Path Dependence and Urban History: Is a Marriage Possible?

Martin V. Melosi
Department of History, University of Houston
Houston, Texas USA

The impact of a variety of technologies and infrastructures on the growth of cities is

a central concern of urban environmental history. The implementation of new urban
technologies, however, was not automatic, coincidental, nor inadvertent, but an intentional
effort by decision makers to confront existing problems faced by cities as they extended
upward and outward in the nineteenth and twentieth centuries.

Path dependence is an important approach for exploring the role of decision making
on the development and impact of specific technologies or technical systems. The theory
focuses attention on the means by which choices are made, the connection of those choices
to future options and sequences of events, and to outcomes. The use of path dependence
reinforces the idea that a variety of externalities influence the adoption of technologies. It
can help historians turn from concentrating so heavily on a singularly past-centered
perspective to a present-centered perspective that devotes significant attention to results as
well as constraints on outcomes. This is especially useful if historians are interested in the
policy implications of their work.

Social scientists, especially economists, began exploring path-dependence theory in

the mid-1980s. Although definitions vary, simply put, path dependence exists ‘when the
present state of a system is constrained by its history.’ The theory has been applied to a
variety of technologies, institutions, and policies. In a larger context, path dependency has
challenged neoclassical economics by stressing the importance of institutions (institutional
economics) and through the suggestion that historical events affect present choices
(evolutionary economics). Economic historian Paul A. David has argued that ‘the future of
economics as an intellectually exciting discipline lies in its becoming an historical social
science.’ Given that economics, and other social sciences, have been dominated by a strong
behaviorist philosophy built upon quantitative methodologies, some saw the notion that
history matters as a radical departure. From David’s perspective ‘[H]ow could there be
anything new or radical in the idea that particular sequences of events in the past have had
enduring effects upon current conditions?’ For historians, path dependence is quite
reasonable, given our own sense of causation and the importance of sequencing events to
determine change over time. The radicalism of path dependency, however, lies not so much
in its fundamental claims, but in its challenge to social science orthodoxy in the wake of the
cliometric revolution.
The debate over path dependency is interesting in its own right. That debate has also
been useful in helping to explore the intrinsic value of the theory for historians, including
urban historians. A central question--and an ironic one--is: why borrow such an idea from
colleagues in the social sciences, who borrowed it from us in the first place? This paper
explores the scope of path dependence as developed by its advocates and criticized by its
detractors, and reflects upon its value to the study of urban history. Paul David’s 1985
article, ‘Clio and the Economics of QWERTY,’ brought widespread attention to path
dependency. David illustrated the theory through a case study on the adoption of the
QWERTY standard Englishlanguage typewriter keyboard in the 1870s. The QWERTY had
been designed to reduce the mechanical jamming of the keys and was quickly adopted by
many office managers. Other keyboards, notably the Dvorak Simplified Keyboard
introduced in the late 1930s, appeared to be faster and more efficient, but the QWERTY
had long since dominated the market. In David’s view, a small--or accidental--historical
event had ‘locked-in’ the QWERTY technology and essentially eliminated its competitors.
While the article garnered substantial attention and stimulated additional study of path
dependence, defenders of neoclassical economic theory were unconvinced by David’s
argument and questioned the QWERTY case in particular and the path dependency theory
in general. But as Douglas J. Puffert—a student of David’s—perceptively noted, ‘David
defined path dependence as a dynamic feature of an allocation process, whether or not the
resulting allocation is efficient. Many readers of his article, however, interpreted path
dependence primarily as a source of market failure through lock-in to a suboptimal
technology.’ In essence, supporters and critics were often debating different issues.
In a recent article, and in response to critics, David defined the concept of path
dependence as ‘a property of contingent, non-reversible dynamical processes, including a
wide array of biological and social processes that can properly be described as
“evolutionary.”’ He added that ‘the policy implications of the existence of path dependence
are shown to be more subtle and, as a rule, quite different from those which have been
