centers. The first
study investigates the different possibilities of round trips in a given day
by different means of transport (car, train, boat and airplane) between
each of the urban centers. The maximum length of stay time within the
destination center which still enables a return home the same day is used
to evaluate these possibilities. This first study is divided into two portions:
the travel opportunities for face-to-face contacts from each urban center to
all other centers are reported first. The term outbound maximum stay time
is used in this case. The corresponding reverse opportunities form the second
portion of this investigation. The term inbound maximum stay time is
used in this case.
The second major investigation focuses on the relationship between
the maximum stay time possible within a work day and the size of urban
population. Again two questions are put forward: what is the potential num-
ber of individuals who can be reached from city A during a day visit for x
hours stay, and what is the potential number of individuals who can
travel to city A during a day visit for x hours stay? The data obtained

2. The urban centers are Bilbao, Nice, Nantes, Toulouse, Geneva, Erfurt and Aarhus.
102 CONTACT POTENTIALS IN EUROPEAN CITIES

from these two questions are used as measures of the outbound and
inbound potential contacts for each urban center.

4.3.1 The European maximum stay time maps

The first major study investigates the different possibilities of round trips
in a single day by different means of transport between each pair of
the ninety-eight European urban centers. The stay time is the measure
used to evaluate these possibilities. It is assumed that the departure from
the home center does not occur prior to six a.m. and that return is before
midnight. Furthermore, these day trips must be possible at least three
times between Monday and Friday.
The study examined the transport possibilities in Europe during the
spring of 1976. Data for train and air travel are derived from the Thomas
Cook Continental Timetable and the ABC World Airways Guide which
contain all scheduled international trips by air, train and boat. The work
procedures used to evaluate European outbound and inbound maximum
stay time are:
1. The maximum length of time possible to remain within the destination
center and still return to the home center the same day is calculated
for each pair of the ninety-eight urban centers.
2. The results from the calculations, more than 9,500, are entered into a
98 x 98 matrix.
3. The values of the matrix are summed by row and column.
The row totals in the 98 x 98 matrix are used as measures of outbound
maximum stay time, while the column totals determine the inbound
maximum stay time for each of the ninety-eight cities.

Outbound maximum stay time. Figure 3 shows that the most favourable
zone for outbound maximum stay time occurs within an area delimited
by lines drawn between Paris-London-Hamburg-Munich-Milan-Lyon
-Paris. The only urban centers that show corresponding high values
outside this delineated area are West Berlin, Vienna, Rome and Man-
chester. This high accessibility area is called the Primary European Center.
The values for the urban centers within this area show variations but, as
seen from Figure 3, they constitute a reasonably homogeneous block.
Outside the Primary European Center, cities in Scandinavia, the British
Isles (excluding London and Manchester) and the south and west of
France form three relatively homogeneous accessibility blocks. The urban
centers within these three zones hold an intermediate position in terms of
ACCESSIBILITY IN EUROPEAN CITIES 103

~----------------~------------------------~-----------~
Figure 3. The European stay time map: outbound accessibility. The values are
expressed relative to the most accessible urban center (Paris = 100).

their outbound maximum stay time and are referred to as the Inter-
mediate European Centers.
The only urban area having a corresponding value to the Intermediate
European Centers and falling outside these three blocks is Madrid. Seen
from Figure 3, the variability in outbound maximum stay time in the rest
of Europe is large. The values vary between 0 and 30. Therefore, within
this low accessibility area there are different degrees of secondary posi-
tions. The urban centers within this zone are referred to as the Secondary
European Centers.
Another striking feature emerging from Figure 3 is the major difference
in values between East and West European centers, well illustrated in the
East and West German case.
104 CONTACT POTENTIALS IN EUROPEAN CITIES

Figure 4. The European stay time map: inbound accessibility. The values are
expressed relative to the most accessible urban center (Frankfurt = 100).

Inbound maximum stay time. The inbound maximum stay time map
(Figure 4) shows that most urban centers belonging to the Primary Euro-
pean Center have equally favourable positions regarding their inbound
travel. A striking change in ranking occurs for Copenhagen. It now belongs
to this group. Thus, Copenhagen has considerably better inbound than
outbound maximum stay time within Europe.
The urban centers belonging to the three blocks called the Intermediate
European Centers still keep their positions in relation to the 'secondary'
cities. But as their values overall are lower when compared to the Primary
European Center, the inbound maximum stay time of these blocks is
lower than their outbound maximum stay time. The opposite is true for
some places in the Secondary European Centers category (e.g. Barcelona,
Genoa, Zagreb and Warsaw). The outbound maximum stay time values
ACCESSIBILITY IN EUROPEAN CITIES 105

of these cities are now comparable to those of the Intermediate European

Centers.
Again, the differences in values between East and West European
centers remain.

Accessibility: inbound or outbound? Most models of spatial interaction

and connectivity make no disparity between inbound and outbound
accessibility. As is evident from Figures 3 and 4, however, the inbound
and outbound accessibilities in the European stay time maps are different.
This phenomenon will always occur because of the scheduling of
transport. In spite of differences, the two maps have many similar
features. It is obvious that the Primary European Center, the Intermediate
Centers and the Secondary European Centers basically contain the same
urban centers no matter if outbound or inbound accessibility is considered.
A numerical comparison between the individual values of the two maps
shows, however, some striking differences. One of these is that a majority
of the urban centers have a relatively higher outbound than inbound
accessibility. This is especially true for urban centers in the Intermediate
European Centers and in the Secondary European Centers. This disparity
can be explained to a great extent by the construction of the time tables.
Thus, the existing schedules of train and air traffic in Europe seem to be
planned in such a way that there are better opportunities to travel into the
European Center than in the opposite direction.

4.3.2 The European potential contact maps

In an earlier part of this paper it was mentioned how Tornqvist investigated
the regional balance between Swedish travel facility demand and supply.
The primary purpose of this section is to investigate the relative positions
of the ninety-eight European centers using a method which can be
regarded as a simplified version of that of Tornqvist.
To obtain quantitative expressions of the outbound and inbound
contact potentials of the ninety-eight urban centers, the relationship
between the maximum stay time and the size of each urban population 3 is
investigated. The population of each urban agglomeration is thus used as a
surrogate for the contact requirement, as it is impossible to find data for
the actual number of contact-intensive employees in most of the city

3. The populations of Bonn and The Hague are added to those of Cologne and Am-
sterdam.
106 CONTACT POTENTIALS IN EUROPEAN CITIES

regions. When using this surrogate measure for contact requirement a

linear relation between city size and contact requirement is assumed. In
Sweden we know, however, that reality is not like that; the big cities
normally have a higher percentage of their employees in contact-intensive
jobs than the smaller ones. A more adequate surrogate measure of contact
requirement could have been constructed. One would be to calculate an
index of contact requirement for each city in which factors like being a
capital or not, the presence of headquarters of national firms and
organizations, the presence of multinationals, together with city size,
could be incorporated. To calculate such an index, however, a lot of data
is needed which is not available at the moment.
To evaluate the contact accessibility of Europe two questions need to
be answered:
1. What is the potential number of individuals (out of about 120 million
in the ninety-eight urban centers) who can be reached from a given
urban center during a day visit for x hours stay?
2. What is the potential number of individuals who can travel to a given
urban center during a day visit for x hours stay?
The population figures required to answer these two questions were
collected from different statistical publications. Data concerning the
maximum stay time in each urban center were taken from the 98 x 98
matrix described in the first study.
The answers to the two questions are summarized for each center
through the construction of so-called location profiles. Figures 5 and 6
show, as an example, the location profiles for Paris, Copenhagen and
Malmli. Figure 5 shows the potential number of individuals who can be
reached from Paris, Copenhagen and Malm<:5 during a day visit for varying
lengths of stay. Figure 6 shows the potential number of individuals who
can travel to these cities during a day visit for 1-18 hours stay.
The urban centers' outbound and inbound potential contacts were
analyzed by various methods using their location profiles, the purpose
being to obtain specific measures of contact potentials. In the first experi-
ment, the area below the profiles for each urban center was calculated.
These areas were intended to measure the outbound and inbound contact
potentials, as the areas under the curves represent the total number of
inhabitants that a person living in, or travelling to, a given city could
visit during a single day. The results of this measure were not satisfactory
as identical computed values could represent totally different location
profiles (see Figure 7a).
ACCESSIBILITY IN EUROPEAN CITIES 107

L_r-_r-_r--;:::::::;::::::~=:;:::=::;S;;=;;=t-+ Max stay tIme

10 12 14 16 18 (hours)

Figure 5. The relationship between the maximum stay time and the potential
number of individuals who can be reached from Paris, Copenhagen and Malmo
during a day visit.

PoP. (millions)

80

70

60

50

40

20

10

L_...-~::=:==;:=:::~s::::::::;:::::::::;:::;;;=:;=?-
10 12 14 16 18
... Max. stay tIme
(hours)

Figure 6. The relationship between the maximum stay time and the potential
number of individuals who can travel to Paris, Copenhagen and Malmo
during a day visit.
108 CONT ACT POTENTIALS IN EUROPEAN CITIES

As an alternative to the first method, the following procedure was

utilized. The area between four and eight hours of length of stay for each
location profile was calculated (Figure 7b). This time interval was chosen
since it is assumed to be required in order to accomplish a day's work and
make a single day trip worthwhile.
Pop. (millions)

a. Max. stay time

2 4 6 8 10 12 14 16 18 (hours)

Pop. (millions)

b. '--___,---f:......:.~-L.+--._-r____,-__.----::::.,..-~ Max. s t a y tim e

2 4 6 8 10 12 14 16 18 (hours)

Figure 7. The contact potential calculated as the area (a) below the whole location
profile (b) between four and eight hours of length of stay.
ACCESSIBILITY IN EUROPEAN CITIES 109

This procedure of calculating the outbound and inbound contact

potentials is not an ideal solution, but unlike the first measure, these
results can be meaningfully interpreted. The results of the contact accessi-
bility analysis are mapped in Figures 8 and 9.

Outbound contact potentials. Seen from Figure 8 there are significant

spatial differences in outbound contact potentials within Europe. The
most favourable zone of outbound accessibility, called the Primary
European Center, is delineated by lines drawn between Paris-London-
Hamburg-Munich-Milan-Lyon-Paris. The only urban centers that have
correspondingly high values outside this zone are West Berlin, Vienna,
Rome, Dublin, Birmingham and Manchester.
The Scandinavian urban centers, the rest of the British Isles, and
southern as well as western France form three other relatively homo-

'--_--"590 km

Figure 8. The European potential contact map: outbound accessibility.

The values are expressed relative to the most accessible urban center (Paris = 100).
110 CONTACT POTENTIALS IN EUROPEAN CITIES

geneous outbound contact accessibility blocks. Because of their medial

position in terms of outbound accessibility, these three blocks are referred
to as the Intermediate European Centers. The only urban center having
similar values to the Intermediate European Centers and falling outside
these zones are Madrid and Turin.
The rest of the urban centers all have lower contact accessibility and
are referred to as the Secondary European Centers. Within this overall
low accessibility zone, Moscow, East Berlin and Barcelona are clearly the
most accessible.

Inbound contact potentials. The inbound contact accessibility map

(Figure 9) shows that most urban centers belonging to the Primary
European Center have the same relatively favourable positions in terms of
their inbound contact opportunities. When compared to the outbound

Figure 9. The European potential contact map: inbound accessibility.

The values are expressed relative to the most accessible urban center (Paris = 100).
ACCESSIBILITY IN EUROPEAN CITIES 111

contact accessibility map, Copenhagen has a much higher position and

therefore is placed within the Primary European Center.
A striking change occurs at the intermediate level of inbound accessi-
bility when compared to the European ouibound accessibility patterns.
Only the British cities can be considered to have an intermediate position.
With the exception of Copenhagen, as already mentioned, all of the other
intermediate outbound centers are clearly less accessible regarding their
inbound properties, and thus drop into the Secondary European Centers
category.

Toward a nodal hierarchy of accessibility. By comparing Figures 8 and 9

it is evident that some similar patterns exist between outbound and in-
bound contact potentials. The Primary European Center overall remains
identical in terms of contact accessibility in both cases. But important
differences clearly exist between outbound and inbound potential. The
major change occurs between the Intermediate Centers and the Secondary
Centers. This is indicated by the marked decrease in the number of cities
of the Intermediate European Centers inbound category and the increase
in the number of cities within the Secondary European Centers inbound
category. Thus, a significant property of existing European transport is
that for the last two types of centers outbound potential contacts are
clearly easier than inbound potential contacts.
Some other properties of the two contact potential maps are interest-
ing. The political boundary between East and West Europe is distinctive
and constitutes an important barrier to the development of a truly
integrated European contact system. Clearly, western Europe has higher
contact potentials than eastern Europe. Additionally, within almost all
European countries, one or two urban centers have significantly better
contact potentials than others in the country. That is, one or two cities
within each country have a favoured position regarding opportunities for
especially international face-to-face contacts. This favourable property
must be at the expense of the remaining urban areas.
A more detailed analysis would reveal at least two different kinds of
transport systems within Europe: one international and several national.
The nodes of the international one, that is, the network connecting the
different nations in Europe, consist primarily of one or two national
centers. From an accessibility point of view the cities under study are thus
arranged in a hierarchy in which national centers have a superior position
relative to the other cities. These other cities are primarily nodes in a
112 CONTACT POTENTIALS IN EUROPEAN CITIES

domestic transport system and, therefore, they are connected to the

European network via one or two national centers.

4.3.3 General conclusions

The results of the two investigations dealing with European outbound
and inbound accessibility have numerous similarities. Figures 34 and
Figures 8-9 illustrate that the Primary European Center, the Intermediate
European Centers, and the Secondary European Centers contain almost
the same urban centers no matter if the accessibility is expressed in terms
of maximum stay time during a day visit or in terms of contact possibili-
ties. A reason for this is that stay time is an important element in the
contact potential measure. Another reason for the similarities is that
the travel facilities between different cities are related to the cities'
population. This property results in travel opportunities being more
developed between the more populated city regions than between the less
populated. Of course it is difficult to determine what is cause and what
is effect in this case. The results concerning outbound and inbound
accessibility in Europe generally support these conclusions. It is within the
most densely populated portions of Europe that the urban centers with
best accessibility in both stay time and contact possibilities are found.
Conversely, the least populated areas fall within the least accessible zones.
Specific exceptions from these general conclusions exist. The most
important one occurs when the urban centers are adjacent to the political
boundary between the East and the West. For example, the East German
urban centers and Prague are geographically part of the dense population
concentration of western central Europe, but their accessibility values in
the maximum stay time maps and the contact potential maps are rela-
tively low.

4.4 PERSONAL CONTACT NETWORKS AND REGIONAL DEVELOPMENT

The two investigations discussed in the second part of this paper have
dealt with regional balance between European travel facility demand and
supply in 1976. It was found that different nodes in different parts of
Europe differ greatly with regard to the possibility of maintaining direct
personal contacts.
To judge from a previous study dealing with the development of the
European air transport network during 1965-76, these regional differences
PERSONAL CONTACT NETWORKS 113

appear to have prevailed over a long time (Erlandsson 1977). With

improvements in technology and the transport network the ability to
maintain direct personal contacts has improved between almost all Euro-
pean urban centers. The major exception is the political division between
East and West. But, from the results of the air-transport network study, it
is clear that accessibility for the cities belonging to the Primary European
Centers has increased relative to the less accessible nodes. Thus, the
majority of the new air transport possibilities since 1965 have been
between those nodes that in 1965 already had the highest accessibility
values in Europe.
This trend since 1965 at the international level echoes a similar trend
at the national level in Sweden. Thus it may be suggested that the contact-
intensive employees in Europe have become more and more concentrated
in the most accessible nodes. As no data on the change and the regional
distribution of contact-intensive employees in Europe exist, there is no
possibility of verification. By studying the development and locations of
the headquarters of the biggest national firms and organizations, the
multinationals, and the international organizations, some indication of the
trend could be obtained.
In future research dealing with accessibility in the European system of
cities, the intention is to carry out a series of experiments with opera-
tional 'models of Europe' similar to those which Tornqvist used for
Sweden. The primary purpose of these model experiments is to test how
different possible developments in European trade policy (e.g. enlarged
markets, customs unions, trade agreements) may affect contact possibili-
ties in various parts of Europe, i.e. to develop the European contact
accessibility maps of the future. These experiments will be based on three
changes which can affect accessibility maps:
1. a change in the need for direct personal contacts in business and govern-
ment,
2. a change in the regional distribution of contact-intensive employees,
3. a change in the transport network.
Which changes are probable depends very much, of course, on the time
range chosen for future studies. Radical change in the personal contact
requirements of various occupations is probably something of the fairly
distant future. But, as can be seen from the Swedish studies mentioned
previously, the regional distribution of employment and the transport
system can change comparatively rapidly.
Of course experiments of this kind give rise to a series of both
114 CONTACT POTENTIALS IN EUROPEAN CITIES

methodological and theoretical problems which must be solved in order

to get an accurate understanding of the European contact potential maps.
The biggest problems seem to be those of geographical scale and political
division.
The Tornqvist approach to the investigation of contact potentials seems
to work quite well on an intra-national scale. Starting with the methodo-
logical side of the model, it is today more or less impossible at the inter-
national scale to copy the Tornqvist approach because we do not have
data on the number of contact-intensive employees needed to calculate
the contact requirement for each node in the European system of cities.
Using the size of population as a surrogate measure for the contact
requirement, as has been done in the investigations described in this chapter,
must be regarded as too simple. But, as has been mentioned earlier, the
data problem can be overcome by calculating some kind of index of
contact requirement. But even if data on the actual number of contact-
intensive employees were available for each city region, it is not clear how
far these would express the real contact requirement on an international
scale, because a great part of this requirement is likely to be intra-national
and not international. To obtain adequate expressions for contact require-
ment on an international scale is thus a very difficult task.
To carry out experiments concerning changes in the transport network
will also cause problems, some of which are political. In Sweden, for
example, it was possible to examine the potential interdependence
between changes in employment and changes in travel facilities. Being one
political unit it was also possible to suggest changes in demand and supply
of travel facilities that could decrease inequalities in regional employment
structure. Doing the same in a study of several countries is very difficult,
as goals and means of regional policy in different nations may conflict.
This means that changing the European transport network - particularly
the air transport network - in order to increase accessibility on an inter-
national scale will be very difficult. The transport policy that could be
regarded as the 'best' for Europe as one unit could thus be inappropriate
for the individual nation with its own intentions in reducing spatial
inequalities. The EEC, for example, must face problems of this kind when
regional policy measures made in Brussels are intended for this big area
consisting of several nations.
Another crucial point in this respect is the question of international
agreements on the scope of international air traffic. Without knowledge
about these agreements, which function as constraints for international
PERSONAL CONTACT NETWORKS 115

trips, it would be difficult to integrate national nodes into a wider

international perspective.
On the other hand, regional policy measures adopted within different
countries must consider not just internal development trends but also
trends in international development. Otherwise the risks of good inten-
tions being destroyed by outside factors are evident. This conclusion leads
to other problems which also must be considered in future work.
One of these is the selection of study objects. To understand the results
from the now existing studies and to integrate them with regional develop-
ment in different countries, the experiments should probably include
more urban nodes.
Another factor is the costs of maintaining direct personal contacts and
what effects these can have for regional development in Europe. As the
study area is large, air transport possibilities are of crucial importance
for most nodes in obtaining reasonably 'good' contact potentials. Inherent
in these trips are normally quite high transport costs compared to other
means of transport. As has been mentioned earlier in this paper, employees
with contact-intensive functions account for 80 percent of all trips by air
in Sweden. It is reasonable to assume, in the absence of evidence to the
contrary, that this proportion is about the same elsewhere in Europe.
This means that most air passengers are on expense accounts with com-
panies and organizations. From this point of view costs for the individual
should not be a severe constraint for national and international trips.
In contrast to international air transport services the domestic air
traffic in most European countries is run by one air company with little
competition. This means that the air company can continue to increase
charges for their services as most passengers need to travel fast and many
are, again, on expense accounts with their firms. This policy can continue
as long as demands for travel opportunities exist. Nodes characterized by
'good' contact possibilities also seem to increase this demand and, as a
consequence, they constitute favourable locations for different kinds of
administrative activities. But what about cities where the demand for air
transport service is not regarded as enough from the air company's point
of view? This question seems to be a very critical one for regional develop-
ment within a country, especially for those firms that intend to locate in
problem regions but from a functional point of view are more or less
hindered because of lack of transport facilities. Hence it may be necessary,
in drawing up policies for regional development, to subsidize the air
transport network to provide the sort of travel facilities required by the
116 CONT ACT POTENTIALS IN EUROPEAN CITIES

decision-makers in finns and organizations. If transport facilities cannot

be subsidized, finns will probably continue to locate their central offices
and laboratories close to accessible nodes - especially those nodes which
offer both intra-national and international contact possibilities. This
would increase rather than reduce spatial inequalities in opportunities.

REFERENCES
Engstrom, M.-G., Regional arbetsfordelning. Nya drag i forviirsarbetets geografiska organisation i
Sverige. C.W.K. Gleerup, Lund, 1970.
Engstrom, M.-G., and B. Sahlberg, Travel demand, transport systems and regional development.
Modes in co·ordinated planning. Lund Studies in Geography, Ser. B., no. 39, Lund, 1973.
Engstrom, M.-G., and B. Sahlberg, Kommunikationer och regional utveckling. Nuvarande och
framtida utvecklingslinjer. SOU 1974: 3, bilaga 7. Stockholm, 1974.
Erlandsson, U., Kontakt och resemojligheter i Europa. Department of Social and Economic Geo-
graphy, University of Lund, 1977 (mimeographed).
Hall, P., The world cities. Weidenfeld and Nicholson, London, 1966.
Hedberg, B., Kontaktsystem inom svenskt niiringsliv. En studie av organisationers extema person·
kontaker. C.W.K. Gleerup, Lund, 1970.
Kristensson, F., People, firms and regions. The Economic Research Institute at the Stockholm
School of Economics, Stockholm, 1967.
Sahlberg, B., Interregionala kontaktmonster. Personkontakter inom svenskt niiringsliv - en flyg·
passagerarstudie. C.W.K. Gleerup, Lund, 1970.
Tornqvist, G., Flows of information and the location of economic activities. Lund Studies in
Geography, Ser. B., no. 30, Lund, 1968.
Tornqvist, G., Contact systems and regional development. Lund Studies in Geography, Ser. B., no.
35, Lund, 1970.
Tornqvist, G., Contact requirements and travel facilities - contact models of Sweden and regional
development alternatives in the future. Systems of Cities and Information Flows. Two Essays.
Lund Studies in Geography, Ser. B., no. 38, Lund, 1973.
5. Human reactions to spatial
diversity: Mobility in regional
labour markets
J.A.M. HEIJKE and L.H. KLAASSEN*

It may be useful, particularly because the subject of this contribution is

mobility in regional labour markets, to introduce briefly the concepts
that will be used in the following sections.
The first important distinction to be made is that between geographical
migration and geographical mobility. In quite a number of publications
these two expressions are used for one and the same phenomenon, viz. the
number of people (or the number of people expressed as a fraction of
another number, like that of total population or working population)
moving from one region to another. Following Klaassen and Drewe
(1973), we understand mobility as the propensity to move; the combined
impact of mobility and the reason, or impulse, to move is called migration.
This may be expressed as:

m = p.i (1)

where:
m : the size of the migration flow
p. : mobility
i : the size of the impulse.
Recalling the title of this contribution with these definitions in mind,
we may say that spatial diversity is to be considered as the main factor
behind the impulses, and mobility as the propensity to react to this spatial
diversity.
Spatial diversity concerns such factors as population, employment
positions, housing and living environment; in general, it refers to the
differences between regions in social and physical structure. Mobility is
associated with the preference structure of the population as far as a
change in residence is concerned. The preferences are related among other

* Netherlands Economic Institute, Rotterdam.

118 HUMAN REACTIONS TO SPATIAL DIVERSITY

things to the social group to which people belong, their training level, and
the experience gained with previous migration. In economic models the
preferences are generally found back as parameters, with the above-
mentioned impulses figuring as variables. That is understandable; a popula-
tion's preference structure is pretty constant. From a policy point of view
it is a one-sided approach, however, for it is possible for a government to
affect the population's mobility as well as the above impulses by policy
measures. In that case, one needs to understand both the migration model
and the mobility model superposed on it.
The effectiveness of a combined impulse and mobility policy can be
defined as:

dm = p.di + idp. (2)

The impulse policy (di) causes the migration to grow by p.di. The effec-
tiveness of the total impulse action is increased by idp., however, if
mobility is increased by dp.. So, improving educational facilities could (in
the long run) improve mobility and thus stimulate migration to regions
where a policy of stimulation is conducted.
About the factors behind mobility, little is known as yet. In this con-
tribution we shall confine ourselves for the moment to a discussion of
the structure of mobility as it emerges from two studies, each dealing with
a specific kind of migration flow. The first (Section 5.1) is concerned with
flows on the intra-national level (interregional migration) within Great
Britain, France, the Netherlands, and Sweden; the second (Section 5.2)
analyses flows on the international level, in particular from Mediterranean
countries to Western Europe. We shall compare the results of the two
studies, and analyse the structure of mobility in each case. The factors
that determine mobility will not be gone into here, though in the discus-
sion of the results of the enquiry into international migration an attempt
will be made to introduce into the analysis the concept of psychological
distance. Special attention will be paid to the influence of kinship
between the languages spoken in the countries of origin and destination
(Section 5.3). Finally, some conclusions will be drawn in Section 5.4.
MIGRATION/MOBILITY IN EUROPE 119

5.1 MIGRATION AND MOBILITY IN FOUR EUROPEAN COUNTRIES

As said above, an analysis was made of interregional migration flows

within four European countries, viz. Great Britain, France, the Nether-
lands, and Sweden. The theoretical basis, empirical results, and political
consequences of that analysis were published extensively by Klaassen and
Drewe (1973), but it seems useful to recapitulate them briefly with a view
of comparing them with those of the study on foreign workers.
One and the same equation was tested for all four countries. This
equation was:

(3)

where:
Mij total number of migrants from region i to regionj; (during a given
period - for the countries involved between 1960 through 1968);
total number of jobs in industries and services in region j in the
base year of the period;
total number of people between 15 and 65 years of age in region i,
also in the base year;
D.. : distance between the centres of regions i and j in car kilometres.
If 0 is the factor most frequently cited on the household and individual
level as an impulse to undertake long-distance interregional migration.
This factor represents an indicator of the prospects on the labour market
of the region of destination. These prospects are a function of the number
of job opportunities coming available through the creation of new positions
and the vacancy of old ones. We will assume that these two components of
change are more or less proportional with 0' Naturally we should have
preferred to differentiate between skill groups for this variable as well as
for Li' but lack of data prevented such detail.
Li represents the number of people in the active age groups in the
region of origin. It is the least ambiguous definition that could be given,
ensuring a high degree of comparability between countries. Li then serves
as an indicator of the group of potential migrants. We assume that this
group is approximately proportional to Li . Differentiation between the
sexes would seem advisable, particularly because their participation rates
differ greatly in the various countries, e.g. between Sweden and the
Netherlands. However, such refinement would severely complicate the
analysis, for it is households rather than individuals that account for the
120 HUMAN REACTIONS TO SPATIAL DIVERSITY

bulk of the migration flows, and households tend to consist of both men
and women. In fact, a different kind of analysis altogether would be
required if sex differentiation were to be introduced. The variable chosen,
then, was simply the number of people in the active age groups.
Dij represen ts the distance between the regions i and j. It stands as a
proxy for psychological distance rather than for the costs of bridging the
physical distance. Assumedly, the financial costs of moving a household
from one region to another are, for one thing, virtually insensitive to
distance, and for another, are negligible compared to the net monetary
advantages of the move, though it must be admitted that at short term the
fmancing of a move may constitute a barrier. What really counts is the
pain of leaving one's friends and relations for a completely different
environment with a different social, sometimes also religious and cultural
structure. That is the 'cost' that, as we assume, increases with physical
distance.
But there are other migration impulses beside the factors mentioned.
Important impUlses may come from the housing facilities and the living
environment; the income advantages to be expected may also playa role.
Lack of data made it impossible to assess the impact of such impulses; the
migration enquiries to be discussed here were, indeed, purely labour
market-orien ted.
Furthermore, we want to draw attention to the fact that the variables
in the migration equation are expressed in absolute levels; in that way it is
relatively easy to determine how a changed impulse, e.g. the creation of
job opportunities, affects the number of migrants.
The results of the analysis are presented in Table 1. Looking at mobility
in its relation to each regression coefficient, we think a distinction should
be made between general mobility as represented by the value of the inter-
'cept, and specific mobility related to a specific impulse variable.
General mobility is highest in the Netherlands (- 3.36) and lowest in
France and Great Britain (- 11.37 and - 10.43, respectively). Sweden
occupies an intermediate position (- 6.81). The influence of distance is
obviously highest in the Netherlands (- 0.009) and lowest in Sweden
(- 0.0014), both France and Great Britain taking an intermediary position.
The coefficients of E. Gob opportunities in the immigration region)
J
suggest a certain relation with the coefficients of Li (potentially active
persons in the emigration region). In both cases the Netherlands show a
low mobility, France and Great Britain a high one, while Sweden takes an
intermediary position.
MIGRATION/MOBILITY IN EUROPE 121

Table 1. Regression statistics for France (F), Great Britain (GB), the Netherlands (NL)
and Sweden (S). * Observations between 1960 and 1968. * *

Impulse Ej Li Dij Intercept R2

Country InQo

F 1.00 1.06 -0.0023 -11.37 0.69

(20.2) (18.6) (16.7)

GB 0.86 0.93 -0.0023 -10A3 0.76

(8.8) (8.4) (5.3)

NL 0.59 0.61 -0.009 - 3.36 0.92

(13.5) (12.6) (21.5)

S 0.83 0.77 -0.0014 - 6.81 0.62

(17.0) (13.7) (15.5)

* The number of regions in France is 21, in Great Britain 8, in the Netherlands 11, in Sweden 23;
the numbers are dictated by the structure of available data.
** Numbers in brackets represent t-values.
The results of the analysis are summarized in Table 2 in qualitative terms.
The outcome for general mobility appears to mirror that of the push-and-
pull factors.
In France and Great Britain, while general mobility is low, push factors
(the number of people present in a region) as well as pull factors (job op-

Table 2. Mobility structure in France, Great Britain, the Netherlands and Sweden.

Mobility General Push-and-pull Spatial

Country mobility factors mobility mobility

F,GB low high medium

NL high low low
S medium medium high

portunities elsewhere) have considerable influence, and we should take

such general statements as 'Le Fran~ais ne se dep/ace pas' and 'We British
don't move' with a grain of salt.
The next feature that strikes us is that spatial mobility is considerably
higher in Sweden, where distances are very long, than in the Netherlands,
a country laid out on a much smaller scale. France and Great Britain,
where distances are more modest than in Sweden but still longer than in
the Netherlands, score 'medium' on the spatial-mobility factor. All in all
the conclusion seems justified that the perception of distance in a country
122 HUMAN REACTIONS TO SPATIAL DIVERSITY

depends on that country's size: the smaller the country, the more distance
seems to be a barrier to migration.
The most important feature emerging for the Netherlands is the high
sensitivity of the working population to distances. Job creation, not a very
powerful migration stimulus at all times, is made even more ineffective by
the people's resistance to moving over any sizeable distance. Klaassen
and Drewe (1973) examined the effectiveness of job creation in regions
of destination at various distances. They found that the creation of jobs at
a short distance (say 100 kilometers) was most successful in inducing
migration in the Netherlands, with Sweden coming second. In France and
Great Britain such job creation was a great deal less effective. The impact
of job creation rapidly diminishes in the Netherlands as distances increase;
at 300 kilometers it is very nearly as low as in France and Great Britain.
The effectiveness of job creation at long distances turns out to be
considerably higher in Sweden than in the other countries studied.
On the grounds of the foregoing it may be assumed that attempts under-
taken in the Netherlands to stimulate migration over a large distance (e.g.
from the west to the north), by creating job opportunities in the area of
destination, will not be successful unless mobility is stimulated as well by
suitable policy measures. By the same token, the planned removal of
government agencies from the western to the northern part of the country
may be strongly opposed because of the relatively large distance involved.

5.2 MIGRATION AND MOBILITY OF FOREIGN WORKERS

A second study that seems relevant in the context of this contribution is

one by Heijke (forthcoming) on the migration of foreign workers. In this
study two groups of countries are considered, viz:
1. the immigration countries: Austria (A), France (F), West Germany (D),
Switzerland (CH), Belgium (B), the Netherlands (NL) and Sweden (S);
2. the relevant emigration countIj.es (the Mediterranean countries): Greece
(GR), Italy (I), Portugal (P), Spain (E), Yugoslavia (YU), Turkey (TR),
Algeria (DZ), Morocco (MA) and Tunisia (TU).
The data on which the analysis is based are presented in the table in
Appendix A. It appears from this table that the data base, if not so homo-
geneous as one would wish, is still sufficiently complete for the analysis.
The figures relate to the year 1969.
Three remarks should be made before the results of the analysis are
MIGRATION/MOBILITY OF FOREIGN WORKERS 123

presented. The first is that only migration from Mediterranean countries

to Western European countries was taken into consideration; as the
migration flows in the opposite direction are negligible, their inclusion
did not seem opportune.
The second is that, because incomes in Mediterranean countries and the
Western European countries differ considerably, an income-difference
factor (~Wji) was added to the variables used in the analysis of migration
and mobilIty in the four European countries. The other variables have
been defined similarly to those in the latter analysis.
The third remark refers to the possible effects of institutional factors as
regards the migration from Mediterranean to Western European countries.
While in the case of interregional migration individuals are free to move,
international migration, by contrast, is governed by strict rules. These
rules concern primarily the question of whether migration is permitted or
not. There is no sense in studying migration other than to and from coun-
tries between wnich it is permitted, which is the case for most of the
countries studied in this investigation. Only a few cells of the migra~ion
matrix involved are empty; these have not been included in the analysis.
The second important question is: has migration, as far as it is allowed,
been influenced by the rules imposed on it, and to what extent? It is
difficult to answer that question, among other reasons, because institu-
tions, and the use or abuse made of them, tend to follow developments
instead of controlling or even checking them. Because institutions, once
migration is allowed, presumably are not an independent variable, and
moreover are difficult to operationalize, they have for the moment been
omitted from the analysis.
Two equations were tested. In the first the influence of distance was
expressed, as in the European analysis, as an e-function; in the second as
a power function, which means that in the statistical test applied to the
natural logarithm of the number of migrants, the explaining
factor in the first test was Dij , in the second In Dij .
The results were:
In Mq = 0.68 In ~ Wji + 1.46 In Ej + 1.22 In Li - 0.002Dq - 5.61
(3.57) (6.10) (4.44) (4.71) (3.11)

(4)
and
124 HUMAN REACTIONS TO SPATIAL DIVERSITY

In Mq = 0.72 In ,::H1'/i + 1.S9ln Ef + 1.17 In Li - 2.69 In Dq + 11.07

(3.76) (6.69) (4.31) (S.OO) (2.80)
Rc2 = 0.661 (S)
The results do not differ much for the two approaches. Neither the
regression coefficients nor the t-values or correlation coefficients differ
significantly. The results of the second approach, which are slightly
better than those of the first, are presented graphically as a Tinbergen-
diagram in Figure 1.
Using the first equation (4) as a basis for comparison with the results of
the interregional migration analysis for the four Western European coun-
tries, we find some significant facts:
1. Income differences between Mediterranean countries and Western
European countries do playa role but their contribution to the explan-
ation of the variance in migration is modest (compare Figure 1).
2. The influence of push-and-pull factors is considerably greater for people
from Mediterranean countries than for migrants from any of the Euro-
pean countries. In other words, Mediterranean people are much more
sensitive to the pressure of the population at home and the pull of jobs
available elsewhere.
3. The restrictive influence of distance for Mediterranean migrants is com-
parable to those from Great Britain and France. This influence is
stronger than that in Sweden, but quite a bit weaker than that in the
Netherlands.
4. General mobility in Mediterranean countries as expressed by the loga-
rithm of the constant in the equation is at -S.6l lower than that in
the Netherlands. It is considerably higher than the figure for Great
Britain and France and comes actually close to that for Sweden (- 6.81).
However, it seems that our including income differences in (4) but not in
(3) is now causing us trouble. Because these differences play but a modest
role in the explanation of the migration variance, dropping them from (4)
will probably increase the intercept. In view of the average contribution of
the income differences we suspect that in leaving out the income differ-
ences from (4), thus making (4) identical in form to (3), we shall cause the
intercept to rise to almost zero. The conclusion may be that the general
mobility in Mediterranean countries probably surpasses that in European
countries, in the sense that there is hardly any general migration-
restricting factor as far as migration from Mediterranean to Western
European countries is concerned.
MIGRA TION/MOBILITY OF FOREIGN WORKERS 125

.6 = Ln(M ij )"
.5 = Ln(M ij )
.4
.3
.2
.1

-1
-2
-3
.7 0,72 Ln(aWji)
.6

Ej=working population in
,00""', QOO ~,,;'" ;0 j 1j
-2,69Ln(Dij} Dij= distance (as the crow flies
between i and j

-21
-22

.2

-1

Switzer-
-3 Austria France Germany land Belgium lands Sweden country
4 .
- 124561234567B91234567B91234561234567B9123456B91 234567B9 )
country i
Figure 1. Migrants from Mediterranean countries.
* The countries 'i' are numbered according to the appendix.
126 HUMAN REACTIONS TO SPATIAL DIVERSITY

5.3 PSYCHOLOGICAL DISTANCES

The residuals presented at the bottom of Figure 1 suggest systematic

factors at work to help direct the flows between the two groups of
countries. One hypothesis is that religion plays ~ role, so that, for
instance, migrants from Spain, Portugal and Italy would preferably go to
the Roman Catholic countries: Austria, France and Belgium. The
hypothesis was tested, and the outcome was negative. The hypothesis
that language differences might be significant was the next one to be
tested, with much more promising results.
For the language test, the immigration countries were divided into
three groups, viz.:
1. France as a French-speaking country, and Belgium, where French
is one of the main languages, spoken widely even in its Flemish pro-
vinces;
2. Switzerland as a multi-lingual country;
3. Austria, Germany, the Netherlands, and Sweden, where Germanic
languages are spoken.
The emigration countries were also divided into three groups, viz.:
1. Algeria, Morocco, and Tunisia, as countries with French as important
second language;
2. Spain, Italy, and Portugal, as countries where other Romance languages
are spoken;
3. Greece, Turkey, and Yugoslavia, as countries where neither Germanic
nor Romance languages are spoken.
The residuals per country group are presented in Table 3.
Table 3. Residuals per country group.

