North 1990: Informal and Formal Constraints PDF

Informal constraints In all societies from the most primitive to the most advanced, people impose constraints upon themselves to give a structure to their relations with others. Under conditions of limited information and limited com- putational ability, constraints reduce the costs of huma...

Informal constraints In all societies from the most primitive to the most advanced, people impose constraints upon themselves to give a structure to their relations with others. Under conditions of limited information and limited com- putational ability, constraints reduce the costs of human interaction as compared to a world of no institutions. However, it is much easier to describe and be precise about the formal rules that societies devise than to describe and be precise about the informal ways by which human beings have structured human interaction. But although they defy, for the most part, neat specification and it is extremely difficult to develop unam- biguous tests of their significance, they are important. In the modern Western world, we think of life and the economy as being ordered by formal laws and property rights. Yet formal rules, in even the most developed economy, make up a small (although very impor- tant) part of the sum of constraints that shape choices; a moment's reflec- tion should suggest to us the pervasiveness of informal constraints. In our daily interaction with others, whether within the family, in external social relations, or in business activities, the governing structure is overwhelm- ingly defined by codes of conduct, norms of behavior, and conventions. Underlying these informal constraints are formal rules, but these are seldom the obvious and immediate source of choice in daily interactions. That the informal constraints are important in themselves (and not simply as appendages to formal rules) can be observed from the evidence that the same formal rules and/or constitutions imposed on different societies produce different outcomes. And discontinuous institutional change, such as revolution or military conquest and subjugation, cer- tainly produces new outcomes. But what is most striking (although sel- dom observed, particularly by advocates of revolution) is the persistence of so many aspects of a society in spite of a total change in the rules. Japanese culture survived the U.S. occupation after World War II; the post-revolutionary U.S. society remained much as it had been in colonial  Informal constraints times; Jews, Kurds, and endless other groups have persisted through cen- turies despite endless changes in their formal status. Even the Russian Revolution, perhaps the most complete formal transformation of a soci- ety we know, cannot be completely understood without exploring the survival and persistence of many informal constraints. Where do informal constraints come from? They come from socially transmitted information and are a part of the heritage that we call culture. The way the mind processes information depends "upon the brain's abil- ity to learn by being programmed with one or more elaborately struc- tured natural languages that can code for perceptual, attitudinal and moral (behavioral) as well as factual information" (Johansson, 1988, p. 176). Culture can be defined as the "transmission from one generation to the next, via teaching and imitation, of knowledge, values, and other factors that influence behavior" (Boyd and Richerson, 1985, p. 2). Culture provides a language-based conceptual framework for encoding and interpreting the information that the senses are presenting to the brain. Essentially the argument being made here is an extension of the argu- ment of Chapter 3 that processing information is the key to understand- ing a more complex behavioral pattern than is derived from the expected utility model. But the emphasis in that chapter was on the incompleteness of the information and the consequent need for institutions to structure human interrelations. In this chapter the emphasis is on the way that the cultural filter provides continuity so that the informal solution to ex- change problems in the past carries over into the present and makes those informal constraints important sources of continuity in long-run societal change. I I begin by examining human interaction when there are no formal rules. How is order preserved in stateless societies? The anthropological liter- ature is extensive, and although many of the findings are still controver- sial, it makes important reading not only for the study of historical work and for an analysis of order in primitive societies, but also for its implica- tions for a modern understanding of informal constraints. Robert Bates (1987), building his analysis on Evans-Pritchard's classic study of the Nuer, states the issue of such constraints as follows: The puzzle, from Evans-Pritchard's point of view, was that, despite the potential for theft and disorder, the Nuer in fact tended to live in relative harmony. Insofar as the Nuer raided cattle, they tended to raid the cattle of others; raids within the tribe were relatively rare. Somehow the Nuer appear to have avoided the poten- tially harmful effects arising from the pursuit of self-interest. And they appear to 37  Institutions have done so even while lacking those formal institutions so common in Western societies which specialize in preserving the peace and forestalling violence: the courts, the police, and so on. (Bates, 1987, p. 8) Bates then describes the deterring effects that both compensation among the tribe and the threat of feud posed for preserving order. He shows how this cooperative solution makes sense in game theoretic terms. A one-shot prisoner's dilemma problem, where it would appear that the players must arrive at a violent solution with the result that each family is worse off, is avoided. Instead an iterated game is played, and with the threat of feud it is in the interests of the parties to preserve order and hence not to pursue interfamily cattle raiding. The critical point here is that it is the members of the family themselves who prevent other family members from engaging in raiding, because a feud, once started, would be harmful to all members. The extensive literature that anthropologists have produced on prim- itive societies makes clear that exchange in tribal societies is not simple. In the absence of the state and formal rules, a dense social network leads to the development of informal structures with substantial sta- bility. No one has described this situation better than Elizabeth Colson (i974)- Whether we call them customs, laws, usages, or normative rules seems of little importance. What is important is that communities such as the Tonga do not leave their members free to go their own way and explore every possible avenue of behavior. They operate with a set of rules or standards which define appropriate action under a variety of circumstances. The rules, by and large, operate to eliminate conflict of interests by defining what it is people can expect from certain of their fellows. This has the healthy effect of limiting demands and allowing the public to judge performance.... At another level, however, they would see conflict as endemic to social life because people who live in close juxtaposition use the same space and want support and attention from the same individuals. Rules, even though they may at times produce conflict, reduce the chances for conflict because they reduce the total amount of ambiguity for those concerned by defining specific rather than universalistic claims and obligations. It becomes possible to order one's life with a set of priorities regarded as legitimate.... Among the Tonga I have had to learn that I should not give just because I feel like giving as this is an insult to all who do not receive. Rules do not solve all problems; they only simplify life. They also give a framework for organizing activities. Standards and some means of applying sanctions are necessary complements to the rules if a system of social control is to operate within a community. Among people such as the Tonga, onlookers apply the standards of performance in particular roles in making an overall judgement [sic] about the total person; this in turn