Transaction Costs, Institutions, and Economic History

Author(s): Douglass C. North

Source: Zeitschrift fr die gesamte Staatswissenschaft / Journal of Institutional and
Theoretical Economics, Bd. 140, H. 1., The New Institutional Economics: A Symposium
(Mrz 1984), pp. 7-17
Published by: Mohr Siebeck GmbH & Co. KG
Stable URL: http://www.jstor.org/stable/40750667
Accessed: 29-08-2017 12:23 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Mohr Siebeck GmbH & Co. KG is collaborating with JSTOR to digitize, preserve and extend
access to Zeitschrift fr die gesamte Staatswissenschaft / Journal of Institutional and
Theoretical Economics

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
 Zeitschrift fr die gesamte Staatswissenschaft (ZgS) 140 (1984), 7-17
Journal of Institutional and Theoretical Economics

Transaction Costs, Institutions, and Economic History

by
Douglass C. North

Since the cliometric revolution the building blocks used in the study of
nomic history have been those of neo-classical economics. That body of
was designed to explain and analyze the allocation of resources at a mom
of time under the severely restrictive test condition of zero transaction
In its pristine form, it assumed perfectly operating markets. While recent
sions have devoted attention to positive information costs, uncertainty
transaction costs, they are nowhere integrated into a general framewor
allows us to analyze the changing structure of economies over time,
i what economic history is all about. In this paper, I shall propos
a framework and suggest its implications for rethinking our views on eco
history1.

The bedrock of all economic theory is the gains from trade. From Adam
Smith on, economists have recognized that it is specialization and division
of labor that underlies the productivity of economies and permits and encour-
ages the growth of new technologies and more efficient forms of production.
But this theoretical foundation of economics is really half a theory. It ignores
the costs of trade - that is, the losses that arise because of specialization
and division of labor. Transaction costs are the costs of specifying and enforc-
ing the contracts that underlie exchange and therefore comprise all the costs
of political and economic organization that permit economies to capture the
gains from trade. How big a percentage of gross national product are these
costs? No one has yet measured them, but even a casual examination of
national income accounts suggests that today in advanced western countries
they may comprise as much as 50 percent of GNP. Moreover, they have
grown substantially in the last century. Since it is the combination of produc-
tion and transaction costs that have determined the output of political/eco-
nomic systems throughout history, a theory of transaction costs must be inte-

1 This essay is both a condensation and extension of the argument advanced in

North [1981].

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
8 Douglass C. North ZgS

grated with production theory

economic historian. The connecti
nomic institutions that are sha
are contractual arrangements b
made to maximize their wealth b
of specialization (including specia
ence between a principal and an
authority to a principal - that
or her own decision-making to
ployers, managers, government
college professors who work fo
etc. The contracts may be implici
they must be defined and enforce
them that make up transactions
We may say that institutions
in the form of rules and regulat
from the rules and regulations;
norms which define the contour
and regulations are specified and
rests on three fundamental ass
individualistic behavioral assump
and enforcing the rules that und
tion that ideology modifies maxi
The behavioral postulate of weal
nomic theory. It is also the corne
This postulate assumes that indiv
mize at any and all margins. Tha
indeed, live in that classic Hob
brutish, and short. Constraints,
limiting certain types of behav
we must note carefully that w
economic theory, very frequentl
together and form an organizatio
of the theory of institutions, ho
key to understanding the form
gain by forming organizations t
and within an economic frame
detectbehavior, then individuals
they, themselves, agree are im
we recognize the individual beh

2 Because of the costs of agency, e

in this framework and, accordingly,
agreements that are capable of being

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
140/1 (1984) Transaction Costs, Institutions, and Economic History 9

