Jei. An Evaluation of the Role of the State and Property Rights in Douglass North s Analysis. Ronaldo Fiani

Transcription

1 Jei JOURNAL OF ECONOMIC ISSUES Vol. XXXVIII No. 4 December 2004 An Evaluation of the Role of the State and Property Rights in Douglass North s Analysis Ronaldo Fiani This paper intends to show that the role Douglass North has attributed in his latest works to the modern State in the working of economic system permits the development of a political economy theory which joins the political and economic dimensions with abstract generalizations quite different from and more interesting than the ahistorical and asocial neoclassical ones, at the same time avoiding a narrow descriptive character. Notwithstanding the fact that some of these ideas have been presented elsewhere, 1 more development is necessary, especially concerning the possibility of spontaneous cooperative behavior to dismiss the role of the State in property rights assignment and enforcement as a necessary condition for modern market functioning. Thus, the next section explores in North s analysis the links between transaction costs, multidimensional property rights, and third-party enforcement in modern societies, which demands the State s intervention not only in the enforcement but also in the definition of property rights, as a necessary condition for market operation. The third section discusses the modern institutional analysis of spontaneous emergence of social cooperation essentially founded on Robert Axelrod s contribution, for its approach seems to contradict the role of the State as a third-party agent to define and enforce property rights. The fourth section discusses the determinants of the State s institutional action to define and enforce property rights according to North s analysis. Then, it will be demonstrated that North s analysis of these determinants is unsatisfactory but states a future research agenda for the study of the relationship between political and economic systems. The last section summarizes the most important questions for future research. The author is Associate Professor in the Economics Institute, Federal University of Rio de Janeiro. He is indebted to Geoffrey Hodgson, Glen Atkinson, and two anonymous referees for valuable comments on an earlier version of this paper. 2004, Journal of Economic Issues 1003

2 1004 Ronaldo Fiani Development and Institutions in Douglass North s Analysis The evolution of North s ideas is quite complex, and it is not our fundamental matter of concern here. We are interested in comparing North s approach in his latest work (during the 1990s) to the present trend of studying only the spontaneous emergence of norms through very abstract models in more orthodox approaches to institutions. However, some very brief remarks on the evolution of North s ideas are necessary, at least to let the unaware reader know about some quite dramatic changes in North s point of viewinthecourseofhiswork. North s first work (The Economic Growth of the United States from 1790 to 1860)was fundamentally concerned with an empirical analysis of the functioning of markets in a process of economic growth. This work would build the link between North and what have been called the cliometric school (also known as the new economic history ), which was characterized by applying quantitative methods and (neoclassical) economic theory to historical analysis. This initial link with neoclassical theory would prove very difficult to break in North s subsequent works. The most important difficulty with early North s neoclassical analysis would be to explain the permanence of inefficient institutions, that is, institutions that do not favor economic growth. 2 The problem of the definition of property rights, its assignments, and enforcement was directly related to the nature (efficient or inefficient) of institutions and has a central role in North s analysis. The problem of the permanence of inefficient institutions would produce perhaps the most important changes in North s thought. In his early work on this issue (North and Thomas 1973), the definition, assignment, and enforcement of property rights always changed in an efficient way, in response to the variations in relative prices of land and labor. These variations were caused by the European demographic cycles during the Middle Ages, in a Malthusian-like cycle. Thus, a reduced population increased the relative price of labor in comparison to land and this, in turn, produced simultaneously a decrease in feudal obligations related to direct labor extraction and an increase in market-oriented usages of land, such as land tenancy. This movement produced population increase and the consequent expansion of agricultural frontier and the growth of commerce, which was caused by the diversity of agricultural products. However, as the stock of potential agricultural land was exhausted, diminishing returns started to operate producing inflation and increasing the price of land relative to labor: the whole cycle was reverted with landlords reducing the number of tenants on their lands and increasing the extraction of direct labor by the means of the imposition of feudal obligations. In this early analysis of the process of economic growth and institutional change, the economic growth causes institutional change mechanically through individuals facing relative scarcity of labor and land. This is quite amazing when one considers that in the same book, North and Robert Thomas discussed the differences between the successful development in England and the Netherlands and the economic failure of Spain and France from the sixteenth to the seventeenth centuries (1973). While in the same

