What Are Institutions? Geoffrey M. Hodgson
The use of the term institution has become widespread in the social sciences in recent
years, reflecting the growth in institutional economics and the use of the institution con-
cept in several other disciplines, including philosophy, sociology, politics, and geogra-
phy. The term has a long history of usage in the social sciences, dating back at least to
Giambattista Vico in his Scienza Nuova of 1725. However, even today, there is no una-
nimity in the definition of this concept.
Furthermore, endless disputes over the definitions of key terms such as institution
and organization have led some writers to give up matters of definition and to propose
getting down somehow to practical matters instead. But it is not possible to carry out any
empirical or theoretical analysis of how institutions or organizations work without hav-
ing some adequate conception of what an institution or an organization is.
This paper proposes that those that give up are acting in haste; potentially consen-
sual definitions of these terms are possible, once we overcome a few obstacles and diffi-
culties in the way. It is also important to avoid some biases in the study of institutions,
where institutions and characteristics of a particular type are overgeneralized to the set
of institutions as a whole. This paper outlines some dangers with regard to an excessive
relative stress on self-organization and agent-insensitive institutions.
This paper draws on insights from several academic disciplines and is organized in
six sections. The first three sections are devoted to the definition and understanding of
institutions in general terms. The first section explores the meaning of key terms such as
institution, convention, and rule. The second discusses some general issues concerning
how institutions function and how they interact with individual agents, their habits,
The author is a Research Professor in Business Studies at The Business School, University of Hertfordshire, De HavillandCampus, Hatfield, U.K. He is very grateful to Margaret Archer, Ana Celia Castro, Raul Espejo, Ronaldo Fiani, JaneHardy, Anthony Kasozi, Uskali Mäki, Douglass North, Pavel Pelikan, John Searle, Irene van Staveren, Viktor Vanberg,anonymous referees, and others for comments and discussions. This essay also makes use of some material from Hodgson
2006, Journal of Economic Issues
and their beliefs. The third examines the difference between organizations and institu-
tions and what may be meant by the term formal when applied to institutions or rules, by
focusing on some of Douglass North’s statements on these themes. The fourth identifies
an excessive bias in the discussion of institutions toward those of the self-organizing
type, showing theoretically that these are a special case. The fifth argues that institutions
also differ with regard to their degree of sensitivity to changes in the personalities of the
agents involved. Finally, the sixth concludes the essay. On Institutions, Conventions, and Rules
Institutions are the kinds of structures that matter most in the social realm: they
make up the stuff of social life. The increasing acknowledgement of the role of institu-
tions in social life involves the recognition that much of human interaction and activity
is structured in terms of overt or implicit rules. Without doing much violence to the rel-
evant literature, we may define institutions as systems of established and prevalent social
rules that structure social interactions.1 Language, money, law, systems of weights and
measures, table manners, and firms (and other organizations) are thus all institutions.
Fol owing Robert Sugden (1986), John Searle (1995), and others, we may usefully
define a convention as a particular instance of an institutional rule. For example, all
countries have traffic rules, but it is a matter of (arbitrary) convention whether the rule is
to drive on the left or on the right. So in regard to the (say) British institutional system of
traffic rules, the specific convention is to drive on the left.2
At some stage we need to consider how institutions structure social interactions
and in what senses they are established and embedded. In part, the durability of institu-
tions stems from the fact that they can usefully create stable expectations of the behavior
of others. Generally, institutions enable ordered thought, expectation, and action by
imposing form and consistency on human activities. They depend upon the thoughts
and activities of individuals but are not reducible to them.
Institutions both constrain and enable behavior. The existence of rules implies con-
straints. However, such a constraint can open up possibilities: it may enable choices and
actions that otherwise would not exist. For example: the rules of language allow us to
communicate; traffic rules help traffic to flow more easily and safely; the rule of law can
increase personal safety. Regulation is not always the antithesis of freedom; it can be its
ally. AsAlanWel s(1970,3)putit,“Socialinstitutionsformanelementinamoregen-
eral concept, known as social structure.” The original institutional economists, in the
tradition of Thorstein Veblen and John R. Commons, understood institutions as a spe-
cial type of social structure with the potential to change agents, including changes to
However, some institutionalists such as John Fagg Foster (1981, 908) have mislead-
ingly defined institutions as “prescribed patterns of correlated behavior.”3 Defining
institutions as behavior would mislead us into presuming that institutions no longer
existed if their associated behaviors were interrupted. Does the British monarchy cease
to exist when the members of the royal family are all asleep and no royal ceremony is tak-
ing place? Of course not: royal prerogatives and powers remain, even when they are not
enacted. It is these powers, not the behaviors themselves, which mean that the institu-
tion exists. Nevertheless, such powers may lapse, and institutional dispositions may
fade, if they are not exercised with sufficient frequency. Furthermore, the only way in
which we can observe institutions is through manifest behavior.4
Not all social structures are institutions. Social structures include sets of relations
that may not be codified in discourse, such as demographic structures in animal species
or in human societies before any understanding of demography. Demographic struc-
tures may limit social potentialities in terms of the number of infants or elderly requir-
ing care and the number of able-bodied adults available to care, produce, and procreate.
But they do not necessarily do this through the operation of rules.5
The term rule is broadly understood as a socially transmitted and customary norma-
tive injunction or immanently normative disposition, that in circumstances X do Y.6 A
prohibition rule would involve a large class of actions Y, from which the prohibited out-
comes are excluded. Other rules may involve requirements to perform a smaller class of
actions in Y. A rule may be considered, acknowledged, or followed without much
thought. The phrase immanently normative requires that if the rule is scrutinized or con-
tested, then normative issues will emerge.
The term socially transmitted means that the replication of such rules depends upon
a developed social culture and some use of language. Such dispositions do not appear
simply as a result of inherited genes or instincts; they depend upon contingent social
structures and may have no direct or obvious representation in our genetic makeup.
Rules include norms of behavior and social conventions as well as legal rules. Such
rules are potentially codifiable. Members of the relevant community share tacit or
explicit knowledge of these rules. This criterion of codifiability is important because it
means that breaches of the rule can be identified explicitly. It also helps to define the
community that shares and understands the rules involved.
The normative aspect of a rule would not be so relevant, and would have no com-
pelling reason to be passed on from generation to generation, if physical and natural cir-
cumstances allowed only one option Y* in circumstances X. If we were compelled by the
laws of nature to do Y* in circumstances X, then there would be no need for normative
compulsions or sanctions. In contrast, multiple options can typically be imagined for
the form of a rule. One culture may uphold in circumstances X do Y; another may
require in circumstances X do Z. Nevertheless, the laws of nature constrain the set of
possible rules that may be formulated. A feasible rule cannot ask us to defy the laws of
gravity or to become Julius Caesar. The set of possible rules can be enlarged by techno-
logical and other institutional developments. For example, the technology of writing
makes feasible the rule that a valid contract on paper must be signed.7
As Searle (1995, 2005) has argued, the mental representations of an institution or
its rules are partly constitutive of that institution, since an institution can exist only if
people have particular and related beliefs and mental attitudes. Hence an institution is a
special type of social structure that involves potentially codifiable and (evidently or
immanently) normative rules of interpretation and behavior. Some of these rules con-
cern commonly accepted tokens or meanings, as is obviously the case with money or lan-
guage. However, as Max Weber pointed out in 1907, some rules are followed “without
any subjective formulation in thought of the ‘rule’” (1978, 105). For example, few of us
could specify ful y the grammatical rules of the language that we use regularly or com-
pletely specify in detail some practical skills. Nevertheless, institutional rules are in prin-
ciple codifiable, so that breaches of these rules can become subjects of discourse.
Even with this criterion of potential codifiability, a problem arises as to how far we
can stretch the meaning of the term rule in the definition of an institution. Friedrich
Hayek (1973, 11), for example, emphasized that “[m]an is as much a rule-following ani-
mal as a purpose-seeking one.” However, his notion of a rule was extremely broad. For
Hayek (1967, 67) the term rule is “used for a statement by which a regularity of the con-
duct of individuals can be described, irrespective of whether such a rule is ‘known’ to
the individuals in any other sense than they normally act in accordance with it.” Hayek
(1979, 159) entertained rules that emanate from the “little changing foundation of
genetically inherited, ‘instinctive’ drives” as well as from reason and human interaction.