problems and costs of organization mainly involve attempts to define and

enforce rules that it is in the individual's interest to disobey.
This brings us to the second assumption, that it is extremely costly to specify
and enforce the contractual rules that underlie institutions. We can begin
to understand this problem by reassessing a basic part of economic theory
which never was founded upon fact. Exchange in standard theory assumed
an instantaneous exchange of a uni-dimensional good or service. But in reality,
goods and services have multiple valued attributes from which we get utility;
and it is extremely costly to measure each attribute in contractual exchange.
Take something as simple as the purchase of oranges for orange juice. What
I would really like to purchase is a quantity of juice with a tart flavor and
a certain amount of vitamin C; but the amount of juice, the flavor and vitamin
C are all valued attributes that are costly to measure separately. Instead,
we buy oranges by number or by weight; and these are poor surrogates for
what we really are looking for. To the degree that any valued attribute is
too costly to be able to measure perfectly then, we have some dissipation
of income as a result of a failure to be able to measure it. When we move
from oranges to more complex commodities, such as automobiles or medical
services, the variety of valued attributes is obviously greater; and to the degree
that these attributes are measured imperfectly, the participants in exchange
have the opportunity to take advantage of each other. Therefore, what con-
tracts attempt to do is to specify as precisely as possible the valued attributes
involved in the exchange process; but as long as some of those are imperfectly
measured, then the problems that we have described will continually obtain.
If it is difficult to measure the valued attributes in exchange for goods and
services, the problems between principals and agents are even more serious.
The principal attempts to constrain the behavior of the agent in such a way
as to fulfill or to carry out the principal's objective of realizing the wealth
from the gains of trade. To the extent that the agent is not perfectly con-
strained, he may cheat, shirk, or otherwise engage in activities which may
be valuable to him but certainly not valuable to the principal. The problems
of constraining the agents and the costliness of measuring the output of the
agent in terms of the many aspects that are involved in the contract pose
the fundamental dilemma of agency and determine the shape of institutions
dealing with hierarchy.
The costs of measurement constitute the first part of the total costliness
of contractual arrangements. Second are the costs of enforcing contractual
arrangements. If we had perfect measurement - that is, we could clearly specify
what a contractual arrangement consisted of- then we would be in a position
to know always when a contract had been fulfilled, or when abrogated by
one party or the other. In that case, if enforcement were perfect, there would
be no problem. That is, if one party violated the contract and the damages
to the other party could be clearly measured, then the enforcement mechanism
of the courts or some other part of the judicial system would simply award

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
 10 Douglass C. North ZgS

to the injured party the amount o

tunism, which plays such a large
never exist.
But as
a matter of fact, enforceme
fect because it is costly to measur
therefore, parties can argue convi
contract; but, more seriously, it is
themselves are agents and, therefo
Judges, juries, mediators, arbitrat
in turn, influences the way they
is more to it than that. Even th
justice makes clear that judges,
also influenced by their view of t
which party has been injured, if t
justice view the contract as being
contractual arrangement itself sho
to the third assumption; and we
influences the framework of instit

//

No theory of institutions would be complete if it excluded ideology. The reason

is straightforward. It is costly to measure performance and people's behavior
and therefore performance can be influenced by their conviction about the
fairness or justice of contracts. Accordingly, we must try to understand how
people arrive at views about the fairness of institutional arrangements. We
begin by recognizing that ideology is ubiquitous: individuals have a need
to rationalize the world around them. In that sense, ideology, as a device
by which to deal with the multiple problems of everyday living that confront
one, economizes on the amount of information that people must have. But
it is more than that - it also involves a judgment about the fairness or legiti-
macy of the contractual arrangements or the institutional arrangements within
which individuals live and act.
Ideologies are derived from the experiences people have. If all people have
the same experiences, we could expect that their perspectives about the world
around them would tend to be the same. Consensus ideologies, such as one
will find in tribal societies, exist because the members of the society essentially
have gone through the same experiences and, therefore, have a common set
of beliefs, myths, taboos that guide their views and perceptions. Divergent
ideologies grow up because individuals have diverse experiences that lead them
to have diffrent views about the world around them. Divergent views stem
partly from geographical and occupational specialization. Ideology thus hinges
upon the very same theoretical basis that underlies the gains from trade -