3 Role of the State and Property Rights in Douglass North s Analysis 1005 period, first in the Netherlands and then in England, institutions were created to safeguard and enforce property rights, particularly in capital markets; in France and Spain, absolutist kingdoms made property rights insecure. The result was that while the Netherlands and England succeeded in overcoming the Malthusian check, France and Spain continued to be submitted to population growth cycles. After reading the book, the inevitable conclusion is that it poses at least two problems: 1. Why did France and Spain not imitate the pattern of property rights protection of the Netherlands and England? 2. If the answer to problem 1 involves the State and its political structure, what were the differences? These two problems are quite difficult to handle without a theoretical model, and it is worse when the analytic frame preserves a clear neoclassical nature. The historical evidence points to the survival of property rights which do not favor economic growth for long periods. That could not be reconciled with the notion that property rights change spontaneously, in an efficient way, when factors relative prices change. In his 1981 book, Structure and Change in Economic History, North tried to respond to both problems. North asserted then that [t]he reason for the differential growth rates among the merging nation-states of Europe during the seventeenth century is to be found in the nature of property rights that had developed in each. The type of property rights established was the outgrowth of the particular way each nation-state developed (148). The importance of North s change of approach must not be minimized here. He states that: (a) The concept of property rights is essential to understanding economic growth. (b) The development of property rights is specific to each nation, and it is not subject to a general, ahistorical law of change as it was before, when North considered development a spontaneous response to relative prices change. Before we develop the discussion of (a) and (b), it is important to properly understand the role of property rights in North s new approach. In effect, as we are going to realize, the concept of property rights is central to North s analysis of institutions through the impact of the costs of transacting in long-term growth. The first step is to carefully consider North s definition of transaction costs: It is the cost of measuring the valuable attributes of the goods and services or the performance of agents in exchange that is the fundamental key to the cost of transacting (1992, 7). This definition of transaction costs inherently relates these costs to the measurement and enforcement of property rights, 3 a kind of association not unanimously shared by authors of what became known as the new institutional economics. For example, Ronald H. Coase in his first

4 1006 Ronaldo Fiani article (1937), which coined the term transaction costs, characterized transaction costs loosely as comprising all the costs incurred when a transaction is performed through the market, with no direct reference to what is really being transacted. More radically, Oliver E. Williamson (1990) explicitly denied any relevance to the concept of property rights in studying transaction costs: Albeit illuminating, the property rights approach to economic organization pulls up short. Akin to the economic theory of socialism,whichheldthatthecentralexercisewasto getthepricesright...theeconomics of property rights maintains that the basic need is to get the property rights straight after which markets will reliably assign resources to high valued uses. Arguments of both kinds truncate the study of economic organization unnecessarily (109). Williamson s critique of the concept of property rights is unfair. While it is true that some authors have discussed property rights in a simplistic way, it is also true that others such as Armen Alchian, Yoram Barzel, and North himself have not. The latter understoodthat,aspropertyrightsalwayscomprisemultipledimensionswhichare costly to measure, some of them remain unmeasured and so there is no such thing as perfectly assigned property rights. Property rights assignment and enforcement are and will always be problematic. For that reason institutions are so important: political and economic institutions assign and enforce property rights. This is the first analytical advantage of studying transaction costs by means of the concept of property rights. It must be clear that North acknowledges goods and services multiple attributes, taking into account the difficulty of measuring them and the agents performance as the source of transaction costs. Traditional economical analysis treats goods, services, and performances as essentially one-dimension objects to be reduced to their quantity while solving equations to get their prices what excludes most of the problems economic institutions are concerned with. To recognize that goods, services, and performances have multiple dimensions makes a quite different picture emerge. Yoram Barzel offered strong evidence about the importance of considering multiple dimensions in economic analysis (1997). He discussed the interesting case that occurred in the United States in August Contrary to what the conventional theory would predict, gasoline price control did not result in offer shortage. Instead of reducing quantity, as traditional theory would have predicted, gasoline sellers changed gasoline attributes. In effect, it is not gasoline that gas stations sell. Instead, what a consumer buys in gas stations is a variable bundle of rights associated with the physical good gasoline: the right to buy it twenty-four hours a day, the right to use credit card instead of cash, the right to have convenience services, and so forth. As a result of the price control imposed in August 1971, gas stations kept the quantities sold but reduced the attributes of the bundle of rights associated to it. Gas stations started to open only twelve hours a day, to reject credit card payments, to reduce convenience services and so on, resulting in no offer shortage, to the disappointment of economic previsions (21 30).