allows them to predict future behavior. Judgment is an ongoing process through which consensus is finally reached. (Colson, 1974, PP- 5 I ~3) 38  Informal constraints Several implications are clear from this review of work by Colson and other anthropologists. Order in the societies they describe is the result of a dense social network where people have an intimate understanding of each other and the threat of violence is a continuous force for preserving order because of its implications for other members of society. Deviant behavior cannot be tolerated in such a situation, because it is a fundamen- tal threat to the stability and insurance features of the tribal group. Richard Posner's model of primitive society (1980), which generates an explanation of many institutional features of such societies, is similar to the one I develop here (although mine has none of the maximizing social wealth or efficiency implications that are explicit in Posner's work). In Posner's model, high information costs, the absence of effective govern- ment, limited numbers of goods and limited trade, limited food preserva- tion, and negligible gains from innovation produce a set of common characteristics: Weak government, ascription of rights and duties on the basis of family mem- bership, gift-giving as a fundamental mode of exchange, strict liability for injuries, emphasis on generosity and honor as high ethical norms, collective guilt - these and other features of social organization recur with such frequency in accounts of primitive and archaic societies as to suggest that a simple model of primitive society, which abstracts from many of the particular features of specific societies, may nonetheless explain much of the structure of primitive social institutions. (Posner, 1980, p. 8) Posner's essay emphasizes the importance of kinship ties as the central insurance, protection, and law enforcement mechanisms of primitive so- cieties. Bates' study of Kenya (1989) equally focuses on the changing pattern of kinship ties in the context of political/economic conditions as the key to understanding the evolving institutional constraints of a society in rapid transition from a tribal society to a market economy. II Informal constraints are pervasive features of modern economies as well. In order to dispel the assertion by law and economics scholars of the centrality of legal doctrine, Robert Ellickson did a field study of the way in which rural residents of Shasta County, California, resolved disputes arising over trespass damage done by stray livestock.1 He found that the residents almost never resorted to legal redress, but instead relied on an elaborate structure of informal constraints to resolve disputes. In a subse- quent article (1987) and forthcoming book, Ellickson provides a great Appropriately titled "Of Coase and Cattle: Dispute Resolution Among Neighbors in Shasta County" (1986). 39  Institutions deal of additional empirical evidence of the pervasiveness of informal constraints. Even the most casual introspection suggests the pervasiveness of infor- mal constraints. Arising to coordinate repeated human interaction, they are (i) extensions, elaborations, and modifications of formal rules, (2) socially sanctioned norms of behavior, and (3) internally enforced stan- dards of conduct. I elaborate on each of these aspects of informal con- straints. 1. In a study of the institutional foundations of committee power Shep- sle and Weingast (1987) demonstrate that the power of congressional committees that is not explained by the formal rules is a result of a set of informal unwritten constraints that have evolved in the context of repeat- ed interaction (exchange) among the players. These constraints evolved from the formal rules to deal with specific problems of exchange and became established as recognized institutional constraints even though they were never made a part of the formal rules. Committee chairs and committees consequently have an influence over legislative choices that could not be derived from the formal structure. 2. Robert Axelrod (1986) provides a vivid illustration of a socially sanctioned norm of behavior. The night before he was to engage in a duel with Aaron Burr, Alexander Hamilton sat down and wrote out all the reasons why he should not accept this challenge; a crucial one, of course, was that he was likely to get killed. Yet, in spite of the overwhelming rational bases for not dueling, he felt that his effectiveness in the public arena would be significantly diminished by such a decision because duel- ing was the accepted way to settle disputes among gentlemen. Social norms dictated the choice, not formal rules. 3. Both of the first two types of informal constraints can be modeled in the context of wealth-maximizing models and therefore lend themselves to treatment in neoclassical (and game theory) frameworks. But internally enforced codes of conduct only have meaning in terms of informal con- straints, altering choices when the individual gives up wealth or income for some other value in his or her utility function. Numerous essays explore voting behavior by legislators and conclude that one cannot ex- plain legislative voting behavior by an interest group model (in which the legislator faithfully mirrors the interests of his or her constituents), but must take into account the subjective, personal preferences of the legisla- tor (Kalt and Zupan, 1984). This literature is controversial because of the statistical problems in getting unambiguous answers, but there is abun- dant qualitative and quantitative evidence that the lower the price of ideas, ideologies, and convictions, the more they matter and affect choices (for empirical support see Nelson and Silberberg, 1987). 40  Informal constraints III How do we explain the emergence and persistence of informal con- straints? A pervasive but relatively simple to explain form of such con- straints is conventions that solve coordination problems: "These are rules that have never been consciously designed and that it is in everyone's interest to keep" (Sugden, 1986, p. 54). The usual illustration of such a convention is rules of the road. The important characteristic of conven- tions is that, given the costs of exchange (Chapter 4), both parties have a stake in minimizing the costliness of measurement and the exchanges are self-enforcing. In terms of the total resources that go into transacting in an economy, conventions that solve coordination probably account for a larger proportion of the costs of transacting than the other informal con- straints described later in this chapter (although in many instances the transaction costs in fact reflect a combination of sources of informal constraints). Informal constraints that arise in the context of exchange but are not self-enforcing are more complex because they necessarily entail features that make the exchange viable by reducing measurement and enforce- ment costs. In the absence of constraints, asymmetric information and the consequent distribution of the gains will lead to devoting excessive re- sources to measurement or indeed can lead to exchange not taking place at all because the exchange is unenforceable. Informal constraints can take the form of agreed upon lower cost forms of measurement (standard- ized weights and measures, for example) and make second- and third- party enforcement effective by specific sanctioning devices or information networks that acquaint third parties with exchange performance (credit ratings, better business bureaus, etc.). Such organizations and instru- ments that make norms of cooperative behavior (informal constraints) effective are not only a major part of the story of more complex exchange through history, but are strikingly paralleled by the game theoretic mod- els that produce cooperative outcomes through features that alter dis- count rates and increase information. The growth of more complex forms of exchange in later medieval and early modern Europe was made pos- sible by a variety of informal institutions such as the early law merchant's publicized codes of merchant conduct. Prices current and the develop- ment of auditing and accounting techniques lowered critical (i.e., mea- surable) information and enforcement costs. These can be