presumed by critics of the concept.’ These alterations in his definition were meant to
suggest that history matters in economic processes, but not always in the same way. In
addition, the concept of historical lock-in should not be taken too literally, but viewed as ‘a
vivid way to describe the entry of a system into a trapping region—the basin of attraction
that surrounds a locally (or globally) stable equilibrium.’ Thus lock-in is not rigidly
inflexible, but the result of making choices in the past that constrain options in the present
or future.
David utilized path dependence theory to question economic orthodoxy. ‘Indeed,
for too long,’ he stated, ‘it has seemingly been our collective educational purpose to
extirpate from the minds of neophyte economists all but the most fundamental human
intuitions about the role of the past in present (economic) affairs.’ ‘Path dependence, at least
to my way of thinking,’ he also stated, ‘is therefore about much more than the processes of
technological change, or institutional evolution, or hysteresis effects and unit roots in
macroeconomic growth. The concepts associated with this term have implications for
epistemology, for the sociology of knowledge, and cognitive science as well.’ While David
set a foundation for path dependence and attracted an initial firestorm of criticism, others
have attempted to refine the theory and broaden its applicability within the field of
economics as well as in the other social sciences. Particularly noteworthy is the work of W.
Brian Arthur, Douglass C. North, and Paul Pierson. among his contributions to refining
path dependency, Arthur has been particularly noteworthy in the area of increasing returns
economics, by demonstrating how small historical events can be amplified by positive
feedbacks. This approach stands in opposition to conventional economic theory that
emphasizes diminishing returns, that is, economic actions that produce negative feedbacks
resulting in a state of equilibrium for prices and market shares resulting in stabilizing the
economy. Arthur and others found this latter approach to defy reality in several cases.
Under increasing returns, multiple outcomes are possible. As a consequence, tracing the
way in which small events accrue to cause a particular system to move to one particular
outcome instead of another becomes important. In examining choices made among
technologies, Arthur noted that ‘modern, complex technologies often display increasing
returns to adoption in that the more they are adopted, the more experience is gained with
them, and the more they are improved.’ When technologies compete for a market, therefore,
‘insignificant events’ may give one an initial advantage for adoption. The technology may
then improve more than the others, and thus appeal to a wider group of adopters, and so on.
In this scenario the technology that by chance gained an early lead may ‘corner the market’
of adopters, and other technologies may become ‘locked-out.’ Such an approach highlights
two new properties in the adoption process: historical lock-in of a given technology
(inflexibility) and path dependence (non-ergodicity). Arthur admitted that not all
technologies enjoyed increasing returns with adoption, but he believes that it might be
useful to determine to what degree the actual economy might be locked-in to inferior
technology paths. This perspective, of course, is in direct opposition to the concept of
diminishing returns, where small events cannot determine outcomes and where a laissez-
faire approach leads to the success of a superior technology. Arthur’s increasing returns
approach challenges those assumptions.
Arthur’s focus on modern technologies, however, does not take ample account of
an institutional context for path dependence. Nobel laureate Douglass North clearly added
that dimension to the theory: ‘History matters. It matters not just because we can learn from
the past, but because the present and the future are connected to the past by the continuity of
a society’s institutions.’ North regards institutions as ‘the rules of the game in a society or,
more formerly, are the humanly devised constraints that shape human interaction.’ His
concern is that there is no analytical framework to integrate institutional analysis into
economics or economic history. North is particularly sensitive to the gulf between
neoclassical theory and its concern over ‘the allocation of resources at a moment of time,’
and how much of ‘a devastatingly limiting feature’ that is to historians ‘whose central
question is to account for change over time.’ ‘Moreover,’ he adds, ‘the allocation was
assumed to occur in a frictionless world, that is, one in which institutions either did not
exist or did not matter.’North made clear that institutional analysis could make the U.S.
economic history (in this case) ‘a truly historical story,’ and much of that history he
believed was path dependent—‘simply by nature of constraints from the past imposing
limits on current choices and therefore making the current choice set intelligible.’ Focusing