5 Role of the State and Property Rights in Douglass North s Analysis 1007 In other words, this evidence shows that once something is transacted what is really sold is a bundle of rights related to the good s or service s multiple dimensions. Since some of these dimensions are not completely specified, some of those rights are poorly defined, not clearly attributed (to use Barzel s words, it is in the public domain) and, consequently, imperfectly enforced, which may cause agents to spend resources in order to capture these rights. The less specified and enforced those rights are, the fewer resources will be spent. That brings the cost problem of identifying and measuring all the relevant attributes of goods and services to the parties involved in the transaction to the forefront of the discussion, since these two steps are previous to the definition, assignment, and enforcement of the rights associated with those attributes. The cost problem belongs to the core of transacting and, more specifically, to the core of making a contract. If the expected cost of identifying and measuring all the relevant attributes of goods and services to the parties involved in the transaction is substantial, property rights will be poorly defined, assigned, and enforced. As a result transaction costs will increase and the expected gains of trade will reduce proportionally. There is a second analytical advantage of North s approach to transaction costs through property rights: this kind of approach analyzes transactions as essentially a transference of property by parties who are allowed to do so, and so does not consider transactions as mere transfers of physical goods. Using John R. Commons terminology, North s institutional analysis is concerned with transactions (transfer of property) and not with exchange (delivery of physical goods) (see Commons [1934] 1990, 55 60). Although North does not distinguish between transactions and exchange as Commons did, North is clearly talking about property transfer. Thus, he is not subject to Commons criticism of orthodox theory, which mixes the property dimension with the physical dimension of trade as if they were the same thing. Therefore, trade gains are directly associated with the definition, assignment, and enforcement of the multiple attributes which constitute a bundle of property rights. The more those attributes are well defined, assigned, and enforced, the more institutions will work as incentive systems to nations economic development. According to North, the simplest and cheapest way of achieving that is through close repetitive personal relationships, as game theory has demonstrated: By personal exchange, I refer to a world in which we deal with each other over and over again in small-scale economic, political and social activity, where everybody knows everybody, and where under those conditions, to use a simple illustration from game theory, it pays to co-operate. That is, game theory says that human beings co-operate with each other when they play a game over and over again, when there is no end game, when they know the other parties to exchange, and when there are small numbers (North 1999, 21). However, North advises us not to expect this kind of social relationship, which emerged since the development of German chemical industry in the second half of the nineteenth century, to be the standard relationship in modern society: The world that it has produced is characterized by impersonal exchange. It is a world in which our

6 1008 Ronaldo Fiani dependence rests upon people all over the world, whom we do not know; there are no repeated dealings; and large numbers of players are involved. Therefore it is a world in which the game is played differently. In game theory, we say such a world is one in which it pays to defect (1992, 21). North s distinction between economic transactions based on personal relationship versus impersonal relationship is even more important to properly understand the institutional economic role North concedes to the modern State. In a previous work, North had already pointed to the importance of that distinction and its consequences: The measured transaction cost of a society where there is a dense social network of interactionisverylow...undersuchconditions,normsofbehaviorareseldomwritten down. Formal contracting does not exist, and there are few formal specific rules. However,whilemeasuredtransactioncostsarelow...productioncostsarehigh,because specialization and division of labour are limited to the extent of markets that can be defined by personal exchange (1989, 1320). For North, the economic system representation, as something ruled by a spontaneous order operating through social norms and culturally consolidated conventions of which individuals are persistently ignorant, seems to fit to a society where the division of labor has not progressed enough, thus making costs of transaction low but costs of production high. This is not the kind of description suitable for modern economic societies: At the other extreme from personal exchange is a world of specialized interdependence in which the well-being of individuals depends upon a complex structure characterized by individual specialization and hence by exchange ties that extend both in time and space. A pure model of this world of impersonal exchange is one in which goods and services or the performance of agents is characterized by many valued attributes, in which exchange takes place over time, and in which there are not repeated dealings. Under these forms of exchange, the costs of transacting can be high, because there are problems both in measuring the attributes of what is being exchanged and problems of enforcing the terms of exchange; in consequence there are gains to be realized by engaging in cheating, shirking, opportunism etc.... Asaresult, in modern Western societies we have devised formal contracts, bonding of participants, guarantees, brand names, elaborate monitoring systems, and effective enforcement mechanisms. In short, we have well-specified and well-enforced property rights. (1989, 1320) In spite of being lengthy, the quotation above is very important to properly settle the limits of economic spontaneous order and its secondary importance to modern developed economic societies. The developed labor division in our time demands a particular institutional frame, which considers the importance of the progressive complexity of economic interactions. This progressive complexity is more institutionally demanding as we recognize that what is transacted is a bundle of property rights with