modeled in a game theoretic framework by raising the gains from cooperative ac- tion or raising the costs of defecting (see Milgrom, North, and Weingast, 1990). Much more difficult to deal with in theoretical terms than wealth  Institutions maximizing informal constraints are internally enforced codes of conduct that modify behavior. It is difficult because one must devise a model that predicts choices in the context of the trade-off between wealth and other values. But strong religious beliefs or commitment to communism, for example, provide us with historical accounts of the sacrifices individuals have made for beliefs. As noted earlier, experimental economics provides evidence that individuals do not always free-ride and a study by Frank (1988) provides both a large body of evidence and a model of such behavior. The literature cited above and the earlier chapter of this book dealing with human behavior make clear that motivation is more complicated than the simple expected utility model. Chapter 3 also emphasized that under certain conditions traits like honesty, integrity, and living up to a reputation pay off in strictly wealth-maximizing terms. Still unexplained is a very large residual. We simply do not have any convincing theory of the sociology of knowledge that accounts for the effectiveness (or ineffec- tiveness) of organized ideologies or accounts for choices made when the payoffs to honesty, integrity, working hard, or voting are negative. Two partial explanations are Howard Margolis's (1982) dual utility model (mentioned in Chapter 2) and Robert Sugden's argument about conventions acquiring moral force. Margolis's argument is that indi- viduals possess not one but two utility functions: S preferences are gov- erned by the usual self-interest preference function, whereas G prefer- ences are purely social (group interested). Margolis attempts to give empirical content to the argument by developing a model with assigned weights given self-interest preferences versus weights given group-interest preferences and by exploring the conditions under which the weights change. Sugden (1986) maintains that a convention acquires moral force when almost everyone in the community follows it, and it is in the in- terests of each individual that people with whom he or she deals follow the rule providing that the individual does too. What evolves according to Sugden is a "morality of cooperation" (Sugden, 1986, p. 173). IV It is time to pull together and summarize the argument of this chapter. The way by which the mind processes information not only is the basis for the existence of institutions, but is a key to understanding the way infor- mal constraints play an important role in the makeup of the choice set both in the short-run and in the long-run evolution of societies. In the short run, culture defines the way individuals process and utilize information and hence may affect the way informal constraints get spec- ified. Conventions are culture specific, as indeed are norms. However 42-  Informal constraints norms pose some still unexplained problems. What is it that makes norms evolve or disappear — for example, dueling as a solution to gentlemanly differences? Even if we do not possess a good explanation for social norms, we can model wealth-maximizing norms in a game theoretic context. That is, we can explore and test, empirically, what sorts of informal constraints are most likely to produce cooperative behavior or how incremental changes in such informal constraints will alter the game to increase (or decrease) cooperative outcomes. This approach may increase our understanding of the development of more complex forms of exchange, such as the early evolution of financial markets.2 A transaction cost framework equally offers promise of exploring infor- mal constraints. Although the informal institutional constraints are not directly observable, the contracts that are written, and sometimes the actual costs of transacting, provide us with indirect evidence of changes in informal constraints. The striking decline in interest rates in the Dutch capital market in the seventeenth century and in the English capital mar- ket in the early eighteenth century provides evidence of the increasing security of property rights as a consequence of the effective interaction of a variety of both formal and informal institutional constraints. For exam- ple, the enforcement of contracts that evolved from merchant codes of behavior included ostracism of those who violated agreements and the eventual encoding of customary practices into the formal law.3 The importance of self-imposed codes of behavior in constraining max- imizing behavior in many contexts also is evident. Our understanding of the source of such behavior is deficient, but we can frequently measure its significance in choices by empirically examining marginal changes in the cost of expressing convictions. Such analysis opens the door to explaining the power of subjective perceptions in affecting choices. If the demand function is negatively sloped (i.e., the lower the cost of expressing one's convictions the more important will the convictions be as a determinant of choice) and formal institutions make it possible for individuals to express preferences at little cost to themselves, then indeed the subjective preferences that individuals hold play a big part in determining choices. Voting, hierarchies that produce slack in the principal/agent relationship in legislatures, and lifetime tenure for judges are formal institutional constraints that lower the cost of acting on one's convictions. It is simply impossible to make sense out of history (or contemporary 2 For an interesting game theoretic application, see John Veitch, "Repudiations and Confiscations by the Medieval State" (1986). 3 See Douglass C. North, "Institutions, Transaction Costs, and the Rise of Merchant Empires," in J. Tracy, editor, The Political Economy of Merchant Empires, Cambridge University Press (forthcoming). 43  Institutions economies) without recognizing the central role that subjective prefer- ences play in the context of formal institutional constraints that enable us to express our convictions at zero or very little cost. Ideas, organized ideologies, and even religious zealotry play major roles in shaping so- cieties and economies. Nineteenth-century U.S. economic history, briefly described in Chapter i, is full of illustrations. Whether we trace the history and consequences of the abolitionist movement, or examine the reasoning of Supreme Court justices that explicitly undergirded the decisions they handed down, or explore the organization, policies, and legislative enactments of the Greenback, Granger, and Populist movements of the U.S. farmer, they all only make sense in the context of subjective perceptions of the actors in the context of formal institutional structures that altered the price indi- viduals paid for their convictions and hence enabled their choices to become effective. In the first case the religious zealotry of the abolitionist groups that activated them to organize politically, together with the Northern electo- rate's growing conviction of the immorality of slavery and the i860 elec- tion, led to the Civil War and the elimination of slavery (Fogel, 1989). In the second case the lifetime tenure specified for justices shielded them from interest group pressures and permitted - encouraged - them to vote their convictions. Their convictions were derived from their subjective construction of the issues. From the Marshall Court (1801 to 1835) to the Rehnquist Court, the justices have interpreted and reinterpreted essen- tially the same set of rules. The Court reverses itself 180 degrees over time because the judges' subjective modeling of the issues changes. The third case reflects farmers' persistent beliefs that they were being wronged by monetary policies, railroads, grain elevators, bankers, and others. They acted on these convictions by forming organizations with the objective of enacting corrective legislation first in state legislatures, then through the Populist party and the Democratic party in the U.S. Congress. What determines how much people will pay to express and act on their convictions ? We seldom know much about the elasticity