Origin DZ E GR Total
MA I TR
Destination TV P YV

F,B + 6.7 + 3.9 -2.7 + 7.9

CH + 2.8 + 1.6 +4.4
A,D,NL,S -8.8 -5.0 +1.5 -12.3

Total -2.1 + 1.7 +0.4 o

The first conclusion that may be drawn from the table is that the sum of
the residuals for each group of emigration countries is relatively small
compared to the residuals for each group of immigration countries
FINAL REMARKS 127

separately. That means that there are hardly any significant differences
between groups of Mediterranean countries as far as their total emigration
is concerned.
A second conclusion is that there seems to be a strong preference for
France and Belgium as immigration countries aad a negative preference
for the Germanic countries, with Switzerland taking an intermediary
position. An explanation for this preference could be the distribution of
language-groups among immigration and emigration countries. The coun-
tries with French as important second language (DZ, MA and TU) show
considerable positive preference for migration to France and Belgium and
a strong negative preference for the Germanic countries. The same, but to
a lesser extent, holds for the other Romance-languages-speaking countries
(E, I and P), where we find positive preference for France, Belgium and
Switzerland (possibly also due to the fact that in Switzerland Italian is
also spoken) and a negative preference for the Germanic countries. The
remaining group of countries, whose languages have no direct kinship with
either Romance or Germanic tongues and who therefore can be con-
sidered to have a free choice, show a slight preference for the Germanic
countries.
The results confirm the expectation that language barriers are an
important element of psychological distance, and affect migration
behaviour significantly. Indeed, further analysis in that direction might
well prove fruitful. On the whole it seems that in migration analyses, more
attention should be given to social and cultural factors, and'in particular
to language differences, than has been given so far.

5.4 FINAL REMARKS

The analysis of migration and mobility behaviour is an essential element

of regional science; that kind of analysis tries to find out why people
move from one region to another. Behaviour patterns tend to change
over time; to study the changes properly it would be necessary to
carry out investigations like those presented in this chapter regularly for a
number of consecutive years. We could then study how coefficients
gradually change over time, finding perhaps general mobility increasing
and distance becoming less and less of a barrier to migration. Unfortunately,
given the lack of appropriate data, it is unlikely that results of such studies
will become available for quite a few years.
128 HUMAN REACTIONS TO SPATIAL DIVERSITY

At any rate, as distance is becoming less important in migration, we

shall have to be careful how to define regions for the purpose of the
analysis. It is doubtful already that data based on the administrative divi-
sion of the Netherlands into provinces are still· adequate for migration
analysis. Among the migrants could be those who change their residence,
crossing a provincial border in the process, while keeping their old jobs.
So, a proper analysis would have to consider at least two different types
of migrants: those who change their job location, and because the distance
from their home to their new place of work is too great, also move house,
and those who change their residence simply because they prefer living
somewhere else, but keep their old job. Consequently, particularly in a
small country like the Netherlands, commuting and migration tend to
become objects of one set of decisions.
Another factor to be mentioned in relation to distance mobility is the
density of population and socioeconomic activities. Greater density means
many and manifold opportunities for living and working within easy
reach; it makes for shorter migration distances and may even make migra-
tion altogether unnecessary. That is indeed in concordance with the low
distance mobility found in the densely populated Netherlands and the
high distance mobility prevailing in a sparsely populated country like
Sweden. The clusters of population and socioeconomic activities are
unequal in size and unevenly spread across space; our regional division is
probably inadequate to allow a satisfactory representation of these
inequalities. An attempt to include the density aspect in the analysis will
make high demands upon the statistics available and their spatial division.
Which is not to say that the problem does not merit closer attention; on
the contrary.
FOREIGN WORKERS IN EUROPEAN CITIES 129

APPENDIX.Foreigners working or staying in a number of Western European

countries (by country of origin, 1969)

Countries of Austria France Western Switzer- Belgium Nether Sweden

destination (only on Germany land on lands (only
workers) January on on October on workers)
Countries on 1, 1969 September January 1, 1969 July 1, on
of August 30,1969 1, 1969 1968 January
origin 1,1969 1,1969
a a b a a c a

Greece 275 10,885 271,300 8,000 13,760 2,300 5,940

Italy 860 632,080 514,600 522,600 176,250 14,500d 5,325
Portugal 367,425 37,500 2,200 4,435 3,100 680
Spain 170 667,610 206,900 87,700 47,745 18,300 3,185
Yugoslavia 48,105 43,340 331,600 16,100 3,415 2,400 14,015
Turkey 8,060 7,160 322,400 7,800 11,335 13,500 1,785
Algeria 562,000 3,300 3,460 155
Morocco 119,520 9,100 20,980 12,800d 470
Tunisia 73,260 3,100 1,290 100 135

Subtotal
Mediterranean area 57,470 2,483,250 1,699,800 644,465 282,670 67,000 31,690
Other countries of
origin 10,310 516,750 e 681,300 288,675 172,935 f 137 ,OO~ 107,000h

Total immigrants 67,780 3,000,000i 2,381,100 933,14oi 455,605 204,000 138,690

Population countries
of destination
(in millions) 7.4 50.3 60.8 6.2 9.6 12.7 8.0

Immigrants from
Mediterranean area
as a percentage of
the popUlation 0.8 4.9 2.8 10.5 2.9 0.5 0.4
Total number of
immigrants as a
percentage of
the population 0.9 6.0 3.9 15.2 4.7 1.6k 1.7

a Source: Migration today, nr. 14, Geneva 1970.

b Source: Wirtschaft und Statistik, August 1972.
c Source: Statistisch Zakboek, CBS, 1969.
d NEI's own calculations based on CBS data end of 1967 and end of 1972 concerning the number
of foreigners present and migration balances in intermediate years.
e Among whom approximately 131,000 Poles, appro 250,000 Africans (including 60,000 from
Senegal, Mali and Mauretania), 35,000 North and South Americans and West Indians.
f Among whom 3,175 Congolese (Kinshasha).
g Exclusive of Surinamese, Antillians (35,000) and Moluccans (25,000).
h Among whom 19,000 Danes, 75,000 Finns, and 13,000 Norwegians.
130 HUMAN REACTIONS TO SPATIAL DIVERSITY

Exclusive of 129,858 seasonal workers (including 119,301 Spaniards, 3,110 Portuguese, 2,408
Italians and 2,079 Moroccans).
Exclusive of seasonal workers (114,081 in August 1968 and 14,233 in December 1968) and
frontier workers (63,062).
k Inclusive of Surinamese, Antillians, and Moluccans this percentage becomes 2.1.

REFERENCES
Heijke, J.A.M., L.H. Klaassen, and C.J. Offereins, Towards a labour market model. H.D. Tjeenk
Willink bv, Groningen, 1975 (published in Dutch).
Heijke, J .AM., Socio-economic aspects of foreign labour. (Forthcoming, to be published in Dutch.)
Klaassen, LB. and P. Drewe, Migration policy in Europe: a comparative study. Saxon House,
Lexington Books, Famborough and Lexington, 1973 (and literature quoted therein).
Part II: Policy-oriented studies in spatial
inequalities and regional development
6. Spatial equity: Some anti-theses to
current regional development doctrine
WALTER STOHR and FRANZ TODTLING*

6.1 SPATIAL EQUITY AND REGIONAL POLICY

Spatial equity or the reduction of spatial disparities of living levels is a

key objective of most national urban and regional development policies.
The present paper deals mainly with experiences in market and mixed
economies, although its basic conclusions would seem to be valid also
for centrally planned economies.
In some of the countries analyzed this objective is formulated in general
terms such as 'balanced development between regions' or 'reduction of
disparities between West and East' (France, Rapport 1975) or 'the preven-
tion of regional imbalance' (Great Britain, Cameron in Sant 1974, p. 15).
More explicit formulations are used in Sweden (Swedish Ministry 1973)
and in the German Federal Republic (Raumordnungsprogramm 1975;
Bundesminister 1968, 1970, 1972 and 1974) where a stepwise definition
of the 'attainment of more equal living conditions' is undertaken through
the specification of components of basic living conditions (employment
facilities, social, commercial and cultural services, 'good' environment,
etc.) and by defining minimum standards and maximum accessibility
ranges for them.
The results of these policies with regard to (1) objective material
indicators of living levels on the whole have been poor, sometimes
counterproductive, as the analyses surveyed below show, while with
regard to (2) the more subjective perception of - frequently immaterial

* University of Economics, Vienna, Austria. This is a revised version of a paper origin-

ally presented at the 13th European Regional Science Congress, Copenhagen,
August 1976, and published in Vol. 38 of the Papers of the Regional Science
Association. The paper was presented further at the University of Groningen in
December 1976 and has also been tabled for discussion in panel form by the
organizer of the 14th European Regional Science Congress at Cracow in August
1977, from which this revised version has benefited.
134 SPATIAL EQUITY: SOME ANTI-THESES

- living conditions by specific regional communities, their results in

general also seem rather unsatisfactory, but have as yet hardly been
analyzed systematically.
The poor results of regional policy in terms of (1) objective material
living levels are expressed by the fact that spatial disparities in material
living levels in most market and mixed economies have not decreased, or if
they have decreased at one level (e.g. the interregional), they have usually
increased at other spatial levels (e.g. the intra-regional or urban level, as is
shown in Section 6.3.1). This although material living levels have been the
major concern of regional policies. Results in material living levels may
have been better in centrally-planned economies. The (2) subjective
perception of conditions of living by concrete regional communities seems
to show growing discontent on the part of subnational social groups
(including local and regional ones) about the increasing impact upon them
of exogenous economic and political determinants (large-scale functional
changes) and their diminishing ability to resist them and to shape their
own destiny within large and still expanding economic and political
systems. The increasing dissatisfaction of ethnic minorities in many
European and other countries may be a symptom of this diminishing
ability as well as local resistance against urban freeways, massive urban
renewal schemes, major airport projects, nuclear power plants, etc. This
applies to market and mixed economies but even more so to centrally-
planned economies where the self-determination of local communities
usually is smaller still.
Both these facts have to be seen against the background of an increasing
functional specialization and integration (between sectors as well as
between regions and nations) and of the increasing mutual interdepen-
dence resulting from it. Large-scale functional changes in demand, in
technology, raw material and energy prices, etc., which are transmitted
vertically within or between sectors and trans-regional or trans-national
organisations are causes of major changes in territorial space (Pred 1976
and 1977). Along with the increasing scale of private and public
functional decision-making units, such vertically initiated changes cause
local and regional disturbances and fall-outs of increasing complexity in
the fonn of environmental pollution, unemployment, congestion, idle
regional resources, etc. which are to a great extent left to territorially
organized communities 'to be mopped up' (Friedmann 1976; 1977).
Regional policies in our present highly interrelated and complex social
systems are increasingly concerned with mitigating the fall-out of func-
SPATIAL EQUITY AND REGIONAL POLICY 135

tional (vertical) changes upon territorial communities. Spatial develop-

ment policies can therefore be considered a 'countervailing power' to
sectoral or other functionally oriented trends and policies. With the
increasing scale of functional (sectoral, organisational, etc.) interrelations
and dependencies, the emphasis of spatial development policies has
shifted from the (formerly dominant) local and regional level to inter-
regional, national and even multi-national levels (such as the European
Community's regional policy). The greater the territorial scale, however,
the more the territorially-organized interests at this scale will approach the
aggregate functional ones (at the world scale they become identical) and
the more they will tend to lose sight of the perceived requirements of
small- and intermediate-sized territorially-organized social groups. The
less, therefore, that regional policy will be able to fulfil the function of
countervailing powers in the interest of the abovementioned local and
regional communities.
This chapter is about the relations between the well-being of small groups
and regional planning at multi-regional, national, and multi-national levels.
Spatial equity will be considered, in somewhat broader terms than is
usually the case in planning analysis and practice, not only as spatial
equality of living levels but also as equality in options of group develop-
ment and human self-realisation over all populated parts of the country.
Besides the criteria usually considered, such as regional product, regional
income, access to employment opportunities, to basic services, or to
minimum environmental quality, there will also be included the right of
individuals and small groups to determine their immediate natural and
human environment and to exert adequate control on the influence of
external economic, technological, cultural and other influences which
affect their well-being. Equity is therefore considered not only in terms of
equal socioeconomic levels but also in terms of equal opportunities of
individuals and groups for diversity and for being different (Matzner
1976). In view of the existing diversity of individual and group disposi-
tions and aspirations this is the only way of facilitating equal opportunities
to all for a maximum degree of human self-realization.
These opportunities have, in the past decades in many - developed and
developing - countries, been reduced considerably by increasing func-
tional and spatial integration of interacting systems of growing size,
complexity and lack of controllability. All these factors have led to
'back-wash effects' not only in economic terms but also in social, cultural
and political terms. The determinants of change have become vested in a
136 SPATIAL EQUITY: SOME ANTI-THESES

few functional and geographical centres, on the impulses of which the rest
of the system has increasingly become dependent. It is maintained that
satisfactory solutions of existing problems at intermediate and small social
scales will only be possible if, along with the presently dominating
strategies for system-wide spatial integration (and regional openness),
explicit instruments for selective spatial closure at various levels are
applied. Essentially this would imply devolving some of the decision-
making powers which have become vested in functionally-organized
(vertical) units back to territorially-organized (horizontal) units at
different spatial scales.
Taking into account the impact which changes in large-scale functional
relations exert upon territorial structures, there are theoretically three
possible strategies for resolving emerging conflicts:
1. Complex systems management. This requires full knowledge of the
systematic interrelations between functional and territorial changes and
viCe versa. In view of the complexity of these interrelations this should
remain a long-term research item but it is doubtful whether - except
for selected problem areas - applicable research results may be avail-
able in the near future. Also, such results may render technocratic, but
not necessarily socially feasible solutions.
2. Priority given to functional change. This strategy has been implicitly
applied during the past half century or more. Large-systems oriented
criteria (technological change, efficiency, etc.) prevailed and the result-
ing territorial changes were either neglected or left to the individual
household, firm or local and regional community to cope with;1 at
best, extreme consequences of the disruption of territorial relations
were looked after by large-scale functional (public or private) institu-
tions, including central government.
3. Priority given to territorial integrity. This would mean that functional
changes are allowed to take place only to the extent that no major
disruption of territorial living conditions is caused, or that substantial
aids are given to territorial communities to adapt their structures to the
functional changes they consider desirable.
The choice between these three theoretical alternatives depends on
whether importance for human and social well-being is attributed to large-
scale functional (mainly market- and institution-based) relations as against

1. Large-scale individual motorization, freeways, large-scale urban renewal, environ-

mentally-dangerous industrial or energy plants are cases in point.
CONCEPTUAL BASES OF CURRENT PRACTICE 137

small-scale, environmental and social relations. It also depends on whether

functional processes such as the introduction of new technologies and
scale economies are considered major objectives in ruling over small-scale
human and physical environments or whether the latter contain the
objectives and are to rule over the former as instruments. A major re-
thinking of what are objectives and what are instruments in our society,
in fact, is involved.

6.2 CONCEPTUAL BASES OF CURRENT REGIONAL DEVELOPMENT

PRACTICE

Many of the shortcomings of current regional development practice seem

to stem from its conceptual bases, which again are influenced consider-
ably by the limitations of current regional development theory. Major
factors in this respect are the heavy reliance on neo-classical economics,
the strong concentration on large-scale vertically (sectorally) organized
private and public institutions, the heavy reliance on market- and
institution-based processes (neglecting non-market and informal processes),
and the strong emphasis on economic processes and the neglect of social
and political processes.

6.2.1 Heavy reliance on traditional theoretical notions

A major strategy of most policies for regional development is the increase
of spatial mobility by transport and communications integration. The
underlying idea - as in neo-classical economics - is that a reduction of
distance friction will make factors and commodities sufficiently mobile to
move to the locations of their highest return; factor and commodity prices
would consequently equalize over space and lead to a convergence of
regional per capita income (Richardson 1973, p. 24; and Section 7.2). By
promoting the areal specialization of activities this would at the same time
increase national economic .efficiency.
While the latter often, in fact, occurred, a convergence of regional dis-
parities in living levels in most cases did not materialize (see Section 6.3.2).
This is similar to the international scene: increasing interaction between
rich and poor countries usually did not reduce but rather widened
existing disparities. In addition, the integrated 'coexistence of rich and
poor societies renders more difficult or impossible the efforts of the poor
societies to choose their style of development' (Streeten 1977). In order
138 SPATIAL EQUITY: SOME ANTI-THESES

to reduce regional disparities in spite of this, additional policy instruments

were used as crutches to try to make the neo-classical model work: a
manipulation of factor prices (e.g. regional capital and employment
incentives), a redistribution of external economies (through public infra-
structure investment), etc. But the persistence of certain immobilities
(of power, of population, of natural resources, etc.) continued to produce
biased spread and back-wash effects; it has recently been increasingly
realized that many of these immobilities are not only unavoidable but
even sought for as an important social and political objective. Such objec-
tives are related to what Allardt has called conditions of 'loving' and of
'being', which we shall discuss in Section 6.2.4.
Sector theory is another important ingredient of both current regional
development theory (Richardson 1973, p. 24) and its practice. In accord
with this theory, regional development policies try to attract sectors
with high productivity increases and with high demand elasticity to less-
developed areas. These are usually industrial sectors able to use a high
degree of scale economies. Although called 'leading' sectors, there is no
indication that they actually are the sectors which yield greatest individual
or group satisfaction in the less-developed areas or countries concerned -
their frequent decline in the provision of basic regional requirements such
as food and key services shows this. Nor were these sectors usually the
most appropriate ones for the optimal utilisation of a region's resources:
heavy reliance on capital and technology in fact often set regional natural
resources and labour free.
Export-base theory is a further important component of current regional
development theory. It is based on the assumption that regional income is
essentially a function of regional export performance. Apart from the
theoretical limitations of export base theory (Hilhorst 1971; Richardson
1973) it has contributed to the surrender of regional development to
factors and decisions outside the region and to large-scale functional inter-
relations.
Growth centre theory, another major element of current regional devel-
opment theory and practice, is essentially a combination of export-base and
sector theory applied to a point economy. Since overall mobility of
factors and commodities had turned out to be unfeasible (and in many
respects undesirable) the implicit attempt was now to collapse reality
back into a set of disaggregate point economies - one of the initial
assumptions of neo-classical economics - and to thereby make regional
equity and national growth compatible. Growth centre theory has led to a
CONCEPTUAL BASES OF CURRENT PRACTICE 139

high emphasis on urban-industrial growth based essentially on extra-

regional determinants. But it resulted in the lack of broader regional devel-
opment observed in most countries as shown below (Section 6.3.2). In
spite of the empirically-found deficiency of the growth centre concept,
major critical evaluation of this strategy has so far only been made on a
few occasions (Conroy 1973; Nagoya Centre 1975; Chapter 7).2
This parallel movement of regional development theory and practice may
have been mutually dependent: regional development practice challenged
the formulation of theories of regional growth which confirmed its initial
policy-approaches; on the other hand the body of theory emerging in this
process reassured regional development practice of the correctness of its
initial steps. In this way development and practice reinforced each other
by circular causation.
The theoretical approaches mentioned above were essentially developed
to explain past spatial patterns which had emerged under conditions of
increasing functional and spatial integration, specialization, industrializa-
tion, the use of agglomeration and scale economies, and of accelerated
innovation. Since it was these factors which essentially brought about
spatial inequalities, the theories could hardly be expected to contribute
much to the eradication of this problem.

6.2.2 Strong concentration on large-scale vertically organised (private or

public) institutions
Regional development practice has relied heavily on large-scale vertically-
organized institutions for implementing its policies. Horizontally-
organized structures - integrated territorially at the local or regional
level - have thereby often been disintegrated or have not been able to
emerge. This applies to directly productive activities, to infra-structure
sectors as well as to human and to man-environment relations. In the
directly productive sectors refuge was usually taken in large-scale multi-
plant (and often multi-national) enterprises - private or nationalized -
not only because they often operated 'leading' industries but because they

2. Kamal Salih in his concluding remarks to the Nagoya seminar said he was not sure
whether he had come 'to a burial of the growth pole idea or to a celebration of the
new agropolitan approach' . . . and . . . 'whether this symposium will mark the
demise of central planning of growth-oriented approaches and ·of accelerated
industrialization as the prime mover of development, or the birth of a self-reliant
planning, of the goal of distribution and of accelerated rural development' (Nagoya
Centre, 1975, p. 416).
140 SPATIAL EQUITY: SOME ANTI-THESES

could readily be identified by regional development agencies and be dealt

with more easily than could a great number of small firms (Hansen 1974,
p. 55 ff, for France; Holland 1976, for Great Britain and France). The role
of these large firms was often further promoted by strong informal ties
between leading personalities of large-scale industry and government
(Sundqvist 1975, for France).
In developing countries multi-national companies were relied upon
heavily for both national as well as for regional development in
programmes such as that for the Northeast of Brazil. In infrastructure
sectors as well, reliance was put mainly on large-scale centralized agencies
to provide investment and organizational skills for the implementation of
projects in less-developed areas. This made these areas even more depen-
dent on formal structures and on outside inputs. Similarly, the relocation
of offices of large-scale government institutions to less-developed areas
was considered an easily manageable instrument of regional policy
(Sweden, Great Britain, France and the Netherlands).
Lasuen (1973) has stressed the space-bridging capability of multi-
plant and mUlti-product firms in transferring development inpulses to
remote areas, particularly by facilitating transfers of capital and techno-
logical innovations through their intra-organizational channels without
having to rely on public transfer channels, which are usually scarce in
such areas. Large-scale organizations, however, tend to uniformly apply
their central decision-making criteria to less-developed areas too. They
will normally apply their own (high) technology in these areas. There-
fore employment effects are often disappointing and the region's resources
will be utilized very selectively and with a bias to the interests of
decision-makers external to the region. Similarly, the employment effects
of relocating public offices to less-developed areas have usually been
disappointing since, along with the offices, usually a substantial part of
the employees were also transferred from highly-developed areas.
A further problem with the space-bridging capability of large-scale
organizations is that it can work both ways: whereas developmental
impulses can be transmitted relatively easily to less-developed areas, they
can also (particularly in the private sector) be withdrawn with equal
facility. With multi-national firms this has an additional dimension: shifts
are possible not only between developed and under-developed areas of
industrialized countries but also to Third World countries with their
much lower wage levels (Holland 1976, p. 43, for Great Britain, France
and Italy; see also Chapter 8).
CONCEPTUAL BASES OF CURRENT PRACTICE 141

In summary one can say that, since most problems to which national
policies for spatial development address themselves were caused by large-
scale influences transmitted through vertical functional channels, the use
of large-scale, vertically-integrated institutions turned out to be a rather
evasive instrument for controlling these problems and in the long-run must
be expected to operate in the interest of these large-scale organizations
rather than in the interest of the respective regions for which the policies
were designed.

6.2.3 Heavy reliance on market-based and formal institutional processes

Both regional development theory and policy have primarily been con-
cerned with market-oriented processes and formal institutional aspects of
development. Indicators used to measure the economic level of develop-
ment or to formulate targets of development (such as regional product,
regional income or regional employment) were, in practice, exclusively
geared towards that part of these variables which passed through the
market (e.g. market production, market derived income, and employment
in market activities); production and employment related to (informal)
non-market activities were rarely measured and hardly ever incorporated
into planning targets. Yet these activities are of great economic and
societal importance. Even for the United States, a country with a relative-
ly small share of non-market-oriented economic activities, the value of the
latter has been estimated at 65 percent of national income which is tradi-
tionally measured by market transactions only (Scitovsky 1976, p. 102).
Scitovsky shows that with the emphasis of economic development
policy on increased specialization, mobility and large-scale functional
activities which rely on the market mechanism, a steady reduction of the
production of non-market goods and services has taken place (e.g. those
rendered within social groups like the family, neighbourhood or clan,
including mutual help and stimulation, advice and work as self-realization).
These non-market goods and services, however, according to Scitovsky,
provide 'comforts' essential for human satisfaction such as the comfort
of 'belonging', of 'being useful' and of cultural continuity or 'sticking to
our habits' (Scitovsky, pp. 114 ff; see also Section 6.2.4).
Indicators of social development widely used, such as access to
hospitals, schools, administrative offices, etc., also measure only the
access to formal institutionalized social services which, particularly in less-
developed areas, are complemented to a considerable extent by informal
activities fulfilling similar functions. In the present process of growing
142 SPATIAL EQUITY: SOME ANTI-THESES

specialization, institutionalization, industrialization and urbanization

these informal functions are increasingly being substituted by formal ones.
Yet there is a wide-spread feeling that this replacement process should not
continue to an unlimited degree and in part should even be reversed
(Illich 1973; Matzner 1976). The argument is that for reasons both of
efficiency and of group involvement many of these institutionalized
public services should be devolved to informal self-organizing groups.
Greater emphasis of small-scale social relations is considered an essential
condition for human happiness. These small-scale human interactions 'do
not go through the market ... are (usually) rendered free, their reci-
procity and equitable distribution being assured by custom, tradition,
social pressure, family discipline or law' (Scitovsky 1976, p. 86). They are
culture specific: 'a determinant of culture and at the same time an expres-
sion of it . . . they contain the substructure which may help to thus
protect human culture in all its splendid diversity in the face of inevitable
change' (Greenbie 1976, pp. 80-81). Relations in small-scale territorial
space, according to Greenbie: 'make human life ultimately satisfying'
as they are 'specific to culture, personality, time, place, and circumstance'
(p. 93). The concept of selective spatial closure which we suggest below
(Section 6.4) is very closely related to these ideas.
Greenbie (1976) uses Hall's concept (1966) of 'proxemic' or small-
scale space,3 and contrasts it with 'distemic' or large-scale space. Com-
munication within 'proxemic' space, according to Greenbie, provides
the individual with security, a sense of identification (by differentiating
the individual from other individuals and groups): 'behaviour in proxemic
contexts ... require(s) a life-time of experience to develop' ... with 'a
clear understanding as to the conventions of its use' (Greenbie 1976,
p. 84) and consequently entails a relatively high degree of immobility.
Behaviour in 'distemic' space on the other hand 'can be learned
abstractly, consciously and much more rapidly than proxemic behaviour'
and therefore facilitates much greater mobility. Conceptual relations in
distemic space can be kept up with a much larger number of people than
sensual relations in proxemic space. Rather than culturally dermed, dis-
temic space therefore can be trans-cultural or 'super-cultural' (Greenbie
1976, p. 83).
3. Hall calls 'proxemic' space that part of a person's or group's environment within
which sensual (non-rational, non-verbal) communication dominates; it extends over
contiguous space and is usually limited by the radius of regular direct personal
contacts.
CONCEPTUAL BASES OF CURRENT PRACTICE 143

In the terms used above, proxemic space can be understood to be

predominantly territorially or horizontally organized, entailing a high
degree of cultural specificity and immobility. Distemic space on the other
hand is predominantly functionally or vertically organized, trans-cultural
and entails a relatively high degree of mobility. Furthermore, distemic
large-scale systems require a comparatively high degree of formal organiza-
tion, their legal and administrative systems must be guided by superior
authorities (whose decisions mayor may not correspond to small-scale
requirements) while proxemic small-scale systems usually operate suffi-
ciently well by informal self-organization and self-policing (Greenbie
1976, p. 92).
Greenbie maintains that a balance of human interaction in both these
spheres - proxemic (small-scale) and distemic (large-scale) - is essential
not only for individual and group well-being but also for the self-
protection of small-scale territorially-organized groups against large-scale
(functional) influences. Yona Friedman (1977) coins this in a different
way by speaking of the illusion of a possibility of outright global com-
munication:

We have seen that such (outright) global communication is not possible. Each
organization exceeding the 'critical group' size corresponding to its structure will
necessarily falter due to a collapse in internal communications. The unfeasibility of
communication beyond a certain scale favours the system of small organizations and
small groups. These small groups (or at least a large part of them) can survive in a
'poor' world if it foregoes specialization, individual storage and transport.
(author's translation)

These arguments all seem to stress the importance of retaining and

explicitly providing for the development of small-scale, non-market, non-
institutionalized informal activities, the importance of territorially
organized (horizontal) besides functionally organized (vertical) relations,
as well as the retention of cultural specificity and the intentional mainten-
ance of a certain degree of geographical immobility. Most of these factors
were either absent from or at best at the non-operational margin of
current regional development theory and practice.

6.2.4 Strong emphasis on economic and material human needs

Both current regional development theory and practice have so far
concentrated on economic and material needs or what Eric Allardt (1973)
calls conditions of 'having', while greatly neglecting (apart from occasional
verbal statements of intent) immaterial human needs which Allardt calls
144 SPATIAL EQUITY: SOME ANTI-THESES

conditions of 'loving' and of 'being'. This is essentially a replica of what

has happened in national development theory and practice for the past
three decades, not only in market and mixed economies but equally
in centrally planned economies. Mainly as a reaction to that narrow
economic orientation, the social indicator movement has tried, since the
mid-l 960s, to introduce the broader concept of 'quality of life' at the
national level. While it must be admitted that important theoretical and
operational problems still need to be solved for national policies, the
concept has still hardly even started to touch on the theory and practice
of spatial development planning. The first tentative approaches in this
direction are works by Fox (1974), Knox (1975 and 1976) and Coates,
Johnston and Knox (1977).
Allardt (1973) distinguishes between three groups of human needs:
1. The above-mentioned dimwsion of 'having', income, density of dwell-
ings, employment conditions, health, education, to which most of the
currently used indicators of regional development relate.
2. The dimension of 'loving', relationships between individuals measured
by components such as local solidarity, family solidarity and friendship.
3. The dimension of 'being', referring to the degree of self-realization
versus alienation of the individual in society expressed by components
such as the degree of irreplaceability, the amount of political resources
or access to decision-making available to the individual.
Galtung (1973) adds two further components of interest to regional
development planning: the diversity of possible life-styles for individuals
to choose from, and the degree of autonomy of groups of individuals to
set their own cultural goals in the face of external control.
Objectives of the latter groups are very rarely to be found in opera-
tional terms in either regional development theory or in regional policies:
conservation of specific cultural and historical values is defined as an
objective of regional policy in West Germany (Bundesraumordnungsgesetz
1965, § 2), the provision of conditions for a wide range of life-styles and
for effective citizen participation and decentralization of decision-making
in Australia (Logan 1974, p. 137, referring to the last labour government's
policy). But indications for operationalization of these objectives are very
scarce.
Reasons for the virtual absence of such objectives and their operation-
alization may be:
1. Some of these objectives (particularly the conditions of 'being') would
require shifts in the distribution of power; while it may be compara-
CONCEPTUAL BASES OF CURRENT PRACTICE 145

tively easy to redistribute material resources, a shift in power usually is

much more difficult to achieve, particularly if it requires existing
centres to devolve their own powers; this is true for market and mixed
economies and may be even more so for centrally-planned economies.
2. There is little knowledge about the functional interdependencies of
these variables amongst each other and with the conditions of 'having',
as well as about the instruments for influencing them.
3. These variables are more difficult to measure than the economic or
material ones nonnally used in regional development programmes and
are therefore less accessible as indicators of political achievement.
4. There may exist a feeling that these 'conditions' or 'comforts' depend
on small-scale community relations which cannot - and should not -
be planned from higher levels.
However, if these small-scale (usually territorially-defmed) conditions
and comforts are heavily influenced and often disrupted by large-scale
functional processes (see above), then national and interregional policies
should provide for defense mechanisms by which small-scale groups are
enabled to fend off consequences of large-scale functional processes they
consider undesirable for their own living conditions. Such defense
mechanisms would increase the resilience of territorial systems to external
shocks or provide the possibility to regionally control their consequences
in order to maintain the functioning of integrated territorial systems.
Examples for such mechanisms will be given in the last section in con-
nection with the concept of 'selective spatial closure'.
There exists an additional danger in operationalizing only objectives
pertaining to Allardt's conditions of 'having'. Since such indicators are
mainly used for measuring the existing level of development and the
progress of regional development programmes, one will tend to under-
rate existing levels· and advances of development, particularly in less
developed areas where non-market and non-institutional spheres still play
a relatively large role. Furthennore, if objectives of spatial development
are operationalized mainly in tenns of conditions of 'having', the dis-
placement of infonnal by fonnal functions will be accelerated, although at
present serious concern about such a process gains increasing ground.
Finally, if the evaluation of regional development programmes takes place
mainly in tenns of indicators of 'having' , the priority (often unconsciously)
assigned to them in one period will tend to perpetuate itself also for
following periods and lead to a long-range priority of material values.
Whichever policy is adopted in this respect, the measurement of
146 SPATIAL EQUITY: SOME ANTI-THESES

development levels and of effects of regional development programmes

should in any case also attempt to include, besides the currently used
variables, estimates of changes of external material effects (pollution, etc.)
and of non-material effects (such as increased distance of individuals from
decision-making, increased frustration, decrease in self-fulfilment, etc.).

