The New Institutional Economics: Its Start, Its Meaning, Its Prospects

That institutions matter for economic performance is an old and inherently plausible intellectual position. However, during the first half of the twentieth century, as the mathematical development of neoclassical theory has progressed and economic models have become increasingly abstract, institutional phenomena have received less and less attention. Thus, institutions play virtually no role at all in what may be regarded as mainstream theory through the 1980s (exemplified by welfare economics and the general equilibrium models of Arrow-Debreu). In other words, it does not matter whether goods and services are exchanged by the use of money or otherwise (Samuelson 1968) or whether production is organized by the price mechanism across markets or within a hierarchically organized firm (Coase 1937) and so forth. Not surprisingly, these extreme views were soon opposed by various strands of a renewed kind of economic institutionalism. Among its outstanding contributors were Coase (1937, 1960), Alchian (1961), Buchanan and Tullock (1962), Olson (1965), Williamson (1971, 1975), and North and Thomas (1973), to name but a few of the pacemakers. Linked with these names are theories such as property rights analysis, the economic analysis of law, public choice theory, constitutional economics, the theory of collective action, transaction cost economics, the principal-agent approach, the theory of relational contracts, and comparative economic systems. Common to all these approaches is that, unlike neoclassical economics, they do not assume the institutional framework as given but make it into the object of research and also seek to consider the implications for economic behaviour of any given institutional arrangements. They do not ask “What would self-interested rational actors do?”, but “How could rational actors be constrained (or constrain themselves) not to pursue their self-interest?” (Miller 1997: 1196). The term “new institutional economics” is used as a generic term by different authors for different combinations of the above-mentioned approaches as well as other fields. In this chapter, we briefly survey the history of the kaleidoscopic use of this term to illustrate the evolution of a new and, as we are convinced, widely applicable bundle of methods of economic research.

Oliver Williamson, creator of the term “new institutional economics”, interprets the term rather broadly. He subsumes under the NIE: aspects of mainline micro theory, economic history, the economics of property rights, comparative systems, labour economics, and industrial organization.

The common threads tying these studies together were:

A few pages later Williamson (1975, 7) describes the principal differences between the earlier literature and his approach as follows:

In his work, Williamson concentrates on what he later calls transaction cost economics (TCE), which he says is ‘part of new institutional economics’ (Williamson 1985b: 16).

The term NIE seems to have remained largely dormant during the following 5 years or so. However, Leonard Silk mentioned the term in the New York Times of September 24, 1980 as a possible ‘new direction that will gradually draw economists away from their tired repetition of stale and sterile arguments.’² I myself ‘discovered’ it around the same time on page one of Williamson’s 1975 book. Shortly before, I had become editor of the time-honoured Zeitschrift für gesamte Staatswissenschaft (founded in 1844), at that time still a German language journal on general economics.³ I wanted to ‘internationalize’ the Zeitschrift and looked for a suitable specialty, a niche that remained close to its original field, the “entire science of the state.”⁴ Public choice and law and economics were already well covered by journals, so the new institutional economics in the sense of Oliver Williamson appeared to be a promising choice. Eirik G. Furubotn was prepared to help me and, without thinking twice, we jointly started a series of international seminars on the NIE with the aim of helping the old Zeitschrift on its feet. That was in 1983. The series became known as “Wallerfangen Conference” and is continued to being organised after my retirement since 1995 by different people and in different locations every year. Papers and discussions are published in the Journal of Institutional and Theoretical Economics (JITE) since 1984.⁵

Among the various approaches to institutional economics, two strands of thought are of interest in relation to the assessment of the NIE. Roughly speaking, they are as follows:

The two lines of thought are by no means to be taken as an indication of the historical roots of the NIE. That is a rather complicated issue, which cannot be dealt with here. As for the history of the NIE itself we limit ourselves to the single question, namely how the term ‘new institutional economics’ evolved to a unifying standard.