on the process of economic growth, he postulated that the primary source of economic
growth is ‘the institutional/organizational structure of a political economy;’ and that
economic growth is dependent on ‘stable political/economic institutions. Furthermore, the
‘belief systems of societies and the way they evolve’ that is the ‘underlying determinant of
institutions and their evolution.’ Understanding the nature of path dependence, he
concluded, ‘is the key to understanding the success or failure of economies in altering their
competitive positions.’ North also observed the ability of several actors to influence
outcomes within an ‘institutional matrix’ through, among other things, a variety of
increasing returns, through inducing incremental change, and through informal constraints
such as the transmission of values. A theory of institutional change incorporating path
dependence, he concluded, can become a central feature in understanding economic growth.
Political scientist Paul Pierson has been influential in utilizing path dependence to
analyze politics, political institutions, and decision making. In a recent article, he explored
path dependence as a social process ‘grounded in a dynamic of “increasing returns.”’ Like
Arthur and others, he argued that increasing returns processes ‘are likely to be prevalent’
and can provide a strong framework for developing the study of historical institutionalism.
He is particularly interested in developing path dependence theory with more rigour and
beyond the ‘loose and not very helpful assertion’ that history matters. Utilization of an
increasing returns process to elaborate path dependency is especially important because in
this process ‘the probability of further steps along the same path increases with each move
down that path.’ In addition, he is interested in exploring a narrow conception of path
dependence (social processes that exhibit increasing returns is one approach) so that the
theory does not suffer ‘concept stretching’ in which different types of temporally linked
sequences are identified under a single banner. Pierson is aware that applying tools of
economic analysis to politics can be ‘treacherous’ unless applied carefully and
systematically. He believes that doing so produces salutary outcomes because ‘The political
world is unusually prone to increasing returns.’ He is particularly concerned that in
identifying initial causes influencing future patterns, the object of study needs to become the
critical juncture or ‘triggering events’ which set development along a certain path. In
addition, he cautioned that path dependent analyses need not presume that a particular
alternative is permanently locked-in. With these caveats, he strongly asserts that increasing
returns arguments may produce exciting results in political science. Skeptics of path
dependency have challenged vigorously this analytical tool and, in turn, questioned the
heightened assaults on neoclassical theory. For those interested in applying path
dependence to a variety of historical problems, the critiques are valuable for an array of
pertinent issues requiring further consideration. The best-known criticism of path
dependence comes from economists Stan J. Liebowitz and Stephen E. Margolis. The
primary thrust of their argument is that path dependence merely presents a false claim of
market failure. They suggest that economists have found imperfections in free markets, and
that improvement was difficult, but also that alleged market imperfections may prove later
not to exist at all. Moreover, there may not be realistic means to overcome supposed
market imperfections and ways may be found to improve upon independent decision
making (by firms and individuals) which is how resources are allocated in a free market.
In defining path dependence, Liebowitz and Margolis assume that the theory
‘spilled over’ to economics from other fields or disciplines—in physics and mathematics in
the form of chaos theory (a non-linear model ‘with sensitive dependence on initial
conditions’) and from biology via contingency (‘the irreversible character of natural
selection’). In comparing chaos theory and path dependence, they suggest that in chaos
theory ‘small events or perturbations tend to cause a system to evolve in very different
ways but the system never settles down in any repeatable path or fixed equilibrium.’ In the
case of path dependence in economics, the view that minor initial perturbations are
important is imported from chaos theory, ‘but has grafted this on to a theory where there
are a finite number of perfectly stable alternative states, one of which will arise based on the
particular initial conditions.’ Perpetual disequilibrium is thus missing from path
dependence.Defense of traditional market mechanisms led Liebowitz and Margolis to
dismiss David’s claims of path dependence as demonstrated in his various case studies,
including the QWERTY example. Liebowitz and Margolis assert that David’s evidence is
shaky, especially the notion that QWERTY became the standard because of the lack of
sufficient challengers. They question the assumptions that the Dvorak keyboard was