For Hayek, therefore, a rule is any behavioral disposition, including instincts and hab-
its, which can lead to “a regularity of the conduct of individuals.”
This excessively broad definition would include such behavioral regularities as
breathing or the pulsation of the heart. This stretches the notion of rule following to
unacceptable extremes (Kley 1994). Despite Hayek’s general emphasis on purposeful
behavior and his rejection of behaviorist psychology, Hayek ended up with a definition
of rule that hinges solely on behavioral regularities, here neglecting the ontology of rules
and the mechanisms involved in their creation and replication.
Essentially, social rules are replicated through mechanisms other than the genes.
However, while rules are not in the DNA, it would be a mistake to go to the other
extreme and regard rule following as something entirely deliberative. Michael Polanyi
(1967) argued convincingly that there is always and unavoidably a tacit substratum of
knowledge that can never be fully articulated, even with the most deliberative of acts.
Rules, to be effective in the social context, can never be purely or fully matters of
The tacit dimension of knowledge creates a problem when we attempt to draw the
line between instinctive, or autonomic, behavioral regularities on one hand and genu-
ine rule following on the other. Some authors refer to the latter but not the former cate-
gory of behavior as “intentional.” A problem here is that the concept of intentionality is
sometimes stretched to cover cases of behavior that are not deliberative (e.g., Bhaskar
1989; Searle 1995; Lawson 1997). Arguing that such an unconscious “intentional” state
“has to be in principle accessible to consciousness” (Searle 1995, 5) creates boundaries
for this enlarged concept of intentionality but extends its territory to some autonomic or
instinctive behaviors, such as breathing and blinking (but not heartbeats), which to
some degree on some occasions can be placed under conscious control. Searle’s crite-
rion would thus suggest that breathing and blinking were always intentional.8
An alternative strategy, preferred by the present author, consists of two elements.
First, the concept of intentionality is reserved for conscious prefiguration and self-reflex-
ive reasoning, with regard to future events or outcomes. As Hans Joas (1996, 158) put it,
intentionality “consists in a self-reflective control which we exercise over our current
behavior.” Unintended acts lack any such conscious deliberation and prefiguration.
Second, rules are regarded as socially or culturally transmitted dispositions, with actual
or potential normative content. An often serviceable test of socio-cultural rather than
genetic transmission is the potential or actual existence of very different rule systems,
even in similar natural environments.
Raimo Tuomela (1995) made a distinction between rules and norms, depending
on the manner of their enforcement. To establish this, he developed a notion of collec-
tive intentionality similar to that of Searle (1995).9 Collective intentionality arises when
an individual attributes an intention to the group in which he or she belongs while hold-
ing that intention and believing that other group members hold it, too. We act thus
because we believe that others have a similar aim. Clearly, many behavioral regularities
develop in society because of such reciprocating intentions and expectations. Tuomela
described such regularities as norms. They involve a network of mutual beliefs rather
than actual agreements between individuals. Norms involve approval or disapproval. In
contrast, for Tuomela, rules are the product of explicit agreement brought about by
some authority, and they imply sanctions. Rules and norms thus differ by virtue of the
different ways they enforce tasks on individuals.
However, such a hard and fast distinction is difficult to maintain. Reciprocating
mutual beliefs become explicit agreements with the addition of single and shared signs
or words of assent. Some behavioral regularities may emerge originally without external
enforcement, but later some external authority may impose sanctions. The difference
between such enforced sanctions and the perceived threat of disapproval by others is
eroded when one considers that both involve some discomfort for the individual con-
cerned. Sugden (2000) went further down this road, arguing that both are explicable in
terms of preferences alone. But even if we reject the utilitarian conflation of values and
preferences, neither external sanctions nor social disapproval is devoid of questions of
value. External sanctions and laws have a capacity to promote their own moral author-
ity, and their transgression may also involve social disapproval. People thus obey laws
not simply because of the sanctions involved but also because legal systems can acquire
the force of moral legitimacy and the moral support of others. On How Institutions Work
Generally, how do people understand rules and choose to follow them? We have to
explain not only the incentives and disincentives involved but also how people interpret
and value them. This appreciation and valuation of rules is unavoidably a process of
social interaction. As Ludwig Wittgenstein (1958, 80) pointed out, “a person goes by a
sign-post only in so far as there exists a regular use of sign-posts, a custom.”
Such considerations are important when we address the special case of legal rules.
For laws to become rules in the sense discussed here, they have to become customary. As
discussed later in this essay, there are examples of laws that are widely ignored and have
not acquired the customary or dispositional status of a rule. Ignored laws are not rules.
For new laws to become rules, they have to be enforced to the point that the avoidance
or performance of the behavior in question becomes customary and acquires a
Institutional economists in the Veblenian tradition, and modern and original prag-
matist philosophers, argue that institutions work only because the rules involved are
embedded in shared habits of thought and behavior (James 1892; Veblen 1899; Dewey
1922; Joas 1993, 1996; Kilpinen 2000). However, there has been some ambiguity in the
definition of habit. Veblen and the pragmatist philosophers regarded habit as an
acquired proclivity or capacity, which may or may not be actually expressed in current
behavior. Repeated behavior is important in establishing a habit. But habit and behav-
ior are not the same. If we acquire a habit we do not necessarily use it all the time. A
habit is a disposition to engage in previously adopted or acquired behavior or thoughts,
triggered by an appropriate stimulus or context.10
Accordingly, the pragmatist sociologists William Thomas and Florian Znaniecki
(1920, 1851) criticized “the indistinct use of the term ‘habit’ to indicate any uniformi-
ties of behavior. . . . A habit . . . is the tendency to repeat the same act in similar material
conditions.” Also treating habit as a propensity, William McDougall (1908, 37) wrote of
“acquired habits of thought and action” as “springs of action” and saw “habit as a source
of impulse or motive power.” As John Dewey (1922, 42) put it, “[t]he essence of habit is
an acquired predisposition to ways or modes of response.” Many habits are uncon-
scious. Habits are submerged repertoires of potential thought or behavior; they can be
triggered or reinforced by an appropriate stimulus or context.11
The acquisition of habits (or habituation) is the psychological mechanism that
forms the basis of much rule-following behavior. For a habit to acquire the status of a
rule, it has to acquire some inherent normative content, to be potentially codifiable,
and to be prevalent among a group. Persistent and shared habits are the bases of cus-
toms. William James (1892, 143) proclaimed: “Habit is thus the enormous fly-wheel of
society, its most precious conservative agent.”
The prevailing rule structure provides incentives and constraints for individual
actions. Channeling behavior in this way, accordant habits are further developed and
reinforced among the population. Hence the rule structure helps to create habits and
preferences that are consistent with its reproduction. Habits are the constitutive mate-
rial of institutions, providing them with enhanced durability, power, and normative
authority. In turn, by reproducing shared habits of thought, institutions create strong
mechanisms of conformism and normative agreement. As Charles Sanders Peirce
(1878, 294) declared, the “essence of belief is the establishment of habit.” Accordingly,
habit is not the negation of deliberation but its necessary foundation. Reasons and
beliefs are often the rationalizations of deep-seated feelings and emotions that spring
from habits laid down by repeated behaviors (Kilpinen 2000; Wood et al. 2002). This
interplay of behavior, habit, emotion, and rationalization helps to explain the norma-
tive power of custom in human society. Hence “custom reconciles us to everything”—as
Edmund Burke wrote in 1757—and customary rules can acquire the force of moral
authority. In turn, these moral norms help to further reinforce the institution in
Habits are acquired in a social context and not genetically transmitted. By accept-
ing the foundational role of habit in sustaining rule-following behavior, we can begin to
build an alternative ontology of institutions in which we avoid the conceptual problems
of an account based primarily on intentionality. This is not to deny the importance of
intentionality but to regard it as a consequence as much as a cause and to place it in the
broader and ubiquitous context of other, nondeliberative behaviors.12
By structuring, constraining, and enabling individual behaviors, institutions have
the power to mold the capacities and behavior of agents in fundamental ways: they have
a capacity to change aspirations instead of merely enabling or constraining them. Habit
is the key mechanism in this transformation. Institutions are social structures that can
involve reconstitutive downward causation, acting to some degree upon individual hab-
its of thought and action. The existence of reconstitutive downward causation does not
mean that institutions directly, entirely, or uniformly determine individual aspirations,
merely that there can be significant downward effects. Insofar as institutions lead to reg-
ularities of behavior, concordant habits are laid down among the population, leading to
congruent purposes and beliefs. In this way the institutional structure is further sus-
Because institutions simultaneously depend upon the activities of individuals and
constrain and mold them, through this positive feedback they have strong self-reinforc-
ing and self-perpetuating characteristics. Institutions are perpetuated not simply
through the convenient coordination rules that they offer. They are perpetuated
because they confine and mold individual aspirations and create a foundation for their
existence upon the many individual minds that they taint with their conventions.