7 Role of the State and Property Rights in Douglass North s Analysis 1009 multiple dimensions, costly to measure but essential to define the gains of trade. The institutional frame prerequisites to cope with this complex situation is analyzed by North: The institutional requirements that are necessary in order to be able to realize the productivity gains associated with the model of impersonal exchange outlined above entail both the development of efficient products and factor markets and of a medium of exchange with reliable features. The establishment of such a set of property rights will then allow individuals in highly complex interdependent situations to be able to have confidence in their dealings with individuals of whom they have no personal knowledge and with whom they have no reciprocal and ongoing exchange relationships. This is only possible as the result, first, of the development of a third party to exchanges, namely government, which specifies property rights and enforces contracts; and second of the existence of norms of behavior to constraint the parties in interaction. (1989, 1320, emphasis added) From the passage above it is clear that, in North s analysis, the complex social environment characteristic of modern societies makes necessary the presence of the State, not only to enforce property rights but also, and specially, to specify those rights. It begs the question of what the role of social norms and conventions in property rights specification and enforcement is. Since Institutions, Institutional Change, and Economic Performance (1990) North has endeavored to properly set the meaning of social norms and conventions spontaneously developed relating to property rights. In another work, North made it clear that [h]owever they are formed, and however they evolve, norms play a critical role in constraining the choice set at a moment of time as well as in the evolution of institutions through time. They are important at a moment of time precisely because of the costliness of measurement and the imperfect enforcement of rules (1993, 14). It means that ideas and values embodied in social norms and conventions matter; moreover, North observed that [t]hey do so because of slack in the system, agency costs, consumption at the job, and so forth, all of which result from the costliness of measurement and enforcement (1993, 15). It means that norms are important not in absolute terms but in relation to the costs of formally designed institutions. Social norms that claim cooperative behavior are significant for economic growth and development as they reduce the costs associated with designing formal restraints to prevent anti-cooperative actions from individuals. But in modern societies based on impersonal transactions, one should not expect that norms and values alone would solve the problem of assigning, measuring, and enforcing property rights. So, the existence of norms of behavior constraining individuals is only secondary to the confidence necessary to economic relations. It is important to clarify that for North there is no equivalence between the State s action and spontaneous social norms (and conventions). The role of the State is more significant. The relatively growing impor-

8 1010 Ronaldo Fiani tance of the State specification and enforcement in modern societies property rights gives the answer to the permanence of ineffective rights: The answer is quite clear: The breakdown of personal exchange is not just the breakdown of a dense communication network, but it is the breakdown of communities of common ideologies and a common set of rules in which all believe. The rise of impersonal rules and contracts means the rise of the state and with it unequal distribution of coercive power. This provides the opportunity for individuals with superior coercive power to enforce the rules to their advantage, regardless of their effects on efficiency (1989, 1321). ThisprocessbringstheState sactioninpropertyrightstotheforefrontandmakes it possible for individuals and groups to use the State s coercive power in their favor, and the ensuing political bargaining has no connection with economic efficiency. To illustrate this point North gave several historical examples of inefficiently defined property rights that lasted for long periods of time. One of those examples is the case of Spanish Mesta, a shepherd s guild with the privilege of free passage of its sheep through any land property in the country. North argued that privilege had hindered agriculture development in Spain for centuries (1981, ). But the really important point here is that North perceives the progressive formal nature of those institutions, defining and enforcing property rights as a direct consequence of the increasing division of labor. That result moves the State to the center of the institutional analysis of modern societies. But before discussing the role of the State, it is necessary to consider first the very popular analysis of spontaneous emergence of social cooperation, for it seems to deny the role of the State in defining and enforcing rights.thisisournextsubject. Spontaneous Social Cooperation and Third-Party Enforcement The importance of third-party enforcement to achieve social cooperation is not consensual among modern theorists who study the role of institutions. More precisely, the importance North concedes to the State in the working of markets is far from being an established truth in modern institutional analysis. In this section, effort will be made to summarize in a few words the alternative approach to market and society functioning that rejects third-party enforcement as necessary to economic and social interactions. Axelrod s analysis of spontaneous cooperation, which has become the reference for social cooperation analysis, will be discussed. It does not mean that Axelrod s analysis is the only one of this kind, or the only significant one. There is an enormous quantity of papers working on the same kind of approach; the interested reader may refer to Bicchieri 1993, Schelling 1960, Schotter 1986, Sugden 1989, and Young 1998, just to name a few. The reason I decided to concentrate on Axelrod s work is his option for an evolutionary method. This option has attracted a legion of admirers of his work, even critical analysts like Sean P. Hargreaves Heap and Yannis Varoufakis (1995). Axelrod s