inherently superior, and that the sustaining influence of the QWERTY clearly indicated a
market failure. Their critique challenges David’s historical data and also his drift away from
neoclassical theory to make the case for market failure. They conclude that ‘what credence
can possibly be given to a keyboard that has nothing to accredit it but the trials of a group
of mechanics and its adoption by millions of typists? If we use only sterilized models of
markets, or ignore the vitality of the rivalry that confronts institutions, we should not be
surprised that the historical interpretations that result are not graced with truth.’ In a rather
clever condemnation of path dependence--again relying heavily on an unwavering reliance
on neoclassical economics--Liebowitz and Margolis developed taxonomy of path
dependence. A minimal form—or first-degree path dependence—is an element of
‘persistence or durability’ in a decision with no apparent harm done at a later time. In
second-degree path dependence, past conditions lead to outcomes that are ‘regrettable and
costly to change,’ but where an individual failed to predict the future perfectly because of
imperfect information. Third-degree path dependence—the strongest form—suggests that
known feasible and preferable alternatives exist at the time the initial decision was made and
thus an error that arises was avoidable. Their conclusion is that first and second-degree path
dependence are common, have always been a part of economic thought, and that traditional
theories of neoclassical economics can explain them. Third-degree path dependence—the
kind associated with the critics of neoclassical theory—does not exist in the real world.
Such argumentation effectively dismisses path dependence. In the concluding chapter of
their recent book, Winners, Losers & Microsoft, Liebowitz and Margolis conclude‘our
message is simple: Good products win. Whether they are lowly mousetraps or high-tech
networks, better products prevail in the marketplace. People choose what they want, and
what they want survives, at least for a while. Surviving products are imitated and become
the norm…Eventually, when something decidedly better comes along, there is a transition
to the new product.’ And as Liebowitz was quoted as saying in a 1998 email exchange:
‘Finally…there is the claim that path dependence enhances the value of economic history.
We think not.’
 To a large extent, Liebowitz and Margolis are defending methods of neoclassical
economic theory against an interloper. It is quite acceptable that path dependence (and
economic history in general) be used as a tool for understanding some past actions, but
when it is also used as a way of explaining and influencing the present, critics—especially
in economics--fall back on disciplinary orthodoxy. As Puffert suggested, ‘[M]any
neoclassical economists view history as little more than a source of data for testing theories.
At the most, they restrict the role of history to determining the fundamental, a priori,
“exogenous” parameters that presumably then determine a unique equilibrium—such
parameters as institutions, technology, factor endowments, tastes, and information.’ To be
fair, of course, it must be added that economists and historians do not always ask the same
questions, although they may have, as economic historian G.R. Hawke stated, ‘overlapping
rather than identical interests.’ He perceptively noted, ‘historians frequently use expressions
like “understanding the past in its own terms,” while economists want to employ concepts
developed more recently in their studies of past societies. The point might be described as a
conflict between “past centred” and “present centred” studies.’ Hawke point outs that the
difference is one of degree, but the difference often exists nonetheless. Since his 1985
article, David has explained and refined his views on path dependence on several
occasions. Ultimately, he has promoted some basic precepts of causation that historians
should find easily compatible with their own views: ‘effects follow causes in temporal
succession’ and ‘there are outcomes which simply cannot be achieved—or, are extremely
unlikely to arise—except through some particular dynamic of intervening events.’ In many
respects, path dependence puts a different name—and at times a different perspective--on a
process that is basic to historical inquiry. Indeed, it is a concept originally borrowed from
the discipline of history itself. What appears most useful about path dependence for
historians, however, is not just the reconfirmation of historical causation, but a shift in
focus from past to present, that is, greater attention to outcomes of past events rather than
simply tracking change over time, which often emphasizes the look back as opposed to the
look ahead or the impact of choices made. In this sense, critiques like that of Liebowitz and
Margolis are valuable in forcing proponents of path dependence to provide more precise
definitions of the theory. As economic historians Lars Magnusson and Jan Ottosson have