6.3. SOME EMPIRICAL EVIDENCE OF RECENT SPATIAL DEVELOPMENT

TRENDS: THE MATERIAL EFFECTS OF REGIONAL DEVELOPMENT
POLICIES

The following section attempts to synthesize the evaluation by various

authors of spatial development trends and policy effects in market and
mixed economies. The countries discussed were limited by the availability
of relevant case studies. A further limitation with regard to the topic of
the present chapter was that most authors found it difficult to distinguish
clearly between autonomous and policy-induced trends of spatial develop-
ment. Most case studies therefore evaluate the cumulative result of both.
Only in a few cases were the policy effects calculated in quantitative
terms (e.g. by spatial or temporal correlation) and in most instances they
were estimated by rather subjective criteria. A third limitation is that most
of the indicators used in these case studies refer to material indicators of
living levels (changes in income trends, in employment trends and
migration, spatial linkages of the input-output type; in few cases broader
socioeconomic variables were aggregated by factor analysis). Evaluations
of indicators of non-material conditions are hardly available in spatially
disaggregated form and by no means on a similarly broad comparative
international basis.
We have distinguished between overall trends of inter- and intra-regional
disparities (Section 6.3.1) and specific intra-regional development trends
in relation of 'growth centre policies' for specific regions (Section 6.3.2).
A more detailed analysis of spatial development trends in relation to
specific regional policy instruments has been undertaken elsewhere (Stohr
and TOdtling 1977).

6.3.1 Overall trends o/inter-regional and intra-regional disparities

We shall start with countries where considerable spatial inequalities have
subsisted or even increased in spite of substantial regional policy inputs,
and then continue with countries which showed a certain reduction of
interregional disparities.
RECENT SPATIAL DEVELOPMENT TRENDS 147

For Brazil detailed investigations of trends in interregional and intra-

regional income disparities have been undertaken by Gilbert and
Goodman (1976) with special reference to the Northeast. Testing
Williamson's hypothesis of a convergence of interregional income dis-
parities in the course of national development (for a further evaluation
see Section 7.3) for the period 1939-68, the authors found that the
statistical evidence was inconclusive and suggested that '(inter)regional
income differentials have remained fairly stable' (p. 129). Conclusive
results became evident, however, in the case of income inequalities within
the Northeast of Brazil where the authors found sharply increasing inter-
personal income inequality during the 1960s, especially within the urban
areas.
In Spain there also seem to exist increasing intra-regional income
disparities. Lasuen (1974), evaluating Spain's regional policy, points out
that the consequences of policies pursued during the past two decades
have been a reduction in interurban income differentials and an increase
in urban-rural income differentials,4 the net effect of which has been a
reduction of income differentials among regions. Similar conclusions for
Spain have been arrived at by Richardson (1975).
For Italy, the persistence of sizeable spatial disparities in such indica-
tors as per capita income within the Mezzogiomo (for which a major
development programme has been in operation for about two decades)
is pointed out by Allen and McLennan (1970, p. 119). Sundqvist (1975,
pp. 171 ff) even indicates that these intra-regional disparities have been
increasing rather than decreasing and states that the 'developing' dispari-
ties have led to a modification of the regional development strategy away
from the growth center policy in 1971. Regarding interregional disparities,
the following trends, similar to the fmdings of Sundqvist, were found by
the OECD (1976, pp. 30 ff): although there has been some improvement
with regard to the South's share in gross industrial investment, heavy
migration from the South continued in the late 1960s (with higher
absolute numbers in 1968-70 than in previous years), and the South's
share of national employment fell from 33 percent in 1951 to 30.6 per-
cent in 1970. With regard to per capita income, the South could keep pace

4. Lasuen (1974) states further that 'until the early 1960s both rural-urban and inter-
regional income differentials were widening. It seems that only from the early
1960s onwards have the interurban and interregional income differentials been
ameliorated, in spite of the continuous worsening of rural-urban differentials due to
the growing weight of urban income in the underdeveloped areas'.
148 SPATIAL EQUITY: SOME ANTI-THESES

with the national growth rate but the gap towards the rest of Italy was
not noticeably diminished (similarly Rodgers 1970; Cao-Pinna 1974).
In France regional economic policy at first sight seems to have been
quite successful: the regional distribution of population and economic
activities (particularly industrial employment) was modified in the desired
direction, the century-old flow of people from the provinces to Paris had
been halted and possibly reversed (Prud'homme, in Hansen 1974, pp.
56 ff). A closer look however, shows that industrial employment increased
in the Western regions mainly around Paris, but not at the greater distance
from it where the most backward areas of France are. Similar findings
relate to the distribution of industrial employment (Sundqvist 1975,
pp. 119 ff). 5 Concerning income disparities between regions both Prud'
homme (1974), pp. 59 ff) and the OECD (1976, p. 26) indicate that these
disparities have not been significantly reduced.
Lack of success of regional policy measures is also indicated for
Belgium (Ruehmann 1968; Davin 1969).

There seem to exist also some examples of at least partial goal fulfilment
of regional policies with regard to reducing spatial disparities of living
levels.
For the United States, Cumberland and Thoman evaluate the Appala-
chian programme as successful in at least some respects. The large gap
between Appalachia and the rest of the nation in terms of employment
rates, migration balance and per capita income narrowed slightly in the
period 1965 to 1973 (Thoman 1976, pp. 20 ff; Cumberland 1973, pp.
101 ff). There is no clear answer to the questions of how much of this
trend is due to regional policy or how intra-regional disparities developed.
For the United Kingdom, Cameron (1974, p. 24) states that regional
policy has contributed to an improvement in the relative unemployment
level and per capita income growth. Moore and Rhodes (1974, p. 49) state
in this connection, however, that although disparities would have been
worse without regional policy they have remained at unacceptably high
levels.
In the Federal Republic of Germany about 550,000 jobs have been
subsidized by regional policy measures from 1969-74 (Hotger 1974, pp.
187 ff). Other investigations indicate, however, that the causal relation-

5. Sundqvist, however, states also a slight improvement of the more distant provinces
with regard to industrial employment in the late 1960s (1975, p. 120).
RECENT SPATIAL DEVELOPMENT TRENDS 149

ship between subsidies and job creation is rather weak and that the
diversity and qualification of jobs created is small (Wolf 1975, pp. 431 ff).
Von Boventer (1969) indicates that regional trends in Western Germany
are very much due to market forces rather than to regional policy measures.
In Austria, inter- and intra-regional material disparities seem to have
been decreasing during the 1950s and 1960s. Berentsen (1976) found that
regional inequalities of per capita product during the period 1961-71/2
have considerably declined both at the Bundesliinder (province) level
and at the Bezirk (county) level. Similarly a broader analysis of nine
indicators of regional levels of living by factor analysis for 1957-71 shows
that a considerable decline in the factor score inequalities was observed
(pp. 99 ff). Berentsen hypothesizes that Austrian regional policy
(although little coordinated) has certainly positively influenced the partial
goal fulfilment of reducing interregional and intra-regional disparities.
But it is difficult to say to what extent the little-coordinated regional
development policy or other specific Austrian conditions have contributed
to this phenomenon. 6
For Japan, Mera (Nagoya Centre 1975) also shows evidence of decreas-
ing income disparities for the forty-six prefectures of Japan during the
period 1961-72. He further analyzes how much regional policy, particularly
the growth pole strategy in existence since 1962, has contributed to the
reduction of income disparities. He concludes, from a comparison of the
spatial distribution of reductions in income disparities and the govern-
ment's industrial decentralization policy at selected growth poles, that this
policy seems to neither have contributed to a reduction in income dis-

6. Such specific conditions might be (some of these points were indicated by Berent-
sen, others by the present authors):
- the small geographical size of Austria;
the expansion of commuting radii from the main employment centres to cover
the majority of populated areas (Berentsen 1976, p. 11);
the long-standing traditions of tourist activities in many rural areas;
the stagnation and peripheral Eastern location of the richest province (Vienna);
the fact that the second richest province lies at the other extreme periphery
(Vorarlberg) ;
certain factor scarcities (land, labour) within the major cities, particularly in
Vienna, reinforced still by rigid land-use regulations and a fragmented land-
market;
the entrenched federalist character of the country;
a very rigid and all-embracing legal and administrative system acting as a forceful
brake for the rapid adoption of innovations.
150 SPATIAL EQUITY: SOME ANTI-THESES

parities among prefectures nor to a reduction in the trend of population

concentration (pp. 260 ft).
The findings of the abovementioned case studies can be summarized
as follows:
1. In most of the countries analysed there is no clear indication of a major
convergence of regional disparities of income or other indicators of
material living levels. This seems the case particularly in countries with
sizeable regional problems (Italy, France, Brazil). For most coun-
tries it is difficult to say to what extent the trend is due to the 'autono-
mous' working of the market mechanism or to explicit policies of
spatial development.
2. From more detailed analyses available for some countries it seems that
spatial development policies in general were not able to change spatial
inequalities in material living levels significantly. In cases where it was
possible to reduce disparities at one scale (e.g. the inter-regional one),
this was usually accompanied by an increase in disparities at other
scales (e.g. the intra-regional or inter-personal ones). Such shifts in
disparities from one geographical scale to another could be observed
particularly where policies of 'concentrated regional development' were
applied, usually combined with sectorally-unbalanced development
(mainly industry) and a strong emphasis on overall efficiency (Spain
and Brazil, to a lesser extent France).
3. In most countries where a reduction of spatial disparities, at least in
some respects, seems to have taken place (Austria, Japan, West
Germanty, USA) either initial regional inequalities have been relatively
small (Austria, West Germany) and spatial development policies were
fuzzy and little articulated (Austria, USA) or were considered to have
had little effect upon the reduction of spatial disparities (Japan). The
reduction in spatial disparities were then either attributed to market
forces (von B6venter for West Germany) or to specific national geo-
graphical or historical conditions (Austria).
4. There are no broader comparative analyses available on the impact of
spatial development policies on non-material indicators of living condi-
tions (such as Allardt's conditions of 'loving' and of 'being').
This lack of information with a wider base is deplorable because there
are indications that the traditional regional policies of large-scale func-
tional integration - even if they improved quantitative aspects of
economic living levels in certain cases - were often accompanied by a
deterioration in qualitative living conditions. This was caused by such
RECENT SPATIAL DEVELOPMENT TRENDS 151

factors as the establishment of predominantly low-skill routine jobs with

relatively high closure rates and cyclical fluctuations in less-developed
areas. These new jobs were to a high degree created by interregional or
international firms in the form of 'extended work-benches', for which the
decision-making, administrative, research and other more qualified func-
tions were retained in the highly-developed core regions. In this context,
there emerged in many cases an increasingly unilateral dependence of
less-developed areas on external decisions, production factors and on
external demand, along with a disvinculation of these new activities from
autochthonous regional economic, social and political structures.
Frequently this was accompanied by a highly selective utilization - and
partial idleness - of regional natural and human resources and by a partial
displacement of region-serving basic services by export-base activities
which often created insufficiencies in basic service accessibility, particularly
for the less mobile population strata. Further consequences were an
increased dependence of peripheral areas on central government subsidies,
the imposition of core-region institutional forms and central decision-
making criteria upon them along with a weakening of autochthonous
regional institutions and (often informally functioning) social support
systems. More detailed empirical material on these points is presented
for market and mixed economies in St5hr and Todtling (1977).
Although no similar empirical analyses have been accessible to the
authors on the effects of regional policies in centrally-planned countries,
it is likely that some of these problems are even more pronounced there
due to the highly centralized planning and decision-making structures, to
the restricted degree of self-determination of local communities in most
centrally-planned economies, and to the relatively low priority given in
many of these countries to the (necessarily territorially-organized) service
sector. The frequently-made assertion that spatial disparities of living
levels are negligible in centrally-planned Socialist countries does not
generally seem to hold (see, e.g., Lacko 1976) and may rather be due to a
lack of relevant data and analyses.

6.3.2 Growth centre - hinterland effects

The following is a summary of a review undertaken for a number of case
studies on the impact of growth centres upon their hinterland (Stohr and
Todtling 1977). Some of these case studies relate in part to what Alonso
and Medrich (1972) called 'induced' growth centres (Richardson 1975 and
Buttler 1973, on Spain; Penouill969, Moseley 1974, and Sundqvist 1975,
152 SPATIAL EQUITY: SOME ANTI-THESES

on France; Hansen 1975a, on the USA; Allen and McLennan 1970 and
Sundqvist 1975, on Italy; Appa1raju and Safier 1975, on 'Third-world coun-
tries'). Others relate to 'spontaneous' growth centres (Moseley 1973,
on Rennes, France and East Anglia, United Kingdom; Gilbert 1975, on
Medellin, Collunibia; Waller 1974, on Peru).
The major findings of the abovementioned and other studies and
investigations can be summarized as follows:
1. Spread-effects from growth centres were usually smaller than expected,
or less than backwash-effects and therefore had a negative net result
on the hinterland. They were narrowly limited in geographical extent,
usually restricted to the commuting area, often as a function of the size
of the centre (Morrill 1974).
2. Increases in income of lower order centres or rural areas create strong
income multipliers in higher order centres but not the other way round
(Nichols 1969; Moseley 1973 and 1974). They seem to move upward
rather than downward within the urban hierarchy.
3. In the context of policies for broad spatial development it is difficult
to justify growth-centre policies for lagging areas due to their lack of
spread-effects in the urban hierarchy downward or from the growth
centre to a broader hinterland (Hansen 1975, Nichols 1969; Moseley
1973 and 1974).

6.4 SELECTIVE SPATIAL CLOSURE AS A STRATEGY FOR INCREASING

THE RESILIENCE OF SPATIAL SYSTEMS?

In the light of the strategy options discussed in Section 6.1, outright

integration will tend to increase initial spatial disparities if functional
change is given priority over territorial integrity. Under such conditions,
a reduction of spatial disparities would require a forceful central redis-
tribution mechanism (such as in centrally-planned economies) which may
be successful in reducing economic disparities but at the same time is
likely to impair immaterial living conditions in peripheral areas (such as
social interaction at smaller scales, family, local and regional solidarity -
Allardt's conditions of 'loving' - as well as individual and small-group
self-realization, individual and local access to relevant decision-making
powers - Allardt's conditions of 'being'). It is understandable that a
powerful central redistributive mechanism in a system of outright
functional integration may be able to reduce spatial disparities in
A SELECTIVE SPATIAL CLOSURE STRATEGY? 153

economic tenns, but it will at the same time tend to increase disparities
in other respects such as in access to decision-making powers, social
amenities, etc. One reason is that economic interaction and redistribution
in general can take place at much larger scales than sociopolitical inter-
action and redistribution.
In view of the problems which policies of outright functional and
spatial integration have created - both at the international and at the
interregional level - we should like to discuss an alternative strategy of
regional development by selective spatial closure. The following ideas
by no means are to suggest a policy of regional autarchy. Both from an
efficiency and an equity standpoint this would be unthinkable today.
Suggested, however, are policies which pennit the channeling of today's
widely uncontrolled economic, social and political 'backwash' effects and
thereby facilitate greater spatial equity of living conditions in all the
respects discussed above. This requires a number of preconditions:
1. The broadening of explicit spatial development policy beyond presently
mainly economic (materially-oriented) consideration to a more explicit
inclusion of social and political processes.
2. The refonnulation of distance friction from a negative concept (to be
diminished as an obstacle to large-scale integration and spatial equili-
brium) to a positive one for the structuring of a spatially-dis aggregate
interaction and decision system (Isard 1976).
3. Greater attention to be paid to non-market- and non-institution-based
activities and to the requirements of small-scale human and man-
environment relations.
4. A shift of decision-making powers from today's mainly functional or
vertical (sectoral) units to horizontal (territorial) units at various levels.
The scale of the territorial decision-making level should ideally be the
one within which a maximum of the repercussions or external effects of
the respective decisions can be internalized. This means a closer
approximation of decision-making scales with corresponding spatial
impact scales. In case of doubt the lower decision-making levels should
be given preference.
We shall not propose specific instruments for such policies here,
particularly because they have to be carefully adapted to the specific
historical, institutional and political conditions in each country or region.
We shall only give examples of policy instruments already in use which
operate (often intuitively) in such directions.
Strategies of selective spatial closure can, in current economic tennin-
154 SPATIAL EQUITY: SOME ANTI-THESES

ology, be applied from the supply side (regional resources, Section 6.4.1),
the demand side (guidance of regional preference patterns, Section 6.4.2)
and through policy instruments modifying distance friction and thereby
changing accessibility to inputs, to markets, to social contacts and to
decision-making power as well as by redistributing the effects of external
and scale economies (Section 6.4.3). If integration between highly- and
little-developed areas takes place in an uncontrolled way, the absence of
these strategies would tend to increase existing spatial inequalities.

6.4.1 Selective regional closure from the supply side

External vertical influences (changes in external demand, technology,
intra-organizational relations, etc.) frequently lead to under-employment
or over-employment of (relatively immobile) regional resources, particu-
larly of labour and natural resources. In order to avoid abrupt changes
in factor employment, a higher degree of co-determination on the transfer
of such regional changes by the affected regional communities may be
necessary. Such regional co-determination could take place in several areas:
1. Application of new technology. Regional co-determination essentially
means that technology would be reverted to the role of an instrument
- rather than a determinant - of (territorially-organized) society. An
extreme case in this respect is China where differentiated levels of
technology are normatively determined ex ante by assigning specific
shares of a sector's production to local, regional, and national industries
(Weiss 1976). More flexible policies might include the negotiated step-
wise introduction of new technology, of 'regionally adequate' rather
than 'highest' technology, or the compensation for negative external
effects which it causes.
2. Natural resource use. Regional co-determination could range all the way
from regional ownership of natural resources to communal control on
the degree and kind of the exploitation, on the location of their pro-
cessing, on waste disposal and other forms of pollution caused by their
use, etc.; it might extend to over-exploitation as well as to actions
required to mitigate the consequences of externally induced under-
exploitation of natural resources. Examples at the international level
are today to be found in most developing countries; at the sub-national
level an increasing number of regional communities are demanding
more decision-making power on their own natural resources (Wales on
its water resources, Scotland on its off-shore oil, the North of Chile on
A SELECTIVE SPATIAL CLOSURE STRATEGY? 155

its copper, etc.); usually the regional bargaining position will be greater
the higher the demand for its natural resources.
3. Employment of human resources. Regional co-determination will vary
widely according to the respective sociopolitical system: from com-
munal decision-making on migration (e.g. in China) on one extreme,
to more liberal policies of regionally differentiated labour exchange
systems and regionally differentiated school and training facilities
geared to specific regional development potentials and needs rather
than to criteria of a uniform national education system.

6.4.2 Selective regional closure from the demand side

A major element of regional closure are regionally differentiated preference
patterns. In the past decades innovation diffusion - i.e. the adoption of
new uniform production or demand patterns - has in a rather simplistic
way been equated with development. The spatial extension of com-
munications and transport media has promoted a trend towards increas-
ing uniformity of preference patterns. This trend towards uniformity has
benefited those regions or organizations which were able to take
advantage of scale and external economies and has relegated the others
to cumulative disadvantage. Diversity in regional preference patterns
(particularly if they are oriented to the use of regional resources) not
only increases the competitive position of peripheral, less developed,
or small regions but also contributes to such non-material human needs
as local and regional identity and others of Allardt's conditions of 'loving'
and 'being'.
The economic advantages of differentiation can be explained by the
fact of decreased marginal returns to innovation once it has spread to a
certain extent. Beyond this point non-innovation (retention of traditional
customs and production methods, historical built form, untouched
natural environment, etc.) usually again becomes an economic advantage.
This can bring advantages induced both from the regional demand side
(greater regional demand for specific regional services and products) and
from the regional supply side (supply of regionally characteristic products
to the outside).
This differentiation of preference patterns can also be expressed by a
different weighting of non-material as against material objectives. Non-
material objectives seem to be promoted by a higher degree of regional
closure where status is often attained in non-material terms, through the
position one holds in the proxemic space of a specific culturally-defmed
156 SPATIAL EQUITY: SOME ANTI-THESES

social system in 'relative ranking hierarchies' (Greenbie 1976, p. 94);

material objectives on the other hand seem to be promoted by regional
openness where, in competitive 'absolute ranking hierarchies', status is
measured in quantitative terms, for instance, by income or consumption.
Specific value systems therefore seem to be related to different degrees
of regional openness or closure. The latter on their part again seem to
influence the value systems.
The maintenance of regionally differentiated preference patterns
depends greatly on the regional disaggregation of the administrative and
decision-making system. Federal systems seem to promote it more than
unitary ones. The more decisions are devolved to regional and local
communities the more they can articulate differentiated preference
patterns.
A further prerequisite. for regionally differentiated preference patterns
is a relatively high degree of intra-regional (as against interregional) inter-
action and communications and transport integration. Instruments for this
purpose are regionalized (cable) TV networks, regional newspapers and
other media, as well as a regionally disaggregated transport network (see
below).
Another frequently utilized instrument to disaggregate regional demand
are regional contract premiums whereby regional suppliers are granted
higher price margins for public tenders on the grounds that this helps to
'lock in' local and regional demand created by public expenditure.

6.4.3 Other measures for selective spatial closure

Biased increase of accessibility for less-developed areas. While a mutual
increase of accessibility between areas of different development levels
normally leads to increasing instead of decreasing disparities (see above),
a unilateral improvement of the access of less-developed areas to core
region markets (without vice versa improving the access of core regions to
less developed areas) can help to overcome these disequilibrating effects.
In Sweden, for example, transport subsidies for long-distance shipments of
regional products are being given unilaterally to manufacturing firms in
the less developed North (Bourne 1975). Similar subsidies are granted for
telephone costs in the North of Sweden.
While the above measures are related to interregional accessibility, such
a biased policy can also be applied to the internal accessibility of - or
between - less developed areas. In such a regionally disaggregated
transport and communications policy, priority would be given to connec-
A SELECTIVE SPATIAL CLOSURE STRATEGY? 157

tions within and between less developed areas (rather than between these
and highly developed core regions). Such a policy would unilaterally
improve the possibilities of less-developed areas to improve their internal
factor market, to increase internal demand, to make use of scale
economies and thereby improve their competitive position also vis-a-vis
the outside. In Sweden extensive proposals for an interperipheral trans-
port network have been made (Tornqvist 1973). In Britain proposals have
been made for intra-regional commuting subsidies for less-developed
sparsely populated areas in order to increase access to labour at individual
peripheral locations (Moseley 1974, p. 146). In France explicit priority
has been given to strengthening the transport connections between
regional equilibrium metropolises (Rapport 1971, pp. 46 ff; Friedly 1974,
pp. 158 fO.
Compensation for spatially-differentiated external and potential scale
economies. Within specific sectors, scale economies are to a great extent
a function of market accessibility. Accessibility again is - apart from
geographic location - considerably determined by public transport invest-
ment, an economy external to the firm. In peripheral areas in which
sufficient transport improvement is either not feasible or not possible
(not all locations can nor should be touched by freeways, etc.), compensa-
tion at a freeway exit, and thereby able to reap scale economies. It is the
maintain basic needs services. In areas with too low population and/or
income density, measures such as subsidies, negative income tax, etc.
might be justified in order to compensate for the lack of scale economies
and thereby facilitate the required private (or public) basic services. It will
depend on the specific economic and political system whether this com-
pensation has to be derived from general taxation or whether it can be
drawn from enterprises gaining high external economies, e.g. due to loca-
tion at a freeway exit, and thereby able to reap scale economies. It is the
typical question of how to compensate for differential locational
advantages between, for instance, a (freeway-oriented) supermarket and a
neighbourhood grocer, if the latter is needed for the supply of less mobile
population strata (children, aged, poor, etc.) in a local or regional context.
These were examples of policy instruments often intuitively developed
to control spatial 'backwash' effects. If strategies of selective spatial
closure were adopted as explicit components of regional development
policy, coherent sets of such policy instruments adapted to the specific
conditions of the respective country or region would have to be elaborated
and subjected to empirical testing. The fact that such instruments might
158 SPATIAL EQUITY: SOME ANTI-THESES

be incompatible with presently practiced regional development strategies

should not be considered a prohibitive obstacle but rather a challenge to
reformulate present strategies.
The guiding objective would be to design consistent policies of regional
development which apply spatial closure at different levels in such a way
that the less-developed areas are put into a position to fully utilize their
own development potential and to lock in, to the maximum possible,
the development impulses they receive from higher developed areas. Only
empirical tests will show whether such alternative strategies are better
able than current regional development practice to contribute to estab-
lished objectives of spatial equity.

REFERENCES
Allardt, E., About dimensions of welfare. Research Group for Comparative Sociology, Research
Report no. 1, University of Helsinki, 1973.
Allen, K. and M.C. McLennan, Regional problems and policies in Italy and France. George Allen
and Unwin, London, 1970.
Alonso, W. and E. Medrich, 'Spontaneous growth centres in twentieth-century American urbaniza-
tion', in N.M. Hansen (ed.), Growth centres in regional economic development. The Free Press,
New York, 1972.
Appalraju, 1. and M. Safier, 'Growth centre strategies in less developed countries', in A. Gilbert
(1976).
Berentsen, B., 'Regional policy and regional inequalities in Austria: the impact of policy upon the
achievement of planning goals'. Unpublished PhD dissertation, Ohio State University, 1976.
Bourne, L.S., Urban systems: strategies for regulation. A comparison of policies in Britain, Sweden,
Australia and Canada. Clarendon Press, Oxford, 1975.
Boventer, E. von, 'Regional economic problems in West Germany', in Robinson (1969).
Bundesraumordnunsgesetz 1965 (GFR), Bad Godesberg.
Bundesminister fiir Raumordnung, Bauwesen und Stadtebau: Raumordnungsberichte 1968, 1970,
1972: and 1974 (GFR), Bonn-Bad Godesberg.
Buttler, F., Entwicklungspole und riiumliches Wirtschaftswachstum. J.C.B. Mohr, Tiibingen, 1973.
Cameron, G.C., 'Regional economic policy in the United Kingdom', in Sant (1974).
Cao-Pinna, V., 'Regional policy in Italy', in Hansen (1974).
Coates, B.E., R.J. Johnston and P.L. Knox, Geography and inequality. Oxford University Press,
1977.
Conroy, M.E., 'Rejection of growth center strategy in Latin American regional development
planning', Land Economics, Nov. 1973, 371-380.
Cumberland, J.H., Regional development experiences and prospects in the United States of
America. Mouton, Paris/The Hague, 1973.
Davin, L.E., 'The structural crisis of a regional economy. A case study: the Walloon Area', in
Robinson (1969).
Fox, K.A., Social indicators and social theory: elements of an operational system. John Wiley and
Sons, New York/London, 1974.
Friedly, Ph., National policy responses to urban growth. Lexington, Westmead, 1974.
Friedmann, J., 'The evolution of doctrine in regional development and planning. An American
view'. Series of seminars at the Interdisciplinary Institute of Urban and Regional Studies (UR),
Wirtschaftsuniversitat, Vienna, 1976.
Friedmann, J., 'Territory and function', Comparative Urbanization Studies, School of Architecture
and Urban Planning, University of California, Los Angeles, 1977.
REFERENCES 159

Friedman, Yona, Machbare Utopien: Absage an geliiufige Zukunftsmodelle. Fischer, Frankfurt am

Main, 1977.
Galtung, J., 'From value dimensions for social analysis to social indicators. Some ideas about world
accounting'. Unpublished manuscript, Oslo, 1973.
Gilbert, A., 'A note on the incidence of development in the vicinity of a growth centre', Regional
Studies, vol. 9, 325-333, 1975.
Gilbert, A.G. and D.E. Goodman, 'Regional income disparities and economic development: A
critique' in A. Gilbert (ed.), Development planning and spatial structure. John Wiley and Sons,
London, 1976.
Greenbie, B.B., Design for diversity. Elsevier, Amsterdam, 1976.
Hall, E.T., The hidden dimension. Doubleday, New York, 1966.
Hansen, N.M. (ed.), Public policy and regional economic development. Ballinger, Cambridge/Mass.,
1974.
Hansen, N.M., The challenge of urban growth: the basic economics of city size and structure.
Lexington, Toronto, London, 1975a.
Hansen, N.M., 'An evaluation of growth centre theory and practice', Environment and Planning,
vol. 7, no. 7,821-832, 1975b.
Helleiner, F.M. and W. Stohr, Spatial aspects of the development process. Proceedings of the
Commission on Regional Aspects of Development of the International Geographical Union,
vol. II, Allister Typesetting and Graphics, Toronto, 1974.
Hilhorst, J.G.M.,Regional planning: a systems approach. Rotterdam University Press, 1971.
Holland, S., The regional problem. MacMillan, London, 1976.
Hotger, H.E., 'Die Erfolge der regionalen Wirtschaftspolitik', Informationen zur Raumentwicklung,
vol. 5,187-192, Bonn Bad-Codesberg, 1974.
Illich, I., Tools for conviviality. Harper and Row, New York, 1973.
Isard, W., 'Oral summary of the sixteenth European regional· science congress', Copenhagen, 1976.
Knox, P.L., Social well-being: a spatial perspective. Oxford University Press, 1975.
Knox, P.L., 'Social well-being and North Sea oil: an application of subjective social indicators',
Regional Studies, vol. 10, no. 4,423432,1976.
Lacko, L., 'The hamlet and farmstead regions of Hungary: planning problems', Rural Transforma-
tion in Hungary, Publishing House of the Hungarian Academy of Sciences, Budapest, 1976.
Lasuen, J.R., 'Urbanization and development: the temporal interaction between geographical
clusters', Urban Studies, vol. 10, 163-188, 1973.
Lasuen, J.R., 'Spain's regional growth', in Hansen (1974).
Logan, M.J. and D. Wilmoth, 'Australian initiatives in urban and regional development', in H.
Swain (ed.), National settlement strategies east and west. Selected papers from a IIASA Con-
ference on National Settlement Systems and Strategies, Laxenburg, 1974 .
.Matzner, E., Notizen zur Gesellschaftsreform. Europaverlag, Vienna, 1976.
Moore, B. and J. Rodes, 'The effects of regional economic policy in the United Kingdom', in Sant
(1974).
Morrill, R.L., 'Growth centre - hinterland relations', in Helleiner and Stohr (1974).
Moseley, M.J., 'The impact of growth centres in rural regions - I and II',Regional Studies, vol. 7,
57-94,1973.
Moseley, M.J., Growth centres in spatial planning. Pergamon Press, Oxford/New York, 1974.
Nagoya Centre, Growth strategy and regional development planning in Asia. Proceedings of the
Seminar on Industrialization Strategies and the Growth Pole Approach to Regional Planning
and Development: The Asian Experience, Nagoya, Japan, 1975.
Nichols, V., 'Growth poles: an eValuation of their propulsive effect', Environment and Planning,
vol. 1, no. 2, 193-208, 1969.
Penouil, M., 'An appraisal of regional development policy in the Aquitaine region', in Robinson
(1969).
Pred, A., 'The interurban transmission of growth in advanced economies: empirical f"mdings versus
regional planning assumption', IIASA, RR-764, Laxenburg, Austria, 1976.
Pred, A., City systems in advanced economies: past growth, present processes and future develop-
ment options. Hutchinson, London, 1977.
160 REFERENCES

Prud'homme, R., 'Regional economic policy in France, 1961-1972', in Hansen (1974).

OECD, Regional problems and policies in OEeD countries. Paris, 1976.
Rapport de la commission 'Amenagement du territoire', Commissariat General du Plan: Prepara-
tion du VIe Plan, Le Documentation Francaise, Paris, 1971.
Rapport sur l'orientation preliminaire du VIle Plan, Annexe a la loi No. 75616 du 10 juillet 1975,
Imprirnerie des Journaux Officiels, Paris, 1075.
Raumordnungsprogramm fur die grossraumige Entwicklung des Bundesgebietes (Bundesraumord-
nungsprogramm), Schriftenreihe 'Raumordnung' des Bundesministers fUr Raumordnung,
Bauwesen und Stiidtebau, Bonn-Bag Godesberg, 1975.
Richardson, H.W., Regional growth theory. MacMillan, London, 1973.
Richardson, H.W., Regional development policy and planning in Spain. Saxon House, Westmead,
1975.
Robinson, E.A.G. (ed.), Backward areas in advanced countries. Proceedings of a Conference held
by the International Economic Association, MacMillan, London, 1969.
Rodgers, A., 'Migration and industrial development: the southern Italian experience', Economic
Geography, vol. 46,111-135,1970.
Ruehmann, P., Die regionale Wirtschaftspolitik Belgiens. Kieler Studien No. 93, M6hr, Tiibingen,
1968.
Sant, M., Regional policy and planning for Europe. Lexington, Westmead, 1974.
Scitovsky, T., The joyless economy, an inquiry into human satisfaction and consumer's dissatis-
faction. Oxford University Press, 1976.
Stohr, W. and F. Todtling, 'Evaluation of regional policies, experiences in market and mixed
economies'. Interdisiplinary Institute for Urban and Regional Studies, University of Economics,
Vienna, Discussion Paper 1. Published also in resumed form in N. Hansen, Human settlement
systems. International Institute for Applied Systems Analysis (IIASA), Ballinger Press, Cam-
bridge, 1978.
Streeten, P., 'Changing perceptions of development', Finance and Development. The World Bank,
vol. 14, no. 3, p. 15, 1977.
Sundqvist, J.L., Dispersing population: what America can learn from Europe. The Brookings
Institution, Washington, 1975.
Swedish Ministry of Labour and Housing and Ministry of Physical Planning and Local Government,
Planning Sweden. Stockholm, 1973.
Thoman, R.S., 'Appalachia after ten years', paper presented to the Third Symposium, Commission
on Regional Aspects of Development, International Geographical Union, Dushanbe, 1976.
TlSrnqvist, G.E., Systems of cities and information flows. Lund Studies in Geography, Senes B,
No. 28, Lund, 1973.
Waller, P., 'The spread of a growth pole - a case study of Arequipa (peru)', in Helleiner and
Stohr (1974).
Weiss, U., 'Industrielle Dezentalisierungspolitik in China - ein Entwicklungsmodell fliI die Dritte
Welt?' paper presented at the University of Economics, Vienna, 1976.
Wolf, F., 'Wie effJZient ist die regionale WirtschaftsfOrderung', Informationen zur Raumentwick-
lung, vol. 9,431438, Bonn, Bad Godesberg, 1975.
7. Aggregate efficiency and
interregional equity
HARRY W. RICHARDSON*

The potential conflict between aggregate efficiency and interregional

equity is perhaps the most crucial dilemma in regional economics and a
major obstacle in the way of effective implementation of regional policies.
The aims of this chapter are to identify cases where this conflict exists,
to elaborate on the nature of the trade-off, and to offer suggestions for
regional policy that help to minimize its effects. However, efficiency and
spatial equity are sometimes compatible, and the paper will also examine
these cases.

7.1 DEFINITIONAL PROBLEMS AND RELATED CONCEPTUAL ISSUES

Since much of the argument of this paper implies that there may be no
universally acceptable definition of either aggregate efficiency or inter-
regional equity, it is proposed to elaborate these concepts as the analysis
proceeds rather than to start with strict definitions and to stand by them.

7.1.1 Definitional problems

The concept of equity itself is, of course, very fuzzy. This should be men-
tioned but not explored too deeply to avoid the risk of the analysis
becoming bogged down in philosophy and ethics. One example of the
problems involved may serve as an illustration. Ko1m (1972) defined a
state of society as equitable if no individual would prefer to exchange his
personal situation with that of anybody else. This implies that an equit-
able state is unanimously preferred to any of its permutations. This works
fine if all individuals have compatible preferences. But if preferences con-
flict it will be difficult to find a solution in which welfare levels that
satisfy some individuals are not envied by others.

* University of Southern California.