This does not mean, however, that institutions stand separately from the group of
individuals involved; institutions depend for their existence on individuals, their inter-
actions, and particular shared patterns of thought. Nevertheless, any single individual is
born into a pre-existing institutional world which confronts him or her with its rules
and norms.14 The institutions that we face reside in the dispositions of other individuals
but also depend on the structured interactions between them, often also involving mate-
rial artefacts or instruments. History provides the resources and constraints, in each case
both material and cognitive, in which we think, act, and create.
Accordingly, institutions are simultaneously both objective structures “out there”
and subjective springs of human agency “in the human head.” Institutions are in this
respect like Klein bottles: the subjective “inside” is simultaneously the objective “out-
side.” The institution thus offers a link between the ideal and the real. The twin con-
cepts of habit and institution may thus help to overcome the philosophical dilemma
between realism and subjectivism in social science. Actor and institutional structure,
although distinct, are thus connected in a circle of mutual interaction and
Commons (1934, 69) noted that “[s]ometimes an institution seems analogous to a
building, a sort of framework of laws and regulations, within which individuals act like
inmates. Sometimes it seems to mean the ‘behavior’ of the inmates themselves.” This
dilemma of viewpoint persists today. For example, North’s (1990, 3) definition of institu-tions as “rules of the game . . . or . . . humanly devised constraints” stresses the restraints
of the metaphorical prison in which the “inmates” act. In contrast, Veblen’s (1909, 626)
description of institutions as “settled habits of thought common to the generality of
men” seems to start not from the objective constraints but from “the inmates them-
selves.” However, as Commons hinted and Veblen (1909, 628–30) argued in more
depth, behavioral habit and institutional structure are mutually entwined and mutually
reinforcing: both aspects are relevant to the full picture. A dual stress on both agency
and institutional structure is required, in which it is understood that institutions them-
selves are the outcomes of human interactions and aspirations, without being con-
sciously designed in every detail by any individual or group, while historically given
institutions precede any one individual. Some Problems with Douglass North’s Exposition
Starting with a definition of institutions as socially embedded systems of rules, it is
evident that organizations are a special kind of institution, with additional features.
Organizations are special institutions that involve (a) criteria to establish their bound-
aries and to distinguish their members from nonmembers, (b) principles of sovereignty
concerning who is in charge, and (c) chains of command delineating responsibilities
However, in several influential statements by North, he has characterized institu-
tions and organizations in a different way. The purpose of this section is to expose some
difficulties in his account and to maintain my alternative definitions. These difficulties
concern North’s apparent distinctions (a) between institutions and organizations and
(b) between “formal rules” and “informal constraints.” North has been insufficiently
clear. Consequently, many people misinterpret him as suggesting that organizations are
not a type of institution. He is also misinterpreted as making a distinction between for-
mal and informal institutions. Strictly, North upholds neither of these distinctions. I
also argue that North has insufficiently elaborated the nature and functioning of social
rules that he rightly identifies as the essence of institutions. His emphasis on the
rule-like character of institutions is consistent with my definition, but I believe that
something else needs to be added. Concerning institutions in general, North (1990,
Institutions are the rules of the game in society or, more formally, are the
humanly devised constraints that shape human interaction. In consequence
they structure incentives in human exchange, whether political, social, or eco-
nomic. . . . Conceptually, what must be clearly differentiated are the rules from
the players. The purpose of the rules is to define the way the game is played. But
the objective of the team within that set of rules is to win the game. . . . Model-
ing the strategies and skills of the team as it develops is a separate process from
modeling the creation, evolution, and consequences of the rules.
North rightly insists that rules must be “clearly differentiated . . . from the players.” The
distinction between players and rules is similar to the distinction between agents and
structures, as discussed elsewhere (Archer 1995; Lawson 1997; Hodgson 2004). Struc-
tures depend upon agents, but the two are different and distinct. North (1994, 361) also
It is the interaction between institutions and organizations that shapes the insti-
tutional evolution of an economy. If institutions are the rules of the game, orga-
nizations and their entrepreneurs are the players. Organizations are made up of
groups of individuals bound together by some common purpose to achieve
North reasonably sees organizations as including political parties, firms, trade unions,
schools universities, and so on. People have interpreted North as saying that organiza-
tions are not institutions. But North does not actually write this. He simply establishes
his own primary interest in economic systems rather than the internal functioning of
individual organizations. He is not so interested in the social rules that are internal to
organizations because he wants to treat them as unitary players and focus on interactions
at the national or other higher levels.
There is nothing in principle wrong with the idea that under some conditions orga-
nizations can be treated as single actors, such as when there are procedures for members
of an organization to express a common or majority decision. As Barry Hindess (1989,
89) argued, organizations can be treated as social actors as long as “they have means of
reaching decisions and of acting on some of them.” James Coleman (1982) came to a
similar conclusion. Interestingly, the criteria that sometimes allow us to treat organiza-
tions as actors require an understanding of organizations as social systems with
However, a problem arises if we define organizations as actors. This would amount
to an unwarranted conflation of individual agency and organization. Organiza-
tions—such as firms and trade unions—are structures made up of individual actors, often
with conflicting objectives. Even if mechanisms for “reaching decisions and of acting on
some of them” (Hindess 1989) are ubiquitous, the treatment of an organization as a
social actor should not ignore the potential conflict within the organization. The treat-
ment of the organization as a social actor abstracts from such internal conflicts, but an
abstraction should not become a fixed principle or definition that would block all con-
siderations of internal conflict or structure.
Abstraction and definition are entirely different analytical procedures. When
mathematicians calculate the trajectory of a vehicle or satellite through space, they often
treat it as a singular particle. In other words, they ignore the internal structure and rota-
tion of the vehicle or satellite. But this does not mean that the vehicle or satellite is
North does not make it sufficiently clear whether he is defining organizations as play-
ers or regarding organizations as players as an analytical abstraction. This has created
much confusion, with other authors insisting that organizations should be defined as
players. However, North has made it clear in letters to the author (September 10, 2002,
and October 7, 2002) that he treats organizations as players simply for the purpose of
analysis of the socio-economic system as a whole and that he does not regard organiza-
tions as essentially the same thing as players in all circumstances. In saying that “organi-
zations are players,” North is making an abstraction rather than defining organizations
When North (1994, 361) wrote that organizations “are made up of groups of indi-
viduals bound together by some common purpose,” he simply ignored instances when
this may not be the case. He is less interested in the internal mechanisms by which orga-
nizations coerce or persuade members to act together to some degree. Crucially, these
mechanisms always involve systems of embedded rules. Organizations involve structures
or networks, and these cannot function without rules of communication, membership,
or sovereignty. The unavoidable existence of rules within organizations means that,
even by North’s own definition, organizations must be regarded as a type of institution.
Indeed, North has essentially accepted that organizations themselves have internal play-
ers and systems of rules, and hence by implication organizations are a special type of
institution (letter to the author October 7, 2002).15
As North acknowledged, it is possible for organizations to be treated as actors in
some circumstances and generally to be regarded as institutions. Individual agents act
within the organizational rule-system. In turn, under some conditions, organizations
may be treated as actors within other, encompassing institutional rule-systems. There
are multiple levels, in which organizations provide institutional rules for individuals,
and possibly in turn these organizations can also be treated as actors within broader
institutional frameworks. For example, the individual acts within the nation, but in
turn the nation can sometimes be treated as an actor within an international framework
Further ambiguities arise with North’s distinction between formal “rules” and
informal “constraints.” Some distinction between the formal and the informal is impor-
tant, but this distinction is attempted in different and confusing ways by various
authors. Some identify the formal with the legal and see informal rules as nonlegal, even
if they may be written down. In turn, if “formal” means “legal,” then it is not clear
whether “informal” should mean illegal or nonlegal (i.e., not expressed in law). Another
possibility is to make the formal/informal distinction one of explicit versus tacit rules.