As mentioned above, the term ‘new institutional economics’ became a more widely-known term around 1980, i.e., 5 years after it had been coined by Williamson (1975: 1). It took another 3 or 4 years until economists used it in the topics or titles of their publications. In EconLit, the term NIE does not appear before 1984, but after that it appears increasingly in titles of journal articles, books, papers in collective volumes and conference publications.⁷ One simple way to check the pulse of our profession is to read editorial prefaces. Editors are supposedly the brokers (or switchmen) of our profession, so why not apply their judgments as a measuring instrument for the development of the use of the term NIE. Following this strategy, I am going to offer a brief report and evaluation of the editors’ comments to six successively published collective volumes, starting with the first publication in which the term NIE appeared, edited by Furubotn and Richter (1984a), and continuing with the collective volumes edited by Langlois (1986), Nabli and Nugent (1989a, b), Harriss et al. (1995a, b), Drobak and Nye (1997a, b), and Clague (1997).

With the foundation of the International Society for New Institutional Economics (ISNIE), whose first conference was held at Washington University, St. Louis, Missouri, on 19–21 September 1997,¹⁴ the NIE came of age. The organizers advertised in the Internet the conference to scholars ‘working on transaction costs, contracting, political rules of the game, the rules of law, norms and culture, and who pursue these interests using standard scientific methods.’ (Furubotn and Richter 1997: 780).

At the inaugural conference, North, Williamson and Coase (in that order) basically described the NIE as has been described above. Although the immediate objective of the NIE is to replace the abstract, static models of neoclassical economics, Coase cautioned that a frontal attack on neoclassic models would be neither needed nor desirable. For appropriable theoretical development to be achieved, it is sufficient to focus on factual matters. Somewhat later Coase (1999: 1 ff.) noted that ISNIE would have a mission, namely ‘to replace the current analysis with something better, the new institutional economics…..The Influence of the new institutional economics will be exerted in the various sub-disciplines [of economics]. Guerrilla actions will take place, which will result in the new institutional economics dominating first one and than other of these sub-disciplines, as indeed is beginning to happen.’ But he added: ‘We will not replace price theory (supply and demand and all that) but will make it vastly more fruitful.’

A selection of the papers presented at the second annual conference of ISNIE at Paris in September 1998 has been published by Ménard (2000a, b), among these papers are contributions by the old guard—Ronald Coase, Douglass North, Oliver Williamson, Harold Demsetz, Yoram Barzel—but also by Masahiko Aoki who supports the ‘institution-as-equilibrium-of-a-game’ concept. Coase and North stress again that the representatives of the NIE don’t want to replace neoclassical theory but are trying to use these analytical tools ‘to study the working of the economic system’ (Coase 2000: 4 and 1999: 5 as quoted above). But that appears to contradict North’s concept of the ‘cognitive science/institutional approach to history’ or Williamson’s casuistic technique to broach bounded rationality. Why not admit that a change in paradigm is already under way?

Williamson (1975: 1) introduced the NIE as a generic term for a divers group of already existing modern economic studies of institutions. The term was later mentioned here and there, but it took 8 years until conference organizers first used it. After 10 or 15 years it became a standard under which economists who are interested in aspects of the economics of institutions assembled. As it turns out, the term NIE is now generally used as a name of what we call the ‘visible-hand’ approach to institutional economics.

Table 1.1 shows the nine modern institutional economic fields, which have been described as part of the NIE in the above-mentioned literature. These fields were developed by different scholars during the 1960s and early 1970s. Core fields of the NIE are:

There are reasons to follow Nabli and Nugent (1989a, b) and to include the theory of collective action (Olson 1965) in the NIE. While property rights economics and the formal parts of contract theory still assume perfect rationality, this is not the case with transaction cost economics and the new institutional approach to history of North.

Williamson (1975: 4) assumes men to be only boundedly rational, North (1995a, b: 18 f.) writes that a theory of institutions has to begin with a ‘modification of the instrumental rationality assumption’ and Coase (1984: 231) regards the assumption ‘that man is a rational utility maximizer’ to be both ‘unnecessary and misleading.’ The assumption of perfect rationality is thus being abandoned by leading neo-institutionalists.