162 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

Hudson (1974) argues that there are three faces of equity: equality
of incomes, equal reward for effort, and the distributional impact of the
supply of public services. The first is the aim of an egalitarian society, the
second,. in its emphasis on equality of opportunity rather than on equality
of incomes, provides a rationale for the market economy (with all imper-
fections such as discrimination eliminated), while the third is an important
but disguised influence on equity in mixed economies. With regard to
public service supply, first impressions may be incorrect. For instance, it
may be wrong to designate the supply of a public service as progressive
merely because it is targeted towards the poor. It is also necessary to
examine who pays for it (possibly the taxpaying poor) and who obtains
the maximum benefit (probably the suppliers themselves rather than the
recipients). The suppliers may be a strong and effective interest group
(e.g. defence contractors) with a very clear knowledge about the value of
their own activities to them, while consumers may have no idea about
how they should value public goods.
A major problem with any discussion of interregional equity is the
extreme vagueness of the concept. First, most of the analyses use average
per capita incomes within a prescribed geographical area as the basic
indicator, then typically use measures of dispersion among regions or
ratios to present the state of interregional equity. 1
Suspicion about using income or gross product per capita data as an
index of welfare has always been strong, and is increasing. 2 Since the
interest in interregional equity is motivated by distributional concerns,
the analyst will not be indifferent to the intraregional distribution about

1. The problems arising from how to measure interregional equity can be illustrated
by a debate about the trade-off between national growth and interregional equity
in Japan. Fukuchi and Nobukini (1970), using a balanced growth model in which
all regions had the same linear homogeneous production function, argued that the
growth rate in the 1950s and early 1960s was much too high to allow progress in
interregional equity. However, Sakashita and Kamoike (1973) showed that the
Fukuchi-Nobukini model did not predict a trade-off if a relative variance of per
capita output was substituted for an absolute variance index. (This modification
is sensible, of course, since if per capita income doubled in every region, the
absolute variance index would become four times larger.) Instead, they suggested
that the lack of movement towards interregional income equalization in Japan was
due to regional differences in production functions.
2. See, for example, the concept of MEW (measure of economic welfare) per capita
developed by Nordhaus and Tobin (1972).
DEFINITIONAL PROBLEMS AND RELATED ISSUES 163

which the average income level says nothing. 3 Using Gini coefficients
sheds no light on the impact of policies on the 'target' lower income
groups. Reliance upon income criteria requires acceptance of the mone-
tary standard as an accurate reflector of relative social value. This is not
always acceptable, especially where there is a mix of public and private
goods. In other words, different components of regional income may have
implicit weights that do not match their relative monetary values. As an
example, $ 10 million spent on a.regional hospital may increase societal
welfare much more than the same dollar increase in industrial output.
Moreover, many variables influencing welfare are not reflected in
money incomes at all. For example, many social indicators such as educa-
tional achievement, health indices, welfare facility scores (e.g. patients
per doctor) and amenities fall into this category. There is no reason why
the distribution of these welfare measures among regions should be con-
sistent with the interregional distribution of income. In Canada, welfare
measures appear to converge more among regions than income measures
(Firestone 1974). This may be true of many developed countries, whereas
the converse may hold in developing countries. The advantage of welfare
disparities over income measures is that they often correct for income
distribution bias since they force a mass penetration approach (Le. a hori-
zontal rather than a vertical aid strategy). In other words, social indicator
scores can be high only by catering to the whole population, whereas
average incomes per capita may be boosted even via the most regressive
policies. Some analysts go further than the social indicators approach and
attempt to develop an aggregate quality of life (QOL) index for a region
(Liu 1975a; 1975b). In a study of United States regions, the two lowest
income regions (East and West South Central) also had the lowest QOL
scores. 4
3. Another weakness of the average per capita income measure was pointed out by
Alonso (1968, p. 10): 'It is perfectly possible for the per capita income of a
depressed region to drop, but for all individuals in it to be better off. The condi-
tions are that the depressed region be subject to diminishing returns and that the
higher-income members leave for other regions where their skills are better re-
warded. Since the region is, in a sense, overpopulated (diminishing returns), the
leaving of some raises the income of those who remain.'
4. The aggregate QOL index approach is questionable. There is a subjective element in
the perception of quality of life (e.g. an individual's attachment to an area) which
is difficult to measure. Also the analyst's value judgments are likely to distort the
index. For example, out of 100 variables Liu chose only two leisure indicators, one
of which was symphony orchestras (rather than, say, a team in a major sports
league).
164 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

Another problem is that the meaning of spatial units is ambiguous, not

least because spatial boundaries are arbitrary. If equity is treated as a
dispersion measure of average per capita incomes among geographical
units (however defmed), there is no way of knowing whether a decline in
the dispersion coefficient means more, the same or even less income
equity. On top of the traditional problem that average incomes per capita
tell us nothing about the distribution of income, the average income per
capita index is obviously non-neutral over space. For instance, average
income per capita tends to diminish with distance from a region's metro-
politan core. The larger a region is defined (especially in the monocentric
case), the lower its average per capita income will tend to be. 5 Even
worse, in interregional comparisons the differentials in per capita income
between regions will appear wider if highly urbanized regions are com-
pared with less urbanized regions. Even if the regions are the same size,
the more urban region will have a larger urbanized core. A more satis-
factory measure of income disparities is obtained if the analysis
standardizes for the degree of urbanization. In the absence of such
standardization the most direct way of achieving interregional income
convergence is to strengthen the urban centres in the less developed
region. This implies an urban bias in regional policy. The point is that
interregional income convergence obtained in this way is frequently
associated with a deterioration in the personal income distribution.
There are many reasons why this may happen. Policies to promote the
cities probably aid existing residents (with higher incomes) more than
migrants while the benefits to the migrants help those who were among
the better off (in incomes, education and skills) in the rural areas. More-
over, the benefits of rapid urban growth tend not to be diffuse but, due
to their capitalization in land values associated with supply inelasticities
in land and housing markets, accrue disproportionately to landowners,
landlords and property developers. If the instruments of policy are either
industrial infrastructure provision or investment subsidies, the chief
beneficiaries are businesses and their owners, while the 'trickling down'
effects to the labour force may be weak or nonexistent. In other cases,
the aid takes the form of intergovernmental transfers, and there is no
guarantee (except where funds are specifically earmarked for equity
programmes) that the recipient government will spend the funds pro-

5. This assumes that the region is larger than a city. Obviously, this statement is false
if a region is defined as a central city and then compared with a metropolitan area.
DEFINITIONAL PROBLEMS AND RELATED ISSUES 165

gressively (cf. the experience in some areas of the United States with
general revenue sharing).
Although efficiency is usually regarded as a much simpler concept,
there is a parallel argument for a broader view of 'efficiency' than output
maximization subject to given resource constraints that, if accepted, may
affect the efficiency-equity analysis. The rumblings about the inadequacy
of maximizing GNP (and GRP) as an efficiency criterion have been loud
in recent years. Consider for instance the problem of environmental
quality. The relocation of polluting industries from a densely populated
core region to an underdeveloped periphery may lower the rate of return
to capital if the new location is a high-cost site, but it may promote inter-
regional income convergence and it may be 'efficient' in a broader sense
in that the relocation may create substantial net social benefits. The
implication is that cost-benefit analysis is superior to output maximiza-
tion as an efficiency criterion. 6
In the developing world in general, and Asia in particular, another
paradigm of regional development is evolving which treats efficiency and
equity in quite a different way. Efficiency is defined in terms of satisfying
human needs rather than via the optimal allocation of a scarce resource
such as capital. The focus is not on generating new resources by capital
accumulation but on making the best use of existing resources, pre-

6. Unfortunately, this conclusion raises new problems. These are connected with the
well-known fact that cost-benefit analysis ignores the income distribution question.
Thus, in a traditional treatment the income equalization effects of a strategy would
have to be evaluated ex post rather than included as an intrinsic element in the cost-
benefit analysis. In fact, this issue has been the subject of a minor controversy.
Goldfarb and Woglom (1974) draw a distinction between the 'efficiency rule' and
a 'welfare rule'. The efficiency rule states that investment (and policy) decisions
should be based on pure efficiency grounds, leaving the question of income re-
distribution to fiscal transfers, assumed costless and instantaneous (Musgrave 1969).
Weisbrod (1968) and Sosnow (1974), on the other hand, argue that income
distribution gains are a benefit and should be explicitly included in the cost-benefit
analysis. Thus a welfare rule is needed to help select projects and strategies. The
justification for this position is that the political system and the institutional frame-
work impose severe constraints on the scope for and speed of direct income
redistribution. McGuire and Garn (1969) suggested a decision rule in which all
projects are evaluated by weighting benefit-cost ratios by a welfare index. Projects
are then selected by maximizing the sum of scores subject to a budget constraint.
A compromise between these conflicting standpoints might be to take account of
the redistributional context of each programme, policy and strategy in the alloca-
tion of the overall budget for spatial policies, but then to apply the efficiency rule
within each programme.
166 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

eminenjly human resources, at the local level. Equity becomes the domi-
nant element in regional planning by promoting a 'bottom-up' approach
in which the target group itself is mobilized to participate in the planning
process at all phases (identification of the problems and their priorities,
finding solutions and execution of projects). Equity criteria are satisfied
since the people themselves are not only involved in planning but are the
main recipients of its benefits, while efficiency is redefined so that it
almost coincides with equity. Thus, the trade-off problem is avoided.
This interesting idea, as yet imperfectly developed, raises certain
difficulties. It is easy to demonstrate its effectiveness in simple cases such
as when a health expert works out with local villagers the solution to a
polluted water supply. However, if the dominant objective is to make the
best use of existing resources, the implication is that material living
standards in rural areas will remain low, absolutely and relatively to urban
dwellers. Also, village participatory democracy may be tolerated so long
as its scope is confined to designing small-scale projects, but beyond a
certain level there is a contradiction between this idea and non-democratic
political structures that so often prevail at the national level. The assump-
tion that the existing elites will give up power rather than face the alterna-
tives of unrest and revolution may be unfounded. Furthermore, it is
insufficient to expound this approach without designing policy instruments
that can implement it. Finally, the revulsion against the dogma of maxi-
mum capital accumulation and discrimination in favour of highly capital-
intensive projects should not be allowed to lead to the equally false dogma
that capital has no role in economic, social and human development. Since
capital remains in short supply, its allocative efficiency remains important
if new allocation criteria are substituted for the old.

7.1.2 Convergence and the polarization-dispersion process

A source of confusion is to equate the 'convergence-divergence' alternative
with 'dispersion-polarization'. In the standard regional growth model, the
implementation of regional policies (implying the dispersion of economic
activity) is assumed to be associated with a convergence in interregional
per capita incomes. Conversely, the polarization of mobile resources to
the core regions is often assumed to imply the divergence of regional
income differentials. Growth in space implies changes in the relative
importance of polarization and dispersion forces over space and through
time. However, the dominance of polarization forces at any time does not
necessarily imply divergence. Indeed, there may be very little inter-
DEFINITIONAL PROBLEMS AND RELATED ISSUES 167

relationship between changes in the spatial distribution of economic

activity and shifts in interregional per capita incomes.
A simple illustration will clarify this point. A convenient, if simplistic,
way of measuring polarization (and dispersion) is by changes in densities
(of population, industry, income, etc.). These have nothing to do with
incomes per capita. An increase in income per capita may be associated
with increasing, constant or declining densities. Income density will rise
relative to population density, but whether the income density increases
absolutely depends upon what happens to population. In a backward
region, it is possible for income per capita to increase in spite of a fall in
income density provided population falls fast enough. Analogously, in
the core regions polarization (increasing income density) may be
associated with a declining income per capita if population increases fast
enough.
This example is neither a theoretical abstraction nor mere arithmetical
logic. It occurs frequently in practice. The most common real-world case
is where rural-urban migration promotes polarization and convergence.
The migrants reinforce agglomeration economies in industry, but (unless
there are increasing returns to labour alone) they also may dampen wages
and per capita income growth. Rural outmigration often implies rising
rural per capita incomes due to improved agricultural productivity
(elimination of hidden underemployment) and the sharing of an expanded
output (and the income generated) among fewer people. These effects are
more likely to be found in developing countries. However, Spain offers a
West European example. During the years 1962-71, the share of the ten
richest Spanish provinces in total output increased from 53.7 to 56.7
percent while that of the ten poorest fell from 5.8 to 5.0 percent. Over
the same period, the per capita income ratio of the two groups narrowed
from 2.6 to 2.1 (Alcaide Inchausti 1974). The explanation of the paradox
is clear, given the above argument. The polarization of population was
stronger than that of economic activity. 7

7. This statement refers to interregional equity. However, if low income migrants in-
crease their productivity by moving, spatial concentration is quite consistent with
an improvement in interpersonal equity. It is only fair to mention the alternative
view. Lasuen (1974) argues that the income convergence is only an apparition due
to arithmetic (additions to population here, reductions there), not economics.
More important, the critical impact of interregional migration is the dynamic
effects of selective migration on the economies of lagging regions.
168 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

7.1.3 Place prosperity versus people prosperity revisited

As suggested by the familiar 'place prosperity versus people prosperity'
controversy (Winnick 1966; Whitman 1972), regional policies tend to be
indirect equity instruments. Aid is concentrated on certain areas in the
hope that the populations of these areas will be helped. As a means of
raising the incomes and welfare levels of individuals, these instruments
may be costly and inefficient. For example, Niles Hansen argues that the
strategy of pumping Federal resources into Appalachia in the 1960s was
less effective than subsidizing outmigration from the region (Hansen
1973). Of course, the question is very complex. Much depends on the
locational preferences of Appalachian residents for staying in Appalachia,
and on whether it is thought that such a psychic income benefit should
enter the policymaker's calculus. Also, the type of instruments adopted
makes a difference. Health and welfare measures, income-maintenance
subsidies to households, human capital investments and training pro-
grammes, all have a higher probability of helping individuals than the
bulldozer - the chosen instrument for Appalachia. 8
Assuming that the aim of equity strategies is to assist the poorest seg-
ment of society, the key issue is how to reach the 'target' population by
aid to places. Regional policies typically benefit industrialists, contractors
and landowners more than the unemployed, the unskilled and the im-
poverished. Indeed, the appeal of spatial redistribution policies in
countries with a highly skewed income distribution is precisely because
they do not redistribute income to the poor (cf. Brazil). In the worst cases
(typically in developing countries) subsidies are given to rich industrialists
based in the capital city (or abroad) to promote capital-intensive industry
in the backward regions, the profits of which (if any) are remitted to the
core region. Here, the commitment to spatial policies merely gives an
egalitarian gloss to a strategy which perpetuates, and may even reinforce,
the existing income distribution and power structure.
In spite of these words of caution, aid to places rather than to people
may be justified in certain circumstances. If the marginal utility of income
is lower in rich regions, regional policies may increase total welfare even if
aggregate output is not maximized. A similar argument is that the emer-
gence of a 'no-growth' philosophy in some developed countries, especially

8. In the first five years of the Appalachian Regional Commission, more than 60
percent of the programme's expenditures were for highways (Cumberland 1971,
96-97).
THEORETICAL CONSIDERATIONS 169

in their prosperous regions and amenity-rich zones, may make residents in

wealthy areas tolerant, and even supportive, of measures to aid less-
developed regions. 9 'Place prosperity' makes sense where the labour force
in a region is immobile and industry is footloose. Even where labour is
mobile, selective migration from a depressed region could lead to a fall in
output greater than the increase in output in the regions of destination. 10
This is a familiar hypothesis, but it needs thorough testing.

7.2 THEORETICAL CONSIDERATIONS

7.2.1 The static neo-classical model

Within the framework of the static neo-classical model, the conflict
between aggregate efficiency and interregional equity, when both are
narrowly defined, will not arise. Assuming identical production functions
subject to constant returns in all regions and perfect competition,l1 an
efficient allocation of resources will result in flows of capital from the rich
(high capital-labour ratio) regions to the poor (low capital-labour ratio)
regions with migration in the opposite direction. As a consequence, factor
returns will converge, thereby promoting interregional equity. It is even
possible to drop the assumption of perfect competition by allowing for
inertia, since this justifies regional policies in the form of relocation
subsidies to capital (and, more rarely, migration assistance).
Unfortunately, this scenario has three major defects. First, the factor
flows will probably be too small to generate much income equalization.
Second, production functions are likely to differ among regions, with
those in the prosperous regions subject to increasing returns due to

9. A related argument is that governmental income transfers from rich to poor regions
may produce direct benefits for tax-donors in the wealthy areas through inter-
dependent utility functions. In other words, tax donors derive utility from the fact
that their revenues become public transfers equivalent to charitable donations
(Hochman and Rodgers 1973).
1O.The effects of outmigration on the region of origin may include: (1) the impact of
its selectivity in terms of age, education and skills on the composition of the popu-
lation (e.g. a higher dependency ratio); (2) a reduction in employment; (3) an
increase in wage levels; (4) the possibility of skill shortages in specific sectors; and
(5) the impact of population decline on the tax base, real estate values and the size
of local markets (Kabaya 1971).
11. These are the critical assumptions. Secondary assumptions include full employ-
ment, a single commodity, a fixed labour supply, no technical progress and zero
transport costs.
170 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

agglomeration economies (von B6venter 1975). Thus, at best, regional

policies will do no more than offset the polarization of capital flows to
the core regions. Third, in a dynamic model innovations will be a source
of additional disturbance and an attraction for capital and possibly
labour. The theory of innovation generation and diffusion suggests that
innovations will not occur evenly over space but will be concentrated in
or near the major urban centres, a tendency favouring the more developed
regions. Thus, the 'base case' may be assumed to be the one in which
policymakers have to 'trade-off' aggregate efficiency for interregional
equity.

7.2.2 The trade-off function

The simplest way of formalizing the possibility that spatial policymakers
will have to compromise between the pursuit of national efficiency and
interregional equity is to hypothesize a trade-off curve such as tt' in
Figure 1. Using the crudest of definitions, efficiency in a dynamic context
may be represented by a higher national growth rate g, while z is a
dispersion measure of regional per capita incomes interpreted as an index
of interregional equity.12 The curve slopes downwards from left to right,
reflecting the hypothesis that policymakers have to permit the national
growth rate to fall to achieve a higher degree of interregional equity.
There are some points to notice about the trade-off curve tt'. First, its
shape is very unclear. It has been drawn as concave to represent the
assumption of diminishing marginal rates of substitution between effi-
ciency and equity. The economic rationale for this is that cutbacks on
growth imply sacrifices in economies of scale. Second, the slope of the
curve is steep because z is inelastic with respect to changes in g. The
explanation is that progress in interregional equity requires persistent
differences in regional growth rates, and hence long-run concessions in
the national growth rate, and cannot be achieved by resource reallocation
in the short run. Finally, the trade-off curve is 'objective' in the sense that
it plots the combinations of national growth rates and interregional equity
indices determined by technical conditions, defined broadly to include
institutional constraints. For instance, the trade-off curve will tend to be
much flatter in a developed than in a developing country. There are at

12.It should be obvious from the earlier discussion that these definitions are much too
simplistic. However, it is pedagogically convenient to use them here to illustrate the
concept of the trade-off function.
THEORETICAL CONSIDERATIONS 171

least two reasons for this: many of the 'backward' regions of developed
countries have some growth potential because of their established infra-
structure, skilled labour pools and urban agglomeration economies; and in
such countries many industries are 'footloose' due to an extensive trans-
portation and communications network and relatively insignificant
locational cost differentials. Steering industries to the lagging regions may,
therefore, tend to be equity-promoting but efficiency-neutral. Situations
of this kind will be defined in this paper as effiency-equity compatible.
g

Figure 1. Trade-off between national growth and interregional equity.

172 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

The above argument implies that the trade-off curve tells us nothing
about the policymaker's own trade-off. This requires specification of the
policymaker's preference function. The preference functions are a kind of
indifference curve with each point on anyone 'curve' representing iso-
utility combinations of growth rates and equity indices. The simplest
approach is to assume that policymakers correctly identify and intend to
implement society's preferences, and that society's preference substitu-
tions between growth and equity are linear. In this case, the preference
functions become a set of straight lines sloping downwards to the right.
The linearity assumption means that society prefers a constant improve-
ment in equity -fwm each one percent decline in the growth rate (needless
to say, this assumption is not critical to the analysis). The optimal choice
for society is to select the point on the 'objective' trade-off curve tt'
which is tangential to the highest preference function PF 1 • This implies
that the combination of gl and ZI is optimal for society. The steeper the
set of preference functions the more society is inclined to favour inter-
regional redistribution. A flat preference function, on the other hand,
implies a growth-oriented society unwilling to consider sacrifices in
national efficiency even for major interregional equity gains. In an equity-
oriented society, i.e. one with steep preference functions (e.g. PF2 ), it is
possible that the tangency point may be located on the section of the
trade-off curve corresponding to a negative growth rate.
The most systematic analysis of the relationship between aggregate
efficiency and interregional equity has been a theoretical and empirical
study of Japan by Mera (1975). The transition of an economy from an
underdeveloped to a developed state is accompanied by spatial polariza-
tion because, although there are diminishing returns to density in the
primary sector, there are increasing returns to density in secondary and
tertiary industries. Thus, the conflict between macro-efficiency and spatial
equity springs from the fact that income maximization requires a higher
degree of spatial concentration of economic sectors in space while income
equalization demands an even distribution of urban-oriented industries.
If direct income transfers are politically feasible (and if taxing the resi-
dents of the rich regions to finance such transfers does not damage their
productivity), the conflict can be avoided. If not, the spatial reallocation
of physical social capital (e.g. via a growth centre strategy) may be the
most appropriate instrument for promoting income equity in space.
As an exercise shedding light on the potential trade-off, Mera calculates
the efficiency loss of the reallocation of social capital to achieve income
THE WILLIAMSON CURVE 173

equalization among Japan's nine regions. In the short run income equaliza-
tion would involve sacrificing 30 percent of national output; in the longer
run when capital and labour can be treated as mobile between regions,13
the loss runs to 12 percent of output. Finally, Mera demonstrates
theoretically that in a two-region system without mobility of labour the
transfer of investment resources in the form of a lump-sum subsidy from
the rich to the poor region may not be harmful to efficiency within a
certain range, i.e. the transfer should not be so large as to throw the
subsidized region off its optimal path of capital accumulation.

7.3 THE WILLIAMSON CURVE

In a dynamic context, the case for no trade-off, at least in developed

countries, was best made by Williamson (1965). He hypothesized inter-
regional income divergence in the early phases of rapid economic develop-
ment with subsequent convergence as the economy matures. His primary
measure (Vw ) was a weighted (according to regional population shares)
coefficient of variation. in regional per capita incomes relative to the
national average. Graphing this coefficient against a level of development
indicator yields the 'Williamson curve', an inverted U. If this hypothesis
is valid, developed countries (and developing countries in the future) can
expect a narrowing in regional economic disparities over- time, in effect a
reconciliation between interregional equity and aggregate efficiency.
Williamson's analysis, though seminal, has been criticized over the years
(for three of the more recent critiques see Metwally and Jensen 1973;
Parr 1976; and Gilbert and Goodman 1976). His own results were am-
biguous, partly because of the sparsity of data - in only five countries
(the United States, Canada, Sweden, Italy and Germany) were time
series data available for over twenty years. A recent supplementary study
suggests no relationship between Vw and GNP per capita in a cross-
sectional analysis of fifteen countries with a wide 'level of development'
spread. Similarly, analysis of time series data on ten countries revealed
five 'divergers' (Chile, Tanzania, Argentina, Thailand and the Philippines)
and five 'convergers' (Brazil, India, Spain, Colombia and Mexico) (Gilbert

13.In particular, the migration of popUlation will tend both to improve the distribu-
tion of income (by narrowing wage differentials) and raise aggregate efficiency in
production (if factor mobility is interpreted as a search for higher returns).
174 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

and Goodman 1976, 118-119). Moreover, the coefficient of variation

measure adopted takes on very different values according to whether the
national metropolis is treated as a seperate region or whether it is merged
in a larger (typically rural) region.
The hypothesis probably relies too heavily on United States experi-
ence 14 and on a belief in historical determinism. In fact, development
patterns vary widely among countries. Unless the turning point in the
Williamson curve can be identified a priori, it is easy to explain away
exceptions to the hypothesis in terms of it being either too soon for
interregional diffusion or as a result of imperfections in the statistics.
A major problem with Williamson's study, especially from the point of
view of the efficiency-equity controversy, is its lack of a theoretical
framework. Although one explanation of convergence during maturity
might be regional l'0licies, introduced in response to political pressures
from backward regions and because rich countries are more able to afford
the opportunity costs of resource diversion, these are probably a rein-
forcing rather than the main factor. There are several important economic
forces that may account for interregional convergence. Even during the
polarization phase of development, as suggested above, rural-urban migra-
tion may raise per capita incomes in the rural areas and dampen income
growth in the core regions (the latter possibility depends on the extent of
increasing returns to production and the functional distribution of income
in the urban sector). Later, as population and incomes continue to rise in
the less developed regions, economies of scale eventually justify
investment there. This is particularly attractive in sectors where transport
costs create a protected local market. The balance between 'spread' and
'backwash' effects may begin to shift in favour of spread, as the effects
of continued polarization flows are offset by the diffusion of technology
and managerial expertise and by a rising demand for the lagging regions'
products that are complementary to the economic structure of the core
regions. The establishment of branch plants may be the first, and possibly
the key, instrument of diffusion. Changes in locational costs, such as
lower interregional transport costs, changes in relative costs of energy,

14.Spontaneous diffusion (and equalizing) forces in the United States have two
dimensions: shifts in locational advantages attracting industries out of the North-
east to the South and the Southwest; changes in household preferences and life-
styles resulting in a decentralization of popUlation out of the large cities into
smaller towns and rural areas in the intermetropolitan periphery (counterurbaniza-
tion; see Berry 1976).
THE WILLIAMSON CURVE 175

mobile external economies, lower wages, and the pull of amenities (e.g.
climate, recreational resources), may favour the less developed regions in
the location decisions of new and the relocation decisions of existing
firms. If export markets are volatile and very competitive, the exploitation
of the resources and demand potential of the lagging regions may offer a
more stable means of maintaining, or raising, the national growth rate.
Finally, the agglomeration economies that were responsible for the earlier
development of the core regions may begin to dissipate because of 'con-
gestion' costs such as rising land values, soaring labour costs and environ-
mental deterioration and symptoms of senescence such as obsolete
plants, a decline in entrepreneurial and managerial talent and debilitation
of the profit motive.
Since many of these arguments are speCUlative and the related hypo-
theses are untested, while the empirical base remains weak, the Williamson
curve remains a tantalizing question mark. For example, there has been no
discussion of how much inequality can be expected in the divergence
phase. In other words, is the curve very steep or almost flat?15 The pre-
diction that developing countries will eventually enter the convergence
phase implicitly assumes that they will become developed to levels com-
parable with those of the advanced industrial countries. This is by no
means self-evident. Also, the regional disparities exhibited by the develop-
ing countries are far wider than those of the developed countries at the
time of their maximum divergence. Of the many possible reasons for this
fact, the restriction of externally-induced growth (e.g. by multinationals
and earlier by colonialist powers) to the core regions is among the most
compelling. Finally, to use the Williamson curve to shed light on the
efficiency-equity trade-off and its possible non-existence, it would be
necessary to be able to separate out the regional policy effect from the
'spontaneous dispersion' effect. The difficulties in isolating regional policy
impacts by comparing 'actual' and 'expected' performance are notorious
(Moore and Rhodes 1973, 1976).

IS. Parr (1976) raised the question of whether there was any systematic variation
among countries in the height and base length of the curve. Noone has attempted
to answer it.
176 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

7.4 OTHER EFFICIENCY-EQUITY COMPATIBILITIES

There are a few other cases, apart from those illustrating the Williamson
hypothesis, that may be efficiency-equity compatible, thus avoiding the
trade-off. One was mentioned earlier (Section 7.2.2); the possibility that
in certain developed countries the combination of footloose industries
and an almost ubiquitous infrastructure and adaptable labour force may
allow policymakers to promote the poorer regions without any loss in
efficiency. A more subtle example, advanced by Higgins (1973, 1975), is
that narrow regional disparities are a precondition for avoiding rampant
inflation. He argues that Phillips' curves in countries with small regional
gaps are lower and to the left of the curves in countries where regional
disparities are wide. The explanation is in terms of differential labour
market pressure and the 'price-setting' characteristics of firms in the
developed regions. The point is that labour markets are regional rather
than national, especially in the high disparity countries, whereas price
increases diffuse very quickly throughout the national economy. This
argument is suggestive rather than convincing. A contrary view is that
because of massive interregional leakages, strong action to develop the
backward regions might even accelerate inflation by stimulating demand
for the capital goods of the core regions. Even before there are direct
empirical tests of these hypotheses, however, price stability should
be considered as an element in efficiency in addition to more obvious
criteria such as maximizing GNP.
The existence of space implies frictions on mobility. If resources are
underutilized in less developed regions (e.g. the evidence of low labour
participation rates) but are immobile, it may be more efficient to invest in
such areas to boost growth, and incidentally to raise incomes, than to
leave the outcome to the imperfections of the market. This important
point was made by Borts and Stein (1964) when they argued that the
appropriate rate of return in such a case is the gross, not the net, marginal
product of capital, since the value of immobile and unemployed labour
would otherwise be close to zero.
Yet another instance of efficiency-equity compatibility may arise
under certain assumptions about the existence of externalities. This
implies yet again that efficiency should be measured in terms of the social
rather than private rates of return. It is not difficult to outline a scenario
with externalities of the kind fitting the efficiency-equity compatibility
case. For instance, if the relocation of industry from advanced to lagging
EQUITY IMPACTS OF REGIONAL POLICIES 177

regions dampens rising congestion costs at the core and builds up the
industrial base of the periphery enough to ensure its future growth poten-
tial, these dynamic social efficiency gains might easily outweigh any loss
in profitability measured by market criteria.
Furthermore, examples of efficiency-equity compatibility might be
based on the hypotheses that the marginal capital-output ratio is lower
in poor regions (implying a higher marginal productivity of investment)
and/or the exhaustion of agglomeration economies of the core region
(with the probable implication of an optimal urban size). In fact, neither
argument is convincing. The marginal capital-output ratio is higher in
lagging regions because of their poverty of infrastructure, absence of
external economies and location disadvantages. The concept of an optimal
city size is based on assertion and value judgment and is virtually untest-
able (Richardson 1973b).

7.5 EQUITY IMPACTS OF REGIONAL POLICIES

7.5.1 Growth centre policies

It is quite common for policymakers to disguise an efficiency-oriented
spatial policy such as a growth centre strategy as an equity strategy.
For example, the development poles policy pursued by Spain after 1964
was 'sold' as a policy to reduce regional income disparities. In fact, the
selected poles were ideal locations for specific types of industrial develop-
ment and several of the poles were situated outside the backward regions
(Leira 1973). The so-called regional policy was little more than the
locational implications of a sectoral strategy.
Leira's critique is even harsher. The strategy was conservative and risk-
averting, because 'natural' poles were chosen which were bound to grow
anyway. Perhaps even more important was that a growth centre strategy
was politically palatable to the government, especially when compared
with alternatives such as agrarian reform or direct income redistribution.
As implemented, the policy had a low fiscal cost, and was compatible with
the ideology of the planners who saw planning as a vehicle for mobilizing
private enterprise, not as a framework for optimizing the allocation of
resources.
The Brazilian case illustrates the same tendency in a more extreme
fashion (Gilbert and Goodman 1976). The main instrument used to
promote economic expansion in the underdeveloped Northeast region,
178 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

the well-known 34/18 investment incentive scheme, was successful in

promoting industrial growth in the regions' leading urban centres (espe-
cially Recife). At first sight, the equity consequences appear far from
negligible with the value of net regional product per capita relative to the
national average improving from 0.42 in the late 1950s to 0.48 in the
1960s. However, the evidence is clear that these trends were accompanied
by a marked deterioration in the distribution of income, especially in the
urban sector. One explanation is the capital-intensive bias of the 34/18
programme so that the employment impacts of regional policies were
minimized. More generally, measures to redistribute economic activities
among places do nothing to improve the welfare of the 'target' poor
unless explicit policies to redistribute income among persons are adopted.
The results in the Brazilian Northeast are striking and dismal. The bottom
30 percent o~ the urban population receives 6 percent of the income,
while the top 30 percent receives 71 percent. The Gini coefficient rose in
the 1960s from 0.49 to 0.56. The real income oflow-income households
in Recife declined absolutely. Less than the minimum wage (defined as
a subsistence income) was earned by 60 percent of the urban employed
(1972), malnutrition was rife, and life expectancy was 21 percent below
the national average. Yet the choice of other indicators, e.g. growth in
industrial output, might suggest that the performance of the Northeast in
the 1960s was impressive.
Of course, all locational policies are spatially discriminatory. But the
central axiom of growth centre theory is that the greater the degree of
discrimination, i.e. the more resources are spatially concentrated, the
higher the overall rate of regional development. However, as the examples
just discussed illustrate, this strategy, whatever its merits, implies
favouring the urban centres, usually but not necessarily the major cities,
in the allocation of infrastructure and creating the bulk of new jobs in or
near these centres,16 and yet often does not avoid a deterioration in the
intra-urban distribution of incomeY In addition, since the 'backwash'
16.A recent study of developing countries by a team at Cornell University found that
those countries with the most decentralized systems of local organization and with
the more impressive rural development performance (higher agricultural productiv-
ity, successful land reform, more equal access to public services, and more labour-
intensive production techniques) also had the more equitable income distributions
(quoted in United Nations 1977).
17.An analysis of SMSAs larger than 250,000 in the United States found that rapid
urban growth was associated with increasing income inequality (Haworth, Rasmus-
sen and Long 1975).
EQUITY IMPACTS OF REGIONAL POLICIES 179

effects of growth centre promotion on its hinterland are generated much

earlier than the 'spread' effects (Richardson 1976), the 'enclave' charac-
teristic of the growth centre, at least in the short run, is almost impossible
to avoid. If the consequences of a growth centre policy, still the most
widely accepted regional development strategy, imply a high, but un-
equally distributed income level at the pole and a uniformly low, and
possibly diverging income level throughout the hinterland, the role of
regional equity in spatial policies requires redefinition.