9 Role of the State and Property Rights in Douglass North s Analysis 1011 analysis of the evolutionary development of norms of cooperation is perhaps the most popular of its kind today. The plausibility that economic interactions regulation might be provided exclusively by spontaneous mechanisms has been sustained in the economic literature chiefly through formal abstract demonstration (generally derived from game theory concepts) of the spontaneous emergence of cooperation. If cooperation emerges spontaneously then it is straightforward that property rights will be devised (somehow) without State intervention to assure cooperation in the economic sphere of social relations. The consequence is that the analysis of property rights becomes of less importance itself for the efficient property rights structure will be generated more or less automatically. It is important to observe that this kind of answer solves the problem of property rights determination indirectly through the analysis of the evolution of rules of conduct. If social cooperative rules of conduct necessarily evolve, it is possible to expect cooperation in definition and enforcement of the property rights necessary to make the economic system work. Axelrod presented the famous computer tournament of sixty-three strategies proposed by specialists in a repeated game (1984). 4 Axelrod then argued that the superior performance of tit-for-tat 5 in the tournament is a clear demonstration that tit-for-tat is the best strategy and not only for a given environment but also if one admits that the environment evolves in a particular way, when populations of strategies with best performance are allowed to grow their populations according to their accumulated payoffs. To distinguish this particular model of evolution from the generally accepted one, which admits mutation and combination of genes, Axelrod considered that his approach was an ecological and not an evolutionary one (51). Axelrod (1984) was emphatic about the importance of tit-for-tat being nice, niceness meaning that a strategy is never being the first to defect, what in his opinion was a strong evidence of the optimality of being cooperative. Cooperative behavior has a rather specific meaning for Axelrod: you must never be the first one to defect, but you must be always ready to retaliate against a defection. For Axelrod, cooperation admits retaliation, but again you must be always ready to forgive any defection if the other player cooperates again. If you are not ready to retaliate against a defection, you may be vulnerable to exploitative strategies. However, if you are not ready to forgive, you may be trapped in a succession of endless retaliations that would lower both your and the opponent s payoffs. These conclusions about spontaneous emergence of cooperative behavior were reinforced under more sophisticated hypothesis in Axelrod As it was mentioned before, Axelrod (1984) did not develop a true evolutionary model and this most recent work was devised to elaborate a true evolutionary analysis, among other relevant extensions of his previous results. The elaboration of a true evolutionary analysis demands some mechanisms of combination and mutation of strategies. That was accomplished through genetic algorithms (1997, 17 20). Axelrod described the results so obtained: The results were quite remarkable: from a strict random start, the genetic algorithm

10 1012 Ronaldo Fiani evolved populations whose median member was just as successful as the best rule in the tournament, tit for tat. Most of the strategies that evolved in the simulation actually resemble tit for tat, having many of the properties that make tit for tat so successful (1997, 20). The conclusion seems clear: for Axelrod tit-for-tat was not only the best strategy in simplified environments but also a benchmark for performance evaluation even in a complex world with mutations and crossover. The more a rule of behavior resembles tit-for-tat, the more successful it will be in an evolutionary selection process. Thus, for Axelrod, the viability of spontaneous social cooperation emergence through evolutionary mechanisms was demonstrated, even in the unfavorable environment of prisoners dilemma. Axelrod s conclusions became a milestone in institutional analysis and tit-for-tat a reference for the study of spontaneous cooperation emergence beyond the strict realm of computer tournaments, as Christina Bicchieri clearly asserted: The very characteristics that make tit for tat successful in computer tournaments are also likely to play an important role in all those prisoner s dilemma like circumstances in which there is repeated interaction but the players are uncertain as to the opponent s character. Especially when the number of repetitions is small, and experimenting more costly, tit for tat seems likely to be chosen by a player who wants unambiguously to signal his intentions and benefit from the possibility of joint cooperation (1993, 245 6). If Axelrod s simulations were decisive in demonstrating that the emergence of spontaneous social cooperation is a strong possibility, North s dichotomy between small and less developed societies with no third-party enforcement of property rights and modern societies with a necessary third-party enforcement of rights (through State s action) would become a serious mistake, but this does not seem to be the case. Actually, game theorists have strongly criticized Axelrod s simulations. Kenneth Binmore (1994, 1998) surveyed the worries about Axelrod s results. The most important are: 1. In Axelrod 1984 the selection process could have converged to any of the large number of Nash equilibria of the game resulting from the sixty-three strategies proposed, provided that the initial strategy distribution was selected suitably. 2. Actually, the good performance of tit-for-tat is significantly dependent on the initial population of strategies, for with a different starting set of strategies (and a random initial distribution of strategies) tit-for-tat becomes completely overcome by other strategies, which do not necessarily show nice properties. 3. Contrary to Axelrod s (1984) reasoning, it is perfectly possible that mean strategies overcome strategies with nice properties as tit-for-tat. For example, if a population of dove strategies (meaning always cooperate ) is invaded by a population of tat-for-tit (the same of tit-for-tat except for that tat-for-tit begins defecting and tries to exploit its opponent), the final population may consist only of tat-for-tits, which cannot be invaded and is not a nice strategy.