of the function or shifts in the function, but we do have abundant evidence that the function is negatively sloped and that the price incurred for acting on one's convic- tions is frequently very low (and hence convictions are significant) in many institutional settings. The long-run implication of the cultural processing of information that underlies informal constraints is that it plays an important role in the incremental way by which institutions evolve and hence is a source of path dependence. We still are a long way from having any neat models of cultural evolution (although see Cavalli-Sforza and Feldman, 1981, and Boyd and Richerson, 1985, for some interesting attempts), but we do 44  Informal constraints know that cultural traits have tenacious survival ability and that most cultural changes are incremental. Equally important is the fact that the informal constraints that are culturally derived will not change immediately in reaction to changes in the formal rules. As a result the tension between altered formal rules and the persisting informal constraints produces outcomes that have impor- tant implications for the way economies change, which is the subject of Part II. 45  Formal constraints The difference between informal and formal constraints is one of degree. Envision a continuum from taboos, customs, and traditions at one end to written constitutions at the other. The move, lengthy and uneven, from unwritten traditions and customs to written laws has been unidirectional as we have moved from less to more complex societies and is clearly related to the increasing specialization and division of labor associated with more complex societies.1 The increasing complexity of societies would naturally raise the rate of return to the formalization of constraints (which became possible with the development of writing), and technological change tended to lower measurement costs and encourage precise, standardized weights and measures. The creation of formal legal systems to handle more complex disputes entails formal rules; hierarchies that evolve with more complex organization entail formal structures to specify principal/agent rela- tionships. The general characteristics of the shift from status to contract have been amply discussed, but it is worth emphasizing the following. Formal rules can complement and increase the effectiveness of informal constraints. They may lower information, monitoring, and enforcement costs and hence make informal constraints possible solutions to more a For a lengthy and thoughtful discussion of what we mean by formal rules, see Elinor Ostrom (1986). Ostrom breaks down the rule structure into the following specifics: position rules that specify a set of positions and how many participants hold each position, boundary rules that specify how participants are chosen to hold these positions and how participants leave these positions, scope rules that specify the set of outcomes that may be affected and the external inducements and/or costs assigned to each of these outcomes, authority rules that specify the set of actions assigned to a position at a particular node, aggregation rules that specify the decision functions to be used at a particular node to map action into intermediate or final outcomes, and information rules that authorize channels of communication among participants in positions and specify the language and form in which the communication will take place. 46  Formal constraints complex exchange (see Milgrom, North, and Weingast, 1990, and Chap- ter 7 for elaboration). Formal rules also may be enacted to modify, revise, or replace informal constraints. A change in the bargaining strength of parties may lead to an effective demand for a different institutional frame- work for exchange, but the informal constraints stand in the way of accomplishing it. Sometimes (but not always) it is possible to supersede the existing informal constraints with new formal rules (this point will be elaborated and qualified in Chapter 10). i Formal rules include political (and judicial) rules, economic rules, and contracts. The hierarchy of such rules, from constitutions, to statute and common laws, to specific bylaws, and finally to individual contracts defines constraints, from general rules to particular specifications. And typically constitutions are designed to be more costly to alter than statute laws, just as a statute law is more costly to alter than individual contracts. Political rules broadly define the hierarchical structure of the polity, its basic decision structure, and the explicit characteristics of agenda control. Economic rules define property rights, that is the bundle of rights over the use and the income to be derived from property and the ability to alienate an asset or a resource. Contracts contain the provisions specific to a particular agreement in exchange. Given the initial bargaining strength of the decision-making parties, the function of rules is to facilitate exchange, political or economic. The existing structure of rights (and the character of their enforcement) de- fines the existing wealth-maximizing opportunities of the players, which can be realized by forming either economic or political exchanges. Ex- change involves bargains made within the existing set of institutions, but equally the players at times find it worthwhile to devote resources to altering the more basic structure of the polity to reassign rights. The extent of economic and political diversity of interests will, given relative bargaining strength, influence the rules' structure. The immediate reason is that the more numerous the interests, the less likely the simple majority (in the polity) will obtain and the more likely exchange will be structured to facilitate complex forms of exchange (partly formal but also partly informal) and other ways of solving problems by coalition forma- tion. It is important to note, however, that the function of formal rules is to promote certain kinds of exchange but not all exchange. Thus Madi- son, in Federalist Paper Number 10, maintained that the constitutional structure was devised in 1787 not only to facilitate certain kinds of ex- change, but also to raise the costs of those kinds of exchange that promote the interests of factions. Similarly, in economic exchange patent laws and 47  Institutions trade secret laws are designed to raise the costs of those kinds of exchange deemed to inhibit innovation. Before going further, it is important to stress that there is nothing in my argument so far about rules that implies efficiency. As stressed above, rules are, at least in good part, devised in the interests of private well- being rather than social well-being. Hence, rules that deny franchise, restrict entry, or prevent factor mobility are everywhere evident. This is not to deny that ideas and norms matter, but to establish that as a first approximation, rules are derived from self-interest. Rules are generally devised with compliance costs in mind, which means that methods must be devised to ascertain that a rule has been violated, to measure the extent of the violation (and consequent damages to the party to exchange), and to apprehend the violator. The costs of compliance include measuring the multiple attributes of the goods or services being exchanged and measuring the performance of agents. In many cases, the costs of measurement, given the technology of the time, exceed the gains, and rules are not worth devising and ownership rights are not delineated. Changes in technology or relative prices will alter the relative gains from devising rules. With these generalizations as background, we can now use the frame- work derived from Chapters 3 and 4 to describe more closely political rules, property rights (economic rules), and contracts. II Broadly speaking, political rules in place lead to economic rules, though the causality runs both ways. That is, property rights and hence individual contracts are specified and enforced by political decision-making, but the structure of economic interests will also influence the political structure. In equilibrium, a given structure of property rights (and their enforce- ment) will be consistent with a particular set of political rules (and their enforcement). Changes in one will induce changes in the other. But be- cause of the priority of political rules, we will analyze the structure of the political system first. We start with a simplified model of a polity made up of a ruler and constituents.2 In such a simple setting, the ruler acts like a discriminating monopolist, offering to different groups of constituents protection and justice or at least the reduction of internal disorder and the protection of property rights in return for tax revenue. Because different constituent groups have different opportunity costs and bargaining power with the 2 This simple model is developed in much more detail in Chapter 3, "A Neo-classical Theory of the State" in North (1981). 