7.5.2 Generative growth policies

Some other considerations arise when the spatial dimension is explicitly
introduced. Some years ago I contrasted two concepts of regional devel-
opment, 'competitive growth' and ' generative growth' (Richardson
1973a, 86-88~ In competitive growth the national growth rate is assumed
given and interregional development is treated as a zero-sum game. To
stimulate activity in some regions, production must be reduced in others.
If marginal returns are lower in the lagging regions, the trade-off is un-
avoidable. Generative growth, on the other hand, treats the national
growth rate as the resultant of the growth rates of individual regions.
Agglomeration economies and spatial concentration of industry may
result in higher output than if production is dispersed. The reorganiza-
tion of the regional transportation system, a more efficient settlement
pattern, the adoption of growth-inducing innovations, implementation
of a growth centre strategy, all these are examples of changes that may
boost regional, and hence national, growth. The improvement in spatial
efficiency in a previously poorly integrated region may generate a higher
long-run social rate of return than if resources had been invested in the
core regions (i.e. discounted net opportunity costs may be non-positive).
In some cases the efficiency gains may require no additional resources
(e.g. for infrastructure) to what would have been allocated to the region
in any event; the gains accrue from a better spatial arrangement. Thus,
intra-regional spatial efficiency may have a positive feedback on the
growth rate of the aggregate economy, hence the description 'generative'.
The policy implication of generative models is that regional planners
should be committed to fostering favourable growth conditions within
a region rather than relying solely on the diversion of resources from
other regions.
180 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

7.6 POLICY IMPLICATIONS

Since in some circumstances there is a trade-off between aggregate effi-

ciency and interregional equity while in others there is efficiency-equity
compatibility, it is not surprising that this analysis does not suggest a set
of universal policy prescriptions. The most cost-effective policies will
vary from one situation to the next. Nevertheless, the most critical
question refers to the policies that should be adopted in those cases where
the existence of a trade-off is indisputable. The analysis in this paper
offers some general guidelines to minimize these trade-off costs.
The policy objective then becomes how to promote interregional equity
at the least sacrifice in terms of economic efficiency. The simplest, some
may say naive, strategy is to allow market forces to determine the alloca-
tion of investment resources, with appropriate intervention to correct for
any misallocation due to unpriced externalities. Even if this leads to a
concentration of economic activity in already developed regions, the
equity goal is then dealt with via direct income transfers from the rich to
the poor regions. Unfortunately, the world is not that simple. This solu-
tion raises many objections, several of them of a practical nature. Inter-
regional income transfers on the scale required may be difficult to
introduce politically, even after persuading the rich regions that some of
the payments will leak back in the form of demand for the regions'
exports and even after allowing for the donor regions' utility gains from
helping out. Perhaps more serious is the possibility that the higher taxes
needed to finance these transfers may discourage investment and pro-
ductivity improvement, thereby depressing the economy's overall
performance. Also, as argued earlier, the direct income transfer strategy
makes more sense if it is from persons to persons rather than from area to
area, since there are poor people in rich regions and rich people in poor
regions. The biggest problem of all is that the income transfer solution
does very little to promote the development of the lagging regions so
that the 'drain' becomes permanent. A strategy that stresses investment
in the less-developed regions may facilitate a subsequent 'bootstraps' kind
of local development and improve the long-run aggregate growth perform-
ance of the economy.
Although the 'backstop' argument in favour of regional policy may be
phrased in terms of social objectives, it is usually possible to marshal an
efficiency case. Its precise characteristics may vary from one situation to
the next, but prominent elements include the pursuit of long-term as
POLICY IMPLICATIONS 181

opposed to short-term efficiency, the maximization of the social rate of

return (i.e. taking full account of negative externalities in prosperous
regions and positive externalities in underdeveloped regions), exploiting
underutilized or immobile resources, mitigating inflation and minimizing
environmental disruption (un challenge ably a dimension in the efficiency
of a modern economy).
Given that some efficiency arguments exist, the problem then becomes
how to design regional policies that promote equity, and within regions
even more than between regions. Although the implications of this point
are quite obvious, they are often neglected by regional policymakers. The
reasons for this neglect are unclear. There is probably a quite unjustified
belief in the efficacy of capital subsidies (investment grants, loans and
tax credits) and in the role of capital-intensive industries as a generator of
regional-wide development. But for every capital subsidy there is a labour
subsidy that offers an equivalent amount of aid (of course, in a general
formula scheme some firms and industries may find a capital subsidy more
beneficial while others would favour a labour subsidy). Also, the provision
of social infrastructure may help to build up a region's agglomeration
economies in a way that may attract new industry as effectively as a direct
subsidy.
Apart from measures to stimulate the establishment of branch plants
(an important instrument of diffusion, and hence an equalizing force,
though unfortunately applicable largely in developed countries where
national demand levels justify multiple plants), the most sensible strategy
is a mix of social infrastructure provision and labour subsidies. Social
infrastructure emphasizes the improvement in non-income components of
welfare, and usually implies 'horizontal aid'. Criteria stressing the increase
in social indicator scores avoid the regressive bias of many measures that
raise average per capita incomes but fail to reach down to the target poor.
The provision of public services, especially in the fields of health, welfare
and human capital, offers a good chance of improving equity, provided
that the policymaker watches out for the potentially pernicious influence
of the public service supplier and for regressive elements in how the
services are financed.
Measures to promote industrial development in lagging regions have a
poor record as equity instruments. In most cases subsidies have been
heavily biased to favour capital-intensive technologies, and even in the few
cases where labour subsidies were operative (e.g. the Regional Employ-
ment Premium in Britain) they were often outweighed by capital sub-
182 AGGREGATE EFFICIENCY AND INTERREGIONAL EQUITY

sidies. Since lagging regions almost always have higher unemployment

rates and lower labour participation rates, the 'trickling down' effects of
the investment subsidy approach have been much weaker than if an
alternative strategy had been adopted. Labour subsidies, on the other
hand, imply higher wages (unless none of the subsidy is passed on, a
requirement which requires strong assumptions about labour market
elasticities) and more jobs (via effects on the capital-labour ratio and,
because of lower costs, on the level of output).
To conclude, in some cases aggregate efficiency and interregional
equity are compatible, especially when a long-run view is taken and when
efficiency is defined more broadly than maximization of private rates of
return. In other cases, however, a trade-off between efficiency and equity
is unavoidable. The task facing policymakers then becomes one of instru-
ment design so as to minimize efficiency losses and to assure an improve-
ment in interpersonal equity as well as spatial equity (i.e. convergence in
interregional per capita incomes and welfare indices). In spite of a poor
record in many countries, this design problem is quite tractable.

REFERENCES
Alcaide Inchausti, J., La Distribucfon de la Riqueza y de la Renta en la Sociedad Espanola: Balance
de Tres Planes de Des/mollo. Madrid, 1974 (mimeographed).
Alonso, W., Equity and its relation to efficiency in urbanization. Center for Planning and Develop-
ment Research, University of California, WP 78, Berkeley, 1968.
Berry, B.J.L., The counter-urbanization process; how general? International Institute for Applied
Systems Analysis Conference, Schloss Laxenburg, Austria, 1976.
Borts, G.H. and J.L. Stein, Economic growth in a/ree market. Columbia University Press, New
York,1964.
Boventer, E.G. von, 'Regional growth theory', Urban Studies, vol. 12,1-29,1975.
Cumberland, J .H., Regional development: experiences and prospects in the United States of
America. Mouton, The Hague,1971.
Firestone, O.J., 'Regional economic and social disparity', 205-267, in O.J. Firestone (ed.), Regional
economic development. University of Ottawa Press, 1974.
Fukuchi, T. and M. Nobukini, 'An econometric analysis of national growth and regional income
inequality ',In ternational Economic Review, vol. 11,84-100,1970.
Gilbert, A.G. and D.E. Goodman, 'Regional income disparities and economic development: a cri-
tique', 113-141, in A.G. Gilbert (ed.), Development planning and spatiJZI structure. Wiley, New
York,1976.
Goldfarb, R.S. and G. Woglom, 'Government investment decisions and institutional constraints on
income redistribution',Journal of Public Economics, vol. 3, 171-180, 1974.
Hansen, N.M., Locational preferences, migration and regional growth. Praeger, New York, 1973.
Haworth, C.T., D.W. Rasmussen and J.E. Long, 'Income redistnoution effects of urban population
growth', Department of Economics, Florida State University, Tallahassee, 1975 (mimeo-
graphed).
Higgins, B., 'Trade-off curves and regional gaps', 152-177, in J. Bhagwati and R.S. Eckaus (eds.),
Development and planning: essays in honor of Paul Rosenstein Roden. Allen and Unwin, 1973.
Higgins, B., 'Trade-off curves, trends and regional disparities: the case of Quebec', Economie
Appliquee, vol. 28, 331-360,1975.
REFERENCES 183

Hochman, H.M. and J.D. Rodgers, 'Utility interdependence and income transfers through charity',
63-77, in K.E. Boulding, M. Pfaff and A. Pfaff (eds.), Transfers in an urbanized economy. Wads-
worth, Belmont, 1973.
Hudson, B.M., 'The three faces of equity', School of Architecture and Urban Planning, University
of California at Los Angeles, DP 47,1974.
Kabaya, R., 'Development of poor regions: general considerations and the case of Japan', Institute
of Urban and Regional Development, University of California, WP 159, Berkeley, 1971.
Kolm, S.C., Justice et Equiti. Edition du Centre National de la Recherche Scientifique, Paris,
1972.
Lasuen, J.R., 'Spain's regional growth', in N.M. Hansen (ed.), Public policy and regional develop-
ment. Ballinger, Cambridge, Mass., 1974.
Leira, J., 'Growth poles in Spain: a legitimizing instrument for an efficiency-oriented policy',
University of California at Berkeley, 1973 (mimeographed).
Liu, B.C., 'Differential net migration and the quality of life', Review of Economics and Statistics,
vol. 57, 329-337, 1975a.
Liu, B.C., 'Quality of life: concept, measure and results', American Journal of Economics and
Sociology, vol. 34, 1-13, 197 5b.
McGuire, M.C. and H.A. Gam, 'The integration of equity and efficiency criteria in public project
selection', Economic Journal, vol. 79, 883-893, 1969.
Mera, K.,!ncome distribution and regional development. University of Tokyo Press, 1975.
Metwally, M.M. and RC:Jensen, 'A note on the measurement of regional income dispersion',
Economic development and cultural change, vol. 22, 135-136, 1973.
Moore, B.C. and J. Rhodes, 'Evaluating the effects of Britain's regional economic policy', Econo-
mic Journal, vol. 83, 87-100,1973.
Musgrave, R., 'Cost-benefit analysis and the theory of public finance', Journal of Economic Litera·
ture, vol. 7,443-444, 1969.
Nordhaus, W.D. and J. Tobin, 'Is growth obsolete?', 4-17 in National Bureau of Economic Re-
search,Economic growth, 50th Anniversary Colloquium, vol. 5, New York, 1972.
Parr, J.B., 'Welfare differences within a nation: a comment ,Papers, Regional Science Association,
vol. 32, 83-91, 1976.
Rhodes, J. and B.C. Moore, 'Regional economic policy and the movement of manufacturing firms
to development areas', Economica, vol. 43,17-31,1976.
Richardson, H.W., Regional growth theory. Macmillan, London, 1973a.
Richardson, H.W., The economics of urban size. Saxon House, Farnborough, Harts, 1973b.
Richardson, H.W., 'Growth pole spillovers: the dynamics of backwash and spread', Regional
Studies, vol. 10,1-9,1976.
Sakashita, N. and O. Kamoike, 'National growth and regional income inequities: a consistent
model',!nternational Economic Review, vol. 14,372-382,1973.
Sosnow, N.D., 'Optimal policies for income redistribution', Journal of Public Economics, vol. 3,
156-169,1974.
United Nations, 'Population and equality',Development Forum, vol. 5, no. 1, p. 5,1977.
Weisbrod, 'Income distribution effects and benefit cost analysis', 177-222 in S.B. Chase (ed.),
Problems in public expenditure analysis. Brookings Institution, Washington, D.C., 1968.
Whitman, M. von N., 'Place prosperity and people prosperity: the delineation of optimum policy
areas', 359-393, in M. Perlman, C.L. Leven and B Chinitz (eds.), Spatial, regional and popula·
tion economics; essays in honor of Edgar M. Hoover. Gordon and Breach, New York, 1972.
Williamson, J.G., 'Regional inequalities and the process of national development', Economic
Development and Cultural Change, vol. 13,3-45,1965.
Winnick, L., 'Place prosperity vs people prosperity: welfare considerations in the geographic re-
distribution of economic activity', 273-283, in Real Estate Research Program, Essays in urban
land economics in honor of Leo Grebler. University of California, Los Angeles, 1966.
8. Capital, labour and the regions: Aspects
of economic, social and political inequality
in regional theory and policy
STUART HOLLAND*

Much of the theory of regional development has a pronouncedly Alice-in-

Wonderland character. Students who take courses on regional and urban
policies because they feel this will equip them to understand or tackle
actual regional and urban problems tend to be confronted by the kings
and queens of the subject with a series of 'as ifs' and 'but suppose' hypo-
theses which turn the subject upside down or back to front in a series of
mystifications. This is quite apart from the issue that if they fail to give
the right answers to the wrong questions in assessment they may find
that someone has flunked them on the course - the metaphorical equiva-
lent of 'off with their heads'.
If this is not uniformly typical of assessment, artificiality is certainly a
uniform enough feature of neo-classical self-balance models of regional
resource allocation. One might draw on Gestalt psychology to explain the
persistent emphasis on harmonious market mechanisms in so much
regional theory. In other words, with the same data on regional disparities
in economic structure and social distribution of jobs and incomes, some
specialists will persist in seeing a departure from underlying convergence
through the market mechanism, while others see symptoms of an inherent
inequality in the working of the market itself.
In principle, this could itself be seen in terms of the analytical and
theoretical assumptions which specialists bring to the analysis of regional
inequality, and which influence their perception of allegedly neutral
'facts'. In practice, this will include the nature of the training and
education of the educ;ltors themselves. But this in turn means much more
than the narrow confines of regional and urban theory. It will involve the
implicit assumptions or basic presuppositions of the theorists with regard
to questions which are not necessarily raised in explicit form in regional
or urban analysis, including the nature of the capitalist economy and

* University of Sussex. Copyright by Stuart Holland. All rights reserved.

186 CAPITAL, LABOUR AND THE REGIONS

society which underlies the market system, the perception of legitimate

or illegitimate state intervention, and so forth.
It is the framework of such implicit assumptions, reflecting a nexus of
values, concepts, ideas and, frequently, dogmas, which constitutes the
dominant ideology for individual specialists and - in combination - a
collective ideology of a particular school of thought and analysis. On
occasion, over time, the implicit ideology is made explicit, as has been the
case with Isard's later work (1969). This assumes not only a pluralistic
competitive economy but also a pluralistic competitive political system.
Such an unreal world of atomistic producers, subject to consumer sover-
eignty and maximising profits and benefits in an incremental manner, very
much suits the incremental marginalist techniques of analysis which pass
for regional science in the Isardian school.
This Isardian approach has been criticised elsewhere (Holland 1976a).
The present contribution seeks to extend some of my previous analysis,
especially by a more diagrammatic treatment of issues in the contemporary
political economy of regional inequality and their relation to the current
prolonged crises in capitalism. A central feature of this reality has been
the emergence since the war of giant multi-product, multi-plant, multi-
sectoral and multinational companies in between the micro-economic
units of much regional theory and the macro-economic aggregates of
national and regional economics. It .is these companies which I have
categorised as meso-economic, i.e. in between micro- and macro-economic
categories.

8.1 THE FALLACY OF SELF-BALANCE THEORY AND CURRENT REGIONAL

POLICY

If one takes the self-balance theory of interregional factor movements,

which assume a long run trend to equality through the free working of the
market, it is clear that they assume the kind of symmetry represented in
Figure 1. In other words, given a more developed and less developed
region (MOR and LOR respectively), it is assumed that labour will migrate
from the more labour abundant LOR to the more capital abundant MOR
(in order to gain either work or higher paid work), while capital will
migrate to the more labour abundant LOR from the MOR (to gain lower
cost labour).
THE FALLACY OF SELF-BALANCE THEORY 187

Labour Capital
(Investment)

Figure 1. The neo-classical case of factor flows.

Such a model, with its assumed equalisation of factor returns, is cer-

tainly vitiated by the twentieth century reality of trades unions, especially
when taken in conjunction with the further reality of domination of
manufacturing industry - the key regionally mobile sector - by big
business. In other words, trades unions will tend to demand comparable
wage rates in both the MOR and the LOR from the larger scale fIrms
which in tum tend to be those mainly capable of organising interregional
initiatives. For some economists such as Vera Lutz (1962) this represents
a prima facie case for blaming the regional problem on trades unions.
However, as Myrdal (1957) has argued, and as is represented by Figure
2, the very fact of relatively more capital intensive production in the more
capital abundant MOR will tend to mean a higher rate of return or profIt.
This in tum will tend to mean that not only labour but also savings
are attracted to the MOR from the LOR. Thus there is a basic asymmetry
or inequality in interregional factor flows which is only marginally offset
by any 'trickle down' or spread-effect from the more developed to the
less developed region. Marina Von Neumann Whitman (1967) has
admirably illustrated the reinforcement of this cumulative inequality
through the application of Keynesian growth theory to initial inter-
regional disparities.
188 CAPITAL, LABOUR AND THE REGIONS

Labour Capital
(Savings)

Figure 2. The Myrdal case of factor flows.

8.1.1 The role of multinational activity

The rise of multinational capital on a major scale, especially over the last
fifteen years, has added a new dimension to this inherent asymmetry or
inequality. In practice, relatively labour intensive industry based in more
developed countries has increasingly located plants in countries such as
Singapore, Taiwan, South Korea, Thailand, Indonesia and the Philippines,
plus a Crown Colony such as Hong Kong, and several Latin American
countries, where the cost of labour is a tenth to a twentieth the cost of
European and North American labour. Such countries have the additional
advantage for such companies of representing what amount to both tax
havens (with tax exemption for up to fifteen years) and union havens
(in the sense of no normal trade union freedoms) (cf. inter alia Pacific
Asian Resources Center 1977).
The result has tended to follow the pattern represented in Figure 3,
whereby capital investment in relatively labour intensive manufacturing
industry has increasingly located in a less developed country (LDC) rather
than a less developed region of the more developed country (MDR). At
the same time, in the European case, at least until the post-1973 recession,
labour tended to be drawn increasingly from peripheral countries in the
Mediterranean area, or intermediate development countries (IMC) (with a
return of much emigrant labour from the MDRs to underemployment in
THE FALLACY OF SELF-BALANCE THEORY 189

the LDRs). This was the case for both relatively capitai intensive invest-
ment, and for the lower paid and relatively insecure (seasonal) jobs in con-
struction .and services. This pattern could be seen in immigration to
countries such as France of workers from Spain, Portugal and Algeria,
and in the case of Germany of workers from Greece, Yugoslavia and
Turkey (cf. Castels and Kosack 1973, part II; see also Section 5.3).

\
\
\
J
I Capital
I (Savings)
J
/

""
Capital
(Investment)

Figure 3. The adapted inter- and intra-national case of factor flows.

In such cases, regions such as western and southern France were relatively
neglected both by French domestic capital, and by foreign multinationals
locating in France (D.A.T. 1974, Table 6). In the Italian case, domestic
capital such as Fiat, with minor exceptions, located increasingly in Latin
America rather than the Mezzogiorno, while most foreign multinational
capital locating in what is nominally southern Italy did so in the northern-
most belt close to either Rome or Napels (Benetti et al. 1975, Table 1).
In the West German case, relatively full employment levels and the relative
190 CAPITAL, LABOUR AND THE REGIONS

centrality of the country as a whole within Europe, plus the historical

dispersion of administrative, social and cultural centres in the capital cities
of the Lander, tended to mean an exceptional degree of job dispersion.
In Britain, again with few exceptions, big business has tended to locate in
LDCs in relatively labour intensive production, although the lower rate of
postwar expansion has meant a lower rate of foreign immigration from
abroad than in some areas in continental Europe (H.C.E.C. 1974).
Such qualifications of the basic Myrdal model are matched by others
which similarly reflect changes which have occurred in the structure of
big business and its postwar operations. Essentially, Myrdal assumed that
the mechanism of syphoning savings from a less to a more developed
region would occur through the stock market and the banking system.
But in several European countries, now including not only Italy - where
the stock market was never strong - but also the UK, share capital con-
stitutes an insignificant fraction of total manufacturing investment (less
than a twentieth in the UK in the early 1970s). Bank finance through
medium term lending is similarly insignificant in the UK, though
important in Germany, France and Italy (in the latter cases through both
state banks and special credit institutes). It seems probable that a high
proportion of the continuing capital syphoning to MDRs from LDRs
takes the form of pension and insurance funds placed by national financial
intermediaries with big business, Le. via the broken line in Figure 3. Much
of this will go into property and other unproductive investment in MDRs
rather than into productive manufacturing investment in LDRs.
One reason is the extent to which most finance for manufacturing
investment in a country such as Britain now comes through retained earn-
ings or Eurodollar and Eurobond issues rather than through the stock
market, deposit bank lending or pension and insurance funds. To the
extent that this trend is matched in other developed countries, the less
developed regions become increasingly peripheral to capital accumula-
tion, even as sources of investment funds. In a country such as Italy, the
LDR of the south has not been a net provider of funds for some time.
The savings syphoning or capital depletion effect rightly enough described
for earlier stages or underdevelopment by Myrdal has given way to net
capital imports from the MDR, via public agencies and state intervention
(Holland 1971). Put again in terms of Figure 3, the LDC has increasingly
been marginalised and by-passed on both the capital and labour
migration accounts, substantially through the new financing and location
patterns of multinational big business.
THE FALLACY OF SELF-BALANCE THEORY 191

There are exceptions to these patterns, such as the recent highly publi-
cised decision by Ford (UK) to locate a new engine plant in South Wales.
But such ventures tend to be either in operations which employ relatively
capital widening and deepening investment, or to represent incidents in
the frontier relations between big business and the state. In other words,
where there is a danger that the state under public pressure might inter-
vene against other aspects of a multinational's activity, such as the
increased intra-company imports of components and finished vehicles
from abroad by Ford to the UK, big business will tend to be sensitive to
the public relations advantage of visible gestures by which they appear to
be cooperating significantly with the public interest in regional develop-
ment.
Certainly, the broad evidence from the post-war period in economies
such as those of Britain, France and Italy shows that big business is hardly
influenced in its routine location decisions in more developed countries by
the incentives on either the capital or labour account offered by govern-
ments (cf. H.C.E.H. 1972 and 1974). Nor are they significantly influenced
by the assumed advantages of the spatial clustering of interrelated activity,
as will be shown below. Both factors are critically important granted the
extent to which (1) state intervention and regional policy is based on such
instruments, complemented by infrastructure provision, in countries
such as Britain, France and Italy; (2) big business and the multinational
sector now dominates half or more of manufacturing output and employ-
ment in such countries (cf. EEC commission 1976c).

8.1.2 The inSignificance of the growth-pole philosophy

The Perroux growth-pole concept, at least in Perroux's original exposi-
tions, depended essentially on externalities, as did Myrdal's original
exposition of cumulative spread effects in a more developed region. This
may be clarified diagrammatically. Thus in terms of Figure 4 it would
mean that firm A needing only x units of a particular component could
produce it for itself at no less than a unit cost of 20, whereas firm B
specialising in large scale production for other firms could produce the
component at a unit cost of 10 with a production of y units. If firm A
were marketing the product at x it would need to charge 24 price units
to make a 20 percent profit, whereas firm B can charge half as much for
the same rate of return.
192 CAPITAL, LABOUR AND THE REGIONS

costs per
unit

24 r-------~~~I---------------------­
_________ I LAC 1
20

12r-------------r-~~----~~----~---
_________ .1 ___ _
10 I MDC cost
I
6 ---------1------------ LDC cost
I
I
&.-._ _ _ _ _ _ _ _ _ _ _ _........._ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _..1.-_ _ _ _ _ production
x y volume

Figure 4. The effects of economies of scale.

This is no more than standard external economies theory. In the regional
context, it needs the addition of relatively high transport costs and other
locational economies to support the spatial concentration or growth
centre case. These assumed locational economies have already been
criticised in depth elsewhere (Holland 1976a).
However, in this context, it is important to see that when one moves
from the single firm, single market assumptions of the early expositions of
growth-pole theory to the multi-product, vertically-integrated multi-
national of contemporary reality, the basic externalities case breaks down.
It does so essentially because such companies are large enough to internal-
ise most economies in relatively optimal scale production in different
plants. Put simply, for any given component, they are able to produce at
volume y rather than x, and supply themselves.
It is important to stress that such self-provision, rather than external
dependence, arises essentially from the scale of the overall operation of
the big business concerned. Quite apart from the fact that transport
costs probably have shrunk by as much as nine-tenths since Weber's time
at the tum of the century, the need for proximate location would be
offset by larger scale integrated production to the extent that the unit
cost gain represented by the difference between x and y was greater than
THE FALLACY OF SELF-BALANCE THEORY 193

the transport cost from more dispersed location. Granted relative limits to
production economies of scale, it appears from a wide range of evidence
that such economies from scale have widely outstripped shrinking trans-
port costs anyway.
But, in addition, and especially since the early 1960s, big business has
increasingly gone multinational on the lines previously described. This has
introduced an entirely new dimension to the issue of proximate versus dis-
tant location. For instance, in a case where labour costs equal half the total
production costs in manufacturing (which is not untypical), the use of
labour in a Third World labour haven where its cost is a fifth of that in a
more developed country would mean a reduction of labour cost (in terms
of Figure 4) from 5 to 1. Assuming unchanged capital costs, this would
reduce the overall cost of production at y from 10 to 6. In other words,
it would mean a cost reduction of 40 percent, and a total unit cost of
only 6.
Such a reduction of costs will massively outstrip normal transport costs
arising from a more distant location, even when the distance concerned
involves circumnavigation of the globe. It is illustrated in the case of a
high value-low weight product such as watches by the fact that the Bulova
company produces components in Switzerland, flies them to the Philip-
pines for assembly by low-cost labour, and then flies them back to the
United States for final sale (Barnett and Muller 1975, Chapter 11). It is
illustrated in general by the phenomenon of the 'run-away' labour
intensive industries of footwear, clothing, electrical goods, electronics and
now increasingly mechanical engineering to the South East Asian and
Latin America countries already cited.
If such factors completely outdate the proximity case of growth centre
or growth pole policy for the leading enterprises in developed economies,
they also undermine the basis of conventional incentives policies, at least
for the relatively labour intensive firms which governments want to attract
to less developed regions (H.C.E.C. 1974).

8.1.3 The role o/monopolies

There are other respects in which the rise of big business in the postwar
period has contributed to the aggravation of regional disparities. For one
thing, the trend to monopoly structures in manufacturing has meant a
switch from the price-taking of consumer sovereignty to the price-making
of producer sovereignty. For example, in Britain the share of the top 100
companies in manufacturing output increased from around 20 percent
194 CAPITAL, LABOUR AND THE REGIONS

in 1950 to an estimate of some 50 percent by the mid-l 970s (Prais 1976).

Similar concentrations of economic power in France, Germany and Italy
are catching up fast (George and Ward 1975; Franko 1976; Dyas and
Thamheiser 1976). This gives a situation in which two or three firms
typically control from a half to 100 percent of output in given product
markets. It was Adam Smith, rather than Karl Marx, who argued that
'people of the same trade seldom meet together, even for merriment or
diversion, but the conversation ends in a conspiracy against the public
or in some contrivance to raise prices' (Smith 1930, vol. 1, chapter 10).
When two or three boardrooms are gathered together in a structure which
dominates any given industry, price-making power therefore should not
surprise us. However, quite apart from arguments from authority, the case
on the price-making power of meso-economic firms in Western Europe
is corroborated by the findings of the Competition Directorate of the
EEC (1976a, p. 9) which reports that inflation is greater in the more
concentrated sectors of EEC industry.
Even if we abstract from net price inflation in monopolistic industry,
it can be seen from an application of oligopoly analysis that leading firms
can either undercut lagging firms through reducing prices temporarily in
line with cost reductions from larger scale and multinational location, or
alternatively can reap larger gains through self-financing from maintaining
prices at a level which allows smaller less efficient firms to survive, earning
super-normal profits when smaller firms scrape by with lower profits
nearer a profit norm.
For instance, in terms of Figure 4, the larger firm with LAC 2 would
reduce price to P2 in a price competitive situation. If it does so in
practice, the smaller scale firm with LAC I and PI would simply go out
of business. In practice, this would be an illustration of the 'elimination
pricing' argued by Sylos-Labini (1969), whereby bigger business wipes
out a smaller enterprise in the sector and picks up its market share.
A lesser variant of this price tactic is the 'no entry' pricing analysed by
Bain (1956), whereby bigger business would be likely (in our example)
to keep price around 20 rather than 12 to deter the entry of smaller firms,
which at that price could not earn any profit.
If we regionalise the example, which is not unrepresentative of reality,
and argue in terms of larger firms in more developed regions, and smaller
firms in less developed, we can see a link between the nature of unequal
competition between meso- and micro-league firms within industries, and
the nature of unequal industrial structures in more and less developed
THE FALLACY OF SELF-BALANCE THEORY 195

regions. Essentially, the more a small micro-firm grows in terms of market

share in a national market, through government regional assistance, entre-
preneurial initiative, or a combination of both, the more it lines itself up
for takeover or elimination by a more established and more secure meso-
enterprise. Alternatively, whatever the entrepreneurship of the smaller
firm, it never will make it into the modern markets dominated by the
bigger meso-enterprise. These dynamics of unequal competition between
bigger and smaller firms are almost neglected in regional incentive
policies (cf. Holland 1976a and 197 6b).
In practice, under inflationary conditions, larger scale business does not
necessarily have to lower prices to exert no-entry or elimination effects.
If it simply holds prices stable, absorbing inflation in costs by virtue of
its high profit levels - which it could reduce from super-normal to normal
- it thereby would eliminate the actual or potential profit of the smaller
firm. In terms of Figure 5 the lower cost meso-enterprise with a multi-
national location would eliminate the higher cost micro-company with
an LDR location in an MDC at point a. At this point, with the price level
Pl equal to 24, it would be still be making more than a 140 percent
profit, granted that it was operating at a unit cost of less than 10.
costs and prices
per unit
micro

a
24 ~--------~~--------------~-----
I
20 -------------------i
I
I
meso I
I
______ J_
10

production
~------------------------------L---~-----volume

Figure 5. Micro- and meso-company cost schedules.

However, price stability has not been a notable characteristic of big

league business in recent years. Moreover, anti-trust and anti-monopoly
196 CAPITAL, LABOUR AND THE REGIONS

authorities take a dim view of the elimination of at least medium or

micro-enterprise on a major scale (rather than their golden handshake
takeover by the big league). Partly to avoid the iron maiden of the compe-
tition authorities, and essentially because it anyway offers them higher
profits, big league business in the multinational meso-sector will
frequently raise prices to allow medium-sized firms in the market to
survive under inflationary or higher cost conditions.
This effect, which has been called umbrella pricing by Edith Penrose,
is illustrated in Figure 6. At point a the costs of the smaller scale LDR
firm are beginning to rise, while the earlier entrant multinational in an
LDC is still reducing its unit costs. By raising price from Pi to P2, the
larger meso-firm in the capacity of price leader not only furtlier increases
its already super-normal profits but thereby enables the smaller scale LDR
firm to avoid profit squeeze. At point b, where cost inflation on
commodities or components means a rise in the multinational's costs, it
can further aid the smaller scale LDR firm and itself by raising price to P3 .
Clearly the profit levels for the larger scale firm in the meso-economic
sector would, under these conditions, be astronomic. At point b in Figure
6 it would be making a profit of more than 180 percent, and a profit of
220 percent at point c. Such profits may appear entirely unrealistic unless
it is realised that some of the more notorious cases of monopoly profits
in the multinational sector, such as in pharmaceuticals, have run to several
thousand percent. Certainly, in such latter cases, the multinationals
concerned have been able to inflate the final price level through transfer
pricing, or the artificial inflation of imported components from subsidia-
ries abroad. But such techniques reinforce rather than qualify the upwards
price-making of our Pi to P3 example.
Again, such new mechanisms of price-making power, through both new
monopoly domination and multinational range, offset the pull effect
of government incentives to locate in the less-developed regions (LDRs) of
more developed countries (MDCs). They contribute to illustrating why
MDCs are increasingly marginalised in the location decisions of multi-
national enterprise. On the wage front, governments in MDCs would have
to offer labour subsidies of between 80 percent and 90 percent to com-
pete with the costs of labour in some Third World countries. On the capital
front they would have to offer subsidies of more than the total cost of
investment. In the case of the example given in Figure 7, where the price
level P3 is more than three times total LDC costs, and where capital cost is
assumed to represent only a fifth of total cost, the subsidy effect would
THE FALLACY OF SELF-BALANCE THEORY 197
costs and prices-per unit
32 c
28 LDR
~----~-+--------P2

24 ~----------~~----~------~-%~----Pl

20

LDC

10

production
~----------------------------~--~-------vo1ume

Figure 6.. LDR and LDC cost schedules.

need to represent more than fifteen times the cost of the capital
concerned, rather than the fifth at present offered.
Of course this is only a comparative cost illustration. Other factors
enter into the. calculus of multinational companies in their location deci-
sions, including not only public relations with governments, but also the
need to locate some plant in MDCs if governments are to award them
contracts from public expenditure, plus the not negligible tendency to
greater 'political stability' in MDCs and the lower threat - at least until
recently - from nationalisation of relatively capital intensive plants.
Nonetheless, allowing for even a wide margin of error in individual cases,
the argument illustrates some of the reasons why firms in the meso-
economic multinational sector can gain more from either raising prices or
going multinational - or both - rather than going multi-regional within
one country in response to regional incentives.

8.1.4 The inadequacy o[ regional policy under the current crisis

There are other reasons for the undermining of regional incentives in
recent years. One of them reflects the fact that with a fall-off in capital
accumulation and growth of investment from the 1960s - even before the
post-1973 recession - leading firms were increasingly undertaking replace-
ment investment and rationalisation rather than investment in entirely
198 CAPITAL, LABOUR AND THE REGIONS

new plants of the kind which could be located in less-developed regions

and areas. In other words, whereas in the 1950s investment had tended to
be both capital-deepening and -widening, employing not only more
capital per worker but also more workers overall, from the early 1960s it
clearly became more capital-deepening per se.
In fact, whereas the rate of productivity increased for modem manufac-
turing products in the 1960s relative to the 1959s in industrialised Western
Europe as a whole, the rate of growth of employment in modem manu-
facturing was already falling. Extrapolated through the 1970s, it could be
argued with some force that even if there had not been a recession of the
post-OPEC price-rise variety, there would have been a net contraction
of modem manufacturing employment.
As a result of such rationalisation investment, the pull effect of
labour incentives to regional location would anyway tend to be under-
mined. Firms either would be planning for the employment of less labour
per unit of capital, or less labour overall. Potentially, on the capital
account, rising capital intensity could mean a greater pull effect from
capital incentives. But these would be undermined not only for the
previous reasons, but also because a higher proportion of investment,
taking the form of replacement of existing plants, would be located in
existing premises in existing locations. This is especially important if
account is taken of the need for big business to maintain relatively stable
relations with big trades unions, especially when it anyway may be
locating a high proportion of investment in relatively labour intensive pro-
duction abroad. If it were to take advantage of available capital incentives
on replacement investment by locating it in an entirely ~ew regional
initiative, it could mean the closure of existing plants in MDRs and major
confrontation with trades unions.
In the post-OPEC price-rise situation, with a recession amounting to
slump, the industrial and manufacturing sectors of leading developed
economies are anyway heavily unemployed. In such circumstances,
management is not likely to respond to regional incentives on new
investment because it already has -spare capacity on its hands. It
has been argued that even 100 percent capital grants on investment
would not match the kind of profit which can be made by the com-
bination of an LDC location and monopolistic pricing. ~ut even if a
100 percent grant on capital were to represent a net 100 percent gain
for a manufacturing company, it would be unlikely to invest with such a
grant today if it felt it could secure it either tomorrow or in five years
THE FALLACY OF SELF-BALANCE THEORY 199

time. This is basically because technology does not stand still. Investment
today in a major plant using today's techniques of production could leave
a company five years' behind technologically if competitors only choose
to invest in new capacity half a decade from now with the production
techniques of the early 1980s.
Apart from this, the current crisis of capital accumulation is not simply
post-OPEC. Its specific form clearly is due substantially to the defla-
tionary policies imposed by the oil- and commodity-consuming developed
countries following the rise in commodity and oil prices in the 1970s. But
the underlying symptoms of crisis, at least in Western Europe, as already
indicated, were evident earlier. In West Germany, to take the example of
the key fast accumulator of capital in the 1950s, gross domestic fixed
capital formation fell from a high of 9 percent a year from 1950-54 to
6 percent in 1960-64 and less than 1 percent (in fact 0.2 percent) in
1970-74 (Zinn, forthcoming). In a stop-go economy such as Britain's,
capital accumulation in industry had lagged behind the Continent for
twenty years after 1950, but its level today, in 1977, is actually lower
than in 1970. In Italy the crisis of capital accumulation was first registered
in 1963-64, and then intensified from 1968. France alone managed a
sustained rate of growth of output into the 1970s but with increasing
investment crisis.
In these circumstances, governments increasingly tried to sustain
investment by offering capital incentives, irrespective of the location
concerned. The efforts were relatively futile. This stemmed partly from
the Keynesian pall in macro-economic policy-making, whereby govern-
ments tended to treat investment decline as a short term rather than
longer term structural problem. They were neglecting Marxist analysis
of trends to rising organic composition of capital, disproportion between
productive and unproductive sectors and under-consumption, partly
because they thought that Keynes had made Marx redundant. It is
perhaps more surprising that they did not even take account of Schum-
peter on the bunching of innovations and their spread effects, with the
implicit lesson that consumer booms of the postwar kind tend to be
followed by relative consumer saturation. Most surprisingly, they tended
to neglect Keynes' own crucial argument that investment performance
depends on profit expectations which themselves depend on expectations
of future demand rather than costs.
The result was a continual upgrading of incentives in the form of tax
rebates and write-offs - depreciation allowances, investment grants, Rand
200 CAPIT AL, LABOUR AND THE REGIONS

D write-offs against tax and later tax relief measures to offset rising
commodity costs. Figures on the total effect of such subsidies are difficult
to come by. In Britain in February 1974 the incoming Industry Minister
found that it took nearly a year for his officials to calculate just how
much public money had been rebated to the top twenty-five companies
from various sources in recent years. However, if the one measure of tax
relief on stock appreciation alone (totalling nearly £4 billion since its
introduction in 1975) is added to previous estimates of the difference
between effective and nominal corporation tax, it emerges that British big
business in the sector of £50 million turnover or more per year is probably
paying no net tax overall. 1
In such circumstances, it is hardly surprising that there is some force to
the previous claim that even major incentives will not promote new
regional initiatives. In Britain, in the meso-economic sector, even 100
percent tax write-offs are not exerting an effect in promoting net invest-
ment. However, what does emerge from the British case is the extent to
which the primary concern of the state with capital accumulation and the
measures which it has vainly taken to try to promote it to higher levels,
have been undertaken at the cost of the regional incentives programme
and in substantial contradiction to it.
This is quite apart from the tendency for big business during recession
to compensate for fallen sales by raising prices, thereby maintaining cash
flow. Such compensatory upward pressure on prices reinforces the price-
making mechanism outlined in Figure 6 and further decreases any pull
effect from regional incentives.
In these senses, the current regional crisis, and the crisis in regional
policy, is part of a wider economic and ideological crisis in the mature
capitalist economies.
Essentially, regional policies based on incentives assume relatively
normal profits between enterprise and within industries, via the com-
petitive process. Without competitive pricing, monopoly prices and
1. Before 1974, Mervyn King (1973) calculated that the cumulative effect of capital
grants, investment allowances and employment premia had amounted to the
equivalent of half the total company taxation. Add to this (1) the fact that by 1977
the British government had rebated nearly £4 billion in additional tax concession
through relief on stock appreciation; (2) also taking account of the analysis of
Geoffrey Wood (1973) that big business gains a disproportionate share of tax relief
through greater efficiency in internal adjustment and claims; plus (3) the fact that
the big league can understate taxable inco:me through under-stating real profits via
transfer pricing (Holland 1975, Chapter 2).
POWER STRUCTURE AND REGIONAL POLICY 201

monopoly profits undermine the effect of incentives. The evidence on

such undermining has now been available to the British authorities, for
instance, for at least five years. Evidence on increased regional disparities
in the EEC (1 976b), and the purely cosmetic role of the Community's
regional fund, also have been clearly available in the 1970s. Yet national
and international authorities continue to offer little more than ineffective
incentives, increased unemployment and relatively shrinking income to
those in problem regions and areas. In short, they stay with indirect
intervention, and hesitate to implement measures for the direct mobilisa-
tion of investment, jobs and incomes at either the national or regional
level.