argued that ‘…Liebowitz and Margolis are certainly right in that if “path dependency” is to
serve as anything more than a catchword for sweeping criticism of neo-classical economics,
it is important that we try to be more specific regarding its theoretical status.’ For the
purposes of historians, Liebowitz’s and Margolis’s argument too narrowly focus on market
failure, inefficiency, and good and bad (mistaken) technical choices. In some respect, the
argument makes a priori judgment, not necessarily questioning that choices lead to
particular paths, but that results that produce inferior outcomes do not correspond to actual
market forces. ‘Better products prevail in the marketplace.’ They seem to accept a
distinction made by others that choices can be past dependent, rather than path dependent.
In other words, they question the sequencing of events that lead to what they argue are
improbable outcomes. Yet, how can any outcome that is either past or path dependent not
be influenced by what came before? Specifically, with respect to superior/inferior
technologies, Liebowitz and Margolis seem to miss the point that what Arthur and others
are talking about are characteristics of technologies rather than simply technologies
themselves.
It is not the point of this paper—nor is it within the expertise of the writer—to enter
the rather extensive and detailed debate over the value of path dependence within the field
of economics, or whether path dependence undermines neoclassical economics and offers a
major epistemological breakthrough in the field. This discussion hardly does justice to a
thorough discussion of market forces, network effects, diminishing returns, increasing
returns, and so forth. More simply, rather than taking on that large task or even questioning
the more narrow notion of whether an inferior technology could succeed in a market
economy, for example, it is argued here that path dependence may be useful for historians
in addressing questions from a slightly different vantage point. What are the constraints
placed on a current generation because of choices made in the past? What are the limits of a
chosen path, as opposed to what is or is not changeable? What are the implications of
particular choices and potential lock-ins on practice and policy? What kinds of impacts are
possible with path-dependent actions? In many respects, these are qualitative issues, but the
kinds of issues that complement traditional uses of historical causation in understanding
technologies, institutions, and politics, useful to urban historians. Again, ‘market failure’
seems to be a very narrow way to judge the value of path dependence. Historians tend to
think in degrees of change, rather than absolutes. It seems that a major virtue of path
dependence is to determine how future choices are constrained or limited rather than how
they are precluded. Lock-in need not be a permanent condition, nor is it the best issue for
historians to track. Choices are made, what are their consequences? Is there a substantial
impact on the sequencing of events because of initial choices? These seem to be more
useful question for historians rather than debating whether initial choices grow out of
‘insignificant’ or ‘small’ events or even random actions. ‘Insignificant,’ ‘small,’ or
‘random’ have to be evaluated and based on outcomes. Something is hardly insignificant if
it produces a significant outcome. To repeat what was stated earlier, path dependence can
help historians turn from concentrating so heavily on a singularly past-centered perspective
to a present-centered perspective that devote significant attention to results as well as
constraints on outcomes. This is especially useful if historians are interested in the policy
implications of their work.
In my recent book, The Sanitary City, I attempt to use path dependency to help
explain why choices made in American cities about water supply, waste-water, and solid
waste systems in the early- to mid-nineteenth century constrained choices available in the
late-twentieth century. (Some reviews rightly noted that my use of path dependence for
solid waste systems was less convincing than for the other two.) Among other things, I
argued that decisions to seek city-wide water supply and waste-water systems in the
nineteenth century were influenced by several factors, including a growing desire by major
cities to gain increasing control over revenues and services in the name of home rule;
dissatisfaction with private approaches to service delivery that either failed to live up to
expectations or weakened the ability of cities to control their own affairs; rapid population
growth that produced a scale of demand for such services that no longer could be
adequately provided by individuals themselves; for water supply in particular, the fear of
fire; and an abiding dread of epidemic disease that could affect the whole city and its
hinterland. In the wake of these converging issues in the nineteenth century, city leaders
sought to provide water supply and waste-water services ( However, these services did not