11 Role of the State and Property Rights in Douglass North s Analysis Finally, even in the case of all individuals being tit-for-tats, tit-for-tat is not evolutionary stable; it is vulnerable to invasion by simplifying mutations of itself as dove strategy. Whereas tat-for-tit cannot be invaded by similar simplifying mutations, neither dove strategy nor hawk strategy (meaning always defect) may invade tat-for-tit. Considering the above criticism, Axelrod s results seem far from demonstrating the spontaneous emergence of social cooperation. There are neat reasons to doubt the idea that evolution will select a specific pure equilibrium in an infinitely repeated game be it tit-for-tat or other specific strategy. Considering the question about which equilibria will be eventually selected by evolutionary forces, Binmore concluded that: One cannot say anything definitive about precisely which of these evolution will select, because this will depend on accidents of history about which little information is likely to be available. However, the evidence from simulations does strongly suggest that it is unlikely that a pure equilibrium will get selected. One must anticipate the selection of a mixed Nash equilibrium. When such equilibrium is realized, many different rules of behaviour for the indefinitely repeated Prisoner s Dilemma will survive together in a symbiotic relationship. However, as all married folk know, to be part of a symbiotic relationship does not imply that pure harmony reigns. (1994, 203) If it is not demonstrated that a nice rule of conduct will necessarily emerge and dominate through the operation of evolutionary forces, what is the correct approach to the role of norms and conventions in institutional analysis? Geoffrey Hodgson answered the question: [M]any important legal rules, enforcements and structures cannot emerge spontaneously through individual interactions. They require additional third party enforcement by the state or another strong institution. This means that many key institutions and legal relations, arguably including property and markets, exist as a result of a combination of spontaneous and statutory mechanisms. The existence of one institution has to be considered in relation to the others that help to support and sustain it. Institutions are generally and inevitably intertwined, and provide essential mutual support for one another. (2002, 13, emphasis added) The acknowledgement of the interconnection of formal and informal institutions, the last ones spontaneously generated as social norms and conventions, was expressively stated by North when he discussed the pervasiveness of informal constraints and the fact that similar constitutions imposed on different societies result in very distinct outcomes (1990, 36). It is a much richer approach to institutional analysis to admit the interface between formal and informal institutions than to try to demonstrate that the only kind of institution relevant to social analysis is the spontaneously generated one.

12 1014 Ronaldo Fiani Even recognizing the relevance of social norms and conventions, they must be studied in their interaction with legal rules established by the state. As Hodgson stated: Individual property is not mere possession; it involves socially acknowledged and enforcedrights...itisnotsimplyarelationbetweenanindividualandanobject.it requires some kind of customary and legal apparatus of recognition, adjudication and enforcement. Such legal systems made their first substantial appearance within the state apparatuses of ancient civilisation. Since then the state has played a major role in the establishment, enforcement and adjudication of property rights (2002, 6). We have until this point only discussed the relevance of the State for property rights definition and enforcement. Now will be discussed the determinants of the State s action in establishing, adjudicating, and enforcing property rights according to North. State s Institutional Action according to North The historical example of Mesta shows the relevance of the State s intervention in the definition and assignment of property rights. For North, the State is responsible for economic performance through property rights specification: A theory of the state is essential because it is the state that specifies the property rights structure. Ultimately it is the state that is responsible for the efficiency of the property rights structure, which causes growth or stagnation or economic decline (1981, 17). Initsearlierformulation,North sconceptofthestatewasnotconceivedasapolitical arena, where different interests meet and confront themselves, but as an agent with its own objectives, at least between certain limits: In fact, the property rights which emerge are a result of an on-going tension between the desires of the rulers of the state, on the one hand, and the efforts of the parties to exchange to reduce transaction costs, on the other. This simple dichotomy actually is anything but simple, since the parties to an exchange will devote resources to influencing the political decision makers to alter the rules. But at least as an initial starting point for theorizing, it is useful to separate a theory of the state from a transaction cost approach to property rights (1981, 18). To perform that function (to sell justice and protection), the State monopolizes the definition and enforcement of property rights (North and Thomas 1973, 97). The payments for those services are the State s taxes. That exchange is advantageous in the sense that the State shows economies of scale in those services and it would consequently be more expensive for private agents to provide justice and protection for themselves. As long as those economies are not exhausted, the enlargement of the State s definition and enforcement of property rights increases society s revenues and produces savings to be divided between the State and society. Obviously there will be a distributional conflict concerning this division: both state and society will try to enlarge their own share at the expense of the other. In a preliminary analysis of this distributional conflict, North and Thomas observed that [t]he more monopoly power an existing prince could