48  Formal constraints ruler, different bargains result. But there are also economies of scale in the provision of these (semipublic) goods of law and enforcement. Hence, total revenue is increased, but the division of incremental gains between ruler and constituents depends on their relative bargaining power; changes at the margin, either the violence potential of the ruler or the opportunity costs of the constituent, will result in redivisions of the incre- mental revenue. Moreover, the ruler's gross and net revenue differ signifi- cantly as a result of the necessity of developing agents (a bureaucracy) to monitor, meter, and collect the revenue. All the consequences inherent to agency theory obtain here. This model of the polity becomes one step more complicated when we introduce the concept of a representative body reflecting the interests of constituent groups and their role in bargaining with the ruler. This con- cept, consistent with the origin of parliaments, estates general, and cortes in early modern Europe, reflects the needs of the ruler to get more revenue in exchange for which he or she agrees to provide certain services to constituent groups. The representative body facilitates exchange between the parties. On the ruler's side, this leads to the development of a hier- archical structure of agents, which is a major transformation from the simple (if extensive) management of the king's household and estates to a bureaucracy monitoring the wealth and/or income of the king's consti- tuents. When we move from the historical character of representation in early modern Europe to modern representative democracy, our story is compli- cated by the development of multiple interest groups and by a much more complicated institutional structure devised to facilitate (again given rela- tive bargaining strength) the exchange between interest groups.3 This political transaction cost analysis is built on the recognition of the multi- plicity of interest groups reflecting concentrations of voters in particular locations. Thus in the United States polity, there are elderly in Florida and Arizona, miners in Pennsylvania and West Virginia, artichoke growers in California, automobile manufacturers in Michigan, and so forth. Because 3 The development of political theory in the last twenty-five years has paralleled developments in economic theory. The developments began in an a-institutional set- ting, in which the model paralleled the a-institutional model of economics. But the result, in terms of the formal theory, was that no stable equilibrium would evolve and that cycling would be a continuous pattern of political systems (at least in two-party, nonideological models). However, this formal finding was at odds with empirical and descriptive studies that provided no evidence of such disequilibrating characteristics, and it remained to take a further step in political theory to explore the nature of the institutional structure that provided for the evolution of equilibrium states in the political system. For a description of this evolution and a model of structure-induced equilibrium, see Kenneth Shepsle, "Institutional Equilibrium and Equilibrium Institu- tions" (1986). 49  Institutions there are multiple interest groups, no particular interest group that a legislator may represent can form a majority. Therefore, legislators cannot succeed acting alone, but must make agreements with other legislators, with different interests. What kinds of institutions will evolve from exchange relationships be- tween legislators reflecting multiple interest groups? Previous work, be- ginning with Buchanan and Tullock (1962), focused on vote-trading or logrolling. This approach was a step forward in recognizing the way by which legislators can strike bargains that facilitate exchange; however, it is too simple to solve fundamental problems involved in legislative ex- change. It assumes that all bills and payoffs were known in advance, and it has a timeless dimension to it. In fact, a variety of exchanges arise in which today's legislation can only be enacted by commitments made for a future date. To lower the costs of exchange, it was necessary to devise a set of institutional arrangements that would allow for exchange over space and time. As with the economic exchange described in Chapter 4, the problem is to measure and enforce the exchange of rights. How does credible commitment evolve to enable agreements to be reached when the payoffs are in the future and on completely different issues? Self-enforcement is important in such exchange, and in repeat dealings a reputation is a valuable asset. But as in economic exchange, the costs of measurement and enforcement, discovering who is cheating whom, when free-riding will occur, and who should bear the cost of punishing defectors make self-enforcement ineffective in many situations. Hence political institutions constitute ex ante agreements about coopera- tion among politicians. They reduce uncertainty by creating a stable structure of exchange. The result is a complicated system of committee structure, consisting of both formal rules and informal methods of orga- nization. The evolution of this structure in the U.S. Congress is described in a recent study of the structure by Barry Weingast and William Marshall entitled "The Industrial Organization of Congress" (1988). In their conclusion, Weingast and Marshall specify the kind of structure that evolved: Instead of trading votes, legislators exchanged special rights affording the holder of these rights additional influence over well-defined policy jurisdictions. This influence stems from the property rights established over the agenda mechanisms, that is, the means by which alternatives arise for votes. The extra influence over particular policies institutionalizes a specific pattern of trades. When the holders of seats on committees are precisely those individuals who would bid for votes on these issues in a market for votes, policy choice under the committee system parallels that under a more explicit exchange system. Because the exchange is institutionalized, it need not be renegotiated each new legislative session, and it is subject to fewer enforcement problems. (Weingast and Marshall, 1988, p. 157)  Formal constraints The evolution of polities from single absolute rulers to democratic governments is typically conceived as a move toward greater political efficiency. In the sense that democratic government gives a greater and greater percentage of the populace access to the political decision-making process, eliminates the capricious capacity of a ruler to confiscate wealth, and develops third-party enforcement of contracts with an independent judiciary, the result is indeed a move toward greater political efficiency. But it would be wrong to assert that the result is efficient political markets in the same sense as we mean efficient economic markets. The existence of efficient economic markets entails competition so strong that, via ar- bitrage and information feedback, one approximates the Coase zero transaction cost conditions. Such markets are scarce enough in the eco- nomic world and even scarcer in the political world. It is true that the move toward a democratic polity will reduce legislative transaction costs per exchange (as elaborated by Weingast and Marshall, 1988), but not only will the number of exchanges increase so that the size of the total political transaction sector will grow, the agency costs between constitu- ent and legislator and legislator and bureaucrat will be substantial. More- over, rational