appear simultaneously—water supply had the greatest priority). Strongly informing choice
was the ‘sanitary idea’ that had migrated from England in the 1840s. The notion that
disease was caused by filth—or miasmas and smells emanating from putrefying waste—
held sway in the United States until the bacteriological revolution became widely accepted
in the twentieth century, purporting that disease was contagious and was caused by
microscopic agents—or bacteria—and not from environmental causes like miasmas and
putrefaction. The miasmatic theory of disease led to the conclusion that disease could be
reduced and indeed eliminated through environmental sanitation. This meant, in the case of
water supply, developing a technical system that carried clean water (determined largely
through the senses) through pipes to home and businesses throughout the city. For
effluents originating in the home or in businesses, pipes could likewise evacuate and
deposit liquid wastes elsewhere—most typically into nearby rivers and streams or into the
ocean. In essence, these early technologies of sanitation were nothing more than elaborated
transportation systems for water and waste-water, but conformed to the thinking about
disease eradication prevalent at the time.
Through these practices, several forms of epidemic diseases showed sharp declines,
due in part to these closed systems. But diseases itself was not eradicated; in some cases
displaced from one location to another or not confronted at all. Yet bad science had led to
relatively effective technology—at least within the context of the nineteenth century. Future
additions to the systems—filters, treatment facilities, chlorination equipment—emerged in
the wake of the bacteriological revolution and took careful account of the communicable
nature of epidemic diseases and placed increasingly less reliance on environmental
sanitation. Despite the significant breakthrough in the understanding of diseases, the basic
systems changed little from their origins in the nineteenth century in the sense that they
were highly centralized, capital intensive systems focusing on access to pure water supplies
and contending with forms of point pollution. In the later years of the twentieth century,
however, pollution problems became increasingly complex—more pollution from run-off
(non-point pollution) and ground-water contamination. These forms of pollution were not
easily addressed with large-scale fixed systems that were incapable of capturing run-off
from a variety of sources and directly addressing groundwater sources of pollution. This
change in context—along with a growing sensitivity to a range of ecological perspectives
which made our understanding of the environment more sophisticated, but more complex
as well—indicated that our technologies of sanitation derived from the nineteenth century
were not up to the task of protecting cities from current health hazards as they were
believed to be years before. Choices made earlier in the path clearly constrained future
options for no other reason than the existing infrastructure was too extensive, too costly to
replace, or resistant to change.
The weakness of my assessment was not discussing viable alternatives to the
existing systems or outlining carefully the available policy options. Frankly, this is
something I intended to do in a series of articles after the book was written; believing that
the primary thrust of the study was to provide the historical underpinnings for some future
discussion of these questions. In some respects, as well, the book did not deal explicitly
with the theory of path dependence much after the introduction, since I assumed that the
issues were implicit in the narrative. The possible weaknesses of my presentation with
respect to path dependence theory, however, do not undermine the value of the theory to
historical scholarship—in this case urban history. Path dependence theory focused my
attention on the means by which choices are made by decision makers, the connection of
those choices to future options and sequences of events, and to outcomes. The use of the
theory clearly reinforces the idea that a variety of externalities influence the adoption of
technologies, but I am not convinced that the proper way to frame the questions is to rely
on the assumption that choices result essentially from historical accidents or insignificant
events. Many decisions are made without the ability to predict outcomes or to appreciate
potential alternatives. Context is extremely important. The miasmatic theory was not an
historical accident or an insignificant event. It was made significant because of its timing
and the related issues and concerns that produced the desire to develop new technologies of
sanitation in the nineteenth century. It is no little comfort that some economic historians,
economists, political scientists, and other social scientists believe that history matters. It is
for historians to help demonstrate what the phrase actually means in practice. Employing
path- dependence theory—or at least experimenting with it--can advance that effort.