Still another variant in the literature is to identify the formal with designed, and the
informal with spontaneous institutions, along the lines of Carl Menger’s famous dis-
tinction between pragmatic and organic institutions. We have at least three important
distinctions, not one. North, like many other writers, does not make his intended dis-
tinction between “formal” and “informal” sufficiently clear.
The picture is further complicated by North’s use of the different terms rule and con-straint. North (1990, 1991, 1994) has written most often of formal and informal con-
straints, rather than formal and informal rules, but he has not indicated why he
dropped the word rule and whether or not constraints are also rules. North has written
frequently of “formal rules” but not of “informal rules.” But some writers interpret
North as making a distinction between formal and informal rules (e.g., Schout 1991).
North’s (1994, 360) examples of “formal constraints” are “rules, laws, constitutions”
and of “informal constraints” are “norms of behavior, conventions, self-imposed codes
of conduct.” This suggests that rules are a special kind of formal constraint.
This creates a further problem for North. If all rules are formal, and institutions areessentially rules, then all institutions are formal. However, North (1995, 15) subsequently
redefined institutions in the following terms: “Institutions are the constraints that
human beings impose on human interactions.” By redefining institutions essentially as
constraints, rather than rules, this raised the question of a possible distinction between
formal and informal constraints. This 1995 definition of an institution neglects the
enabling aspect of institutions by emphasizing constraints alone. North (1997, 6) then
shifted back to a conception of institutions as “the rules of the game of a society.”
In correspondence with the present author, North identified “formal rules” with
legal rules “enforced by courts” (October 7, 2002; see appendix). In contrast: “Informal
norms are enforced usually by your peers or others who impose costs on you if you do
not live up to them.” Despite the persistent analytical emphasis in North’s work on the
power of informal and customary relations, his definitions dispose him to identify both
rules and institutions with “formal” (i.e., legal) regulations.16
This confinement of the concept of an institution to legal rule systems can be criti-
cized for excluding social orders that are not legally expressed from the category of an
institution. An exclusive emphasis on legal rules can downplay the existence of rules
and institutions that can also constrain and mold human behavior in significant ways.
Important examples include language and powerful social customs such as those per-
taining to class in Britain, caste in India, gender in numerous countries, and many other
phenomena. Some rules and institutions—such as language and some traffic conven-
tions—can emerge largely spontaneously as coordination equilibria, which are repro-
duced principally because it is convenient for agents to conform to them. To some
degree, moral beliefs, sanctions, and constraints operate in all these cases. Not all
powerful rules or institutions are decreed in law.
North rightly emphasizes “informal constraints” but does not admit the category of
informal rules. But all contingent constraints that derive from human action (rather
than the laws of nature) are essentially rules. Accordingly, there is not a clear line
between rules and constraints, as North suggests, and instead social constraints are
Furthermore, an overemphasis on the formal and legal aspects can overlook the
reliance of legal systems themselves on informal rules and norms. As Émile Durkheim
argued in 1893, “in a contract not everything is contractual” (1984, 158). Whenever a
contract exists there are rules and norms that are not necessarily codified in law. The
parties to the agreement are forced to rely on institutional rules and standard patterns of
behavior, which cannot for reasons of practicality and complexity be fully established as
laws. Legal systems are invariably incomplete and give scope for custom and culture to
North fully accepts the importance of the informal sphere and frequently discusses
the informal aspects of formal (i.e., legal) institutions. He emphasizes, for example, the
roles of ideology and custom. But he insufficiently acknowledges informal institutions
that are not decreed in law, including those that arise spontaneously, such as coordina-
tion equilibria. North (1990, 138) has rightly and additionally emphasized “informal
constraints” and the “cultural transmission of values” but unnecessarily confines his
definition of institutions to rules codified in law.
Whether we are dealing with formal or informal rules, we need to consider the ways
in which rules are enacted. While it does not necessarily have to enter into the definition
of an institution or rule, there has to be some account of how rule-systems affect individ-
ual behavior. Pointing to the incentives and sanctions associated with rules is insuffi-
cient because it would not explain how individuals evaluate the sanctions or incentives
involved. We also have to explain why they might, or might not, take incentives or
Clearly, the mere codification, legislation, or proclamation of a rule is insufficient
to make that rule affect social behavior. It might simply be ignored, just as many drivers
break speed limits on roads and many continental Europeans ignore legal restrictions
on smoking in restaurants. In this respect, the unqualified term rule may mislead us.
North fully acknowledges that mere rule proclamation is not enough. But in trying
to understand how behavior is fixed or changed, his attention sometimes shifts to the
“informal constraints” of everyday life. Of course, the informal sphere is vital, but ironi-
cally, according to North’s own definitions, “informal constraints” are not institutions
at all. I prefer a broader conception of institutions that accommodates the informal
basis of all structured and durable behavior. That is why I define institutions as durable
systems of established and embedded social rules that structure social interactions,
rather than rules as such. In short, institutions are social rule-systems, not simply rules.
The ambiguity of the terms formal and informal with regard to institutions and rules
suggests that these words should either be abandoned or used with extreme care. It may
be best to use more precise terms such as legal, nonlegal, and explicit instead.
While broadly subscribing to North’s definitions, Pavel Pelikan (1988, 372; 1992,
45) compared North’s “rules” to the “genotype” within the “phenotype” of the organiza-
tional structure.17 If rules are like genes, then this underlines the importance of consid-
ering their mechanisms of survival and replication and the way in which they can affect
individuals or organizations. Rules do not have the capacity to copy themselves directly;
they replicate through other psychological mechanisms. From a pragmatist perspective,
the gene-like entities behind rules are individual habits, because these habits are the con-
ditional, rule-like dispositions that marshal behavior. Rules generally work only because
they are embedded in shared habits of thought and behavior. Hence it is best to treat
habits rather than rules as social genotypes. Self-enforcement versus External Enforcement
With one possible exception, all institutions depend on other institutions. As
Searle (1995, 60) pointed out, “language is the basic social institution in the sense that
all others presuppose language, but language does not presuppose the others.” Language
is basic because all institutions involve social interaction and interpretation of some
kind. Accordingly, all institutions involve at least rudimentary interpretative rules.
This literature on self-organization and spontaneous orders provides the essential
insight that institutions and other social phenomena can arise in an undesigned way
through structured interactions between agents. The focus on self-organizing aspects of
the social system can be traced back to David Hume and Adam Smith, and it is a major
theme in the Austrian school of economics from Carl Menger to Hayek. This literature
shows that social order can emerge that is not itself an intention or property of any sin-
gle individual or group of individuals.
However, even self-organizing institutions require a (rudimentary) language so,
with the exception of language itself, the concept of self-organization must be qualified
by the acknowledgement of the prior and extrinsic organization of communicative or
Furthermore, the concepts of self-organization or spontaneous order are insuffi-
cient for an understanding of all institutions. Menger ( 1981) himself recognized
a distinction between “organic” (self-organizing) and “pragmatic” (designed) institu-
tions. But many subsequent authors ignore the latter to concentrate on the former.
Indeed, much of the existing literature on institutions exhibits an excessive emphasis on
the (albeit essential) idea of self-organization, to the detriment of other vital mechanisms
of institutional emergence and sustenance.
With institutions that are not self-organizing, there is a stronger dependence on
other institutions that are required to enforce the internal rules. We first examine some
typical mechanisms of self-organization and then move on to cite some cases where such
An archetypical self-organizing configuration is a coordination game. Coordina-
tion rules typically provide incentives for everyone to conform to the convention. Con-
sequently, a coordination equilibrium can be self-policing and highly stable. Language is
an example. In communication we have strong incentives and inclinations to use words
and sounds in a way that conforms as closely as possible to the perceived norm. Norms
of language and pronunciation are thus largely self-policing (Quine 1960).
