Other Articles

Dec 21, 2009 — Neil Boyle

In order to simplify the explanation of the old and new theoretical frameworks on which institutional analysis is constructed, I took the liberty of creating a parallel construction that is analogous to the production function of the firm, which for purposes of elucidation I call an “approximate” governance function.

For the past three quarters of a century, economists were unable to incorporate governance concerns into the theory of economic exchange because it was technically unworkable to do so, powerful incentives pointed elsewhere.[ii] Governance problems were diagnosed and treated exogenously in administrative organization and management terms. No one had yet figured out a way to frame human nature internally in economic theory until Oliver Williamson came along. Williamson framed the two basic assumptions of human behavior—rationality and self-interest—in the more realistic concepts of bounded rationality and opportunism with guile, and with a bit more, framed them in economic cost and benefit terms, and he did so by adhering to the canons of modern economics. I might add that this framing was done with the concept of economic costs and benefits in mind as well as with financial costs and benefits.[iii]

His work — an alliance between law, economics and organization theory—called Transaction Cost Economics is worthy of study and application. TCE makes understanding the governance of institutions intelligible and their performance predictable. While it is said that other theories (e.g., Agency Theory) can make a similar claim, none is as robust as TCE in the generation of refutable hypotheses, a central tenet of scientific methodology – with high explanatory power. The development of a new institutional analysis follows from this feature of TCE .

The governance function accomplishes four things. First, it fills the void that was created by the “black box” of orthodoxy by addressing the confusion between markets and firms. Williamson introduced the concept that the firm is not an extension of a market as orthodoxy would have it, but rather an entirely different way to govern economic exchange. This gave rise to the firm as a governance structure, whereas in orthodoxy and as mentioned above the firm is a production function concerned with the ratio of inputs to outputs and unconcerned with the role economic organization plays in the process. This glossing over of the difference has had harmful policy implications in the past. (See case study on strategic supply chain.)

Second, the governance function relates three empirical governance ideas: transaction costs, mitigation of transaction costs, and vertical integration. As many advocate, the purpose of vertical integration is to monopolize competition. But, this is incorrect. Vertical integration is a means employed by traders to mitigate certain kinds of transaction hazards by internalizing the transaction into the firm. (See the stylized case study about how vertical integration and risk mitigation evolved out of market adaptation to trading hazards in early times in the Case Study Archive.)

Third, the governance function creates a means for analyzing organizations. This is accomplished by avoiding framing the governance function in production function terms, a task that is not as easy as it sounds. Years of doing it the traditional way, we have become accustomed to a traditional mindset. The governance function is framed with the transaction as the unit of analysis; the basic assumptions of human behavior are incorporated in workably realistic terms; discrete structural analysis is the means of analysis rather than marginal analysis; and the systems concept of remediableness is adopted to avoid comparisons of solutions to hypothetical ideals.[iv]  These are not easy hurdles to overcome.  

And fourth, it does all of the above and more by maintaining the relentless rationality that has been the keynote of economics for the past three quarters of a century. The governance function is complementary to the production function of microeconomic theory and practice and not a substitute for it. To understand the governance function, there is no need to unlearn anything of fundamental value that was previously learned in economics. Typical topics, include: the human and environmental risk factors that create transaction costs that magnify and distort financial and economic costs and skew resource allocation policies; the identification and assessment of transaction costs; the creation, modification and impact of human incentives on economic exchange (or trade); the alignment or misalignment of contracts and their costs; the misallocation of economic resources due to failure to factor in governance costs; and the hidden hazards of economic exchange.