Two branches of the NIE developed after Coase (1937, 1960): the transaction cost economics (TCE) of Williamson (1985a, b) and the new institutional economics of history (NIEH) in the sense of Douglass North (1986). Next to Coase, Williamson and North became the two leading representatives of the NIE. The difference between the ‘Williamsonian’ and the ‘Northian’ approach is described as follows by Ensminger (1992: 21 f.).

While Williamson (1981) ‘takes the position that institutions are expressly designed to reduce transaction costs and that, in competitive markets, those that fail to do so will not survive.’ On the other hand, North (1981: Chap. 3, 1990: 8, 52) argues vehemently against the view ‘that institutions are created only to reduce transaction costs and increase economic efficiency.’ The reasons for inefficient institutions are inefficiencies of political markets, ‘democracy in polity is not to be equated with competitive markets in the economy.’ (North 1990: 51).¹⁵

However, both approaches hardly square with the constantly referred to neoclassical economics and seem unwilling to make the break to a fundamentally new paradigm.¹⁶ While this is not achieved, one should at least be ready to apply different, problem related analytical methods for different objects of economic research and define the science of economics by its objects, instead of its method, as is done in other applied sciences. Economics is more appropriately compared to engineering or medicine than to physics or biology (see Varian 1993: 2 f.).

The Williamsonian approach to the NIE is mainly applied in microeconomic fields like the theory of firm, of industrial organization, antitrust and the economics of organization. The Northian approach is used to tackle macroeconomic issues such as the history of national economies, of development economics and transformation economics.

Oliver Williamson points out, inter alia, that non-standard contracts may—but need not—result from monopolistic practices. The reason is that transaction specific-investments can play an essential role after the conclusion of a contract. Williamson illustrates this by using the concept of fundamental transformation. After contract conclusion, the parties find themselves locked into a bilateral monopoly situation, whereas before they were free to choose with whom to trade. Transaction-specific investments of whatever kind (if only in the form of time invested in search, inspection, and bargaining) are the reason for this transformation. In addition, Williamson takes into account the fact that we do not know what the future will bring. Thus, adaptability to the unforeseen becomes an issue. Under Knightian uncertainty it is impossible to write a complete contract that details all possible future contingencies.¹⁸ Therefore, contracts are unavoidably incomplete. They contain loopholes and the lock-in of the parties may invite opportunistic behaviour by the other side because the parties may be unable to verify their case to a third party (e.g., a court) due to information costs (a special kind of transaction costs). Court ordering may thus have to be supplemented or even substituted by private ordering to protect the parties effectively against opportunism of their trade partners. There are various ways to organize the governance structure of a contractual relationship. Their efficacy depends on the particular circumstance, including the size of specific investments and the frequency of transactions between the parties. To sum up, one may say that transaction cost economics (or “the economics of governance” as Williamson 2005 also calls it) joins three fundamental concepts: adaptability, governance, and transaction costs.¹⁹ TCE has been supported by numerous empirical studies (for an overview, see Shelanski and Klein 1995 or Boerner and Macher 2002 ²⁰).

Williamson’s TCE (or economics of governance) is a theory of contracts under conditions of uncertainty and asymmetric information, in which un-programmed adaptations matter.²¹ Legal enforcement and self-enforcement complement each other with the aim ‘to design workable order-preserving mechanisms for adapting to disturbances.’²² Both court ordering and private ordering characterize the governance structure (or ‘organization’) of non-standard contractual relationships. Attentive actors agree before they come to terms on a governance structure that they regard suitable. Market and hierarchy are two of the imaginable ideal types of possible governance structures. It is important to see that the choice of an efficient (or better: efficacious) governance structure results not from optimizing some target function subject to a set of constraints. It may rather be understood as a form of boundedly rational or ‘suitable’ choice from a set of governance structures in the sense of Selten’s hypothesis of the casuistic structure of boundedly rational strategies (see Furubotn and Richter 2005: 180).²³ Which governance structure the parties choose depends on the particular situation. The problem for the parties is then to agree about both the ‘right’ diagnosis (of the situation) and the ‘best’ cure (the governance structure). Williamson’s (1985a, b: 79) table of ‘efficient governance’ may be understood as an example of how to think, not as an answer to the parties’ decision problem. The idea is, to think less like a physicist and more like a physician.