8.2 EXISTING POWER STRUCTURE AND REGIONAL POLICY

The pre-Keynesian perversities of nation states vainly trying to restore the

world economy by cutting imports and hoping to increase exports,
despite the mutual beggar-my-neighbour consequences of such a policy,
would similarly defy rational explanation if we were living in a purely
rational world. Of course, we are not. But the explanations for ignoring
the changing facts of the real world lie less in Gestalt psychology - as
initially suggested - than with the hesitation of those in power to adopt
policies which could challenge the freedom of large-scale capital to deter-
mine what it will produce, why, when and where - on a global scale.
Such freedoms are central to the capitalist mode of production in either
its ·liberal, monopoly or state variants. Capitalism as a power structure is
not simply a matter of class as ownership, but also as control. Control
over the nature and criteria of investment, its timing and its location, still
lies crucially with management - and, in manufacturing, private manage-
ment - rather than with public authorities. Its criteria - essentially profit
- are still private and internal to the enterprise, rather than public, social
or external. Yet regional development policies depend essentially on the
admission of conflict between private and social benefits, i.e externalities.
At an earlier phase in the development of capitalism, in the immediate
postwar period, some governments intervened relatively decisively with
policies designed to reconcile the private and public interests, where
necessary through constraint of private freedoms. The locational controls
of Industrial Development Certificates (IDCs) in the UK were an example
of such intervention. But, retrospectively, it is possible to see that there
202 CAPITAL, LABOUR AND THE REGIONS

were particular conditions for the effectiveness of such intervention.

F or one thing, free market location had been very much discredited by
the concentration of interwar unemployment in particular regions and
areas - essentially those of coal-mining and ship-building. This had been
accompanied by a relative over-development of London and the Southeast
(not least in terms of its vulnerability during the war to bomber attack).
As the Barlow Report, embodying such analysis, was undertaken and
published in the late 1930s and, more importantly, gained the status of
conventional wisdom, by that time the postwar Labour government came
round to implementing locational controls through IDCs via the Town and
Country Planning Act. In other words, the political case for such controls,
when urged at the level of the state, was pushing at an already open door.
The opening had been made already at the ideological level.
However, with the rise of multinational capital on a greater scale in the
postwar period, and its extension through more than half of British manu-
facturing, the premise of locational controls was significantly undermined.
Basically, big business could successfully threaten to locate abroad if not
allowed the location of its choice in a more developed or intermediate
region. Moreover, by and large the switch represented the general under-
mining of national economic sovereignty by multinational capital. From a
brief postwar period where it appeared that, in lcey" respects, the public
interest would be demanded of private capital, the balance of power
moved back from the wartime state to private responsibility for both
capital accumulation and its location. In other words, the state disclaimed
primary responsibility for the shape and distribution of economic activity.
The shift to incentives rather than controls was general throughout
Europe. While the incentives, at least in the 1950s and early 1960s, may
have exerted a marginal effect on national capital, they proved decreas-
ingly effective thereafter. Indicative national plans with regional targets,
in Britain, France or Italy, mainly indicated the limits of indirect state
intervention in favour of the regions. In the current crisis of the system,
the essential class nature of the main European countries emerges in the
extent to which trades unions are blamed for excessive wage costs, as a
cause of the continuing crisis, despite the fact that - internationally -
such wage inflation has been largely restrained from 1975.
The result has been a significant social incidence of the costs and
benefits of economic crisis. Rather than modify the conventional
Keynesian framework to the new combination of recession and inflation,
and using public expenditure, public enterprise and price controls in a
POWER STRUCTURE AND REGIONAL POLICY 203

planned framework for recovery, governments have held back or cut

public expenditure, extended the public sector only into failing or bank-
rupt enterprise, allowed virtually every price-rise requested or imposed by
big business, saturated capital with tax concessions and subsidies, and
restrained wages wherever they could gain the support of enough trade
union leaders. The half lesson learned on the relative ineffectiveness of
the regional employment premium in Britain was followed by its abandon-
ment as a public expenditure saving without substitution of the planned
focus of expenditure and location in the regions which should have been
its counterpart. Concern with regional unemployment was swamped by
the massive increase in total employment and the fiscal crisis of the state.

8.2.1 The fallacy of pluralist models of state in tervention

In the response to this crisis, even governments which at previous elections
had drawn on 'special relationships' with the trades unions chose to adopt
attempted 'solutions' to the crisis which placated capital. Their response
illustrates the wrongness of pluralistic models of the state, and with
them the wrongness of pluralist models of state intervention in regional
development of the kind hypothesised by Isard (1969).
For instance, pluralist models of the state assume a primacy of the
democratic political process. By casting votes in elections, the individual
voter sways the balance of power in a parliament or other representative
assembly. The exercise of power at the level of the state - both govern-
ment itself and the civil service machinery - is supposed both to be
dependent on the political process and to reflect policies and pressures
formulated or expressed at the political level. Further, the state itself is
assumed to represent a neutral agency with regard to the interests of the
two main power groups in society: business and trades unions, or capital
and labour. This is partly because it is supposed to fulfil or express the
changing will of the electorate via the ballot and the balance of parties.
It also is assumed that neither capital nor labour is dominant within the
system, i.e. that neither group can frustrate the will of a democratically
elected government with a mandate from the people. Where big business
or big unions are introduced in the analysis, it is assumed that they tend
to counterbalance or countervail each other, thus representing no essen-
tial threat to the pluralistic mechanism, with the balance of power deter-
mined by the political process and exerted through the fulcrum of a
neutral state. Such an assumed process is represented in Figure 7.
204 CAPITAL, LABOUR AND THE REGIONS

Figure 7. Pluralist model of state intervention.

Neither political life nor the state are quite like that. For one thing, the
pluralist model assumes a degree of rationality in decision-making and the
exercise of power which is notably lacking in too much policy-making
(or the response to events which frequently passes for policy). In other
words, it assumes not only competition for power between parties
representing individual interests (paralleling economic competition
between individual units), but also a responsiveness of the system to
individual ideas or sets of ideas in the form of policies proposed either by
pressure groups or parties, especially when the ideas are those of the
party forming the government. In short, the system is supposed to be
rational.
However, in reality, the political process in the sense of what happens
at the level of the ballot and political institutions does not dominate the
apparatus of the state and the exercise of state power. Nor are capital
and labour relatively balanced pressure groups within society in the
manner hypothesised in the pluralist model and represented in Figure 7.
In practice, the situation is much closer to that represented in Figure 8.
POWER STRUCTURE AND REGIONAL POLICY 205

8 I
I

State

Ideology

Capital

Figure 8. An alternative model of state intervention.

In a capitalist system both politics and the role of labour tend to be

marginalised relative to a close central relationship between capital and
the state, with an intermediate but central role played by ideology in the
sense described earlier, i.e. the framework of implicit assumptions, values,
concepts, ideas and dogmas dominant in a capitalist society.
This more Marxist analysis relates crucially to the concept and reality
of capitalism as a mode of production, distribution and exchange. The
dominance of capital over labour in practice, as opposed to the myth-
ology, is illustrated by Figure 9, which indicates the range of decisions
in which capital rather than labour exercises an effective control over the
allocation of resources in the system. This includes what is produced,
why, when, how, for whom, by whom and - importantly enough in this
context - where. Even the question of what cost, on the wages account,
is largely determined by capital and then contested by labour on rela-
tively unequal terms (especially where the capital concerned can threaten
an alternative multinational location).
At present most governments in the mature capitalist economies leave
206 CAPIT AL, LABOUR AND THE REGIONS

all these decisions to capital, despite the fact that they affect the nature
of production, investment, counter-cyclical policies, export and import
trade, innovation, prices and employment, as well as location. This is not
to say that the state does not try to influence the behaviour of firms, and
thus these macro-economic aggregates. Such attempts at influence are the
substance of economic policy. But the policy framework is almost
exclusively by indirect rather than direct measures, and intended to
influence rather than control - tax changes on consumption, investment
incentives, interest rate changes, and changes in the money supply,
exchange rate changes, Rand D under-writing, price ceiling targets and
employment subsidies.

1. What ... Type of product, service and investment

2. Why ... Private profit and internal corporate criteria

3. When ... Mainly during pressure on capacity and rarely

in recession

4. For whom ... Sectoral, social and spatial market (regional,

national, international)

5. By whom. .. Labour, male or female, employment volume

6. How ... Capital-labour ratios, technology and nature

of the work process

7. What cost. .. Wage share in production cost (regional,

national, multinational)

8. Where ... Location of pl;mt and jobs (local, regIOnal,

national, multinational)

Figure 9. Decision-making: the dominance of capital over labour.

In these circumstances, it is hardly surprising that regional policy also is

mainly indirect and allegedly influential, based on incentives rather than
controls. The character of regional policy in a capitalist system reflects the
general character of state intervention. To change the character of such
regional intervention therefore is not a matter of taking thought either
individually or collectively in international seminars, with rational policies
readily appreciated by civil servants, governments and international bodies
such as the EEC Commission. Policies designed to directly harness the
power of capital in favour of problem regions and areas, including inner
POWER STRUCTURE AND REGIONAL POLICY 207
urban areas, challenge not only capital, but also the prevailing ideology
dominating the broad range of government intervention.
Such new direct policies for mobilising investment, jobs and incomes
in less developed areas are resisted substantially, in other words, not
because they fail to give a rational account of the nature of the regional
and urban problem, but because they threaten a major change in the
relations between capital and the state. In short, they threaten a change
in the system.

8.2.2 The function of prevailing ideology

The resistance to such change in policies is not simply a matter of Gestalt
and personal psychology, but the general framework of social psychology,
including the implicit assumptions and unspoken premises which regional
analysts and policy-makers bring to the regional problem, sharing them
with their peer groups and their social and educational classes.
The resistance is dramatic enough if we take account of the extent to
which the dominant ideology gives a false - in the sense of inaccurate -
representation of the working of the market and resource allocation, both
in the general and regional policy contexts.
For instance, if we translate the central area of analysis in Figure 8
to Figure 10, we can see that the state seeks to implement regional and
urban policy at the levels of central, regional and local government. It
does so in terms of a dominant ideology of competitive, price-taking
liberal capitalism, assuming normal profits and therefore the responsive-
ness of capital to incentives to locate and operate in areas of public need.
Yet in reality, the dominant characteristic of capital itself is monopolistic,
price-making enterprise, multinational in operation, with abnormal
profits, unresponsive for the most part to national government incentives
in favour of problem regions and particular localities. Such dominant
enterprise in the meso-economy, in between macro-aggregates and the
micro-enterprise of smaller national capital, literally contradicts the
above-mentioned prevailing ideology.
To the extent that regional economists continue to operate within the
framework of such a conceptual paradigm, they falsely represent reality
not only to students and themselves, but also to policy-makers and
governments. The misrepresentation may well be accidental, through
failure to grasp the changed structure of the dominant enterprise within
the system, and the extent to which they have undermined the text-
book assumptions of regional theory and policy. Alternatively, it may
208 CAPITAL, LABOUR AND THE REGIONS

State - Central - Regional - Local authorities

Dominant I-- Competitive liberal capitalsm - Incentives - Private

ideology

Dominant I---- Monopolistic - Multinational - Private

capital

Figure 10. Prevailing power structure.

reflect a degree of literal inertia in the revision of teaching courses, or

the lack of exposure of civil servants and policy-makers to new ideas,
at least in some cases.
But in practice, such misrepresentation serves the interest of big
business in disguising the extent to which a handful of unelected and
largely unaccountable giant companies now dominate the effective out-
come of both national and regional policies. Such companies are happy
enough to continue to operate in an ideological framework which assumes
that they need incentives to offset significant locational diseconomies in
one rather than another regional location - partly because they thereby
gain public subsidy for what they would have done anyway, and partly
because it reinforces the image of a competitive process squeezing profits
to an industry norm, which is not unimportant in their relations with
trades unions in the key areas of wage bargaining and resistance to indus-
trial democracy.
On the other hand, it would be a mistake to exaggerate the rationality
of business management, even in endorsing a central irrationality within
the system. Corporate management probably needs myth as much as did
primitive society. The unattractiveness of a new location in a desolate
inner urban site or on the fringe of a declining industrial area is appar-
ently crucial to the wives of many middle managers, as are the lack
of city lights of the kind found in the more sophisticated metropolitan
centres. When the manager's wife now increasingly has a part-time job
offered only in a metropolis, quite apart from changing the children's
schools and leaving friends and relatives, such personal social factors
POWER STRUCTURE AND REGIONAL POLICY 209

become critical to locational policy in the boardroom, even if rarely

mentioned. In such a climate, myths of substantial economic costs from
particular regional locations gain credence.
A more central irrationality is illustrated by the whole present debate
on public expenditure, quite apart from its regional distribution. The
arguments of Milton Friedman swept boardrooms and cabinet rooms
following the post-1973 recession not because managers and ministers
rushed to read his permanent income hypothesis, but because there
clearly was a fiscal crisis of the state which could not be readily explained
by other variants of the conventional wisdom. But in their ritual demands
for public expenditure cuts and a pyre on which to sacrifice social spend-
ing, management and its pressure group representatives overlooked the
fact that although public expenditure in most of the developed countries
had approached or surpassed 50 percent of total spending, public enter-
prise accounts almost uniformly for less than 15 percent of the total
output (C.E.E.P. 1975). In other words, some 85 percent of the supply of
goods and services diminished by public expenditure cuts would affect
private rather than public enterprise and cut in the first instance profit
rather than public tax receipts and spending.
Of course, management and capital in general welcomes Milton Fried-
mann partly because he advocated reduced personal taxation as a
counterpart to reduced public spending. Like himself, they appear to have
seen only one side of the policy bauble proffered and neglected the
extent to which, by supporting cuts in public expenditure, they would in
principle contribute to unemployment themselves. They also clearly, if
one-sidedly, saw that any ensuing unemployment would mainly affect
labour and the working class, thereby contributing to the weakening of
not only wage bargaining power but also the disposition of unions to hear
the siren voices of the advocates of industrial democracy and workers'
control. This presumably explains in part why management and capitalist
pressure groups world-wide neglected to endorse an essential complement
to reduced taxation and public expenditure in Friedmann's prescription,
i.e. his advocacy of free collective bargaining.
The marginalisation, rather than dominance, of political institutions
in resource allocation, as illustrated in Figure 8, should be clear enough
from the relative weakness of regional politics in the current crisis,
despite the fact that the incidence of already high unemployment is fre-
quently doubled in classic problem regions through Western Europe.
210 CAPITAL, LABOUR AND THE REGIONS

8.3 SOCIALISATION AND CHANGE

If either the capitalist system is to recover on a significant scale or indivi-

dual regions attain relatively full employment and income, it is not
likely to be the result of conventional national or regional policies of the
kind previously described. In practice, it is only probable that such ends
will be gained by a transformation of the system itself. In other words,
it is probable only through the kind of change which at present is most
strongly resisted by the most organised power groups in the system -
big business and the state.
Such change cannot benefit the regions or solve the urban crises, unless
it concerns both these main power relations in society and the broad range
of factors outlined in Figure 9, i.e. the what, why, when and how produc-
tive resources are allocated, as well as the where of their location. Unless
more direct pressures are exerted to harness the power of (especially) big
business within the system in the interrelated areas of production, invest-
ment, jobs and pricing, as well as trade (including transfer-pricing from
and to subsidiaries abroad, which both understates profits and permits
capital exports), governments will not exercise a sufficient strategic
control of economic resources to be able to undertake the kind of
planning for regional and urban location necessary to cope with present
problems.
Put more simply, the spatial aspect of economic activity cannot be
divorced from the structure of economic power and its social distribution.
Any such divorce not only mystifies reality but also gives rise to false
hopes for regional development.
Yet ironically enough, if admission is made and account taken of the
new structure of economic power, it similarly opens new dimensions for
transformation of the system in a manner which protects the political
freedoms which rightly are held dear by those wrongly resisting any
radical change.

8.3.1 Selective socialisation of production

A primary reason for the hold of pluralist liberal capitalist models of
resource allocation, both in general and in the regional context, lies in the
implicit premise that only such a liberal economic structure, with
restrained and indirect state intervention, can protect the liberal political
freedoms which distinguish the Western democracies from centrally-
planned and politically illiberal systems. In practice, it is substantially
SOCIALISATION AND CHANGE 211

this conscious or unconscious fear which denies genuine pluralism in

access to power and the allocation of resources in most national and
regional politics in the Western democracies.
But if an appropriate challenge is effectively launched against the
prevailing mystique of liberal capitalist economic orthodoxy, it can be
seen that the highly unliberal concentration of economic power in the
hands of monopolistic multinational companies has opened the means to
a form of democratic planning crucially different from the centralist
planning of the Soviet type.
Essentially, when less than one percent of enterprise commands more
than half of economic activity, the decisive planning of that enterprise
- in the meso-economic sector - can transform the broad range of
macro-economic aggregates, induding the distribution of investment, jobs
and incomes between regions. It also can make possible the allocation of
employment and income at the micro-economic level, of both plants and
firms, in particular in urban areas or peripheral locations.
For instance, the fifty largest firms in Britain represent about a third
of manufacturing output and employment, and the 100 largest about one
half (Aaronovitch 1975). This is against a total of some 60,000 manufac-
turing firms in the British economy. If the government managed to
harness the location of new plants in the few dozen leading firms to
regions and areas where unemployment is high, it could contribute
directly to the reduction of regional and urban inequality. In so doing,
since it was concerning itself primarily with less than one percent of enter-
prise in the system, it could avoid direct involvement with the decision-
making and resource allocation in more than 99 percent of British manu-
facturing enterprise.
A similar mini-max effect could be attained in the other leading indus-
trial economies, where not more than two percent of manufacturing
enterprise account for half or more of output and employment. In other
words, a maximal effect could be registered on macro-economic activity
through intervention in a minimal number of meso-economic enterprises.
The fact that such meso-economic enterprise tends to be vertically
integrated and to diperse its plants world-wide as well as between regions
in one country, without dependence on local externalities or dependence
on nearby suppliers or buyers, means that they could efficiently operate
individual plants either in hitherto neglected and un favoured inner urban
areas, or in dispersed peripheral locations. In this sense, if a growth
centre were designated in regional planning, with a complex of inter-
212 CAPITAL, LABOUR AND THE REGIONS

related industry, it would arise from the case for a balanced and related
job structure in an area of high unemployment rather than from a need
for the proximate location of interrelated firms to gain from assumed
locational economies.
The political pluralism of such planning could actually be ensured in
part through the regionalisation of both planning itself and the new public
enterprise in manufacturing which would be necessary to ensure that the
public authorities disposed of direct means to fulfill national or regional
planning targets. The case for such an extension of public enterprise has
been argued elsewhere (Holland 1972 and 1975). Its rationale as a direct
instrument for intervention, as opposed to indirect measures such as
incentives and designated growth poles dependent on the free working of
the market, will be evident from the previous argument.
If such an extension of the public sector to ensure means for direct
intervention amounted to a re-mix in the present distribution of public
and private ownership in the developed economies, account should be
taken of the fact that at present public enterprise not only contributes on
average less than 15 percent of GDP, but also is mainly concentrated in
regionally immobile basic industry and services. In other words, the activi-
ties which have been nationalised to date, at least in Western Europe,
concern mainly mining and oil (which are geologically tied) or power
generation and distribution, transport and communications (which are
geographically tied). Despite nationalisation in some countries of all
or part of the steel and motor industries, the bulk of manufacturing
remains in private hands.
For such reasons, both in quantitative and qualitative terms, a re-
mix of the economy involving a significant extension of public enterprise
into the regionally mobile sector of manufacturing would not mean an
end to the mixed economy of the conventional wisdom, but a closer
approach to a more evenly mixed system rather than the unequal mix
represented by the contrast between public expenditure and the contribu-
tion to GDP of public enterprise.
In some cases, if an effective integration of the national and regional
economy is to be assured through planning, particular public enterprise
agencies would need to be primarily national in character. This is the
general case for state holding companies, even though these could and
probably should have related regionally-based subsidiaries. In other
cases, there is a strong argument to be made for regional public enter-
prise, and also municipal or borough enterprise, responsible to regional
SOCIALISATION AND CHANGE 213

and local authorities.

In both analytical and policy terms, national public enterprise would
mainly be meso-economic, and regional and municipal enterprise micro-
economic in character. Cooperative enterprise on the model originally
pioneered in Britain but most dynamically developed in continental
Western Europe could span both the meso- and micro-economic sectors,
ranging from large-scale national production, wholesaling and retailing,
to regional and local units on a small scale.
To the extent that such enterprises could follow the precedent of the
private sector in retaining a sizeable proportion of earnings for self-
fmance, they would be assured a considerable basic degree of autonomy
in operation. In fact this is particularly probable with the small-scale local
operation, especially to the degree that it developed a bargaining position
in terms oflocal employment with a local authority.

8.3.2 Base initiatives and democratic planning

The effective pluralism of any such re-mix in the economy clearly would
not be assured by legislation alone. Indeed, legislation itself is improbable
unless sufficient political pressure is mobilised to lever change towards
such a new representation of public, cooperative, regional and municipal
enterprise. If such pressure is to be achieved, it is only likely to be through
a further approach to pluralism in reality rather than fiction, via an
increasingly real involvement of regional pressure groups, trade unions and
parliaments themselves in the decisions taken by the state at national,
regional or local levels. Such change at the level of the state, ideology, and
capital is represented diagrammatically in Figure 11.

State Government - Pressure groups - Unions - Parliament

Social Re-mixed economy - Structural, social, spatial planning

ideology

Socialised New public, cooperative, regional, municipal enterprise

capital

Figure 11. An alternative power structure.

214 CAPITAL, LABOUR AND THE REGIONS

Inasmuch as change of this kind is radical in comparison with the

present distribution of power within capitalist society and the present
bankreupcy of wl!at passes as regional policy, it might well be considered
utopian. But this would only be to the extent that its critics are prepared
to ignore both the patent crisis of the system to date, where crises in both
capital accumulation and regional distribution pre-dated OPEC's price
rises by several years, and to ignore the emergence of a new radicalism in
the main parties of the European Left, with commitments to extension of
public enterprise or new democratic planning of big business, or both, in
the main parties of the Left in such countries as Britain, France and Italy.
If such change is no more probable, as such, than a resolution of the current
crisis within the terms of reference of the prevailing system, it nonetheless
is on the political agenda in a manner hitherto unparalleled since the war.
It has increasing chance of acceptance by both electorates and govern-
ments to the extent that it emphasizes that such a new extension of public
power and planning involves a process of social negotiation between the
main interested parties in society rather than an act imposed from above
by a bureaucracy.
For there is neither a need n<;>r a strongly supported case for such
planning as the technocratic imposition from above of decisions deter-
mined behind closed doors by ministers and civil servants. In practice,
such a process on the lines of existing policy formulation would be
unlikely either to devolve decision-making or transform its focus on the
interests of particular social classes and regions. The transformation
concerned would need to involve both the apparatus of the state and the
role of labour and the political process. Without support and pressure
from organised labour and political p.arties, it is improbable. With it, it is
possible.
The mechanisms for such a democratised planning of resourses, focused
on big business, have already been specified in the public enterprise and
planning proposals of the British Labour Party (1973, pp. 17-18). In this,
the Planning Agreements proposals, so far strenuously resisted for the
most part by the Labour government, spell out the mechanisms whereby
the structural, social and spatial dimensions of activity could be integrated
in a national and regional planning process. The proposals include the
obligatory revelation to unions and government by companies in both the
public and private sectors of their main forecasts for production, invest-
ment, employment, pricing, trade and location. This is intended to be
complemented by annual negotiations (in theory bi-annual) between
SOCIA LISA nON AND CHANGE 215

government, management and unions on those changes in the planning

of big business which would reconcile the corporate and the public
interest.
There is no reason why the Planning Agreements procedure should not
be both regionalised and localised (Holland 1972, pp. 298-300). In this
way, the medium term expenditure of governments, with a specified
medium term location of investment and jobs by big business, could be
made available to regional and local authorities. On the basis of this
information, such authorities could pressure and bargain for the kind of
change in government and corporate planning which would help them to
fulfil regional and local interests.
Having argued as much in principle, there are substantial reasons in
practice why governments and big business, in either private or existing
and new public enterprise, should resist such revelation to regional and
local authorities. For one thing, under present circumstances, they appear
hardly in a position to assure their own forward expenditure plans.
Further, granted in particular the dependence of regional development on
a distribution of manufacturing employment (plus head office services),
there is the previously noted trend to contraction of manufacturing with
the mature capitalist economies. To the extent that this reaches a point
at which recovered output will be associated with declining manufacturing
employment, there is indeed a crisis within the prevailing system.
Within the framework of the prevailing system, there is no obvious
escape from these dimensions of both national and regional crisis.
However, in the framework of a system transformed on the indicated
lines, there is considerable scope for resolution - provided that the links
between the structural, social and spatial distribution of resources are
recognised and pressured via the political process.
Basically, the over-investment, under-consumption and shrinking
manufacturing employment in the present system reflect the nature of the
mode of capitalist development characteristic of the postwar period. This
has been based on the application of new technologies to consumer goods
production, under a relatively unchanged distribution of income between
social groups and classes.
There is in fact some scope for a recovery of output and employment
in manufacturing in the medium term through a substantial redistribution
of income in favour of the lower paid and unemployed, which would have
a sizeable regional effect granted the greater incidence of these groups in
less developed regions and problem areas. Put in other words, while
216 CAPITAL, LABOUR AND THE REGIONS

the professional classes may not need a third automobile and television
or a second refrigerator and deep freeze, many working-class people would
be glad to gain their first of at least some such items. (Not that this would
benefit national employment while the goods at present are produced in
Third World labour and tax havens.)

8.3.3 Socialisation of consumption

The real scope for a recovery of output and employment in the mature
industrialised countries comes not only from a redistribution of income,
nor just from the kind of socialisation of production already outlined, but
from a further socialisation of consumption. In essence, technical progress
and increases in productivity, under capitalism, involve redundancies and
job loss whenever there is no absorption of displaced labour by entirely
new industries on the Schumpeter lines which characterised postwar
expansion. This job loss tends to be concentrated in problem regions and
areas to the extent that they represent a concentration of declining indus-
try in the face of the technical progress and innovation concerned. To the
extent that job loss means a trend to under-consumption through reduced
incomes, this further aggravates the problem under capitalism.
With the socialisation of demand, the vicious circle of under-consump-
tion can be broken. At present, under a capitalist rationale and within a
capitalist ideology, income increases are considered legitimate only if
related directly to increases in productivity achieved by those directly
involved in the work process. But in a socialised economy, where both
income and employment were considered a social right rather than a
private boon, it would be feasible to match the productivity gains through
technical progress and innovation with a wider and more equitable
distribution of income, including interregional income.
In this sense, planning which involves productivity increases through a
reduction of employment - under a capitalist mode of production -
could represent a reduction of working hours and a redistribution of
income under a socialised mode of development. Whereas such reduction
of employment, under capitalism, represents an acute incidence of un-
employment and income loss in particular regions and areas, it could
represent a social gain through shorter working hours with a regionalised
redistribution of employment in a socialised system.
The contrast between oligarchic capitalism and democratic socialism
in this sense is very clear. Under capitalism, technical progress and techno-
logical unemployment represent a social cost for those unemployed, to be
SOCIALISATION AND CHANGE 217

resisted by every force at the diposal of the trades union movement.

Under a socialised or socialist mode of development, with a more equit-
able distribution - social and spatial - of employment and income,
technical progress and innovation could represent a gain to be welcomed
and endorsed by society.
If this represents an indictment of our present method of organising
affairs, the fault may lie substantially with ourselves, at least to the extent
to which we help legitimise rather than challenge the inherent inequality
- structural, social and spatial - of the existing system.
REFERENCES
Aaronovitch, S. and M. Sawyer, Big business. MacMillan, London, 1975.
Barnett, R. and R. Muller, Global reach: the power of the multinational corporations. Simon and
Schuster, London, 1975.
Benetti, M., M. Ferrara and C. Medori, II capitale straniero nel Mezzogiorno. Coines Edizioni,
1975.
Brain, J.S., Barriers to new competition. Harvard University Press, Cambridge, 1956.
Castels, S. and G. Kosack, Immigrant workers ant:. class structure in Western Europe. Oxford Uni-
versity Press, London/New York, 1973.
Centre Europeen de 1'Entreprise Publique, L'impact economique actual des entreprise publiques
dons la Communauti Europeenne. Brussels, 1975.
Delegation It l' Amlmagement du Territoire, Invertissements etrangers et amenagement du territoire.
Paris, 1974.
Dyas, G. and H. Thamheiser, The emerging European enterprise. MacMillan, London, 1976.
EEC Commission, Fifth report on competition policy. Brussels, 19i6a.
EEC Commission, First report on the regional fund. Brussels, 1976b.
EEC Commission, Problems of inflation. (Maldague Report), Brussels, 1976c.
Franko, L.G., The European multinationals. Harper and Row, London, 1976.
George, K. and T.S. Ward, The structure of industry in the EEC. University of Cambridge, Dept. of
Applied Economics, Occasional Paper 43,1975.
Holland, S., 'Regional underdevelopment in a developed economy: The Italian case?', Regional
Studies, vol. 5, 71-90,1971.
Holland, S. (ed.), The state as an entrepreneur. Weidenfeld and Nicolson, London, 1972.
Holland, S., The socialist challenge. Quartet Books, London, 1975.
Holland, S., Capital versus the regions. MacMillan, London, 1976a.
Holland, S., The regional problem. MacMillan, London, 1976b.
House of Commons Expenditure Committee, 6th Report, Public money in the private sector.
Minutes of Evidence, 1972.
House of Commons Expenditure Committee, 2nd Report, Regional development incentives.
Minutes of Evidence, 1974.
Isard, W., General theory: social, political, economic and regional. MIT Press, Cambridge, Mass.,
1969.
King, M., The Guardian, November 14, 1973.
Labour Party, Labour's programme 1973. London.
Lutz, Y., Italy: a study in economic development. Royal Institute of International Affairs, Oxford
University Press, London/New York, 1962.
Myrdal, G., Economic theory and underdeveloped regions. Duckworth, London, 1957.
Neumann Whitman, M. von, International and interregional payments adjustment: a systematic
view. Princeton Studies in International Finance, no. 19,1967.
Pacific Asian Resources Center, 'Free trade zones and industrialisation of Asia'. Tokyo, 1977
(Special Issue AMPO: Japan-Asia Quarterly Review).
Prais, S.J., The evolution of giant firms in Britain. NIERS, London, 1976.
218 REFERENCES

Smith, A., The wealth of nations. Methuen, London, 1930.

Sylos-Labini, P., Oligopoly and technical progress. 2nd ed. Harvard University Press, Cambridge,
Mass., 1969.
Wood, G., The Financial Times, May 8, 1973.
Zinn, K.G., in S. Holland (ed.), Beyond capitalist planning. Basil Blackwell, Oxford, forthcoming.
9. Integrated regional planning:
As applied to the Northern Netherlands
PAULDREWE*

'Et un homme politique est moins recevable que

quiconque: il a pris parti dans des luttes publiques.
11 ne peut dissimuler ses engagements sous la toge de
1'universitaire, ou la tunique du Persan.'
(A. Peyrefitte, Le mal [ran<;ais)

This chapter is about the Integrated Structure Plan for the Northern
Netherlands (hereafter called ISP) as an example of integrated regional
planning.
In order to prepare the ground for a critical examination, we need:
a. some background information on ISP (provided in Section 2),
b. a framework for judging its creation and its products (outlined in
Section 3).
As a framework we have chosen the strategic choice approach, a particular
design of a planning cycle and pursuit of its stages (based on both
theoretical reasoning and practical experience).
In Section 4, we then confront ISP with strategic choice stepwise: from
problem formulation to monitoring and review, including information on,
recent ISP developments. At the end of this section an attempt is made to
summarize 'why integrated planning has failed (so far)'.
Finally, by way of evaluation, we try to link the analysis of ISP and its
results with the general theme of spatial inequalities, regional development
and regional science (Section 5).
Note that this chapter is not written by a participant observer of ISP,
but rather by an interested, though not 'innocent', bystander who had to
rely for data mainly on published sources, e.g. progress reports presented by
the Minister of Economic Affairs to the Dutch parliament and published
reports through which an interested public has been informed about
research results and policy conclusions.

* University of Technology, Delft, the Netherlands.

220 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

9.1 BACKGROUND INFORMA nON ON ISP

The idea of preparing ISP was launced in a report dealing with the devel-
opment of the northern part of the Netherlands back in 1972. While the
northern part usually covers the provinces of Groningen, Friesland and
Drenthe, it was extended in this report to include similar problem areas in
northeastern and northwestern Overijssel, the latter being classified as an
eastern province. The total plan area corresponds to the shaded map area
of Figure 1.
LEGEND: PROVINCES
GR = Groningen
FR = Friesland
DR = Drenthe
o = Overijssel
G cc Gelderland
U = Utrecht
NH = Noord·Holland
ZH = Zuid-Holland
Z = Zeeland
NB = Noord-Brabant
L = Limburg

WEST

..
NB ..
:_. . ;<>. . /t·<
: .. ) ............... ....
...........

/ (:~.'
~.... : ..••.:
SOUTH WEST

.-'
MAASTRICHTC¥ /
f...············:-

Figure 1. Regional demarcation of the Netherlands.