Similarly, some (but not all) legal rules have a strong self-policing element. For
example, there are obvious incentives (apart from avoiding legal sanctions) to stop at red
traffic lights and to drive on the same side of the road as others. Although infringements
will occur, these particular laws can be partly enforced by motorists themselves, because
infringements can increase perceived personal risks.
A coordination equilibrium can be self-enforcing because not only does each player
lack any incentive to change strategy but also each player wishes that other players keep
to their strategy as well (Schotter 1981, 22–3). If agents have compatible preferences and
strategies in this sense, then coordination rules can often emerge spontaneously and be
self-reinforcing. Even if I prefer to drive on the left, when I find myself in a country
where driving on the right is the convention, then I will drive on the right, and others
will prefer that I do this. A coordination equilibrium has characteristics of stability and
self-enforcement, even when the equilibrium is not ideal for everyone involved.
However, coordination games are a special case. Contrasting configurations
include the famous prisoners’ dilemma game, which allegedly represents several types of
social situations, including the socially suboptimal but individually advantageous use of
private cars rather than public transport (Best 1982), the famous “tragedy of the com-
mons” (Hardin 1968), and aspects of the employment contract (Leibenstein 1982).
At least in a one-shot play of the prisoners’ dilemma game, each player has an incen-
tive to defect. The situation of mutual cooperation is not a Nash equilibrium because
each player can gain an advantage by shifting from cooperation to defection. The Nash
equilibrium is where both players defect but each player gets less than he or she would if
both players cooperated. A “spontaneous order” may emerge, but it is clearly suboptimal
Coordination rules are followed primarily because of convenience. By contrast,
suboptimal outcomes in the prisoners’ dilemma game raise normative questions in a
more acute manner. Although all rules involve costs and benefits, there is a big differ-
ence between following a rule simply because it is convenient to do so and following a
rule because of a normative belief. Viktor Vanberg (1994, 65) has rightly pointed out
that writers in the spontaneous order tradition—from Hume and Smith through
Menger to Hayek—acknowledge inadequately the additional moral and legal mecha-
nisms that are required for enforcement in noncoordination games. Walter Schultz
(2001, 64–6) stressed a similar distinction in his powerful discussion of the problem of
Until recently, the problem of enforcement has been relatively neglected in the lit-
erature. As noted above, some rules are largely self-enforcing. In contrast, laws that
restrict behavior, where there are substantial, perceived net advantages to transgression,
are the ones that require the most policing. Hence people frequently evade tax payments
or break speed limits. Without some policing activity, the law itself is likely to be
infringed, debased, and “brought into disrepute.”
For example, there are incentives to debase money. If they can evade detection,
individual agents have an obvious incentive to use less costly, poor quality, or faked ver-
sions of the medium of exchange. If such forgeries or debasements are allowed to
endure, then bad money will drive out the good. Money is not self-policing in the same
way as language and may require some external authority to enforce its rules, as Menger
himself acknowledged (Latzer and Schmitz 2002).
Self-policing mechanisms can be undermined if there is the possibility of unde-
tected variation from the norm and there is sufficient incentive to exert such variations.
Language and money differ in this respect. The argument for enforcement by a third
party such as the state is thus stronger in the case of money and some laws than in the
Attempts to explain the evolution of contract and private property in entirely spon-
taneous terms have failed. Some authors attempt to explain the enforcement of property
rights by means of such devices as trading coalitions (Greif 1993). Itai Sened (1997)
showed that property rights are not entirely self-reinforcing and some external institu-
tion such as the state is required to enforce them. With a larger number of actors, it is
more difficult for individuals to establish mutual and reciprocal arrangements that
ensure contract compliance (Mantzavinos 2001, chap. 8). If trading coalitions do
emerge, then these themselves take upon state-like qualities to enforce agreements and
protect property. In a world of incomplete and imperfect information, high transaction
costs, asymmetrically powerful relations, and agents with limited insight, powerful
institutions are necessary to enforce rights.
It is an open question as to whether another strong institution, apart from the state,
could fulfill this necessary role. I simply note that an important class of institutions
exists in which such institutions depend on other institutions in order to enforce effec-
tively their rules. In the real world, there are many examples where some institutions are
sustained and supported by others. The role of the state in enforcing law and protecting
property rights is but one example. A major agenda for enquiry is to explore the extent
of such complementarities and understand their mechanisms in depth. Agent Sensitive and Agent Insensitive Institutions
Here I introduce a different distinction with the terms agent sensitive and agent insen-sitive institutions. An agent sensitive institution is one in which the reigning equilibria or
conventions can be significantly altered if the preferences or dispositions of some agents
are changed, within a feasible set of personality types. This issue is best approached by
considering some examples of agent insensitive institutions.
In one of his earliest papers, Gary Becker (1962) demonstrated that behavior ruled
by habit or inertia is just as capable as rational optimization of predicting the standard
downward-sloping demand curve and the profit-seeking activity of firms. He showed
how the negatively inclined market demand curve could result from habitual behavior,
up against a moving budget constraint. A constraint means that agents, whether
super-rational or habit driven, have to stay on one side of the line. With agents of each
type, rotations in the budget constraint can bring about downward-sloping demand
curves, irrespective of whether agents, in his terms, are habitual or rational.19
Much later, Dhananjay Gode and Shyam Sunder (1993) showed that experiments
with agents of “zero intelligence” produce behaviors that differ little from those with
human traders. They suggested that structural constraints can produce similar out-
comes, whatever the objectives or behavior of the individual agents. As in Becker’s
model (1962), systemic constraints prevail over micro-variations. Ordered market
behavior can result from the existence of resource and institutional constraints and may
be largely independent of the “rationality,” or otherwise, of the agents. Structural con-
straints, not individuals, do much of the explanatory work. We thus face the possibility
of a study of markets that focuses largely on institutions and structures, to a degree inde-
pendent of the assumptions made about agents.20
These models suggest that ordered and sometimes predictable behavior can some-
times result largely from institutional constraints. The explanatory burden is carried by
system structures rather than the preferences or psychology of individuals. I describe
such cases as “agent insensitive” institutions because outcomes are relatively insensitive
to individual psychology or personality.
Partly on the basis of the Gode and Sunder results (1993), Philip Mirowski (2002)
argued that to understand markets we do not have to pay much attention to the psychol-
ogies, cognitive processes, or computational capacities of the agents involved. Instead he
treated the market itself as a computational entity. His arguments may apply to some
institutional structures, including some markets, and to that extent they are important
and worthwhile. But they do not constitute a general theoretical strategy unless agent
insensitivity is itself general among institutions.
What is common to the Becker (1962) and Gode and Sunder (1993) models is the
existence of hard and insurmountable (budget) constraints. They push the agents into
position and offer them few alternatives, whatever their inclinations. Hence these mod-
els are agent insensitive and the constraints do much of the explanatory work. Such
hard constraints do exist in reality, but they are a rather special case. Other institutional
constraints operate through disincentives or legal penalties. But in such cases it may be
possible to cross the line or break the law. The propensity to break rules or transgress
constraints will in part depend on the preferences and dispositions of each individual
agent. If the constraints were softer, then the agents would have more discretion and it
would be likely that the personalities of the agents would have to be taken into account.
By wrongly suggesting that agent insensitivity is a general case, Mirowski’s research strat-
egy carries the danger of a general conflation of agency into the institutional structure.
Consider the alternative possibility of relatively high incentives to conform to a
convention. A coordination game is ostensibly agent insensitive because the players
have an incentive to conform to the reigning convention, even if it is not their most
favored option. British drivers will drive on the right in America and continental
Europe, even if they find it easier to drive on the left. To a degree, such traffic conven-
tions are agent insensitive. However, a convention can be overturned if a sufficient
number of people defy it. As long as the benefits of coordination are finite, the possibil-
ity exists of a relatively extreme personality type that may be inclined to overturn the
prevalent convention. In the different case of hard constraints, all agents are required to
comply, whatever their inclinations.