A version appeared in Dietor Schott, et all, eds., Resources of the City (Hampshire, UK: Ashgate, 2005):
262-75.

P. A. Gorrigne, ‘Path Dependence –Causes, Consequences and Policy,’ in Grimes, Arthur, et al. eds. Economics for
Policy: Expanding the Boundaries: Essays by Peter Gorringe (Wellington, NZ 2001).

See D. J. Puffert, ‘Path Dependence in Economic History,’ November, 1999, Internet www.vwl.uni-muenchen.de/
ls_komlos/pathe.pdf , based on ‘Pfadabhangigkeit in der Wirtschaftsgeschichte,’ in Herrmann-Pillath, Karsten,
Lehmann-Waffenschmidt, and Marco, eds, Handbuch zur evolutorischen Okonomik (forthcoming). See also L.
Magnusson and J. Ottosson, eds., Evolutionary Economics and Path Dependence (Cheltenham, UK, 1997), 1-2.

P. A. David, ‘Path-Dependence: Putting the Past into the Future of Economics,’ Technical Report No. 533
(Stanford,CA Dec. 1988): abstract.

Ibid., 6.

The name of the keyboard was derived from the first six letters on the top row of the keys.

P. A. David, ‘Clio and the Economics of QWERTY,’ American Economic Review ,75, May, 1985, 332-37. See also
David, ‘Understanding the Economics of QWERTY: The Necessity of History,’ in W.N. Parker, ed., Economic
History and the Modern Economist (New York 1986); David, ‘Path Dependence and the Quest for Historical
Economics: One More Chorus of the Ballad of QWERTY,’ University of Oxford Discussion Papers in Economics
and Social History, No. 20, 1997; David, ‘Path-dependence and Predictability in Dynamic Systems with Local
Network Externalities: A Paradigm for Historical Economics,’ in D. Foray and C. Freeman, eds., Technology and
the Wealth of Nations: The Dynamics of Constructed Advantage (London, 1993).

A well-known rebuttal is S. J. Liebowitz and S. E. Margolis, ‘The Fable of the Keys,’ Journal of Law and
Economics, 33 , 1990, 1-25.

Puffert, ‘Path Dependence in Economic History,’ 2.

P. A. David, ‘Path Dependence, Its Critics and the Quest for “Historical Economics,”’ 1, Internet, www-
econ.stanford.edu/faculty/workp/swp00011.pdf.

Ibid., 10.

David, ‘Path-Dependence: Putting the Past into the Future of Economics,’ 9.

David, ‘Path Dependence, Its Critics and the Quest for “Historical Economics,”’16.

W. B. Arthur, Increasing Returns and Path Dependence in the Economy (Ann Arbor 1994), 1.

Ibid., 28.

W. B. Arthur, ‘Competing Technologies, Increasing Returns, and Lock-In by Historical Events,’ Economic
Journal, 99, March, 1989, 116.

Ibid., 117, 123, 126-28. See also Arthur, ‘Competing Technlogies and Economi Prediction,’ Options,
(Laxenburg, Austria 1984), 10-13; Arthur, ‘Self-reinforcing Mechanisms in Economics,’ in P.W. Anderson, K.J.
Arrow and D. Pines, eds., The Economy as an Evolving Complx System (Redwood City, CA 1988); Arthur, ‘Path
Dependence, Self-reinforcement, and Human Learning,’ American Economic Review, 81, 1991, 133-58; Arthur,
‘Competing Technologies: An Overview,’ in G. Dosi, C. Freeman, R. Nelson, G. Silverberg, and L. Soete, eds.,
Technical Change and Economic Theory (London 1988), 590-607; Arthur, ‘Positive Feedbacks in the Economy,’
Scientific American , 262, 1990, 92-99.

D. C. North, Institutions, Institutional Change and Economic Performance (Cambridge, 1990), vii.

Ibid., 3.

Ibid., 131.

D. C. North, ‘Some Fundamental Puzzles in Economic History/Development,’ in W. B. Arthur, S. N. Durlauf, and

D. A. Lane, eds., The Economy as an Evolving Complex System II (Reading, MA 1997), 224-25.

Ibid., 228.

North, Institutions, Institutional Change and Economic Performance, 137-38.

P. Pierson, ‘Increasing Returns, Path Dependence, and the Study of Politics,’ American Political Science Review,
June 2000, 1. See also ‘Not Just What, but When: Timing and Sequence in Political Processes,’ Studies in
American Political Development, 14, 2000, 72-92; S. E. Robinson, ‘You Can’t Get There from Here: Path
Dependence, Evolutionary Economics, and Public Agencies,’ (Paper presented at the Scientific Study of the
Bureaucracy Conference, College Station, Texas, February 2-3, 2001), 2.

