Transaction Cost Economics and Beyond

Transaction cost economics (TCE) has positioned itself in the center of the economics of organization. From Williamson (1975) onwards TCE has made important progress both in conceptualization and in empirical testing (see Groenewegen and Vromen in this volume). The position of TCE at this moment is outlined in the contribution of Williamson in this volume. The key concepts are of a technical (asset specificity), of a human (bounded rationality) and a behavioral nature (opportunism). The general strategy out of which TCE works can be summarized as follows: After having characterized the transaction, the potential governance structures are discussed in terms of transaction cost minimizing capabilities. The match of transactions with governance structures is formulated in reduced form hypotheses to be confirmed in empirical research.

Viewing the firm as a governance structure, rather than as a production function, has had numerous ramifications. For one thing, the boundary of the firm is no longer defined by technology but is something to be derived (from comparative transaction cost considerations). For another, marginal analysis gives way to discrete structural analysis (Simon, 1978; Williamson, 1991). Also, attention is focused on much more microanalytic features of transactions and organization. The roles of power and authority come under renewed scrutiny in the process.

Reading the above quotations, one might reasonably conclude that no progress at all had occurred in the development and testing of transaction cost arguments in the interval between Fischer’s and Simon’s appraisals. But while Fischer may have been justified in his criticism, Simon clearly is not. Economists’ initial pessimism about the prospects for deriving testable implications from transaction cost reasoning has turned out to be grossly misplaced. Theoretical advances beginning in the 1970s spurred a profusion of empirical research that continues unabated. As recent surveys amply document (see Joskow, 1988; Shelanski, 1992; Crocker and Masten 1994), “the research that must be performed to estimate the exogenous parameters and to test the theory” is well underway.

This chapter is a brief and non-exhaustive review of how transaction cost economics has been used in the field of marketing, where it has diffused rather thoroughly and has achieved a considerable degree of acceptance. This review focuses on how an observer might view the version of TCE that marketers have enfolded within their own eclectic, pragmatic, managerial, and above all, empirical paradigm. Empirical research practice within marketing revolves around what an economist might view as a curious blend of values from econometrics, biometrics, sociometrics, and psychometrics. Thus armed, marketers have boldly taken TCE where no economist has dared to go. After reading this exposition, some may conclude that marketers have baldly taken TCE where no economist ought to go. I, however, intend to argue that marketers have used TCE to derive not only useful insights into marketing problems but valuable perspectives on TCE itself, in particular in terms of the explanatory mechanism underlying TCE.

The traditional structure of public utilities has been one of extensive vertical integration, where multiple stages of production were conducted under the umbrella of the regulated firm. Prices for the final product were generally determined by rate-of-return regulation, which in theory permitted the firms to price in a fashion that generated revenues sufficient to cover variable costs and allowed a fair return to stockholders on their capital investment. Pricing of products at intermediate levels of production became a concern only to the extent that it could be used as a tool to circumvent the regulatory process, as in cases where the firm might set such prices strategically in order to shift costs from regulated to unregulated subsidiaries¹ or when political concerns dictated pricing to implement cross-subsidization between divisions of a regulated firm.² Overall, pricing of intermediate goods tended to be viewed as an internal matter, and left largely to the discretion of the regulated firms.

Transaction cost economics (TCE) as developed by Williamson (1975, 1979, 1985, 1991) focuses on the relationship between attributes of transactions and characteristics of the governance structures used to accommodate these transactions. Transactions vary in many dimensions, the most important of which is the degree of asset specificity. A party incurring relation-specific investments will demand safeguards to prevent its counterpart from attempting to appropriate the quasi-rents associated with these assets (Klein et al., 1978).^l These safeguards can take the form of formal, legally enforceable contracts, or of extra-legal private ordering arrangements.

The convergence of economics and sociology that has been going on for some twenty years by now has also changed organization studies from the core. The main change has been the introduction of a double contracting perspective into organizational analysis. Contracting is now being considered from the point of view of interest alignment (ex ante) and from the point of view of adaptation to unforseen changes (ex post). This perspective was impossible within standard neoclassical economics, because, due to complete information, complex contingent claims were supposedly easily drafted. In traditional sociology, this perspective could also not have been developed because role-playing, norm-conforming man would observe the silent agreements, rendering ex post governance superfluous. Only the convergence of economics and sociology created models of man that were realistic enough to invite a host of problems around the creation and governance of contracts and still stringent enough to allow relatively tight theory building.

The rapidly extending success of transaction cost economics has attracted attention primarily to the decisive role of its concepts in explaining agents’ choices to “make” by themselves, to “buy” through market exchanges, or to enter into transactions through intermediate forms. Because transactions require institutional supports and are costly, there is a continuous trade-off among using the price mechanism, relying on administrative coordination, or having recourse to “hybrid forms.”

This chapter surveys and critiques the economic literature on authority relations within the firm, with the goal of outlining a program for further research. My focus is on environments where compliance with authority is privately enforced. I also compare authoritative decision making with the alternative of bargaining, and discuss the equilibrium determination of organizational form in a market setting. Given space constraints, the discussion is informal, and caveats concerning matters of detail are omitted.

The aim of this chapter is to discuss the limits of the contractual approach to the theory of the firm and to question the notion of wholly circumscribing the employer-employee relationship in a contract. I submit that the employment relationship is the organizational core of the firm.