In contrast to a system with hard constraints, a number of configurations are agent
sensitive. Consider, for example, a reigning pattern of cooperation in a repeated prison-
ers’ dilemma game resulting from a population dominated by units playing the tit-for-tat
strategy. However, they can be invaded by an influx of others who always cooperate. If
this occurs, then the consequent population of cooperators would clearly be vulnerable
to an invasion by a species that consistently defects. In turn, if this invasion were incom-
plete, or subject to a slight amount of error, then a new invasion of tit-for-tat players
could take advantage of the fact that consistent defection was not absolute. Each of
these outcomes is unstable (Kitcher 1987; Lindgren 1992). The prevailing conventions
are sometimes sensitive to the types of player that are involved. Another possibility of
agent sensitivity results from the existence of multiple (Nash) equilibria. Slight differ-
ences between the personalities of agents may matter if there is a choice between two or
If we introduce greater variance in personality and observe the stability of reigning
conventions, then we can gauge the agent sensitivity of the institutional set-up involved.
Institutions exhibit different degrees of agent sensitivity and insensitivity, and investiga-
tions should not be confined to extreme or particular types. Concluding Remarks
This essay has proposed some key definitions, as follows:
• Social structures include all sets of social relations, including the episodic and those
without rules, as well as social institutions.
• Institutions are systems of established and embedded social rules that structure
• Rules in this context are understood as socially transmitted and customary
normative injunctions or immanently normative dispositions, that in
• Conventions are particular instances of institutional rules. • Organizations are special institutions that involve (a) criteria to establish their
boundaries and to distinguish their members from nonmembers, (b) principles of
sovereignty concerning who is in charge, and (c) chains of command delineating
responsibilities within the organization.
• Habituation is the psychological mechanism by which individuals acquire
dispositions to engage in previously adopted or acquired (rule-like) behavior.
Some of the discussion in this essay has involved putting some flesh on these bare-bones
definitions, particularly in the case of institutions and rules. This involves the key con-
cept of habit, which is regarded as a key element in the understanding of how rules are
embedded in social life and how institutional structures are sustained.
Many writers attempt distinctions between “formal” or “informal” institutions or
rules. However, these terms have been used misleadingly and in different ways. Does the
term formal mean legal, written, explicit, codifiable, or something else? The ambiguities
surrounding these terms mean that they cannot be taken for granted. One is required to
specify more clearly what is meant in each case or use more transparent terms such as
legal, nonlegal, and explicit instead.
Generally, the idea that there is a dividing line between institutions that are entirely
“formal” on one hand and entirely “informal” institutions on the other is false, because
“formal” institutions (in any of the above senses) always depend on nonlegal rules and
inexplicit norms in order to operate. If laws or declarations are neither customary nor
embodied in individual dispositions, then—“formal” or not—they have insignificant
effects. They are mere declarations or proclamations, rather than effective social rules.
Some declarations simply codify existing customs. Other proclamations may eventually
become effective rules but only through additional powers, such as persuasion, legitimi-
zation, or enforcement. To put it differently, legal or “formal” institutions that do not
have strong “informal” supports are unsupported legislative declarations rather than
real institutions. This does not mean that legal rules are unimportant but that they
become important by becoming incorporated in custom and habit.
An attempt has also been made here to avoid some overgeneralizations concerning
the nature of institutions. In particular, while self-organization is an extremely impor-
tant phenomenon in both nature and society, it would be a mistake to suggest that all
institutions are of this type. It has been shown here that some institutional rules require
other institutions for their enforcement.
Second, while there are cases where institutional rules or constraints do much of
the explanatory work, and therefore the institutional outcomes are relatively insensitive
to the personalities or psychologies of the agents involved, these instances are not uni-
versal. To regard all institutions as agent insensitive is to lead to the further error of con-
flating individuals into the institutional structure, where the interplay of both is
required to understand how institutions are formed and sustained. Appendix A Extracts from Correspondence between Douglass C. North and Geoffrey M. Hodgson
“First of al . . . organizations, you say, are special institutions. I think that for cer-
tain purposes we can consider organizations as institutions, but for my purposes organi-
zations are to be separated out from institutions. That is, I am interested in the macro
aspect of organization, not in the internal structure of organization. If I was interested in
the latter . . . I would be interested in internal structure, governance and indeed al the
kind of internal problems of structure, organization and conflicts of interest. . . . I am
not interested in that. What I am interested in is that the actors in the process of overall
societal, political economic change, and . . . I can forget about the internal structure,
even how decisions are made internal to the organization and simply look on the entre-
preneurs of organizations as the key players in the process of institutional change. . . .
For certain purposes one can consider organizations as institutions but for the purposes
that I am dealing with—looking at the macro aspects of institutional change—I do not
have to; indeed, I do not want to. . . . As I said at the beginning, I think that we are real y
not in too much disagreement. I think the issue really is the kinds of questions that I am
asking which make me focus in a particular way.”
I understand fully that you are interested primarily in overall, socio-economic
change, rather than the internal structure of organizations. I believe that such a special
focus is entirely legitimate. And I am one of very many who warmly appreciate your
I would presume that you also believe it to be legitimate to study the internal struc-
ture of organizations. . . . What concerns me is the need for conversation and common
understanding between those . . . who concentrate on the internal structure or organiza-
tions and those (like yourself) who concentrate on overall socio-economic change. For a
conversation to take place there must be shared meanings of terms.
In this respect I find a statement like “organizations are players” to be potentially
problematic. If “organizations are players” means “I will treat organizations as players
for the purposes of my kind of analysis” then, given certain conditions, that would be
OK in my view. However, if “organizations are players” means “organizations are
defined as players” or “organizations are essentially the same as individuals or players”
then I would find these formulations to be misleading. . . . When you say “organizations
(a) “for the purposes of analysis of the socio-economic system as a whole it is
legitimate to treat organizations as if they are players”? or
(b) “organizations are essentially the same thing as players,” i.e. in all
If you follow (a)—and your letter of September 10th seems to lean in this
direction—then a definition of organization is still outstanding. So in this case, I
(a*) Would you accept a definition of organization that accepted that organizations
themselves had internal players and systems of rules, and hence organizations
I would very grateful for your help and clarification on these points. If your answers
to (a) and (a*) were in the affirmative then I would be in complete agreement
I would like also to turn to a second issue where I am still unclear. This concerns the
distinction between the formal and the informal. I also think that this issue is
important but I think that there is much confusion in the broad literature on
(c) Does the formal/informal dichotomy refer to a distinction between legal and
nonlegal, or a distinction between explicit and tacit rules/constraints?
(d) Given the answer to (c), are there such things as informal rules?
(e) Is a social/organizational/behavioral constraint also a rule? . . .
I know that I am pressing you a bit but I think that it is important to achieve maxi-
mum clarity, and then hopeful y consensus, in this area. . . . My hope is that we can
increase our agreement and then move on.
In reply to your letter, I am in complete agreement with you on the first part of your
question. That is, I agree that (a) and (a*) are exactly what I have in mind, so we are in
complete agreement. On the other issue, I regard informal norms not as rules but as
norms of behavior which have different kinds of enforcement characteristics than for-
mal rules. Formal rules are enforced by courts and things like that. Informal norms are
enforced usually by your peers or others who will impose costs on you if you do not live
up to them. In that sense, because they have different enforcement characteristics, I do
not consider them to be a rule in the same sense that formal rules are. Notes
1. Jack Knight (1992, 2), for example, similarly defined an institution as “a set of rules that struc-
ture social interactions in particular ways.” However, there is a debate within the new institu-
tional economics whether institutions should be regarded essentially as equilibria, norms, or
rules (Aoki 2001; Crawford and Ostrom 1995). But this interpretative conflict arises essen-
tially within an intellectual tradition that takes individual preferences or purposes as given.