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
140/1 (1984) Transaction Costs, Institutions, and Economic History 1 1

that is, the gains from trade arise because of specialization and division of
labor, but divergent ideologies, conflicting views about the justice and fairness
of the system, also emerge from these sources. They may emerge from geo-
graphic specialization in that geographically different groups evolve different
customs, traditions, languages as they confront different problems. Or they
may emerge from occupational specialization, which results in impersonal ex-
change and a loss of personalized relationship between employer and employee.
The implications of ideological consensus or ideological diversity for our
modeling of institutions should be clear. To the degree that the members
of a society have the same ideological framework, the formal rules of the
society that define the constraints making up institutions will not have to
be defined very clearly and enforcement mechanisms and procedures may
be minimal or even absent altogether. But to the degree that society has diverse
ideologies reflecting the growth of specialization and division of labor, more
resources will have to be devoted, first to defining the rules precisely, and
second to enforcing those rules. Such definition and enforcement is necessary
because, with conflicting ideologies, the individual participants will feel no
necessity to constrain individual maximization (cheating, shirking, etc.) at the
expense of the other party. Given the costliness of measuring performance,
ideological consensus or alienation is a fundamental influence upon the form
of institution.
We can summarize this very brief framework for theorizing about institu-
tions as follows: The gains from trade can only be realized by the establishment
of contracts between principals or principals and agents which form the institu-
tional framework of a political/economic system. Therefore, without some
form of political order, organized trade cannot develop. In effect, transaction
costs are so high that they preclude economic organization. With simple poli-
tical and economic organization, custom may be sufficient to specify the struc-
ture with a consensus ideology. As the gains from trade that arise from speciali-
zation and division of labor result in more and more complex forms of political
and economic organization, the resource costs of capturing those gains from
trade will increase. They will increase not only because the costs of measure-
ment and enforcement of an ever-lengthening chain that result from greater
and greater specialization will grow, but also because ideological alienation
will characterize the relations between participants in the exchange process
and will require greater and greater amounts of resources devoted to specifica-
tion and enforcement of contracts.

///

Let us now apply this analytical framework to the economic history of the
western world. That history has been shaped by two fundamental revolutions.
The first was the development of agriculture, which appears to have first
occurred in the Middle East about ten millenia ago and then developed inde-

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
12 Douglass C. North ZgS

pendently in other parts of the

genesis in the expansion of the eig
but really occurred in the mid-nin
and technology. These were funda
altered the population/resource
of population with at least constan
vided the basis for the unmatched
Realizing their potential, however
political and economic organization
The first economic revolution pro
tated a much more complicated soc
a set of exclusive property right
the community action. The develo
animals and plants involved some f
agriculture itself required that de
plant, when to plant, when to harv
necessary between planting and ha
about who would perform the task
tunity to store goods against perio
of goods necessitated coordination
gation was practiced, this, too, req
ment of tasks and performance.
The institutions which developed
evolved out of the communal decis
Such simple decision-making wou
emerging agricultural communitie
would recognize his or her stake
such community decisions would
could be easily detected. However,
nity would tend to put the commu
Agricultural productivity would t
to store food would reduce the con
changes would tend to favor pop
growth would increase both the
the incentive for any individual to
the measurement of shirking wou
would also increase the costs of co
buting the community product.
The result of all this was the crea
body of rules to order internal stru
those rules and to compete with
fundamental achievement of the an
on its coercive power; and as lon
buted, the rulers of the state can

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
140/1 (1984) Transaction Costs, Institutions, and Economic History 13