13 Role of the State and Property Rights in Douglass North s Analysis 1015 claim that is, the less close or threatening were his rivals the greater the percentage of rents which the state could appropriate (1973, 98). That first approach will be developed more deeply in what North called his neoclassical State model (1981). Its starting point is the hypothesis of State rulers who maximize their utility. North specified the behavior of rulers who will act as discriminator monopolists making distinctions among the interest groups, while defining and attributing property rights in order to maximize his revenue. Nevertheless, any ruler will have limits to his discriminating behavior because there will always be rivals to offer the same kind of services (23). About the ruler s discriminating behavior North stated that: The basic services that the state provides are the underlying rules of the game. Whether evolving as a body of unwritten customs (as in the feudal manor) or as a written constitution, they have two objectives: one, to specify the fundamental rules of competition and cooperation which will provide a structure of property rights (that is, specify the ownership structure in both factor and product markets) for maximizing the rents accruing to the ruler; two, within the framework of the second objective, to reduce transaction costs in order to foster maximum output of the society and, therefore, increase tax revenues accruing to the state. (1981, 24) North observed that (1) those objectives are not obligatorily consistent because the set of property rights rules which maximizes the ruler s revenue is not necessarily the same as that which maximizes social product (1981, 24 25) and (2) there are agency problems between the ruler and his bureaucracy and therefore some rent dissipation will always happen in consequence of a collusion among bureaucrats and constituents (27). The services offered by the ruler have different cost curves; therefore some of them are essentially public goods while others show the conventional U-shaped average cost curve (25). About the State s institutional action, North concluded that [t]he ruler will specify a set of property rights designed to maximize his monopoly rents for each separable part of the economy by monitoring and metering the inputs and outputs of each. The costs of measuring the dimensions of the inputs and outputs will dictate the various property rights structure for the diverse sectors of the economy, which therefore will be dependent on the state of the technology of measurement (26). It is important to understand that North defined the State s objective of maximizing revenue taxes as the factor that explains the relationship between the definition of property rights and measurement technology while recognizing that there are limits to that kind of action: The ruler always has rivals: competing states or potential rulers within his own state. The latter are analogous to the potential rivals to a monopolist. Where there are no close substitutes, the existing ruler characteristically is a despot, a dictator, or an absolute monarch. The closer the substitutes, the fewer degrees of freedom the ruler possesses, and the greater the percentage of incremental income that will be retained by the constituents (1981, 27).

14 1016 Ronaldo Fiani The competition between the ruler and potential substitutes is motivated not only by tax revenue distribution; definitions of property rights and services offered are also a matter of concern to interest groups. They are influenced by interest groups bargaining capacity: Constituents may, at some cost, go over to a competing ruler (that is, another existing political-economic unit) or support a competitor for ruler within the existing state (North 1981, 27). Unfortunately, North did not satisfactorily develop the line of reasoning to the modern State case. At the same time North presented his neoclassical theory of the State, he admitted its weakness when considering modern democracies: Themodelofthestateisdeficientinotherwaystoo,butitisparticularlydeficient when we move from a single ruler to the modern pluralist state. A theory of the resolution of the conflicts in such a state has baffled modern political scientists (68). In his last book, Institutions, Institutional Change, and Economic Performance (1990), North showed a clear effort to enrich his State analysis from the somewhat simplistic scheme of a maximizing ruler. He remarked that [t]he existing structure of rights (and the character of their enforcement) defines the existing wealth-maximizing opportunities of the players, which can be realized by forming either economic or political exchanges. Exchange 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 (47). The preceding statement shows that individuals strategically consider the possibility of investing in political changes in order to redefine property rights and, therefore, existing gains opportunities. It is, of course, completely different from the neoclassical model of State in which a maximizer ruler negotiates property rights definition and public services in exchange for tax revenues. North (1990) presented a new State concept with a multiplicity of agents that may invest in changes in the polity structure, which defines and enforces property rights. He recognized the neoclassical model of State as a first approach and a simplified model (48), and he explicitly admitted the historical limitations of the neoclassical model of State: When we move from the historical character of representation in early modern Europe to modern representative democracy, our story is complicated by the development of multiple interest groups and by a much more complicated structure devised to facilitate (again given relative bargaining strength) the exchange between interest groups (49). So, the representation of the modern State as a tax revenue maximizer ruler acting like a standard economics textbook discriminating monopolist is rejected by North. This is not a convenient model when one studies modern democracies, characterized by multiple interest groups and very complex institutional structure. That institutional complexity and interest groups multiplicity work against the idea of a single maximizer ruler: Because 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 make agreements with other legislators, with different interests (1990, 50).