Being relatively stable, institutions have equilibrium-like qualities, even if their equilibria can
be disturbed. These equilibria are reinforced as preferences or purposes become molded by
the outcomes. Turning to norms and rules, they are not simply the “environment” in which
the (rational) actor must decide and act; they are also internalized in the preferences, and rep-
licated through the behavior, of the individual. Repeated, conditional, rule-like behavior
acquires normative weight as people accept the customary as morally virtuous and thus help
stabilize the institutional equilibrium. Once we see the effects of institutions on individuals,
as well as the effects of individuals upon institutions, the three aspects of institutions become
2. Note that the French economie des conventions school adopts a broad definition of convention
that is closer to the notion of a rule adopted here (Thévenot 1986; Orléan 1994; Favereau and
3. Tony Lawson (2003a, 189–94) listed several behavioral definitions in the institutionalist liter-
ature and rightly criticized them. Walter Hamilton’s famous definition of an institution as “a
way of thought or action of some prevalence and permanence, which is embedded in the hab-
its of a group or the customs of a people” (1932, 84) is preferable to some of the later
institutionalist definitions, as long as habits and customs are interpreted as dispositions
4. While flawed, definitions of institutions in terms of behaviors were understandable during
the positivist era in psychology and the social sciences—from about the 1920s until well after
the Second World War—when it was widely and mistakenly upheld that discussions of
unobservables had no place in science (Hodgson 2004).
5. Margaret Archer’s (1995) useful discussion of demographic structure nevertheless fails to
identify the distinction between structures in general and institutional, rule-based structures.
6. See Ostrom 1986 and Crawford and Ostrom 1995 for detailed analyses of the nature of insti-
7. The definition of technology is itself problematic and is not attempted here. See Nelson and
8. Roy Bhaskar (1989, 80, 85, 112) wrote “intentional human behaviour is . . . always caused by
reasons” and “the reason for the behaviour is itself a belief.” But, it is then admitted, “[b]eliefs
may be unconscious, implicit or tacit.” Consequently, the concept of intentionality is
stretched to cover unconscious behavior and we have no criterion by which we can decide
whether a form of behavior is “action” or “mere movement.”
9. For a critical discussion of John Searle’s treatment of collective intentionality see Vromen
10. Lawson (2003b, 333) has interpreted Veblen differently, without textual evidence, “as using
the term habit to indicate certain (repeated) forms of action.” On the contrary, there are sev-
eral passages in Thorstein Veblen’s works that suggest a view of habits as propensities or dispo-
11. The conception of a habit as a propensity or disposition is also found in modern works such as
Camic 1986, Margolis 1987, Murphy 1994, Ouellette and Wood 1998, Kilpinen 2000,
12. The dispositional treatment of habit here is broadly consistent with Viktor Vanberg’s concept
(2002) of “program-based” activity, which he insisted must be rendered consistent with our
13. For a discussion of the original concept of “downward causation,” see Sperry 1991.
Reconstitutive downward causation is further discussed in Hodgson 2003 and 2004, and the
process is modeled, with a habit-forming mechanism, in Hodgson and Knudsen 2004.
14. See Hodgson 2004 for a discussion of historical accounts of this insight, from Comte,
through Marx, Lewes, Durkheim, Veblen, and others to Archer 1995.
15. See the appendix for the correspondence with Douglass North on this question.
16. Like North, John R. Commons advanced a predominantly legal conception of an institution
(1934), and I have criticized this elsewhere (Hodgson 2004, chap. 13). See Fiori 2002 for a dis-
cussion of the role of the distinction between formal rules and informal constraints in
17. Pavel Pelikan’s definitions (1988, 1992) are similar but not identical to those of North. He
treats institutions as “rules” but also explicitly considers internal “institutions” (rules) of
18. Although mutual cooperation can emerge in repeated prisoners’ dilemma games, Robert
Axelrod’s “tit-for-tat” strategy (1984) can be outcompeted by alternative strategies (Kitcher
19. Note that Gary Becker implied a dichotomy between “habitual” and “rational” behavior
(1962), which he later abandoned by attempting to explain habitual behavior in rational
terms (Becker and Murphy 1988). Becker’s position contrasts with the pragmatist view, where
habit is the grounding rather than the antithesis or outcome of rational deliberation.
20. For useful discussions of these results, see Denzau and North 1994, Mirowski 2002, and
Mirowski and Somefun 1998. Jean-Michel Grandmont (1992) has similarly demonstrated
that aggregate demand can be well behaved under certain distributional restrictions, merely
by assuming that individual behavior satisfies budget constraints, without any reference to
utility maximization. See also Hildenbrand 1994. References
Aoki, Masahiko. Toward a Comparative Institutional Analysis. Cambridge: MIT Press, 2001.
Archer, Margaret S. Realist Social Theory: The Morphogenetic Approach. Cambridge: Cambridge University Press,
Axelrod, Robert M. The Evolution of Cooperation. New York: Basic Books, 1984.
Becker, Gary S. “Irrational Behavior and Economic Theory.” Journal of Political Economy 70, no. 1 (February
Becker, Gary S., and Kevin M. Murphy. “A Theory of Rational Addiction.” Journal of Political Economy 96, no. 4
Best, Michael H. “The Political Economy of Socially Irrational Products.” Cambridge Journal of Economics 6, no.
Bhaskar, Roy. The Possibility of Naturalism: A Philosophic Critique of the Contemporary Human Sciences.
Binmore, Kenneth. Review of Complexity and Cooperation by Robert Axelrod. Journal of Artificial Societies and So-cial Situations 1, no. 1 (1998). http://jasss.soc.surrey.ac.uk/JASSS/1/1/review1.html.
Camic, Charles. “The Matter of Habit.” American Journal of Sociology 91, no. 5 (March 1986): 1039–87.
Coleman, James S. The Asymmetric Society. Syracuse: Syracuse University Press, 1982.
Commons, John R. Institutional Economics—Its Place in Political Economy. New York: Macmillan, 1934.
Crawford, Sue E. S., and Elinor Ostrom. “A Grammar of Institutions.” American Political Science Review 89, no.
Denzau, Arthur T., and Douglass C. North. “Shared Mental Models: Ideologies and Institutions.” Kyklos 47,
Dewey, John. Human Nature and Conduct: An Introduction to Social Psychology. New York: Holt, 1922.
Durkheim, Émile. The Division of Labour in Society, translated from the French edition of 1893 by W. D. Halls
with an introduction by Lewis Coser. London: Macmillan, 1984.
Favereau, Olivier, and Emmanuel Lazega, eds. Conventions and Structures in Economic Organization: Markets, Net-works, and Hierarchies. Cheltenham, U.K., and Northampton, Mass.: Edward Elgar, 2002.
Fiori, Stefano. “Alternative Visions of Change in Douglass North’s New Institutionalism.” Journal of EconomicIssues 36, no. 4 (December 2002): 1025–43.
Foster, John Fagg. “The Papers of J. Fagg Foster.” Journal of Economic Issues 15, no. 4 (December 1981):
Gode, Dhananjay K., and Shyam Sunder. “Allocative Efficiency of Markets with Zero-Intelligence Traders:
Market as a Partial Substitute for Individual Rationality.” Journal of Political Economy 101 (February 1993):
Grandmont, Jean-Michel. “Transformations of the Commodity Space, Behavioral Heterogeneity, and the Ag-
gregation Problem.” Journal of Economic Theory 57, no. 1 (1992): 1–35.
Greif, Avner. “Contract Enforceability and Economic Institutions in Early Trade: The Maghribi Traders’ Co-
alition.” American Economic Review 83, no. 3 (June 1993): 525–48.
Hamilton, Walton H. “Institution.” In Encyclopaedia of the Social Sciences, Vol. 8, edited by Edwin R. A.
Seligman and Alvin Johnson. New York: Macmillan, 1932, 84–89.
Hardin, Garrett. “The Tragedy of the Commons.” Science 162 (1968): 1243–8.
Hayek, Friedrich A. “Notes on the Evolution of Systems of Rules of Conduct.” In Friedrich A. Hayek, Studies inPhilosophy, Politic,s and Economics. London: Routledge and Kegan Paul, 1967, 66–81.
———. Law, Legislation, and Liberty. Vol. 1, Rules and Order. London: Routledge and Kegan Paul, 1973.
———. Law, Legislation, and Liberty. Vol. 3, The Political Order of a Free People. London: Routledge and Kegan
Hildenbrand, Werner. Market Demand: Theory and Empirical Evidence. Princeton, N.J.: Princeton University
Hindess, Barry. Political Choice and Social Structure: An Analysis of Actors, Interests, and Rationality.
Hodgson, Geoffrey M. How Economics Forgot History: The Problem of Historical Specificity in Social Science,
London and New York: Routledge, 2001.
———. “The Evolution of Institutions: An Agenda for Future Theoretical Research.” Constitutional Political Econ-omy 13, no. 2 (June 2002): 111–27.