BACKGROUND INFORMATION ON ISP 221

The North is a prototype rural, industrially less-developed region as it is

known in all highly industrialized societies of the Western European type.
As in all societies of this type, socioeconomic inequalities at the geo-
graphic scale of regions have existed for a long period of time, manifesting
themselves as 'spatial inequalities' between a less-developed periphery
such as the northern region and a more developed center (the western
part of the Netherlands). The general picture of this regional imbalance
has been extensively described by Holland (1976).
Being concerned about 'regional equity', the Dutch government made
efforts to intervene in the regional development process, especially since
World War II. Concern has been aroused usually by a northern un-
employment rate well above the national average. In order to reduce
unemployment in the North, a policy of 'bringing jobs to the workers' has
been adopted, focusing on the improvement of (industrial) infrastructure,
subsidies to stimulate private investment and the like (cf. Cremers 1976).
Furthermore, a policy of redistributing population from the western to
the northern part of the Netherlands has been followed. This policy serves
two purposes simultaneously: to relieve the center from population
pressure and to stimulate welfare at the periphery. Back in 1972 the view
was expressed that, given the complexity of the northern problem in its
economic as well as its non-economic aspects, integrated planning was
called for (Report on the Northern part of the country, 1972).
Plan-making began in May 1973. It has culminated, for the time being,
in a report published in 1976 in which the main directions of a socio-
economic policy are indicated (Report on socioeconomic policy for the
Northern part, 1976). Throughout our analysis this report will be treated
as the major visible product of ISP so far. In Section 9.3.2, attention will
be paid to some of its more recent developments.
Progress in plan-making between May 1973 and January 1977 has been
described in a series of reports, the time order of which is shown in
Figure 2. Each of these progress reports informs us about a number of
activities reflecting a very specific division of labour.
But before we eleborate on this, the organisation of ISP has to be
sketched. It consists, first of all, of a steering group (plus a secretariat)
in which both 'The Hague' and 'the North' participate. 'The Hague' stands
for higher civil servants from a great number of departments at the
national level, while 'the North' is primarily represented by governors and
politicians at the provincial level. Some steering group members have been
commissioned to coordinate the various activities of plan preparation. The
 1/1 1/1 1/1 1/1 1/1
'72 '73 '74 '75 '76 '77

-- I I I ------,-------.~I -- ~ -
1/5 3115 31/8
t 30/4

~
------~
Report on the Progress reporr---
----------~ Progress
------------------ Progress -- Progress Progress Report on
northern part number 1 report report report report socio-economic
of the country number 2 number 3 number 4 number 5 policy for the
nort hern part
of the country

Figure 2. ISP: phases in plan-making.

BACKGROUND INFORMATION ON ISP 223

steering group is not only responsible for producing ISP, it is also in

charge of initiating the necessary technical work and of establishing work-
ing parties to carry out specific research tasks. The activities, summarized
in Figure 3, reveal the complexity of the enterprise called ISP. This com-
plexity is closely linked to particular features of the Dutch political (plan-
ning) system which is structured as a three-tier system: national, regional
(provincial) and local. Only the national and regional level, so far, have
been directly involved with ISP. Local involvement, however, has only
been postponed. Aside from that, a very particular division of labour has
been adopted in the Dutch planning system. There is, first of all, 'sector
planning' with sectors corresponding to departments such as Economic
Affairs, Social Affairs, Housing and Physical Planning, Transport, and so
forth. Sectors are most clearly articulated at the national level, but
regional and local counterparts also exist. Next, the economic aspects of
all sectors (at all levels) are coordinated through economic 'facet
planning'. Spatial (sociocultural) aspects are brought together in spatial
(sociocultural) facet plans. This is reflected in working parties 2, 3 and 4
(Figure 3). 'Integrated planning', finally, means integration of all three
types of facet plans at the national, regional and local level (see working
party 1). Obviously, such a division of labour poses intricate problems of
horizontal coordination between sectors and facets. Additional problems
arise from the necessity of coordinating plans at three spatial levels vertic-
ally and between four provinces horizontally.

9.2 A STRATEGIC CHOICE FRAMEWORK

A course in strategic choice, even a short one, goes beyond the scope of
this contribution. All we need, at any rate, is some kind of a checklist cap-
able of being used as a framework for judging ISP. More detailed informa-
tion on the strategic choice approach and the closely related technique of
AIDA (Analysis of Interconnected Decisions Areas) is provided by Friend
and Jessop (1969) who first developed this approach and, e.g., by Hickling
(1974), Friend et al. (1974), Bather et al. (1975,1976).

9.2.1 A checklist
Our checklist will be described in terms of phases in the planning process
and prominent features of each phase. We shall also indicate our motives
for choosing this particular framework. In describing the checklist we are
 Working parties Additional research Products
Integrative Facet-planning Others Pu blished reports and
others

Progress report Integrated planning 2 Socioeconomic 5 Systematic 5 ISP 3 (1974)

number 1 development inventory: studies 6 ISP 2 (1974)
3 Spatial development and plans concern- 7 Report on popula-
4 Sociocultural ing the North tion aspects (1974)
development 6 Policy goal analysis (a popular version
7 Population aspects of the working
8 Public participation group's report)

Progress report 1 continues 2 continues 7 continues 10 Inventory 8 ISP 4 (1974)

number 2 9 Secretariat's report 3 continues 8 continues environmental
on goal approach 4 continues quality
and organization 11 Administrative
ofISP aspects

Progress report continues 2 continues 7 continues 2 ISP 5 (1975) (preliminary report)

number 3 3 continues 8 continues 3 ISP 6 (1975) (preliminary report)
4 continues 4 ISP 7 (1975) (preliminary report)
Intentieverklaring (formal
commitment of parliament and
legislative bodies at the
provincial level to ISP)

Progress report 1 continues 2 continues 7 continues 10 con tinues 7 ISP 8 (1975) (final report)
number 4 3 continues 8 continues 11 continues Inauguration departmental
4 continues 12 popUlation commission for ISP
development
Northern sub-regions
13 Regionalization Formation ofa
national multi-year socioeconomic
financial budget policy group
Progress report 1 continues 2 continues 10 continues 2 ISP 9a (1976)
number 5 4 continues 11 continues (part 1: economic
sectors)
ISP 9b (1976)
(part 2: provinces)
4 ISP 10 (1976)
Formation of a socio-
cultural policy group
Report on socio-
economic policy for
Northern part (1976):
'policy report'

Progress report 2 continues 10 continues ISP lla (1977)

number 6 (sectoral and 11 continues (part 1: national
sub-regional perspective)
differentiation ISP 11 b (1977)
of 'policy (part 2: departmental
report') sectors)
Implemen tation
'policy report'
Implementation
ISP 10 (sociocultural
policy group)

Figure 3. ISP: activities, products and phases in plan-making.

226 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

mainly dr~wing on the West Berkshire structure plan experience as

reported by Bather et al. (1976). Phases or steps which are to be distin-
guished are:
1. problem formulation,
2. generation of alternatives,
3. evaluation of alternatives and choice of action,
4. policy design,
5. monitoring and review.
Problem formation starts with issues and/or problems. After having been
observed and interpreted, these are grouped in a number of policy (deci-
sion) areas. Grouped policy areas form the so-called strategic focus, which
is usually a very complex set of problems. But not all of these problems
have to be tackled at the same time. It may be decided to select some
areas as an 'initial focus' while the remaining problems are set aside to be
taken into account at a later stage (this is also called 'progressive reform-
ulation'). The selection of an initial focus may be based upon considera-
tions of:
a. potential impact (some problems affect a greater number of people to
a greater extent than others),
b. political importance or public controversy,
c. controllability (available instruments may decide upon the priority of
policy areas),
d. interrelatedness (political choice in some areas may determine the
choice in other decision areas).
A first step in the generation of alternatives is to develop policy options
per decision area (of the initial focus). Options can be bundled and
worked out as 'scenarios'. The development of options per problem area
usually yields a large and unmanageable number of possible combinations
between areas. Phase (2) therefore includes a reduction of combinations
through what is called 'option bars'. Option bars or incompatibility of
options may be caused by:
a. technical factors (e.g. physical constraints),
b. planning experience (certain combinations of policy options may have
proved to be incompatible in practice),
c. other levels of policy (acting as constraints),
d. political judgments,
e. constraints imposed by other agencies,
f. option definitions (a quantification of certain options may reveal
inconsistencies where nominal definitions show none).
A STRATEGIC CHOICE FRAMEWORK 227

Ideally, evaluation of alternatives and choice ofaction involve the follow-

ing sequence of activities. Criteria for judging the merits of alternative
plans are selected and structured. These criteria are related to objectives
pursued. Alternatives, having been judged according to these criteria, are
arranged according to rank. Then criteria are weighted and, based on the
overall score, a 'preferred strategy' is determined.
The preferred strategy then has to be translated into policy. This is
called policy design. Designing a policy means:
a. to provide a policy context,
b. to determine the elements and functions of the policy statement,
c. to provide a reasoned justification for the plan,
d. to examine the monitoring and review aspects of the policy, and
e. to determine how uncertainties will be managed.
A matrix of policy sets versus issues and policy sets versus objectives
may serve purpose (a). Such a matrix combines the outcome of the
problem formulation phase with that of phases (2) (policy sets) and (3)
(objectives used in evaluating alternatives). Under (b), types and content of
policy statements are to be distinguished. Content refers to quantity
(how much?), quality (of what type?), timing (when?), location (where?),
and sectoral impact (who will benefit and who will lose?). The wording of
policy statements deserves special attention too. A reasoned justification
for the plan (c) asks for an assessment of policy effectiveness, consistency
with other policies, and an explanation of policy statements linked to
technical work performed at phase (1) through (3). Monitoring and review
(phase (5) ) have to be prepared beforehand in terms of indications of
how to assess policy effectiveness, circumstances precipitating and
agencies initiating review.
The management of uncertainties is the cornerstone of the strategic
choice approach. Three types of uncertainties are distinguished which
already" had to be coped with in earlier phases, i.e. uncertainties related
to:
- the environment the plan is meant to control (UE),
- policy values to guide this control (UV),
- interrelated decision or policy areas, other attempts at controlling the
same environment (UR).

At phase (4) we are mainly concerned with uncertainties related to plan

implementation the management of which is expressed in immediate
actions, delayed actions and contingency plans (e).
228 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

The emphasis in policy design is on visible products or published

documents. This, however, does not imply a neglect of important invisible
products. Inter-agency involvement in the preparation of a plan may lead
to shared knowledge and perspectives on problems, mutual understanding,
links in communication networks, etc.
In phase (5), proposals worked out in the preceding phase (under (d».
monitoring and review, will be executed.

9.2.2. Why a strategic choice framework?

Why have we chosen a strategic choice framework for judging the enter-
prise of ISP? Generally speaking, we are convinced that there are certain
basic features which make the strategic choice approach suitable to the
task of preparing an integrated structure plan.
First of all, there is the management of uncertainties with remedies,
i.e. ways of reducing each of the three types of uncertainty: specific
technical research with regard to UE, consultations with politicians and
spokesmen for citizen participation (UV), inclusion of or coordination
with other relevant policy (decision) areas (UR). During the preparation
of ISP special importance has been attached to research in the largest
sense. It has been one of the major tasks of the steering group to initiate
research and to establish working parties to carry out various research
projects. Consultations with politicians have taken place, while citizen
participation has received special attention right from the start (it has
culminated in a large-scale participation procedure centered on the socio-
economic policy report from 1976).
As to uncertainties related to interconnected decision areas, it is
through the steering group that no less than ten departments at the
national level and four provinces are involved. At the national level, the
following ministries participate: Economic Affairs; Housing and Physical
Planning; Cultural Affairs, Recreation and Social Welfare; Home Affairs;
Transport, Waterways and Public Works; Finance; Education and Sciences;
Social Affairs; Public Health and Environmental Hygiene; Agriculture and
Fisheries. Thus ISP is burdened with inter-organizational problems.
Approaches to handling the 'inter-corporate dimension' in public planning
such as in Friend et al. (1974) are of paramount importance here.
It is therefore hightly relevant to evaluate the way uncertainties have
actually been managed in ISP. Note that uncertainties which cannot be
reduced can nevertheless be 'accepted', in trying to ensure that the most
'robust' course is adopted.
A STRATEGIC CHOICE FRAMEWORK 229

Given its rather complex task, a critical evaluation of ISP should also
cover the way complexity has been handled. The strategic choice
approach emphasizes the interrelatedness o/issues. To manage complexity
it proceeds through progressive reformulation. This implies a cyclic
instead of a linear-deductive planning process (Figure 4).

1 .......I - - - - - - - - 1..~ 5 1 problem formulation

2 generation of alternatives

I 1
3 evaluation of alternatives and
choice of action

4 pol icy design

5 monitoring and review

Figure 4. A cyclic process of planning.

If we move deeper into specific technical research, another feature of

stategic choice may prove to be useful. The strategic choice approach
stands for selective use 0/ analytical techniques ('adaptive analysis').
Techniques to be applied must fit the questions arising at each stage of
the process. Research is no longer limited to a 'survey phase', as in
traditional plan-making, to be followed by 'analysis' and 'plan' in a
linear-deductive way. Thus a strategic choice framework provides us not
only with a guideline for judging the empirical basis of ISP, but also with a
guideline for assessing the practical potential of today's regional science.
If one expects ISP to be an effective tool in lessening spatial inequali-
ties, then it can be neither a classical long-term master plan, nor a fascinat-
ing intellectual exercise only. Then it should allow for commitment: it
should initiate immediate as well as delayed actions and 'procedures to
be followed in the event that certain contingencies arise' (contingency
plans). We have seen that the strategic choice approach, indeed, is focus-
ing on such a 'commitment package' to be combined with monitoring and
review.
230 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

9.3 ISP: ITS CREATION AND ITS PRODUCTS

9.3.1 [SP stepwise

Problem formulation. In order to fmd out what has happened in the field
of problem formulation, we have to review the whole process summarized
in Figure 3.
Phase (1), a one-year take-off phase, provides plan-makers with a
Northern and a central government view of development problems: one
resulting from a: systematic inventory of studies and plans concerning the
North (ISP 3, 1974), the other from a policy goal analysis at the national
level (ISP 2, 1974).
ISP 3 (1973) informs us about goals, instruments, bottlenecks, plan-
ning and research with regard to four provinces and ten areas of concern.
This information is not just compiled, but also 'evaluated' by a number of
knowledgeable organizations in the North. Although bottlenecks refer
most directly to problems, the latter may also be inferred from informa-
tion on goals, provided that the impediments to goals are known too
(following Chadwick's definition of 'problem = goal + impediment to
that goal').
ISP 2 (1974) reflects the view of 'The Hague'. Special attention is paid
to the regional aspects of policy goals (and instruments) developed by
various departments at the national level. Not all of the departments
are equally involved at the regional level. And there are, of course, con-
siderable difficulties in horizontal and vertical coordination. The repre-
sentants of national government agencies express a firm belief in the
usefulness of the so-called COBA-apporach as a tool in integrating
national and regional planning. This approach has been developed by a
national commission charged with the improvement of policy analysis.
According to COBA, each department should structure its policy in terms
of a programme in which objectives at different levels of concreteness and
corresponding instruments and activities are ordered hierarchically (as
sketched in Figure 5). However, this is not (yet?) common practice,
hence it is not possible to derive an integrated structure plan for the
North straightforwardly from regionalized programme -structures. 'The
Hague', nevertheless, recommends preparing ISP along the lines of the
COBA-approach (as outlined in Beleidsanalyse 1973). The planning
approach adopted includes the following steps:
1. a sketch of regionalized and national development per sector or depart-
ment at the national level, to be prepared by the working party in
ISP: ITS CREATION AND PRODUCTS 231

\ f

I b
I I b
I
/' ~ /t ~
c c c
II c
II c

t t t t t
d d d
II d
II d

t t ~ t ~
e e e
II e
II e

a Prime objective

b First (second,etc.) level objectives

c Operational level objectives

d Instruments

e Activities

Figure 5. The COBA-approach: a schema tical representation.

charge of 'integrated planning',

2. a sketch of the development per facet from a Northern point of view
which explains the formation of working parties on socioeconomic,
spatial and sociocultural development,
3. models of regional population (re)distribution to be developed by the
working party dealing with population aspects (Report on population
aspects, 1974),
4. an integration of information gathered in the preceding steps by the
working party on integrated planning in order to identify long-range deve-
lopment problems of the Northern Netherlands (as well as information
gaps asking for additional research), thus providing a basis for:
5. actual plan-making following the COB A-approach sketched in Figure 5
and starting with 'primary objectives' (for the sake of flexibility, the
working out of alternative courses of action is suggested).
232 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

We notice that problem formulation is postponed to a later stage in the

process. This implies that a problem formulation right from the start,
based on the systematic inventory of studies and plans concerning the
North (ISP 3, 1974), has been rejected.
However, the planning approach did not work out as 'planned'. Steps
(1) and (2) took more time than expected. Aside from that, the sketch
of sectoral development lacked the required regional depth. The working
groups involved had even been asked to specify socioeconomic, spatial,
and sociocultural development by sub-region in the North. Note that
working parties dealing with spatial and sociocultural development have
already started with the formulation of prime and lower objectives during
step (2) (ISP 6, 1975 and ISP 7, 1975). As a result of all this, the working
party on integrated planning was unable to proceed with step (4). Facing
these difficulties and with a fixed deadline (January 1977), the steering
group decided that plan-making no longer could wait. In order to speed
up the process the original approach was given up and replaced by a short-
cut. Subsequent efforts had to be concentrated on a problem area of
strategic importance: the relative socioeconomic backwardness of the
Northern Netherlands. A socioeconomic policy group was formed to
work out short- and medium-term countermeasures. It was also decided
to concentrate on the sociocultural development of the North as a second
area of strategic importance, which gave rise to the formation of a socio-
cultural policy group. A third area of strategic importance has been
indicated, dealing with population distribution among sub-regions or
central places, though no spatial policy group has been created so far.
Problem formulation could have been handled more efficiently. It is
the systematic inventory (ISP 3, 1974) which provides sufficient informa-
tion on problems, at least for the first round of a cyclic process. And these
problems are not only observed, they have already been interpreted by
knowledgeable people. The interpretation being limited to provincial
bodies, it would have been desirable, moreover, to obtain reactions from
special interest groups and the general public at this stage. Such reactions
have only been solicited with regard to questions concerning the future
amount of population growth (ISP 8, 1975), although public participation
had been envisaged with regard to all decisive moments in the making of
ISP, including problem formulation (ISP 4, 1974). Using ISP 3 (1974) as a
starting point, the strategic focus shown in Figure 6 could have been com-
piled. Note that this is just a broad summary of policy of decision areas.
Reactions from Northern organizations have actually yielded more
ISP: ITS CREATION AND PRODUCTS 233

detailed infonnation. It is on the basis of this infonnation that an initial

focus could have been selected, considering potential impact, political
importance, controllability and interrelatedness, as previously mentioned.
Public participation, as advocated by the working party dealing with this
subject, offers an opportunity to identify potential impact and political
importance (or public controversy). The question, however, remains
whether the whole organization of ISP has ensured adequate political
inputs.

Employment & labor market Agriculture & horticulture

(e.g. relief work) (e.g. improvement of
agricultural infrastructure)

1
Industry Services
(e.g. effectiveness (such as retail function
present poliCY in rural areas)
Instruments)

Socio-cultural area in Education

general (investment-and 2 (e.g. integration of
activity- subsidies) facilities)

Physical planning General problems

(e.g. hierarchy of 3 (such as provincial 4
central places) budget problems)

Economic development 1
SOCIO -cultural development 2
Spatial development 3
General problems 4

Figure 6. ISP: strategic focus (based on ISP 3,1974).

As far as controllability is concerned, it is once again the working party

dealing with participation which has emphasized the need of investigating
the feasibility of an integrated structure plan for the North at an early
stage. This could also have influenced the choice of an initial focus. As to
interrelatedness, ISP 3 (1974) suggests that spatial development should
rather serve than shape economic and sociocultural growth, which might
have helped the selection of an initial focus, too. It is the working party
on sociocultural development which points out the necessity of attaching
equal weight to sociocultural aspects, thus questioning the traditional
dominance of economic arguments which leaves miscellaneous unsolved
234 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

problems to be taken care of by 'social planners' (lSP 7,1975).

What about the working party charged with the development of models
of regional population distribution? Eight non-demographic criteria have
been selected in order to evaluate alternative futures of population growth
which implies the relationships indicated by Figure 7.

Figure 7. Evaluating alternative models of regional population distribution:

relationships implied.

We consider this a 'wrong track', as far as the selection of an initial focus

is concerned, because the relationships are based on an untested global
hypothesis likely to be refuted by empirical evidence. A simple relation-
ship between population volume and welfare seems highly implausible.
Summarizing, we fmd it hard to understand why the selection of an
initial focus has been postponed to phase 4 (covered by Progress report
number 4), more than two years after the creation of ISP was begun. The
working group that has provided us with the inventory had already con-
cluded:
a. that objectives related to the relative economic backwardness of the
Northern Netherlands (such as a reduction of the income gap, of re-
gional unemployment and a more balanced employment structure) are
sufficiently supported by existing basic material (ISP 3, 1974, p. 21),
b. that objectives for sociocultural improvements deserve high priority
as the planning of this facet has, hitherto, failed to keep up with
economic and physical planning (ISP 3, 1974, p. 22).
ISP: ITS CREATION AND PRODUCTS 235
Generation of alternatives. What has been done in the field of generation
of alternatives? Does ISP go beyond the minimum alternatives of 'a
change in policy' versus 'no change in policy'? The makers of ISP affirm
the desirability of elaborating alternative policy changes, for the sake of
flexibility (as part of the COBA-approach) and public participation.
Yet, so far no alternative solutions have been generated. In ISP 10 (1976)
this has been postponed to a future stage of experimentation in the
domain of sociocultural development. The Report on socioeconomic
policy for the Northern part does not even refer to alternatives. It pro-
poses just one course of action as against the trend (with present policy
unchanged).
However, the working party on socioeconomic development has pro-
duced an estimate of future unemployment in the North different from
the estimate of the Central Economic Planning Bureau of the Netherlands
(CPB), as shown in Table 1. The divergence has not been interpreted in
terms of alternatives. The Northern projection is based upon an
unchanged national policy (which yields 250,000 unemployed in the
country as a whole), while the CPB assumes a change in national policy
reducing overall unemployment to approximately 150,000 in the year
1980.
Table 1. Jobs, labour force and unemployment (labour surplus) in the Northern part
of the Netherlands 1975, 1980 and 1985.

1975 c 1980 (ISP)c 1980 (CPB)d 1985 c

Total number of jobs 508,450 498,200 502,550

Total labour force a 538,300 551,500 574,900
Total unemployment (labour surplus) 29,850 53,300 ± 25,000 72,350
Total unemployment as % 5.5 9.7 ± 5.0 12.6
of total labour force a
Total unemployment at the national level ± 205,000 > 250,000 ± 150,000
Total number of jobs to be created in b
the North (unemployment rate = 3%) > 73,000 ± 20,000

a Oriented to the North.

b Taking into account hidden unemployment: assuming one hidden unemployed per registered
unemployed.
c Source: ISP 9a, 1976, p. 119.
d Source: Report on socioeconomic policy for the Northern part, 1976, pp. 29-30.

What about alternative models of regional population distribution, based

on different assumptions about Northern net-migration (Table 2)? Do
236 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

they qualify as alternative solutions to the general Northern problem of

relative socioeconomic backwardness? The answer is 'no' because the
causal structure implied (Figure 7) is highly dubious.
Table 2. Regional population distribution in the Netherlands in the year 2000: alter-
native models (x 1,000).

Region a Model Ie Model lid Model me ModellV f

Nortfi 1,700 1,747 1,913 2,070

East 3,213 3,198 3,145 3,095
West 6,529 6,506 6,424 6,347
Southwest 403 403 401 399
South 3,648 3,639 3,610 3,582
Netherlands 15,493 15,493 15,493 15,493

a Compare Figure 1.
b Including Southern Ysselmeerpolders.
c Migration according to trend.
d Zero net-migration.
e Annual net-migration 4,000.
f Net-migration rate 1% per annum.
Source: ISP 8, 1975, p. 4-based on the Third report on physical planning in the Netherlands,
part one (1974).

Potential fields of searching for alternatives have already been revealed by

ISP. First, the existence of major uncertainties can hardly be denied. Is
there a chance for even higher unemployment than the projected quarter
of a million? Will the national government succeed in reducing overall
unemployment to 150,000 in the year 1980? Starting from 150,000
unemployed, the ISP-people are either very optimistic or careless. While
uncertainty at the national level has led to an elaboration of variants as
part of the Central Economic Plan (Van de Klundert 1977), ISP does not
take into account 'alternative economic futures'. An excellent opportun-
ity to work out a contingency plan for worse times to come (Le. more
than 150,000 unemployed) has been wasted.
And what about future population growth which affects (among
others) the future volume of the labour force, hence future labour
surplus? Note that back in 1965 the total Dutch population by the year
2000 had been estimated at 20 million. This estimate has dropped to 14.3
million in eleven years time. This kind of uncertainty is mainly caused by
changes in fertility. Aside from that, alternative models of regional
population (re)distribution (Table 2), first presented in 1974, have been
revised twice in about three years (Report on physical planning in the
ISP: ITS CREATION AND PRODUCTS 237

Netherlands, part one, 1974, part two a, 1976, part two d, 1977; Drewe
1977a). These uncertainties mainly pertain to the long run. As to 1980,
they are far less important compared to those attached to future em-
ployment.
Second, 'The Hague' and 'the North' may propose different courses of
action. A sketch of regionalized development per sector at the national
level does not necessarily coincide with a sketch per facet from a Northern
point of view. Instead of settling for an 'early compromise', divergence of
views could also be translated into alternatives.
Third, 'the Hague' is, of course, concerned about maintaining a national
balance in interregional development (see also Friedmann 1972). This fact
may also account for differences of view between 'nation' and 'region'.
Excluding external migration, population (re)distribution turns into a
zero-sum problem, i.e. there can be no 'winners' without 'losers' (Drewe
1977b). But even if the reduction of unemployment does not present
itself as a 'zero-sum problem', the labour market situation in the rest
of the country (including repercussions of policy intervention in the
North and elsewhere) may give rise to meaningful alternatives.
What we have said about potential fields of searching for alternatives
refers to broad alternatives only. But the ISP work also provides entries
into more detailed alternatives, especially when an attempt is made to
derme policy options per decision area.
Given the overriding priority of dealing with relative socioeconomic
underdevelopment of the North, scenarios of labour market develop-
ment are called for, somewhere along the line of what the working
parties on socioeconomic development and population aspects already
have produced. Once the assumption of 'no policy change' is relaxed,
various meaningful alternative futures open up.

Evaluation of alternatives and choice of action. Having concluded that

in the making of ISP, so far, no alternative solutions have been generated,
there is almost no ground for a critical examination of the evaluation
stage. Models of regional population distribution are the only 'alternatives'
that have been 'evaluated' explicitly. Even if they do not qualify as
solutions to relative economic backwardness, the evaluation procedure
may still teach us something about the meaning plan-makers attach to
evaluation in the case of ISP.
The alternative models have been listed in Table 2, the results of the
evaluation are summarized in Table 3. While members of the working
238 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

party on population aspects originally expressed a slight preference for the

maximum alternative (model IV), they later adopted a preferred strategy
which comes close to 1.8 million people in the North by the year 2000
(which then turned out to be the 'new trend'). The circumstances
accounting for this change in preference are:
- reservations with respect to the preliminary choice including the rela-
tive weakness of the early preference (as shown in Table 3), and reserva-
tions similar to ours (see problem formulation above),
- reactions from the Northern region (including politicians) which are
either indifferent with regard to model IV or divided,
- the fact that a 'new trend' has been computed (replacing both model
I and II) which is, in general, judged favourably.

Table 3. Testing four models of regional population distribution against eight criteria:
summary of the results.

Criterion Preference Rank order of preference Strength of

yes or no preference

Quantity of employment yes III II IV weak

Quality of employment yes IV III II moderately strong
Commuting distance no none
Regional income yes IV III II moderately strong
Amenity level yesa IV III II weak
Urbanity Yes a IV III II subjective
Environmental diversity at
the national level yes II III IV weak
Landscape and ecological
quality yes IV III II Ib unknown

a Depends upon the quality of land use planning.

b If model IV (model III) and 'careful' land-use planning do not go together the rank order of
preference changes to: III, II, I, IV (II, I, III, IV); depending also on the availability of 'reliable'
instruments, if uncertain: I, II, III, IV.
Source: ISP 8,1975, p. 4.

From a viewpoint of advanced evaluation methodology, the original

evaluation procedure can hardly be recommended as 'standard operating
procedure' for plan-making. There are at least two substantial short-
comings concerning the manner in which the alternatives have been
scored on the criteria and the way the criteria have been weighted. We are
not informed about the operational definitions of the criteria and no
attempt has been made to predict the impact of alternative models (for
example, their contribution to the quantity of employment in the North).
Moreover, all criteria have received equal weight.
ISP: ITS CREATION AND PRODUCTS 239

Policy design. To comment on policy design we must subject the

products of plan-making to a close scrutiny. We consider the Report on
socioeconomic policy for the Northern part (1976) - hereafter called
'policy report' - as the major visible product of ISP, at least for the time
being. But there are two policy groups, a socioeconomic and a socio-
cultural policy group. While the former is associated with the policy
report, the latter has been charged with the follow-up on the final report
produced by the working party on sociocultural development (ISP 10,
1976). So ISP 10, after all, may qualify as a second important product of
ISP.
Let us start with the policy report. As can be seen from Figure 8, part
one of the report covers the socioeconomic structure of the North, the
main objective of socioeconomic policy, the relationship between socio-
economic, spatial and natural environment facet planning, regional
differentiation and policy instruments. The 'commitment package'
contains forty-one policy recommendations. Part two deals with the
follow-up, emphasizing the need for elaborating more detailed policy
proposals in terms of sub-regions as well as economic sectors.
Using the 'ingredients' of policy design described in Section 3 as a
checklist, we arrive at the following conclusions:
Policy context. Back in 1974, the working party on public participation
raised the question of the exact legal and political status of ISP, as am-
biguity of status would have reduced participation to sheer tokenism
(lSP 4, 1974). As a result, a formal commitment of parliament and
legislative bodies at the provincial level to ISP was agreed upon and a
Departmental Commission for ISP was inaugurated (compare Figure 3).
However, this has not solved the status problem, as recent discussions
show. The legal and political status of the policy report is not free from
ambiguity either. It has been worded rather carefully as 'progress report
pertaining to the main course of policy'. And in its preface the Depart-
mental Commission for ISP, headed by the Minister of Economic Affairs.
expresses reservations with respect to the feasibility of the target of
reducing unemployment in the North by 20,000 by 1980. This target
has been calculated as the reduction of unemployment needed to bring
the Northern unemployment rate (5 percent) back to the national average
of 3 percent. In other words, it is proposed to reduce the regional
(Northern) component of structural unemployment (a concept known
from shift-and-share analysis) to zero. Compare the policy report (pp.
29-30) and Table 1 for details.
 Policy recommendations
1. Socioeconomic structure
2. Main objective of socio- 2.1 Strategy of socio- 2.1.1 Reduction of regional I, 2, 3, 4
economic policy economic policy component of structural ,.....
unemployment \l:)
t-
2.1.2 Scale of economic
activities
5,6 ....
0'1
'-'
.....
2.1.3 Role of government 7,8
3. Socioeconomic policy and 3.1 Socioeconomic policy 9,10 ~
facet planning and physical planning
3.2 Socioeconomic policy II, 12
5
and natural environment ~z.-:.
4. Regional differentiation 13
.sa
II) II)
5. Policy instruments 5.1 Regional components in 14
selective growth policy "'1::
o <II
at the national level .... p.
>-'-'
5.2 Labour market policy 15, 16 ........
(,) '"
5.3 Reduction of regional 5.3.1 Acquisition of new 17 'Os::
(,) .....s::
P'II)
component in structural economic activities ... 0
unemployment 5.3.2 Investment subsidies 18 S (,)
0 ....
(in general) 5.3.3 Natural gas: price and 19,20,21 s::
0
allocation policy o(,)
11),0
....
II)
5.3.4 Electric power rates o ....
Of) <II
5.4 Socioeconomic policy 22,23,24
and scales of economic ~
s::
activities o
5.5 Socioeconomic policy 25,26,27 1::
o
and the role of g-
government et::
5.6 Socioeconomic policy 5.6.1 Socioeconomic policy 28,29,30,31,32 00
and facet planning and physical planning ~
5.6.2 Socioeconomic policy 33,34
and natural environment
.s,
t:.,
5.7 Regional differentiation 35,36,37,38
in socioeconomic policy
5.8 Infrastructure 39,40
ISP: ITS CREATION AND PRODUCTS 241

Figure 8 leaves no doubt about what facet 'is to be master'. Except for
some policy recommendations related to physical planning and natural
environment (Figure 8, headings 3 and 5.6), the task of making an inte-
grated structure plan is far from being completed. It is not altogether
abandoned either, but the report does not indicate how to achieve
integration. Although a commitment to 'progressive reformulation' is
revealed, it is rather limited to sub-regional and sectoral differentiation
of economic policy instead.
As the policy report sticks to the 'initial focus' of relative economic
backwardness, and, as no alternative solutions have been generated, this
report hardly gives a full policy context.
Elements and functions of the policy statement. The forty-one policy
recommendations are mainly addressed to central government. As to their
implementation one is, therefore, referred to a national Report on
regional socioeconomic policy (Minister van Economische Zaken 1977,
pp. 98-102). It has been suggested that the content of policy statements
asking 'how much?', 'of what type?', 'when?', 'where?' and 'who will
benefit and who will lose?' should be examined. The policy report, how-
ever, contains only few specific statements. The target has been quantified
(the famous 20,000 jobs to be created), and the target year (1980) has
been fixed. Statements about locations are deferred to follow-up planning.
There is a strong emphasis on 'quality' in most of the recommendations.
But the question of 'who will benefit and who will lose?' has not been
discussed at all. Considering the wording of the policy statements, one
notices a striking difference between, on the one hand, the specificity of
the employment target and the ambiguity of most of the policy
recommendations on the other. An indicative figure would fit the policy
proposals much better than a target. The target may quickly be outdated
while the proposed means, allowing for a large amount of discretion,
might also easily be contested unless they are at least defmed.
Reasoned justification for the policy. This part of policy design covers
policy effectiveness, consistency with other policies, and explanation of
policy statements linked to technical work.
Assessing policy effectiveness means checking the feasibility of alterna-
tive solutions prior to the choice of a certain course of action, the assess-
ment based on past experience with policy instruments as well as on
impact prediction. We should not expect this to be a strong point of the
policy report, given our above-mentioned diagnosis. The working par-
ties charged with a systematic inventory and with public participation
242 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

had raised the feasibility question at an early stage of plan preparation

(ISP 3, 1974, p. 13; ISP 4, 1974). One has even initiated additional
research with regard to the regionalization of the national multi-year
budget, but the results were not available on time (which gave rise to
policy recommendation number 26).
It is, for instance, postulated that existing instruments of economic
policy were not less effective in stimulating employment in smaller
centers than they were in larger centers (policy recommendation
number 9). We happen to know of research results which seem to contra-
dict this assumption, economic policy being more effective in larger than
in smaller centers of the Northern Netherlands (compare Nederlands
Economisch Instituut 1972).
As long as the task of preparing an integrated structure plan is far from
being completed, horizontal consistency will remain uncertain. What
about the consistency within the socioeconomic policy area? A thorough
check on the compatibility of some forty-one policy proposals is missing.
The same holds for vertical consistency, given the 'centralistic bias' of
most of the proposals.
Finally, are these policy recommendations well 'explained'? Are they
adequately linked to technical work at different stages of plan
preparation? Major deficiencies in technical work have already been
detected in this Section. Existing technical reports, as enumerated in the
last column of Figure 3, are not explicitly referred to in the policy report
- except for an estimate of future unemployment in the North, presented
by the working party on socioeconomic development. But even this piece
of information is not used directly to justify the proposed policy (com-
pare Table I). Without a reasoned justification, it becomes extremely
difficult for all those who have not been directly involved in preparing
ISP to discuss the policy report in a meaningful way.
Basis for monitoring and review. Without an adequate assessment of
policy effectiveness prior to the choice of a 'final plan', the basis for
monitoring and review is severely limited. Policy design is generally very
important as it specifies what it is exactly that has to be checked once
implementation has started. In the policy report, circumstances and
follow-up activities which might allow one to assess the feasibility of the
target have been indicated rather vaguely. If the target should prove to be
unfeasible, it is by no means clear which part of plan-making will have to
be recycled and who will be responsible for the review. In fact, it is the
proclaimed attempt at reducing the Northern component of structural
ISP: ITS CREATION AND PRODUCTS 243

unemployment to zero which provides the only controllable basis for

monitoring and review.
Management of uncertainty. ISP, as any plan, has to cope with various
forms of uncertainty - as demonstrated during our critical examination
of problem formulation, generation of alternatives, evaluation of
alternatives and choice of course of action. The management (reduction as
well as acceptance) of uncertainty, finally, is reflected in (proposals for)
immediate actions, delayed actions and contingency plans. The way in
which the forty-one policy recommendations have been worded makes it
rather difficult to distinguish clearly between immediate and delayed
actions, though some recommendations can be implemented straightaway
(see also Minister van Economische Zaken 1977). Contingency plans,
however, are completely missing. This is surprising, given the overriding
uncertainty concerning the availability of financial resources (expressed
in the preface of the policy report). Incidentally, the cost of reducing
unemployment in the North by 20,000 until 1980 has been estimated
unofficially at 1,300 million guilders. It is through the commitment
package that management of uncertainties is linked with monitoring and
review. In the case of ISP the basis for monitoring and review, already
weak because of the reasons mentioned above, is further weakened by
an unclear distinction between immediate and delayed actions and by a
complete lack of contingency plans.
What about the Report on sociocultural development (lSP 10, 1976)?
Is it any better than the policy report, as far as policy design is concerned?
Let us take a look at the table of contents (Figure 9) before answering
this question. Part A deals with substantial aspects of sociocultural
development which are sketched as policy recommendations pertaining
to eight sectors. 'Labour' is not only treated as a sector in itself, but also
as a special aspect of other sectors, whenever possible and relevant. Part B
concentrates on the administrative and organizational side of social wel-
fare policy which includes basic principles of social planning, participants,
coordination and cyclic process and, finally, some policy proposals.
From a user's point of view, the relative socioeconomic backwardness of
the North is not only a matter of (un)employment although this ISP-
priority is reflected in the way 'labour' has been treated in ISP 10 (1976).
Planning for social welfare, the reporters seem to be less concerned
about the distinction between socioeconomic, spatial and sociocultural
development as they are primarily interested in promoting human well-
being. Hence they come somewhat closer to the idea of 'integrated
planning', i.e. at the user level.
244 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

PART A PARTB
Development sketch Regional development planning
(substantial aspects of sociocultural (administrative and organizational side
development) of social welfare policy)
Policy recommendations per sector: Planning concept covering:
1. education 1. basic principles of social planning
2. labour 2. participants
3. health (care) 3. coordination and cyclic process
4. social work 4. policy proposals
5. art and culture
6. housing
7. leisure, sport and recreation
8. public transport (appendix)

Figure 9. ISP 10, 1976, table of contents.