15 Role of the State and Property Rights in Douglass North s Analysis 1017 The presence of multiple interest groups and legislators puts up the question of how those legislators might behave. That s a rather difficult question for North as it involves the analysis of free rider behavior and values such as justice and equity. About them North observed that [t]hese issues appear to show in the voting behavior of legislators;itiswidelyobservedthatonecannotexplainthevotingbehavioroflegislators within the narrow confines of a principal/agent model, in which the agent (the legislator) is faithfully pursuing the interests of the principal (constituents). The agent s own utility function his or her own sense of the way the world ought to be appears to play a role in the outcomes (1990, 21). But legislators who assign property rights are not the only agents to constitute the State. These rights have to be enforced, and this is the function of the judiciary for [r]ules are enforced by agents (police, foremen, judges, juries, and so on) (North 1993, 14). He admitted the same kind of difficulty, identified before with regard to legislators, influencing judges decisions: judges with lifetime tenure can and do vote their own convictions, as even the most casual study of courts in general and the Supreme Court in particular testifies (1993, 18). It must be understood that North s rejection of his neoclassical model of State does not mean a rejection of the idea that legislators and judges in a democracy have some degree of autonomy when taking their decisions. North considers the representation implicit in agent-principal political analysis of modern State as a political arena, where the legislator or judge is simply an agent of interest groups (his principals), as incomplete and so incorrect. The legislator/judge has his own normative view of the world and will try to implement it whenever it looks possible and desirable to him. And the possibilities of acting autonomously are not rare, since modern democracy increases the number of political exchanges and what became known as rational ignorance from the voter 6 (1990, 51). It brings us to the last question: the autonomy of legislators and judges and their relationship with each other and with the interest groups supporting them. Rather deceptively, North s ideas about this question are quite superficial. He has limited himself to agreeing with William H. Riker (1976) when the latter showed skepticism about the possibility of constitutional forms preventing arbitrary use of power (North 1990, 60). But this is not an answer to the question of why legislators (and to some degree judges) have arbitrary power in some circumstances, whereas in other ones they must act exactly according to the groups supporting them. It is inevitable to agree with Itai Sened (1997, 47) when he stated, North advanced the transaction costs theory to recognize the importance of political structures but, so far, has fallen short of complementing it with a theory of the state.

16 1018 Ronaldo Fiani Conclusions: Elements for Future Research The recognition of the role of the modern State in property rights specification demands a theoretic analysis of its political activities regarding those rights. North began with his neoclassical model of State which he would later reject, considering it too simplistic to understand modern democracies, which present complex networks of political institutions connecting legislators in a process of political bargain among themselves and interest groups. Jack Knight (1992) criticized North s approach, which considers institutions as being determined by the effort to reduce uncertainty in human interactions. Economic and political institutions would be the result of political bargain among groups with distributive purposes. Knight s criticism brings to the forefront the political process through which economic institutions are created. The acknowledgement of political processes and structures relevance to institutional analysis brings back questions that political scientists have been working on for a long time. Their work gives a picture of the State more complex than North s neoclassical model of the State. Particularly interesting is the work by Joel S. Migdal (1987, 1988), when he discussed why some States seem strong enough to implement their rules while others are so weak that they have to accommodate almost all of its pressure groups and regional leaders. His research points to two different factors. The first factor is the incessant competition between the State and other political organizations to establish their political dominance and the resources the State may employ in this competition. Contrary to North s early view, the State is not necessarily the single national ruler; it has to fight for political power against other political organizations or regional strong leaders, remarkably in the Third World. The second factor is that States are inserted differently but necessarily in the economic and political international arena. It results in relations of trade, borrowing, factor movements, and so on, that affect not only the social structure in each country but also the constraints the State faces and the resources it has to act politically. Political bargain is not usually motivated by economic efficiency and, therefore, the kind of property rights assignment that will result from it is a matter of economic concern. It poses two questions for a future agenda of research when one admits the role of institutions in economic development and growth. First, it is not clear when legislators have to serve interest group demands and when they are free to follow their own values and objectives. Second, when legislators have significant degrees of freedom to follow their own ideals, what are those ideals and how they come into play? Thirdly, when legislators and politicians have to serve interest groups, what determines which groups will be served? The first question relates to legislators legitimacy, a concept largely discussed in political theory but not appropriately considered by economists, even economists who recognize the importance of institutions and political structures to explain economic performanceinthelongrun.why,inthesamesocieties,aretheremomentswhenpoli-