———. “The Hidden Persuaders: Institutions and Individuals in Economic Theory.” Cambridge Journal of Econom-ics 27, no. 2 (March 2003): 159–75.
———. The Evolution of Institutional Economics: Agency, Structure, and Darwinism in American
Institutionalism. London and New York: Routledge, 2004.
Hodgson, Geoffrey M., and Thorbjørn Knudsen. “The Complex Evolution of a Simple Traffic Convention:
The Functions and Implications of Habit.” Journal of Economic Behavior and Organization 54, no. 1 (2004):
James, William. Psychology: Briefer Course. New York and London: Holt and Macmillan, 1892.
Joas, Hans. Pragmatism and Social Theory. Chicago: University of Chicago Press, 1993.
———. The Creativity of Action. Chicago: University of Chicago Press, 1996.
Kilpinen, Erkki. The Enormous Fly-Wheel of Society: Pragmatism’s Habitual Conception of Action and Social
Theory. Helsinki: University of Helsinki, 2000.
Kitcher, Philip “Why Not the Best?” In The Latest on the Best: Essays on Evolution and Optimality, edited by John
A. Dupré. Cambridge: MIT Press, 1987: 77–102.
Kley, Roland. Hayek’s Social and Political Thought. Oxford: Clarendon Press, 1994.
Knight, Jack. Institutions and Social Conflict. Cambridge: Cambridge University Press, 1992.
Latzer, Michael, and Stefan Schmitz, eds. Carl Menger and the Evolution of Payments Systems. Cheltenham, U.K.,
and Northampton, Mass.: Edward Elgar, 2002.
Lawson, Tony. Economics and Reality. London and New York: Routledge, 1997.
———. “Institutionalism: On the Need to Firm up Notions of Social Structure and the Human Subject.” Journalof Economic Issues 37, no. 1 (March 2003a): 175–207.
———. Reorienting Economics. London and New York: Routledge, 2003b.
Leibenstein, Harvey. “The Prisoners’s Dilemma in the Invisible Hand: An Analysis of Intrafirm Productivity.”
American Economic Review (Papers and Proceedings) 72, no. 2 (May 1982): 92–7.
Lindgren, Kristian. “Evolutionary Phenomena in Simple Dynamics.” In Artificial Life II, edited by Christopher
G. Langton, Charles Taylor, J. Doyne Farmer, and Steen Rasmussen. Redwood City, Calif.: Addi-
Mantzavinos, Chris. Individuals, Institutions, and Markets. Cambridge and New York: Cambridge University
Margolis, Howard. Patterns, Thinking, and Cognition: A Theory of Judgment. Chicago: University of Chicago Press,
McDougall, William. An Introduction to Social Psychology. London: Methuen, 1908.
Menger, Carl. Grundsätze der Volkwirtschaftslehre. Tübingen: J. C. B. Mohr, 1871. Published in English as Princi-ples of Economics, edited by J. Dingwall and translated by B. F. Hoselitz. New York: New York University
Mirowski, Philip. Machine Dreams: Economics Becomes a Cyborg Science. Cambridge and New York: Cambridge
Mirowski, Philip, and Koye Somefun. “Markets as Evolving Computational Entities.” Journal of EvolutionaryEconomics 8, no. 4 (1998): 329–56.
Murphy, James Bernard. “The Kinds of Order in Society.” In Natural Images in Economic Thought: “Markets Readin Tooth and Claw,” edited by Philip Mirowski. Cambridge and New York: Cambridge University Press,
Nelson, Richard R., and Bhaven N. Sampat. “Making Sense of Institutions as a Factor Shaping Economic Per-
formance.” Journal of Economic Behavior and Organization 44, no. 1 (2001): 31–54.
North, Douglass C. Institutions, Institutional Change, and Economic Performance. Cambridge: Cambridge Univer-
———. “Institutions.” Journal of Economic Perspectives 5, no. 1 (Winter 1991): 97–112.
———. “Economic Performance through Time.” American Economic Review 84, no. 3 (June 1994): 359–67.
———. “Five Propositions about Institutional Change.” In Explaining Social Institutions, edited by Jack Knight and
Itai Sened. Ann Arbor, Mich.: University of Michigan Press, 1995: 15–26.
———. “Prologue.” In The Frontiers of the New Institutional Economics, edited by John N. Drobak and John V. C.
Nye. San Diego and London: Academic Press, 1997: 3–28.
Orléan, André, ed. Analyse économique des conventions. Paris: Presses Universitaires de France, 1994.
Ostrom, Elinor. “An Agenda for the Study of Institutions.” Public Choice 48, 1986: 3–25.
Ouellette, Judith A., and Wendy Wood. “Habit and Intention in Everyday Life: The Multiple Processes by
which Past Behavior Predicts Future Behavior.” Psychological Bulletin 124 (1998): 54–74.
Peirce, Charles Sanders. “How to Make Our Ideas Clear.” Popular Science Monthly 12 (January 1878): 286–302.
Pelikan, Pavel. “Can the Innovation System of Capitalism be Outperformed?” In Technical Change and EconomicTheory, edited by Giovanni Dosi, Christopher Freeman, Richard Nelson, Gerald Silverberg, and Luc L.
G. Soete. London: Pinter, 1988, 370–98.
———. “The Dynamics of Economic Systems, or How to Transform a Failed Socialist Economy.” Journal of Evolu-tionary Economics 2, no. 1 (March 1992): 39–63.
Polanyi, Michael. The Tacit Dimension. London: Routledge and Kegan Paul, 1967.
Quine, Willard van Orman. Word and Object. Cambridge: Harvard University Press, 1960.
Schotter, Andrew R. The Economic Theory of Social Institutions. Cambridge: Cambridge University Press, 1981.
Schout, Adriaan. Review of Institutions, Institutional Change, and Economic Performance by Douglass C. North. Economic Journal 101, no. 5 (November 1991): 1587–9.
Schultz, Walter J. The Moral Conditions of Economic Efficiency. Cambridge and New York: Cambridge University
Searle, John R. The Construction of Social Reality. London: Allen Lane, 1995.
———. “What Is an Institution?” Journal of Institutional Economics 1, no. 1 (June 2005).: 1–22.
Sened, Itai. The Political Institution of Private Property. Cambridge: Cambridge University Press, 1997.
Sperry, Roger W. “In Defense of Mentalism and Emergent Interaction.” Journal of Mind and Behavior 12, no. 2
Sugden, Robert. The Economics of Rights, Co-operation, and Welfare. Oxford: Basil Blackwell, 1986.
———. “Team Preferences.” Economics and Philosophy 16, no. 2 (2000): 175–204.
Thévenot, Laurent. Conventions economiques. Paris: Presses Universitaires de France, 1986.
Thomas, William, and Florian Znaniecki. The Polish Peasant in Europe and America, vol. 2. New York: Octagon,
Tuomela, Raimo. The Importance of Us: A Philosophical Study of Basic Social Notions. Stanford, Calif.: Stanford
Vanberg, Viktor J. Rules and Choice in Economics. London: Routledge, 1994.
———. “Rational Choice versus Program-Based Behavior: Alternative Theoretical Approaches and Their Rele-
vance for the Study of Institutions.” Rationality and Society 14, no. 1 (Summer 2002): 7–53.
Veblen, Thorstein B. The Theory of the Leisure Class: An Economic Study in the Evolution of Institutions. New York:
———. “The Limitations of Marginal Utility.” Journal of Political Economy 17, no. 9 (November 1909): 620–36.
Vromen, Jack J. “Collective Intentionality, Social Reality, and Evolutionary Biology.” Philosophical Explorations
6, no. 3 (September 2003): 251–64.
Weber, Max. Max Weber: Selections in Translation, edited and introduced by W. G. Runciman. Cambridge: Cam-
Wells, Alan. Social Institutions. London: Heinemann, 1970.
Wittgenstein, Ludwig. Philosophical Investigations. Oxford: Basil Blackwell, 1958.
Wood, Wendy, Jeffrey M. Quinn, and D. Kashy. “Habits in Everyday Life: Thought, Emotion, and Action.”
Journal of Personality and Social Psychology 83 (2002): 1281–97.