ignorance on the part of constituents isgoing to increase the role, in many situations, of incomplete subjective perceptions playing an important part in choices. The atypical informed constituent may indeed know his or her own interest in making choices about familiar local repeated problems, but even the informed constituent is going to be at sea in making choices about the complex nonrepetitive problems of an inter- dependent political and economic world. The point is that formal politi- cal rules, like formal economic rules, are designed to facilitate exchange but democracy in the polity is not to be equated with competitive markets in the economy. The distinction is important with respect to the efficiency of property rights. Ill As a first approximation we can say that property rights will be developed over resources and assets as a simple cost-benefit calculus of the costs of devising and enforcing such rights, as compared to the alternatives under the status quo. Changes in relative prices or relative scarcities of any kind lead to the creation of property rights when it becomes worthwhile to incur the costs of devising such rights. This simple model has been the basis not only for my own early work (North and Thomas, 1973) but also for a substantial amount of the property rights literature, which looks on the development of property rights as a simple function of changes in economic costs and benefits. The simple model of the evolution of proper- ty rights would be consistent with Axelrod's The Evolution of Coopera-  Institutions tion (1984), but such an argument leaves out the role of the polity and the consequent kinds of property rights that will be specified and enforced. In North (1981), I revised the 1973 argument to account for the ob- vious persistence of inefficient property rights. These inefficiencies existed because rulers would not antagonize powerful constituents by enacting efficient rules that were opposed to their interests or because the costs of monitoring, metering, and collecting taxes might very well lead to a situation in which less efficient property rights yielded more tax revenue than efficient property rights. This argument is an improvement over the efficiency argument but needs amplification. The efficiency of the political market is the key to this issue. If political transaction costs are low and the political actors have accurate models to guide them, then efficient property rights will result. But the high transac- tion costs of political markets and subjective perceptions of the actors more often have resulted in property rights that do not induce economic growth, and the consequent organizations may have no incentive to create more productive economic rules. At issue is not only the incremental character of institutional change, but also the problem of devising institu- tions that can provide credible commitment so that more efficient bar- gains can be struck. In Chapter 11, I shall explore how such sufficient paths of development can persist through time. IV The rules descend from polities to property rights to individual contracts. Contf acts will reflect the incentive-disincentive structure imbedded in the property rights structure (and the enforcement characteristics); thus the opportunity set of the players and the forms of organization they devise in specific contracts will be derived from the property rights structure. The contract specified by economic theory is simple, complete, and straightforward. It involves an exchange of a unidimensional product at an instant of time. The contract in modern complex economies both is multidimensional and extends over time. Because there are multiple di- mensions, with respect both to the physical characteristics and to the property rights characteristics of the exchange, of necessity the result is that one must spell out many of the provisions. Moreover, the contract will typically be incomplete, in the sense that there are so many un- knowns over the life of contracts extending over time that the parties will (deliberately) leave to the courts or to some third party the settlement of disputes that arise over the life of the contract.4 4 See Goldberg (1976) for discussion of relational exchange and the complicated contracts that in fact characterize modern exchange. 52-  Formal constraints Contracts provide not only an explicit framework within which to derive empirical evidence about the forms of organization (and hence are the basic empirical source for testing hypotheses about organization), but also clues with respect to the way by which the parties to an exchange will structure more complex forms of organization. That is, the contracts will reflect different ways to facilitate exchange, whether through firms, fran- chising, or other more complex forms of agreement that extend in a continuum from straightforward market exchange to vertically integrated exchange.5 The complex interaction of institutional constraints and the development of organizations are the subject of Chapter 9. I should close this chapter with a word of warning — although explicit rules provide us with a basic source of empirical materials by which to test the performance of economies under varying conditions, the degree to which these rules have unique relationships to performance is limited. That is, a mixture of informal norms, rules, and enforcement charac- teristics together defines the choice set and results in outcomes. Looking only at the formal rules themselves, therefore, gives us an inadequate and frequently misleading notion about the relationship between formal con- straints and performance. 5 The enormous literature that has evolved in the last fifteen or twenty years in the New Industrial Organization has provided us with an immense amount of valuable material on the kinds of organizations that will evolve and the forms of governance that will be reflected as ways to solve problems of complicated exchange. See in particular Oliver Williamson, Markets and Hierarchies (1975), and the subsequent literature that has developed from Williamson's pioneering work. 53  7 Enforcement A good deal of literature on transaction costs takes enforcement as a given, assuming either that it is perfect or that it is constantly imperfect. In fact, enforcement is seldom either, and the structure of enforcement mechanisms and the frequency and severity of imperfection play a major role in the costs of transacting and in the forms that contracts take. There are two reasons why enforcement is typically imperfect. The first takes us back to the preceding chapters that explore the costs of measuring the multiple margins that constitute contract performance. The second rests in the fact that enforcement is undertaken by agents whose own utility functions influence outcomes. In Chapter 4, asymmetries of information held by principals and agents about the valuable attributes of what was being exchanged were exam- ined in the context of the wealth-maximizing behavior of the parties to exchange. In this chapter I wish to extend that analysis to explore the problems that arise in the transfer of rights. Parties to an exchange must be able to enforce compliance at a (transaction) cost such that the ex- change is worthwhile to them. On the face of it, the problem sounds simple. Surely the gains from trade, which economists take to be the bedrock of economic performance, should make it worthwhile to evolve cooperative solutions among parties to capture jointly those gains. Indeed under certain circumstances, as I have noted in earlier chapters, the issues are so resolved. Trade does exist, even in stateless societies. Yet, as empha- sized earlier, the inability of societies to develop effective, low-cost en- forcement of contracts is the most important source of both historical stagnation and contemporary underdevelopment in the Third World.1 a This emphasis upon enforcement is another major difference between Oliver Williamson's approach to transaction costs and the one taken in this study. William- son assumes enforcement to be imperfect (otherwise opportunism would never pay), but does not make it an explicit variable in his analysis. Such an approach simply does not lead the scholar to be able to deal with the