interests. The state is an essential prerequisite for capturing the gains from
trade, but is also the source of exploitation. During the eight millenia from
the development of agriculture to the Roman Empire, there were periods
of sustained growth when the interests of the ruler coincided with productivity
increase: Athens in the fifth century, Rhodes in the fourth and third centuries,
and the Roman Empire in the first two centuries A.D.
But if the state is the necessary prerequisite for economic growth, it is
also the source of man-made economic decline. Ultimately, Athens fell to
Sparta; Rhodes was reduced by the competition and jealousy of Rome; and
the Roman Empire itself finally fell in the fifth century A.D. The inherent
instability of the state has been a basic characteristic of history, but at least
in the millenia to the end of the Roman Empire in the fifth century, A.D.,
it was the basic source of economic growth - at the price of increasing re-
sources going to political organizations, and persistent exploitation by those
in control of coercive power.
The second economic revolution, in contrast, was based on an elastic supply
curve of new knowledge which could be and was translated into the scientific
and technological advances of the modern world. It made the underlying as-
sumption of neoclassical economics realizable. That optimistic assumption
was that new knowledge could be produced at constant cost and that substitu-
tion at all margins made possible sustained growth even with the explosive
population increase that coincided. The price, however, has been an enormous
increase in the resources devoted to transacting and political and economic
institutions that are inherently even more unstable. Let us first look at the
economic consequences of this second economic revolution and then we shall
turn to some of the political consequences.
The technology of the second economic revolution was characterized by
significant indivisibilities in the production process with large fixed capital
investment. The realization of the potential scale economies required large
volume, continuous production and distribution. Alfred Chandler [1977],
has eloquently described the growing complex process of modern mass produc-
tion and distribution which occurred in the late nineteenth century. However,
this process required both occupational and territorial specialization and divi-
sion of labor on an unprecedented scale. The greater the specialization, the
greater the number of exchanges in the production process. Home production
is characterized by complete vertical integration. In this case, there are no
measurement costs, since each person tailors production to his or her own
needs, but the price is the productivity gains from specialization. The second
economic revolution produced just the reverse. It resulted in exponential mul-
tiplication of exchanges, with immense gains in productivity but at the price
of growing transactions costs arising from this exchange. Obviously, the pro-
ductivity gains from specialization have exceeded the increasing transactions
costs. Hence, the quantum leap in living standards that made possible the
modern western world.

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
 14 Douglass C. North ZgS

But the transactions costs associat

amounts of resources. We have on
in society from the late nineteent
1970, the United States labor force
workers, however, increased only
collar workers increased from 5 million to 38 million. While not all white
collar workers were concerned solely with transacting, the great bulk of them
were. And that's not all. If the coordination and integration of the production
process involves an ever-growing proportion of the labor force within manu-
facturing firms, the second economic revolution equally fostered a growing
number of firms specializing in transacting from producer to consumer. These
include finance, banking, insurance, auditing, accounting, etc. In addition gov-
ernment, also primarily concerned with transacting, grew immensely: in this
period from 1 million to 12.5 million employees.
The underlying forces changing the economy were the growing problems
of quality control at each step in the lengthening production chain, and the
growing problems of labor discipline and bureaucracy that accompanied this
radical change in production. Much of the new technology, indeed, was
designed to reduce the attendant transactions costs by substituting capital
for labor, by reducing the degrees of freedom of the worker in the production
process, or by automatically measuring the quality of the intermediate goods.
Underlying problems were, first, the problem of measuring inputs and out-
puts so that one could ascertain the contribution of individual factors input.
Equally, there was room for conflict over the subsequent payment to factors.
Additionally, not only were there residual unpriced outputs, such as waste
and pollutants, but there were complicated costs of specifying the desired
property for the goods or services produced at each stage in the production
process.
Second, there was the problem of large, fixed capital investments that had
long life and low scrap value and that required exchange relationship in con-
tractual agreements over long periods of time. During these periods, there
were uncertainty about prices and costs and abundant possibilities for oppor-
tunistic behavior on the part of one or both parties to exchange.
The ever increasing cost of measuring the quality of goods and services
produced has resulted in increasing resources devoted to sorting, grading,
labeling, trademarks, warranties and licensing - all costly devices to measure
the characteristics of goods and services. Yet despite the resources devoted
to measuring the quality of goods and services, the dissipation of income
is evident all around us: in the difficulty of measuring automobile repairs,
of evaluating the safety characteristics of products or the quality of medical
services or of measuring educational output. A major political consequence
has been the demand for government intervention to provide quality stand-
ards. In addition, while team production permitted economies of scale, it did
so at the cost of increasing shirking. The discipline of the factory system