A similar view is expressed with regard to the concept of regional

development planning (part B) which the reporters also find useful for
socioeconomic and physical planning. Another point which deserves
mention concerns the way policy effectiveness is treated by the 'social
planners'. Having proved themselves aware of administrative and
organizational bottlenecks in welfare planning in their preliminary report
(ISP 7, 1975), they have launched their special concept of regional
development planning. Not convinced that effectiveness goes without
saying, they propose to experiment with the concept in four test areas
(one test area per province). Of course, representing the youngest branch
of planning in the Netherlands, these social planners are less experienced
than their colleagues from economic and physical planning. There is not
much to be said about their planning .concept prior to experimentation
and evaluation. But as some of its features look similar to the strategic
choice framework (e.g. the cyclic character of regional development
planning), it may provide a better basis for monitoring and review and also
for a better management of uncertainties.
However, the working party's fmal report cannot carry more legal and
political weight than the policy report, at least for the time being. Being
a technical report it will probably be even less important (except perhaps
as a basis for sector planning).

Monitoring and review. ISP has not yet reached this stage. But our critical
examination of policy design and the preceding phases show that it is ill-
prepared for it. Among the missing essentials to be monitored are, e.g.:
ISP: ITS CREATION AND PRODUCTS 245

. . . progress of physical change through a development control monitoring system,

the assumptions about social and economic change through published data and
periodic surveys, the means and rate of implementation by public and private agencies
through regular consultations and the continued relevance of the key issues and ob-
jectives through public consultation procedures (Bather et al. 1976, p. 59).

9.3.2 Recent developments of ISP

Meanwhile Pr.ogress rep.ort number 6 has been published, c.overing the
peri.od between the first .of January and the first .of August 1977. This
rep.ort inf.orms us ab.out the f.oll.owing devel.opments:
1. A sect.oral and sub-regi.onal differentiati.on .of s.oci.oec.on.omic p.olicy
rec.ommendati.ons f.or the N.orth (c.ontained in the p.olicy rep.ort) is
underway. The publicati.on .of results (including reacti.ons .obtained
through public participati.on) is scheduled f.or early 1978.
2. S.ome .of the p.olicy rec.ommendati.ons have been implemented, i.e.
pr.op.osals related t.o the impr.ovement .of transp.ort infrastructure and a
number .of subsidies all .of which fit the m.ore traditi.onal regi.onal s.oci.o-
ec.on.omic p.olicy (Minister van Ec.on.omische Zaken 1977).
3. The s.oci.ocultural p.olicy gr.oup has started f.ormulating specific policy
pr.op.osals as well as implementing the experiments in the field .of
'regi.onal devel.opment planning' pr.op.osed in ISP 10 (1976).
4. The p.olicy rep.ort had been subject t.o a public participati.on pr.ocedure
.on a N.orthern scale. N.ot .only has this public 'testing' .of the 1976
rep.ort c.ome t.o an end, but b.oth the .organizat.i.onal set-up and the
c.oncept .of participati.on .on a N.orthern scale have been alt.ogether
aband.oned. Future participati.on c.oncerning the pr.op.osals t.o be
w.orked .out by the s.oci.oec.on.omic and the s.oci.ocultural p.olicy gr.oup
will be limited t.o the pr.ovincial scale.
5. The w.orking party .on integrated planning has published its final rep.ort
in tw.o parts. In ISP lla (1977) future devel.opment .of the N.orthern
Netherlands is described fr.om a ·nati.onal p.oint .of view, while ISP lIb
(1977) pr.ovides a sketch .of its sect.oral devel.opment with sect.ors
c.orresp.onding t.o (eight) departments at the nati.onallevel. S.o, finally,
step 4 .of the .original planning appr.oach has been c.ompleted. ISP Ita
pr.ovides a number .of 'gl.obal p.olicy suggesti.ons' per sect.or t.o be
w.orked .out by the ab.ovementi.oned facet-oriented p.olicy gr.oups.
6. Additi.onal research pr.ojects dealing with envir.onmental quality and
administrative aspects (see Figure 3) have been c.ompleted in the course
.of 1977.
246 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

9.3.3 Why integrated planning has failed (so far)

ISP has been an attempt to integrate provinces, sectors, and facets,
among others. But public participation on a Northern scale has been given
up, the working party in charge of integrated planning has only come up
with 'global policy suggestions' per sector and these suggestions are going
to be worked out by (two) facet-oriented policy groups. Why has the
integration failed? Among the factors that have impeded integration we
find two major ones:
- serious shortcomings in 'the planning of the planning',
- 'surprising' changes in the planning environment.
The term 'process planning', readily adopted in the Netherlands, has often
proved to mean nothing more than a label used to provide plausible, but
untrue reasons for what has actually been an unplanned review neces-
sitated by circumstances not foreseen. Similarly, the term 'experimental',
at first used to characterize the uniqueness of ISP and, later, to justify
trials and errors in preparing it, seems to have drawn away attention from
the need to plan the planning process. ISP, at the outset, was not con-
sidered as research into the possibilities and difficulties of integrated
regional planning, but, after all, as a serious attempt at policy-making for
the Northern Netherlands. Especially ISP, unprecedented in the Nether-
lands as far as its 'comprehensiveness' is concerned, would have required
a thorough, careful choice of inter-organizational arrangement, procedure
and technique. In fact, the coordination between some ten departments
and four provinces has been far from adequate.
Furthermore, the COB A-approach has hardly been examined critically
at the outset, neither with regard to its appropriateness nor to its feasi-
bility. A procedure focusing on a hierarchical ordering of objectives,
instruments and activities tends to neglect, for example, the existence of
conflicting objectives, instruments and activities. Without the interrelated-
ness of issues, there would not even be a need for integrated planning. We
are not suggesting that ISP has failed because one has not applied the
stategic choice approach. One simply needs a framework to discuss a
complex enterprise such as ISP and to examine whether the policy course
chosen is the 'best' alternative within given constraints. However, using
the strategic choice approach does reveal differences between the trial and
error approach to ISP and a more thorough, careful choice of procedure
(the lack of which has turned out to be just one of the factors accounting
for the failure of integrated planning).
Aside from that, ISP has been extremely pretentious on the re-
ISP: ITS CREATION AND PRODUCTS 247

search side. Note that the working party on integrated planning has
presented its final report (ISP 11 a, 1977) as a comprehensive analysis
of development. Given our present state of knowledge and technique,
such an attempt to analyze the way in which 'everything depends on
everything else' is doomed to failure. It certainly neglects the pragmatic,
problem-solving character of plan-making which 'only' requires a selective,
though adapted analysis, implying a selective use of analytical techniques.
As important changes in the planning environment have occurred, ISP
met with a crisis - some time between September 1976 and May 1977
(as documented in Progress report number 4). ISP, which had been
primarily geared into plan-making for the long range, suddenly had to take
into account the economic crisis of the 1970s as well as counteracting
government measures. As a result the original 'comprehensive' approach
has been replaced by a short-cut with special emphasis on the relative
socioeconomic backwardness of the North. This step, which on the one
hand, has brought ISP nearer to commitment, on the other hand also
seems to have provoked the abortion of integrated planning. The Report
on socioeconomic policy for the Northern part, the formation of separate
policy groups and the aftermath (described in Section 9.3.2) testify to
this. Surprising changes in the planning environment are quite normal in
planning practice, especially when it takes several years to prepare a plan.
However, allowing these changes to upset your planning approach in such
a way as to lose its reason for being (reducing 'ISP' to 'SP' so to speak)
once again points to serious shortcomings in the planning of the planning.

9.4 ISP, SPATIAL INEQUALITIES AND REGIONAL DEVELOPMENT

We have judged ISP in terms of planning procedure and technical work.

This does not imply, however, that we consider planning methodology as
an end in itself. What really counts is whether the ISP-approach has solved
the problem. Has ISP really provided us with an effective tool to reduce
spatial inequalities? It is our general impression that [SP is rather part of
the problem than part of the solution.
Too much time has been invested in problem formulation. It took more
than two years to select a priority theme. Even then the priority seems
to have been forced upon plan-makers by changing economic circum-
stances. The relative economic underdevelopment of the Northern region
being a rather 'old' problem which even back in 1973 had been far from
248 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

solved (lSP 5, 1975), there was really no need to start from scratch.
More time should have been spent on investigating the feasibilty of
policy proposals (the effectiveness of policy instruments) instead. Some
of the instruments have been applied for quite some time now (Cremers
1976). The probable impact of newly proposed policy instruments could
at least have been explored or subjected to experimentation. But a feasi-
bility test prior to the choice of action does not seem to be a standard
operating procedure in Dutch planning. The choice of a policy course
almost always precedes a listing of instruments whose effectiveness
seems often to be taken for granted. This practice surely does not stimul-
ate the assessment of policy effectiveness. The 'state-of-the-art' is reflected
in the 'Report on regional socioeconomic policy 1977-1980' (1977, pp.
29-37) dealing with investment subsidies, improvement of infrastructure,
selective investment policy, migration subsidies and relief work. The
gloomier the outcome of a feasibility test, the more pressing the need for
a more direct state intervention (Holland 1976). The latter may imply a
more or less gradual change of the existing socioeconomic order (see
also Chapter 8).
To propagate the target of ceating 20,000 jobs by 1980, hoping that
it turns into a self-fulfilling prophecy (Policy report, p. 28), can hardly
be a substitute for a proper feasibility test. Good intentions alone tend to
breed false hope. George Henry Moulds' plea for 'thinking straighter'
would certainly not be wasted on ISP:
Much advertising and salesmanship, both of goods and of ideas, makes the Appeal to
Practical Consequences. Use this soap and you will be more beautiful. Wear this suit
and you will get ahead in business. What is there that is not sold by such an appeal?
In every case let us ask for proof that the promised benefit will actually be delivered
(Moulds 1966, p. 36).

If doubts about policy effectiveness are not cleared during policy prepara-
tion, then they just live on. Once a course has been adopted, it usually
takes some time before it will be reviewed. If the enterprise of ISP is going
to be perceived as a failure, then it is highly uncertain whether a timely
review along the lines of 'integrated planning' will be launched at all.
As a consequence, regional policy will most probably return to 'where it
belongs', i.e. to the realm of traditional regional economic policy. Time is
running out too. ISP was begun in May 1973, the target year is 1980, and
by the end of 1977 the socioeconomic policy had not even been politic-
ally endorsed.
ISP/SPATIAL INEQUALITIES/REGIONAL DEVELOPMENT 249

Reason enough to sketch an alternative, pessimistic scenario. Living in

the Northern part of the Netherlands is not only disadvantageous in terms
of (un) employment, quality of employment, income, participation in
secondary and higher education, accessibility of urban services. There are
also positive aspects to it such as: environmental quality, territoriality,
sociocultural facilities. In the past the balance of negative and positive
factors has been such as to stimulate selective out-migration, i.e. out-
migration of young and skilled groups of the Northern population. There
is reason enough to expect regional policy, based on the policy report, to
fall short of the employment target. Sociocultural measures alone can
hardly tip the balance in favour of staying for the young and skilled.
Sociocultural policy, in general, will therefore be reduced to its more
traditional role: easing economic backwardness through the supply of
better services in the field of social work, leisure, sport and recreation
etc. and increasing the mobility of Northerners through providing them
better education. Planning for population (re)distribution, spatial planners
are well-advised to look beyond recent total net-immigration into the
Northern Netherlands.
Complexities arising from vertical relations between the national and
the regional level can be reduced by increasing either centralization or
decentralization. It looks as if ISP has suffered from 'misplaced centraliza-
tion' the way plan preparation has been dominated by 'the Hague'. We do
not imd it useful to discuss the .issue of 'centralization versus decentraliza-
tion' in abstract terms in spite of general theories such as those provided by
cybernetics and macro-sociology (Ashby 1956; Hage 1972), pointing out
that decentralization is generally more efficient in coping with great
complexity than centralization. I rather prefer to sketch a plannning pro-
cess at odds with what ISP was meant to be and turned out to be. In order
to identify the gap between the following sketch and the actual regional
planning procedure, it should be compared with the point of view taken
by the central government (Minister van Economische Zaken 1977, pp.
74-76). The Belgian experience, as reported by Bauvir and Jaumotte
(1977), is worth studying in this respect.
Regions such as the North should be encouraged to make their own
development plans without any national interference during plan prepara-
tion, adopting a procedure securing active participation of regional (local)
politicians and of the public at large. As proved in the case at hand by
Northern working parties, there is usually sufficient expertise available at
the regional level to master the technical work involved. Models of
250 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

regional development or other planning methods and techniques available

at the national level should not be used as a substitute for, or to prescribe,
technical work at the regional level. It is only at a later stage that 'the
Hague' should become involved, which does not exclude 'vertical'
consultation at earlier stages. Without any massive revenue sharing (thus
without changing the present political system radically) 'the Hague' would
have to decide upon financial feasibility anyhow. We are by no means
denying the existence of a significant national purpose in regional devel-
opment. But the necessity to choose how to spend a limited amount
of money together with criteria defining national economic efficiency and
regional equity should be applied to the evaluation of regional plans.
National government, instead of aiming at regionalized national plans,
could concentrate on the task of securing the consistency and integra-
tion of regional plans for the country as a whole. Consequently, planning
methodology at the national level must be tailored to fit this task.
Generally speaking, an approach is needed to evaluate the impact of
public policies on regional welfare. Take, for example, the approach of
Haveman (1976), who has formulated four basic requirements:
1. measuring impacts should include both target and non-target regions;
2. impacts should include changes in aggregate and per capita regional
income, in the intra-regional income distribution, in the efficiency in
the use of regional resources and regional environmental effects;
3. basic research should concentrate on the nature of prominent linkages
between policy actions and regional development impacts such as the
discontinuous or strategic induced investment effects of alternative
policies;
4. emphasis should concentrate on the regional effects of major policy
strategies rather than on the regional impacts of individual projects or
small programmes.
What is really at stake here is the maintenance of a national balance in
interregional development. And that is basically a political problem.
It is essential to recognize the political function of plans to provide a basis
for negotiations between the regional and the national level. This also
holds for planning methods and techniques, for instance, for evaluation
techniq ues.

. . . no quantitative method such as economic analysis (cost-benefit analysis) or other

numerical methods can be used as the sole or even dominant basis for determining
which action is 'best' in 'the overall public interest'. Rather, instead of erecting a
ISP/SPATIAL INEQUALITIES/REGIONAL DEVELOPMENT 251
'smokescreen' of elaborate technical methodology, the analysis of alternatives should
seek to catalyze constructive debate about conflicting objectives and priorities (Man-
heim 1977, p. 11).

It looks as if the authors of progress and published reports related to ISP

have instead tried to avoid political connotations, thus suggesting that
they are solely dealing with 'objective' matters such as technical work and
scientific research. If civil servants from 'the Hague' should deny the
political implications of their work, wouldn't that be in itself a political
decision? Fortunately, for those who read newspapers, Northern journ-
alists! have left no doubt about the political background of ISP. Denying
the political character of a plan such as ISP reduces it to an ideology:
a mere suggestion that spatial inequalities are tackled seriously, while
existing inequalities survive or even grow, hence protecting the interests
of 'well-to-do regions' or well-to-do social classes within regions (cf. for
the local level, Drewe 1977c). Other features of ISP may also work to
strengthen its ideological function. This holds for almost all stages of
technical work, but, fIrst of all, for the conspicuous lack of a thorough
assessment of proposed policy instruments. Thus technical work is becom-
ing a substitute for an open political discussion of the socioeconomic
order. In fact, a discussion of the issues raised by Holland (1976) has been
avoided.
As technical work and products have been severely criticized, some
architects of ISP have pointed out that despite rather weak visible
products, 'real' progress has been made in the fIeld of invisible products
such as shared knowledge and perspectives on problems, mutual under-
standing, links in communication networks, especially between partici-
pants from the national and the regional level. To consider these supposed
byproducts as a substitute for a 'good' plan, however, certainly con-
tributes to impairing the political credibility of ISP. It is, after all, the
policy report, a published document, that the Northern population has
been invited to make comments on. There may also be negative by-
products, like participants 'looking back in anger and frustration'.
Finally, there is the question of the meaning of regional science from
the viewpoint of an enterprise such as ISP. Looking at regional science
both in terms of educational programme and research programme, we fIrst

1. We are referring, especially, to articles published in Nieuwsblad van het Noorden in

which the correspondent of this paper in the Hague comments on ISP and related
issues in his column 'Den Haag op de korrel'.
252 INTEGRATED REGIONAL PLANNING: THE NETHERLANDS

will distinguish three types of knowledge (cf. for the following argument,
Working Group CES 1973):
a. community and environment knowledge,
b. organization knowledge,
c. operations knowledge.
Operations knowledge refers to knowledge of techniques and procedures
that help in the operations of regional governance. 'Procedures' refer to
the design of a planning cycle and the pursuit of its stages as, for example,
in the strategic choice framework. By 'techniques' we mean 'the use of
more discrete and specialised elements of procedure? e.g. a labour
market model for the Northern Netherlands used to quantify future mis-
matches between regional labour demand and supply (F.N.E.I. 1978).
Knowledge about the region, including its economic, spatial and
sociocultural aspects, is called community and environment knowledge,
while organization knowledge can be illustrated as knowledge about
horizontal relationships between departments at the national (institutions
at the regional and local) level as well as about vertical relations between
'the Hague' and 'the North'.
Regional science (education and research), in order to be relevant to real
world cases of regional governance, calls for an integration of all three
types of planning knowledge into a problem-solving learning method and a
concept of problem-centered education. It is this type of learning method
and education which provides planning capability. In other words, a
regional science focusing on spatial statistical analysis alone will most
certainly fall short of providing assistance to (integrated) regional
planning. Even 'analytical studies' (combining operations with community
and environment knowledge) alone will not be sufficient. We also need
'policy studies' (integrating organizations and community and environ-
ment knowledge) and 'meta-planning studies' or 'organization-planning'
(organization plus operations knowledge). And there is even more to it.
If regional science would just produce more relevant information, we
would still be facing the gap between planning methodology generated
and disseminated, on the one hand, and, on the other, the utilization of
this methodology in practice. Striving for a 'full entry into the problem-
solving space of all potentially relevant users' (Havelock 1974) demands a
close investigation of the reasons why the results of policy research are not

2. For a recent comparison between available planning methods and techniques and
information needs arising at different stages of plan preparation, see Drewe (1977 d).
ISP/SPATIAL INEQUALITIES/REGIONAL DEVELOPMENT 253

used more widely than they appear to be and, consequently, a close

investigation of approaches directed at a better utilization of knowledge
(details are given by Drewe 1977 d). Or to put it in a nutshelI,-a selective
use of analytical techniques ('adaptive analysis') is called for. Techniques,
to be applied, must correspond to information needs arising at each stage
of plan preparation, and, 'the penalty for not knowing must exceed the
cost of fmding out' (Jones 1973). Simulating adaptive analysis in practice
is one of the basic features of a problem-solving learning method and of
applied research.

REFERENCES
Ashby, W.R.,An introduction to cybernetics. Chapman and Hall, New York, 1956.
Bather, N.J. et al., A strategic choice approach to structure planning. Proceedings of the PTRC
Summer Annual Meeting: Structure Plan Practice, PTRC, London, 1975.
Bather, N.J. et al., Strategic choice in practice: the West Berkshire structure plan experience.
Geographical Papers, University of Reading, 1976.
Bauvir, L. and C. Jaumotte, Le planijication regionale et l'amenagement du territoire en Belgique.
Paper presented at the annual meeting of the French-speaking section of the Regional Science
Association, Aix-en-Provence, September 15-17, 1977.
Beleidsanalyse, Driemaandelijks Bericht van de Commissie voor de Ontwikkeling van Beleids·
analyse ('National commission charged with the improvement of policy analysis'), no. 3, 1973.
Cremers, W.J.M., 'Het Nederlandse regionale beleid sinds 1945' (,Dutch regional policy since
1945'), Intermediair, vol. 12,45-53,1976.
Drewe, P., 'Migration in the Netherlands: policies and models', Delft Progress Report Quarterly,
vol. 2, 343-369, 1977a.
Drewe, P., Inte"egionale migratie en spreidingsbeleid, deelrapport 3, berekening implikaties
spreidingsbeleid (,Population (re) distribution policy and intervention rates implied'). Memoran-
dum, Institute for Town Planning Research, Delft University of Technology, August, 1977b.
Drewe, P., Segregatie in Rotterdam, beleidsanalyse: sociale segregatie en grootstedelijke politiek -
een referentiekader (,Social segregation and urban politics - a framework'). Internal research
paper, Delft University of Technology, 1977c.
Drewe, P., Planning methods and techniques, the state-of·the art. Paper prepared for the Inter-
national Colloquium on Recent Developments in Planning Methodology (Committee on
Housing, Building and Planning of the United Nations Economic Commission for Europe), The
Hague, October, 1977d.
F.N .E.I., De kwalitatieve structuur van de noordelijke arbeidsmarkt: een eerste kwalitatieve analyse
('A labour market model for the Northern Netherlands'}. Federatie van Noordelijke Econo-
mische Instituten, Assen; 1978.
Friedmann, J., 'The implementation of regional development policies: lessons of experience',
International Social Development Review, 45-105, 1972.
Friend, J.K. and W.N. Jessop, Local government and strategic choice. Tavistock, London, 1969.
Friend, J.K. et al.,Public planning: the inter·corporate dimension. Tavistock, London, 1974.
Hage, J., Techniques and problems of theory construction in sociology. Wiley, New York, 1972.
Havelock, R., Research on the utilization of knowledge. Paper presented at the AAAS-ASIS
Symposium, San Francisco, 1974.
Haveman, R.H., 'Evaluating the impact of public policies on regional welfare', Regional Studies,
vol. 10,449-463,1976.
Hickling, A., Managing decisions. The strategic choice approach. Mantec, Rugby, 1974.
Holland, S., Capital versus the regions. MacMillan, London, 1976.
ISP 2, Rapport van de werkgroep beleidsdoelstellingen analyse noorden. January, 1974.
254 REFERENCES

ISP 3, Rapport van de werkgroep SJlstel'Nltische inventarisatie van studies en plannen inzake het
noorden. January, 1974.
ISP 4, Rapport van de werkgroep inspraak. December, 1974.
ISP 5, Interim-rapport van de werkgroep economilche ontwickkeling. March, 1975.
ISP 6, Interim-4'apport van de werkgroep ruimtelijke ontwikkeling. March, 1975.
ISP 7, Interim-4'apport van de werkgroep sociaal-culturele ontwikkeling. March, 1975.
ISP 8, Eindvenlllg van de werkgroep bevolkingsaspecten. October, 1975.
ISP 9a, Eindrapport van de werkgroep economische ontwikkeling, deel I: bedrijfstakvooruit-
zichten. 1976.
ISP 9b, Eindrapport van de werkgroup economische ontwikkeling, deel il: provincia1e vooruit-
zichten. 1976.
ISP 10, Eindrapport van de werkgroep wcilzal-culturele ontwikkeling. October, 1976.
ISP lla, Rapport van de werkgroep integrale planning, deelI: het noorden in nationaal penpectief.
March, 1977.
ISP 11 b, Rapport van de werkgroep integrale planning, deel II: departementale sektor-verkenningen.
August, 1977.
Jones, J.C.,Design methods. Wiley, London, 1973.
Manheim, M.L., The emerging planning proce,,: neither long-range nor short-range, but adaptive
and (hopefully) decisive. Paper presented at the 56th Annual Meeting of the Transportation
Research Board, Washington, D.C., January, 1977.
Minister van Economische Zaken,Nota regionaal socilzal-economilch be1eid 1977-1980 ('Report on
regional socioeconomic policy 1977-1980'). 's-Gravenhage, 1977.
Moulds, G.H., Thinking straighter. Kendall/Hunt. Dubuque, Iowa, 1961i.
Nederlands Economisch Instituut, Kampen en Zwolle na tien jaar stimuliering ('A case study of
regional policy'). Rotterdam, 1972.
Report on population aspects, Het noorden, een venterkte bevolkingsgroei? Of juist niet!. Staats-
uitgeverij, 's-Gravenhage, 1974.
Report on socioeconomic policy for the Northern part, Het socilzal-economish beleid voor het
Noorden des Lands. Staatsuitgeverij, 's-Gravenhage, 1976.
Report on the Northern part of the country, Nota Noorden des Lands. Tweede Kamer der Staten-
Generaal, 's-Gravenhage, 1972.
Third report on physical planning in the Netherlands, part one, Oril!nteringmota. Minister van
VRO, Staatsuitgeverij, 's-Gravenhage, 1974.
Third report on physical planning in the Netherlands, part two a, Verstedelijkingmota. Minister
van VRO, Staatsuitgeverij, 's-Gravenhage, 1976.
Third report on physical planning in the Netherlands, part two d, Ventedelijkingmota. Minister
van VRO, Staatsuitgeverij, 's-Gravenhage, 1977.
Van de Klundert, T., 'Het Centraal Economisch Plan 1977' ('The Central Economic Plan 1977'),
Economisch Statistische Berichten, vol. 62,490495,1977.
Working Group CES, 'Education for planning, the development of knowledge and capability for
urban governance' (Report of a Working Group at the Centre for Environmental Studies,
London), in D. Diamond and J.B. McLoughlin (eds.), Progre" in planning, vol. 1, Pergamon,
Oxford, 1973.
Subject index

abatement 14,15,57,71 possibilities 99-100,112-3,116

accessibility 100,102,105,111-4,156 potentials 93,97,100,108,111,
agglomeration economies 170,175,181 115-6
agriculture 11 requirements 97,100,106,114
Allardts' conditions of 'loving' and systems 95
'being' 138,144,150,155 contradictions
antithesis 24 antagonistic 26
Austria (income differences in) 149 cosmic and historical 24-28
non-antagonistic 26
back-wash effects 135,153,157,174 pseudo resolution of 35-39
Barlow report 202 contro1(s)
basic services 151 framework of indirect 17
'beggar-my-neighbour' 201 political- parameters 81
Brazil costs oflabour 188,193,196,207
growth centre policy in 177-8
income differences in 147 decision-making
business (big) 190-1,193-4,198,203, components of - processes 15
204-5 concentrations of - functions 12
departmental structure 18
capitalism (oligarchic) 216 process 15
choice of action 237 demographic subsystem 4-7
choice set 57, 83 denationalized corporation 42
city development policies 135, 150
functional concept 3 1 dialectical materialism 25
regional 35 diffusion matrix 74
territorial concept 32 disagreement equilibrium 62,67,84
COBA-approach 230, 246 discrepancy index (regional) 78
communication costs 11 diseconomies (Iocational) 208
competitive growth 179 dispersion 167
conflicting goals 17 distance function 7
conflict matrix 70,71 distance (psychological) 7, 118, 120,
congestion costs 157 126-7
consumer sovereignty 193 distemic space 142-3
contact dualistic philosophy 27
accessibility 106
activity 94 economic subsystem 10-12
landscape 97-9 Edgeworth box 60
interregional network 96 efficiency
maps 105, 110
256 SUBJECT INDEX

aggregate - 17,161,170,172-3,180, center policy 146,149,177,193

182 center theory 136, 178
- equity compatibility 171, 176, 180 competitive or generative 179
- equity dilemma 47,174 pole 38,151-2,211
entropy maximising model 6 model 166
enterprise pole philosophy 191
macro 186 potential 171, 177
meso 186,194,200,207,213
micro 195,207,213 Hicks-Kaldor compensation scheme 55
regional public 212 hinterland effects 151, 179
environment(al) horizontal aid 181
externalities 55-6,61,69,70,91
externalities model 82 ideology 207
principles 71 inbound accessibility 112
profile 51 inbound maximal stay time 101-2, 104
protection 47 inbound potential contact 102, 106,
protection policy 68-9 110
quality 47,50,58,67,71,83 income
subsystem 13 distribution 168, 178
welfare index 59 equalisation 169
equilibrium (multiregional) 50 generating factors 10
equity(interregional) 1,17,161-2, regional-differences 13
164,166-7,169-73,180-2 transfer 180
evaluation of alternatives 226,237 indifference curves 59-60
export-base activities 151 inequality (interregional index) 76,
export-base theory 138 78,80
external economies 157 (see also spatial)
externalities 17,48-50, 176,201 innovations 170
input-output matrix 13,75
face-to-face contacts 12, 96 integration
factor flows 189 functional 10,24,29,41,134,139
feasible state of the spatial system territorial 10,29,43
61,84 Intermediate European Center 103,
footloose industry 169,171, 176 105,111-3
France investment subsidy 182
income differences in 148 Isardian school 184
migration in 119-122 Italy (income differences in) 147
Friedman's prescription 209
functional Japan
change 136 efficiency-equity trade-off 172-3
differentiation 97 income differences in 149
organisation 18,43 job dispersion 190
job functions 94-5
generation of alternatives 226, 235
Gestalt psychology 185,201,207 Keynesian pall 199
Gini coefficients 163, 178 knowledge types 252
Great Britain latent variables 3
income differences in 148 leading industries 139
migration in 119-122 level of living components 9
growth labour costs 188,193,196,205
SUBJECT INDEX 257

Lowry model 12 Pareto

optimality 50; 72
management of uncertainties 228,243 optimum 61,65,67-69,84
market-based processes 141 state 61,84
Marxist analysis 205 partial optimum 85
Maslov model 8 people prosperity 38,168
measurement problem 4 personal contact networks 112
Mediterranean countries (migration philosophy (monistic) 27
from) 122-27 physical profile 51
megalopolis 34 Pigou-Baumol equilibrium (PBE) 63-68,
migration 85
behaTIour 119,122,128 PigouTIan solution 55
matrix of gross-rates 5 place prosperity 38, 168-9
selective 169 planning
mini-max effect 221 approach 230
Minkowski p-metric 54,56 cyclical 17,219
mobility democratic 213-4
behaTIour 127 facet 223
geographical 117 hierarchical decentralized 17
ingeneral 119,122,128,176 integrated regional 219,223,246,
policy 118 248
specific 120 integrated structural 16, 17,219,223,
monitoring 226,228,242,244 228,245,247,251
monopolistic pricing 198 sector 223
monopoloid market structures 11, 193 pluralism 186,203,212
multinational actiTIty 188 polarization 167, 174
Myrdal model 190 policy
context 239
neo-classical economics 137 design 226-7,239
neo-classical model 169, 185 framework for regional development
neo-imperialism 39 15
the Netherlands (migration in) 119-22· regional 16,233-5,177,185,197,
nodal hierarchy of accessibility 111 199,206
nodes of opportunity 38 preference pattern (regional) 155-6
'no-entry' pricing 194 production distribution (regional) 64
no-growth philosophy 168 profile
non-institutional sphere 145, 153 locational 106, 108
non-market sphere 145,153 spatial welfare 49
non-place urban realm 41 pollutee-pays principle 66, 69
polluter-pays principle 66, 68-9
objective function 73 pollution 14,50
oligopoly analysis 194 pollution tax 64,79,80,88
ontogenetic unity 28 population preference structure 118
ontology 27 power
operations knowledge 252 distribution 18
option bars 226 structure 188,201
outbound accessibility 110, 112 preference function 172
outbound maximum stay time 101-2 price system 48
outbound potential contact 102, 106, Primary European Center 102,105,
109 113
258 SUBJECT INDEX

problem formulation in planning 226, diversity 117

232 division oflabour 24, 39
process planning 246 dominance 53
producer sovereignty 193 inequality 1,12,47,93,116,139,
production-consumption set 56,57, 146,154,219,221,247
82-3,88 redistribution policy 68-70
production "function 169 welfare patterns 47, 49
propensity to move 14 spatio-temporal block 52
proxemic space 142-3, 155 specialisation
push-and-pull factors 124 areal 137
functional 10, 134
quality of life concept 144 socio-cultural subsystem 1, 7, 8
quality of life index 163 state
matrix 3,15,51
redistribution policy 68 models of behaviour 203-7
Redistribution policy 68 vector 2,3
regional development 1,15,93,99, strategic choice approach 219, 223,
115,137-9,158,185,219 225-8,246
regional science 1,18,251-2 Sweden
relocation subsidies 169 contact systems. in 95-100
"run-away" industries 193 migration in 119-22

scale economies 157 target population 168

scenarios 226, 249 territorial
Secondary European Center 103,105, integrity 136,152
111,2 organisation 18
sector theory 138 threshold level 57
selective spatial closure trade-off (function) 170, 180, 182
as a strategy 1524
from the demand side 155-6 umbrella pricing 196
from the supply side 154-5 uncertainties 227-8,243
other measures of 156-8 United States (income differences in)
self-realisation 154
individual 8 urban bias 164
of groups 8 urban pricing 196
service sector (location) 12
social welfare
costs 43 collective 8
disintegration 38 functions 72
indicators 51,144,163,181 index 59
integration 25,38 individual 8
vector 51 loss 70
welfare 78 measures 163
socialisation of consumption 216 regional 47, 57
socialisation of production 210 West Germany (income differences in)
space bridging capability 140 148
Spain (income differences in) 147, 167 Western European countries (migration
spatial to) 122-6
diffusion matrix 74 Williamson curve 173-5
diffusion patterns 58