problems of historical evolution, where 54  Enforcement I Under what conditions will contracts tend to be self-enforcing? In a wealth-maximizing world, the answer can be stated very simply. Con- tracts will be self-enforcing when it pays the parties to live up to them — that is, in terms of the costliness of measuring and enforcing agreements, the benefits of living up to contracts will exceed the costs. The most like- ly and indeed empirically observable state in which contracts are self- enforcing is that in which the parties to exchange have a great deal of knowledge about each other and are involved in repeat dealings, as de- tailed in the earlier chapter dealing with tribal and primitive societies and with small communities. Under these conditions, it simply pays to live up to agreements. In such a world, the measured costs of transacting are very low because of a dense social network of interaction. Cheating, shirking, opportunism, all problems of modern industrial organization, are limited or indeed absent because they do not pay. Norms of behavior determine exchange and formal contracting does not exist. At the other extreme, the world of impersonal exchange is characterized by specialized interdependence in which the well-being of individuals depends upon the complex structure characterized by individual spe- cialization and hence exchange extends through both time and space. In a pure model of the world of impersonal exchange, goods and services or the performance of agents is characterized by many valued attributes, exchange takes place over time, and there are no repeat dealings. In the context of a wealth-maximizing world, where there are high costs of measurement and no form of enforcement is possible, the gains from cheating and reneging exceed the gains from cooperative behavior. I have stated, of course, an extreme form of impersonal exchange, because in the real world, whether present or past (where impersonal exchange did oc- cur to a degree), we find all kinds of mitigating circumstances by which parties attempt to assure compliance. The exchange of hostages, os- tracism of merchants who reneged on agreements, to name two examples, provide incentives to parties to live up to agreements. Reputations, de- pending on the costs of information, provided parties in long-distance trade and impersonal exchange a mechanism to enforce agreements. Kinship ties, various forms of loyalty, minority groups in societies bound together by common beliefs in a hostile world — all provided frameworks within which living up to agreements was worthwhile. In addition, it should be noted that at times and places ideological commitments to integrity and honesty also played a major role. Nevertheless, the dilemma the key problems of institutional change, of contracting, and of performance turn on the degree to which contracts can be enforced between parties at low cost. 55  Institutions that is posed by impersonal exchange without effective third-party en- forcement is central to the major issues of development. II Let me explain this dilemma more precisely by expanding on the game theory framework briefly outlined in Chapter 2.2 There I started with a very simple prisoner's dilemma situation in which it simply did not pay parties in a single exchange or in a single agreement to live up to the terms of exchange. Under these conditions, rational self-interested individuals will arrive at a Pareto inferior solution, that is, one that leaves both parties worse off than they would have been had they cooperated, because it is determined by one of the parties that he or she will be much worse off if the other party chooses not to cooperate. Now, there are ways for the parties to get around such a dilemma. They might ex ante sign a contract agreeing to live up to some set of standards, and they might hire lawyers to see that the contract is enforced. Note that by so doing, however, they have introduced transaction costs to the argument. And the costs of trans- acting, that is of providing the mechanism for assuring that the parties live up to a cooperative solution, may exceed the gains that the parties can derive from it. If we shift from a once and for all game to a repeated game or an iterated game, then the possibility of a cooperative solution becomes much more evident, as Axelrod has explored (1984). That is, if the game continues indefinitely, it usually pays the parties to live up to the terms of exchange, because the gains from successive iterations exceed the benefits that could be derived from a single defection, from "running off with the profits." Note, however, that a game so conceived must be played in perpetuity. If there is an end of the game or people believe that the game might end, then indeed the discount rate may enter in to determining whether it is worthwhile to continue to cooperate. The smaller the proba- bility of continuing for another round, the greater must be the payoffs to sustain an equilibrium; also, the greater the possibility of short-run gains, the greater must be the payoffs. Note that if the game runs continuously, there are still transaction costs, because one must still acquire information about the other party. However, the assumptions that are involved in this cooperative solution are seldom realized in the real world. They entail the game lasting continuously, they entail that one repeats the game with the 2 This section depends heavily upon a substantial literature in game theory of which the essay "Corporate Culture and Economic Theory" by David M. Kreps (forthcom- ing in Alt and Shepsle, editors, Perspectives on Positive Political Economy) was especially useful. 56  Enforcement same players, and they entail that one can observe compliance on the part of the other party. Observing compliance in terms of the model suggests that one can measure unambiguously the outcomes of contracts, so that one can determine whether a party has in fact not lived up to the terms of the contract. This evolving story of game theory tells us that under very simplified conditions, that is, when the parties acquire perfect information and the game both lasts indefinitely into the future and is played between the same parties, one can reach self-enforcing cooperative solutions. But needless to say, these assumptions not only are strong but are simply not observed in the real world. In a world of impersonal exchange, we are exchanging with multiple individuals and can acquire very little informa- tion about all of them. Our information not only is imperfect, but varies remarkably from one party to another. Many times the exchange is a once and for all exchange and not repeated at all. Under these conditions, it is easy to see why the problems that we have laid out here are simply unresolvable in terms of cooperative solutions that can exist in imperson- al exchange. The inevitable conclusion that one arrives at in a wealth- maximizing world is that complex contracting that would allow one to capture the gains from trade in a world of impersonal exchange must be accompanied by some kind of third-party enforcement. Indeed, this con- clusion clearly mirrors the quotation from Norman Schofield at the end of Chapter 2, describing the conditions necessary for equilibrium solutions to emerge in the context of complex, cooperative games. This game theoretic story can be translated into the framework devel- oped in the preceding chapters. In the straightforward neoclassical story, the gains from trade are realized with zero transaction costs. That is, the parties to exchange costlessly know everything about the other party and enforcement is perfect. No institutions are necessary in a world of com- plete information. With incomplete information, however, cooperative solutions will break down unless institutions are created that provide sufficient information for individuals to police deviations. There are two parts to an institution's assuring cooperation. First, it is necessary to form a communications mechanism that provides the information necessary to know when punishment is required. By making available the relevant information, institutions make possible the policing of defections. Typ- ically they economize on information, so, for example, players need no longer know the