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
140/1 (1984) Transaction Costs, Institutions, and Economic History 15

is nothing more than a response to the control problem of shirking in team

production. From the perspective of the employer, the discipline consists of
rules and regulations, incentives and punishments. Innovations, such as Tay-
lorism, were methods of measuring individual performance. From the point
of view of the worker, they were inhuman devices to foster speed up and
exploitation. Since there was no agreed upon measure of output that constitute
contract performance, both were right.
In addition, the gains from opportunistic behavior were equally enhanced
and led to strategic behavior, both within the firm (in labor-employer relations,
for example) and in contractual behavior between firms. Everywhere in pro-
duct and factor markets the gains from withholding services were altering
the terms of agreement at strategic points. Vertical integration, horizontal
integration, bonding of the participants were efforts to limit such activity.
An increasing appeal to government to act as a third party to contracts under-
lay a great deal of the growth of regulation. There were still further costs -
those associated with the familiar problems of bureaucracy. If the multiplica-
tion of rules and regulations inside large organizations is a device to reduce
shirking and opportunism, the deadweight losses associated with bureaucracy
are so familiar they require no further elaboration here.
Then there were the external effects - the unpriced benefits and costs -
that were a consequence of employing this technology. They too are a familiar
story. The growth of the business firm was a method of internalizing unpriced
benefits. The unpriced costs are reflected in the modern environmental crisis,
in which the problem of measuring and reducing them have both altered
voluntary organization and induced the growth of government intervention
in the twentieth century.
Let us now turn to looking at the over-all political economic consequences.
If the second economic revolution ushered in an era of unequal prosperity
in the western world, it also induced a massive reaction against market econo-
mies and market forms of resource allocation. We observe in the late nine-
teenth and the twentieth centuries the growth of labor movements predomin-
antly dominated by socialist and communist ideologies. Even in economies
where markets tend to dominate them, they have fostered ever-increasing
growth of government intervention and regulation. Third World countries
have shown little enthusiasm for market forms of resource allocation. What
has made the market system tend to self-destruct? It is clear that control
of the state was for a brief period in the hands of groups whose self-interest
promoted the growth of market forms of resource allocation. But it is also
clear that control of the state eventually passed into the hands of groups
who favored elimination or at the very least modification of market forms
of resource allocation. I advance two hypotheses to explain this phenomenon.
The first is that market competition induced massive alienation because of
the particular characteristics of the exchange relationship that energized
groups to overcome the free rider problem and gain control (or at least partici-

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
 16 Douglass C. North ZgS

pate in control) of the state. The s

interest groups to shield themselv
to alter property rights and hence
Let us examine the first hypothe
factor and product markets was an
ductive potential of the second e
massive ideological alienation; and
an ideological superstructure to leg
cal alienation has produced imme
world of personalized exchange th
at an earlier time, the growth o
diverse perceptions of reality whi
ideologies. Informal agreements
A formal set of rules evolved to c
it also created conditions in whi
those rules. Traditional status rela
integrity" were replaced by ubiqu
In particular, the inability to mea
induced conflict over what constit
quences of occupational specializ
down the communication and pe
a consensus ideology and to prod
and conflicting perceptions of rea
occupational specialization. Alienat
trol of the state, to alter their term
The second hypothesis suggests th
ket come from the inherent instab
became worldwide as transporta
led to sharp fluctuations in the ter
to devote resources to attempting
state in the interests of reducing c

IV

The structural transformation of western economies in the past century as

a consequence of the second economic revolution has been the subject of
a vast literature by social scientists. But this literature has been incomplete
because of the failure to understand the basic nature of the transactions cost
growth that has accompanied this tremendous productivity increase. Marxists
have claimed that it was the result of capitalism, but even the most cursory
examination of the problems of the Soviet Union and other socialist societies
makes clear that they face the same problems of cheating, shirking and aliena-
tion that characterize the western world. It is not capitalism, but the growth

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms
140/1 (1984) Transaction Costs, Institutions, and Economic History 17

of a world of specialization and division of labor, that has produced the

problems that we confront. The price of civilization is the growing costs of
political and economic organization and the instability that has accompanied
them. Until we understand them and provide an analytical basis for being
able to realize their inherent nature, we shall not be able to deal with them
adequately. That is the task that confronts the modern social scientist today.

References

North. D.C. f 198 11. Structure and Chanze in Economic Historv. New York.
Chandler, A. [1977], The Visible Hand, Cambridge.

Professor Douglas C. North

Department of Economics
Washington University
St. Louis, Miss. 63130
U.S.A.

This content downloaded from 196.43.135.150 on Tue, 29 Aug 2017 12:23:26 UTC
All use subject to http://about.jstor.org/terms