Ibid., 3.

Ibid., 18.

Ibid., 19, 22-23. Others have made a distinction between trajectories and turning points. The former are
“interlocked and interdependent sequences of events” whereas the latter are ‘events that have the potential to
redirect trajectories along new paths.’ Trajectories are inertial, but turning points can switch trajectories to new
paths. See L. A. and D. Harrison, ‘Technological Trajectories and Path Dependence,’ 4., Internet www.bath.ac.uk/
imp/pdf/18_AraujoHarrison.pdf.

S. J. Liebowitz and S. E. Margolis, Winners, Losers & Microsoft: Competition and Antitrust in High Technology
(Oakland, CA,, 1999), 1.

S. E. Margolis and S. J. Liebowitz, ‘Path Dependence,’ May 28, 2001, 1, Internet, HYPERLINK "http://
www.pub.utdallas.edu/~liebowitz/palgrave/palpd.html" www.pub.utdallas.edu/~liebowitz/palgrave/palpd.html.
See also Liebowitz and Margolis, ‘Path Dependence, Lock-In, and History,’ Internet, HYPERLINK "http://
www.utdallas.edu/~liebowitz/paths.html" www.utdallas.edu/~liebowitz/paths.html .

Liebowitz and Margolis, Winners, Losers & Microsoft, 39.

Ibid., 49-56.

Ibid., 235.

Email, John Conover, December 14, 1998, Internet HYPERLINK "http://www.johncon.com/john/

correspondence/981214003911.12734.html" www.johncon.com/john/correspondence/
981214003911.12734.html. Others support Liebowitz and Margolis with similarly strong language. In a review
of Winners, Losers & Microsoft appearing in Regulation, 23,------ emeritus professor of mineral economics
Richard L. Gordon noted: ‘Liebowitz and Margolis are overly polite is dealing with the fundamental problem of
the path-dependence model: its reliance on a long chain of improbable assumptions to prove that markets can
sustain bad choices.’

Puffert, ‘Path Dependence in Economic History,’ 1.

G.R. Hawke, Economics for Historians (Cambridge 1980), 3.

Ibid., 12-13. Causation is central to historians’ thinking, so much so that, as one study suggests, ‘historians
continuously use causal language. In fact, they are the only cognitive discipline in which it is correct to
emphasize the types of causation present in ordinary, undisciplined, common-sense discourse.’ See P. K. Conkin
and R. N. Stromberg, Heritage and Challenge: The History and Theory of History (Arlington Heights, IL, 1989),
170. For more on historical causation and conceptions of time, see 171-91; A. J. Lichtman and V. French,
Historians and the Living Past: The Theory and Practice of Historical Study (Arlington Heights, IL 1978),
44-72, 249-54; R. F. Berhofer, Jr., A Behavioral Approach to Historical Analysis (New York 1969), 211-69; D. H.
Fischer, Historians’ Fallacies: Toward a Logic of Historical Thought (New York 1970), 164-86.

Magnusson and Ottosson, Evolutionary Economics and Path Dependence, 3.

Something may be past dependent, the argument goes, if events within a specific system can be predicted on the
basis of the state of the system at a later time, independent of how the system arrived at that time. See Araujo and
Harrison, ‘Technological Trajectories and Path Dependence,’ 2.

Pierson, ‘Increasing Returns, Path Dependence, and the Study of Politics,’ 8.

However, it might be useful at some future date to consider the impact of path dependence theory on the Kuhnian
notion of paradigms. Path dependence suggests evolutionary change, while Thomas Kuhn made a case for
paradigm shifts in science—an idea that has been broadly applied to many fields. See Thomas Kuhn, The
Structure of Scientific Revolutions (Chicago 1962).

The Sanitary City: Urban Infrastructure in America from Colonial Times to the Present (Baltimore
2000).
PAGE

PAGE 10

The Sanitary City: Urban Infrastructure in America from Colonial Times to the Present

(Baltimore: John Hopkins University Press, 2000).