There are certain ideas and concepts which in retrospect it is easy to see are liable to misinterpretation and misapplication—often to the understandable annoyance of those who were at the forefront of developing such ideas and concepts. Transaction cost economics certainly appears to suffer from this problem for reasons which are easy to comprehend.

The main thrust of this chapter is to compare and link two separate literatures that claim to provide a coherent account of organizational evolution. On the one hand we have transaction cost theory, arguably dominant within economics, which uses a universal efficiency-seeking logic; on the other hand we have various approaches to the understanding of corporate restructuring that, while being different, all emphasize a shifting strategic logic. It will be argued in this chapter that these two bodies of literature provide differing approaches to organizational evolution that can benefit from the insights offered by each.

In part, what sustains transaction cost and, more generally, contractarian approaches to the understanding of the firm is a particular view of social science. It is the reductionist view that phenomena have to be understood by breaking them down conceptually into smaller and smaller constitutive components: it is said that wholes should be explained in terms of their parts. In the social sciences, these injunctions typically take the form of an atomistic ontology and a challengeable methodological individualism (Hodgson, 1988). Yet it has been repeatedly asserted by philosophers, including Popper (e.g., Popper and Eccles, 1977, p. 18), that complete reductionism has never been achieved in any science. At least in economics, it is arguable that complete reductionism is not possible (Hodgson, 1993b; Udéhn, 1987).

This chapter is motivated by the exciting exchange of ideas during the “Transaction Cost Economics (TCE) and Beyond” Conference in Rotterdam, June 1994. Lengthy discussions with some of the main contributors in the field have helped clarify my views on the main issues, implications, and future directions of the vibrant TCE perspective. It became evident during the Conference that, in contrast to early critiques of the nonoperationaliza-bility and vagueness of the TC concept, proponents of the perspective are currently actively involved in empirical investigation, to the effect that they now believe that enough evidence has been accumulated to throw the ball into the critics’ court (see for example Masten, 1994).

What is the boundary of a franchised restaurant as a firm? Does it consist only of the organization of power exercised by the owner over the employees? Or does it extend to include as well the structure of patrons and long-term suppliers of goods, capital, and franchising rights? Is the network of that restaurant merely an organization (i.e., firm) at a greater scale, or is that network qualitatively different from the organization?

Governance structures are assumed to be cost efficient in the sense that decision makers will select that structure out of a number of alternatives which offers the highest potential for a reduction of production and transaction costs concerning an existing product or service. Low costs reflect the firm’s ability to economize on bounded rationality and to attenuate opportunistic behavior (Williamson, 1985). Cost efficiency of a specific organizational form depends on the characteristics of the product or service to be transacted. Within the transaction cost framework it can be interpreted that a decision to construct or change a specific governance structure derives from the experience with numerous previous transactions in practice.

Following an earlier argument for the development of a dynamic, learning based theory of transactions (Nooteboom, 1992a), the present chapter explores the possibility of a model which represents processes of search and evaluation of transaction partners and investment in transaction relations, from the perspective of learning. The focus is on relations between industrial users of inputs and their sources. Epistemologically, the model is based on a constructivist theory of knowledge. In economics, this yields the concept of “external economy of cognitive scope,” which implies that learning, or dynamic rather than static efficiency, is a crucial dimension of transactions. The aim of the analysis ultimately is to arrive at a simulation model in which the dynamics of search and learning in the context of transactions can be further explored. A tentative specification is given of a model which represents search as the reduction of “cognitive distance”, modeled as a distance in Euclidean space. Investment in reduced cognitive distance between a user and a supplier yields the benefit of improved perception by the user of the supplier’s competence and of the user’s own needs, improved perception by the supplier of the user’s needs, and improved efficiency of transaction between user and source. However, it also yields an increased risk of spill-over of sensitive information to competitors, through the transaction partner, in an extended network of relations. The model reflects the idea that reducing cognitive distance to zero may be inferior to reducing cognitive distance less but with respect to a greater number of partners: beyond some point of proximity, more diversity of learning can be worth more than further accuracy of learning.

In modern economic analysis, the notion of transaction costs is used both as a technical concept in order to measure different kinds of costs related to contracting (search and information costs, contracting costs, monitoring and enforcement costs), as well as in the form of an explanatory device in order to understand economic change. Hence, according to the latter meaning, it is often presupposed by modern institutional theory that institutional change and different levels of transaction costs play a determinate role in the successive evolution of organizational modes. This evolutionary approach is clearly found in “true” institutional theory (known from the so-called “old institutionalists” and also expressed in the recent work of North), as well as in the so-called “neoclassical” institutional tradition from Coase to Williamson (see Eggertsson, 1990).

From the contents of the preceding chapters in this volume, it follows that it is perhaps no exaggeration to conclude that TCE is on the verge of establishing itself as the new orthodoxy in the new theory of the firm, that is to say, in the group of theories in which the firm is considered more than a production function. TCE can be considered the new orthodoxy, because compared to the other theories, the approach is well-established, its theoretical framework is well-developed, with hypotheses in many areas of economics and with positive results of empirical testing. From the preceding chapters, it has also become clear that many critics are arguing that the analysis should go beyond TCE and indicating how that should be done. In these conclusions, we want to address the intricate question: if the current state of the art is thus cursorily but correctly characterized, then what is the best way to proceed?