entire past history of any partner. Second, because punishment is often a public good in which the community benefits but the costs are borne by a small set of individuals, institutions must also provide incentives for those individuals to carry out punishment when called on to do so (see Milgrom, North, and Weingast, 1990, for an elaboration of this argument). It should be stressed that creating an in- 57  Institutions stitutional environment that induces credible commitment entails the complex institutional framework of formal rules, informal constraints, and enforcement that together make possible low-cost transacting. This argument states that the players may devise an institutional frame- work to improve measurement and enforcement and therefore make possible exchange, but the resultant transaction costs raise the costs of exchange above the neoclassical level. The more resources that must be devoted to transacting to assure cooperative outcomes, the more diluted are the gains from trade of the neoclassical model. The more complex the exchange in time and space, the more complex and costly are the institu- tions necessary to realize cooperative outcomes. Quite complex exchange can be realized by creating third-party enforcement via voluntary institu- tions that lower information costs about the other party; ultimately, how- ever, viable impersonal exchange that would realize the gains from trade inherent in the technologies of modern interdependent economies re- quires institutions that can enforce agreements by the threat of coercion. The transaction costs of a purely voluntary system of third-party enforce- ment in such an environment would be prohibitive. In contrast there are immense scale economies in policing and enforcing agreements by a pol- ity that acts as a third party and uses coercion to enforce agreements. But therein lies the fundamental dilemma of economic development. If we cannot do without the state, we cannot do with it either. How does one get the state to behave like an impartial third party? Ill If formal third-party enforcement is essential, it is important to define exactly what one means by it. In principle, third-party enforcement would involve a neutral party with the ability, costlessly, to be able to measure the attributes of a contract and, costlessly, to enforce agreements such that the offending party always had to compensate the injured party to a degree that made it costly to violate the contract. These are strong conditions that obviously are seldom, if ever, met in the real world. It is costly to measure the attributes. The enforcer is an agent and has his or her own utility function, which will dictate his or her perceptions about the issues and therefore will be affected by his or her own interests. Enforcement is costly. Indeed, it is frequently costly even to find out that a contract has been violated, more costly to be able to measure the vio- lation, and still more costly to be able to apprehend and impose penalties on the violator. But achieving third-party enforcement in contracting is a major dilem- ma for economies that would engage in impersonal exchange. The devel- 58  Enforcement opment of credible commitment on the part of political bodies, such that one has assurances that political bodies will not violate contracts of par- ties or engage in conditions that will alter radically the wealth and income of parties, is always relative; even in the most highly developed countries we observe political entities altering the wealth of parties by all manner of changes in the price level or in rules that affect the well-being of indi- viduals. Nevertheless, there is an immense difference in the degree to which we can rely upon contract enforcement between developed coun- tries and Third World countries. In developed countries, effective judicial systems include well-specified bodies of law and agents such as lawyers, arbitrators, and mediators, and one has some confidence that the merits of a case rather than private payoffs will influence outcomes. In contrast, enforcement in Third World economies is uncertain not only because of ambiguity of legal doctrine (a measurement cost), but because of uncertainty with respect to behavior of the agent. Even when enforcement is sufficient to enable elaborate contracts to be made because they are subject to a system of courts that act as a bulwark against the violation of contracts, contractors must take into account those margins at which it is hard to measure whether a contract has been fulfilled. Hence, under conditions of uncertainties with respect to the future or problems of agency for which enforcement is difficult, the con- tracting parties will attempt to structure contracts that will minimize the potential both for contract violation and for rent dissipation by the par- ties. 3 I conclude this analysis of enforcement by pointing out where it is taking us. Third-party enforcement means the development of the state as a coercive force able to monitor property rights and enforce contracts effectively, but no one at this stage in our knowledge knows how to create such an entity. Indeed, with a strictly wealth-maximizing behavioral as- sumption it is hard even to create such a model abstractly. Put simply, if the state has coercive force, then those who run the state will use that force in their own interest at the expense of the rest of the society. Madi- son laid out a solution to this problem in The Federalist Papers and Vincent Ostrom has expanded on it in his theory of the compound re- public (1971); the correct constitutional forms will restrain the tyrannical exercise of political power. William Riker's skepticism, however, still seems appropriate: But another main theme is a celebration of the efficacy of constitutional forms in restraining the tyrannical exercise of political power. This is the theme I wish to 3 See Barzel (1982) for an elaboration of these issues. 59  Institutions address. I guess I want to defend the view he attributes to Woodrow Wilson. My initial tendency, however, as a constitutional theorist raised in the same tradition as Professor Ostrom has always been to agree with his proposition. But everytime I convince myself that I have found an instance in which constitutional forms do make a difference for liberty, my discovery comes apart in my hand. It is, of course, all a matter of the direction of causality. Professor Ostrom believes that at least part of the reason we are a free people is that we have certain constitutional forms; but it may just as easily be the case that the reason we have these constitu- tional forms is that we are a free people. The question is: Does constitutional structure cause a political condition and a state of public opinion or does the political condition and a state of public opinion cause the constitutional struc- ture? This sounds at first like the chicken and egg problem in which there is no causal direction; but I think that usually there is a cause and that constitutional forms are typically derivative. It seems probable to me that public opinion usually causes constitutional structure, and seldom, if ever, the other way around. As Rousseau contended, it is in the end the law that is written in the hearts of the people that counts. (Riker, 1976, p. 13) In a subsequent historical chapter, I briefly describe the way in which such constitutional forms evolved in England in the seventeenth century. But although that story describes a successful outcome, it does not give a definitive answer to the question of how it was achieved. It was surely a mixture of formal and informal constraints. Both respect for the law and the honesty and integrity of judges are an important part of this success story. They are self-enforcing standards of conduct, and I believe that they are important. How does one create such self-enforcing constraints? Part of the answer is that creating a system of effective enforcement and of moral constraints on behavior is a long, slow process that requires time to develop if it is to evolve - a condition markedly absent in the rapid transformation of Africa from tribal societies to market economies. The quote from William Riker goes to the heart of the issue of creating effec- tive institutional constraints. 60