Douglass North aims at an economic explanation of the structure and performance of economies through time.²⁴ He starts from the simple observation that human cooperation requires rules of behaviour, namely institutional constraints, which in the final analysis define the opportunity set of individuals.²⁵ Because of transaction costs, the nature of the institutional framework plays a major role in the performance of an economy.²⁶ It reduces the uncertainty of human interaction and, thus, the costs of cooperation.²⁷ Persistent changes in relative prices, due to lasting exogenous changes (such as changes in total population, knowledge or ideology²⁸ or to competition among organizations²⁹) will lead actors to realize that they could be better off under alternative institutional arrangements and result in institutional change. In general, institutional change will be incremental and path dependent—incremental because large changes will create too many opponents, and path dependent³⁰ because institutional change is governed by the kind of knowledge and skills that actors have invested in.³¹ Due to high (sunk or variable) transaction costs, inefficient institutions may persist for long periods of time.³²

Institutional constraints comprise informal and formal rules of behaviour.³³ Informal rules comprise norms, conventions, and internally held codes of conduct (North 2005, 50). Formal rules consist of political rules (example: the constitution³⁴), economic rules (examples: property rights,³⁵ contract laws) and contractual agreements between actors (example: sales contract). Political rules lead to economic rules, ‘though the causality runs both ways.’³⁶ The enforcement of rules matters. Self-enforcement would be ideal but is frequently not promising.³⁷ In general, legal enforcement is more effective.³⁸ However, the coercive force given to the State may be used by those in power to their own advantage.³⁹ In a simple model, North (1981, 28) interprets the State as a ruler who maximizes his profit subject to two basic constraints: the degree of political competition with rivals and with other States, and transaction costs. For both reasons, the property-rights structure, which maximizes the social product, may not maximize the ruler’s (long-term) monopoly rents. North’s judgment is rather pessimistic. He argues that, in order to stabilize his power, the ruler will agree to a property rights structure that is favourable to those groups with close access to alternative rulers, regardless of its effects upon efficiency. And because of the costs of determining and collecting taxes, a less efficient property-rights structure may be more favourable to the revenue-maximizing ruler.⁴⁰

North’s concept of the NIEH aims at a general theory of the interaction between polity and economy. It is, to this extent, an application of the economic theory of politics to economic history. However, in contrast to public choice theory and the theory of collective action,⁴¹ North assumes imperfect individual rationality and emphasizes the role of ideology. He rejects ‘rational choice and efficient market hypotheses’.⁴² Instead, he opines that because of imperfect individual rationality

mental models, institutions and ideologies all contribute to the process by which human beings interpret and order environment. Mental models are, to some degree, unique to each individual. Ideologies and institutions are created and provide more closely shared perceptions and ordering of the environment. (Denzau and North 1994: 21)

Both approaches deal with the same object: the institution. Williamson (1985a, b) prefers to speak of ‘governance structure,’ North (1981, 1990) uses terms like ‘rules of behaviour’, ‘institutional constraints’ or ‘structure.’ In addition, North (1990) distinguishes between ‘institutional arrangements’ and ‘institutional environment,’ the former being a subset of the latter, which are a set of fundamental political, social, and legal ground rules that govern economic and political activity (Davis and North 1971: 133). North stresses the role of ideology. Thus, we may note as follows (in a strongly simplified manner):

TCE abstracts from the interaction between economic and political decision-making. It takes norms, customs, mores, tradition and so forth as givens, the latter with the argument that ‘institutions at this level change very slowly—in the order of centuries or millennia’ (Williamson 2000