Governance is also an integral part of American capitalism; and American capitalism is about, in my humble opinion, its ingenuity for finding compromise through appropriate trade-offs between the socially good versus the economically efficient markets. “Putting a human face on American capitalism” was never an easy task to pull off.[v] Ingenuity is confirmed because American capitalism has never succumbed completely to one or the other extremes, although it has moved away from the center during different epochs of American history. Mostly it remains somewhere mid-range while slowly nudging the morphology of compromised governance to a new equilibrium. Hence, it is no surprise that the 2007-2008 financial meltdown in the U.S. and elsewhere rivaled the 1929 Great Depression, yet it never succumbed to its level completely. The 2008-2009 real economies have so far not come near the level of public suffering in the 1930s. Economic safety nets have had something to do with this outcome.[vi] But these safety nets are from an older era.  Are they still appropriate or will they fall short of accomplishing their aim; indeed, is their aim still valid in today’s world of global markets where everything is a mouse click away and financial risks are abstracted from their sources and considered transferable? Is something missing in the current microeconomic framework?

The first and second versions of compromised governance.

Ironically, the mid-range of compromised governance was and continues to be achieved despite (or maybe because of) the machine-like rationality of traditional economic theory and regulation. In subsequent parts of this introduction, I describe a second but different version of compromised governance. The former focuses on the polity as in the previous paragraph, from a macro perspective and produces a compromise in governance that bridges opposing political poles. This is good. The latter focuses on the transaction through a microanalytic lens and discovers a faulty or compromised governance system that was designed using production function guidelines. This is not good. Is there something besides relentless rationality contributing to these two outcomes? Further inquiry compels new and deeper answers, which are buried in what was up to the late 1960s probably the most unexplored area of economic research, namely the transaction, where the governance of trade across a market interface between autonomous firms occurs. Traditionally, economists have been reluctant to seriously consider governance of the transaction as a research agenda for doing so would have invited too many inexplicable dead-ends. This impression comes across strongly in Williamson’s literature about TCE.  Many of these dead-ends have now been answered satisfactorily by institutional economics.

The term that aptly describes the outcome of orthodox institutional design is “compromised governance.” The same term aptly describes orthodox institutional analysis. Much theorizing about governance is done in the name of institutional analysis. Institutional design and analysis usually take the form of a mixture of administrative organization, and management problems and compares them against idealized legal or regulatory standards. Such comparisons have little empirical basis and frequently lead to wrong policy conclusions.

This second brand of compromised governance analysis (CGA) is a slippery slope for several reasons. First, as economies evolve and become more developed, transactions become more complex requiring deeper knowledge to execute them efficiently. (See example of PIATCO about how transaction complexity overtook state institutional capacities listed in the Case Study Archive). Second, policy choices based on the firm as production function analysis routinely exclude the costs of governance. This introduces biases into the allocation of resources. Where governance costs are excluded, it is generally an instance when the project has not been implemented in accordance with its plan. Third, because compromised governance analysis is external to the actual phenomena, analysis outcomes are reliant on analyst’s discretion and are thus difficult to replicate scientifically. This raises the question of scientific validity. Fourth, CGA has little grip on the phenomena of governing the transactions between buyers and sellers. Orthodox institutional analysis is able to expand on the technological rules on which engineering, technology, management and administrative organization theories are premised, but unable to do so with regard to the rules of economic governance. In part, this explains much of the shortcomings of policy analysis that is carried out on the basis of the firm as a production function.

Abstracted “out-of-touch-with-phenomena” knowledge is what gives birth to the second variety of compromised governance. This kind of knowledge is different from the “in-touch with phenomena” variety. The differential is what distinguishes traditional analysis and the TCE framework. The structural changes that take place within the internal organization of production firms and the inter-temporal processes that occur in the interfaces between say a production buyer and a supplier seller do not form a part of the production paradigm. These structures and processes are external to it and therefore outside of the analytical gambit that comprises TCE. This is an important point as analysts persist in seeking solutions that are derived from exogenous variables. (See example about the game theoretic nature of regional economic integration listed in the Case Study Archive.)

Compromised governance is just that, a compromised faulty solution based upon the unexplained phenomena of proxy governance variables. Statistical analyses based on large quantities of external time series and panel data from national census, and labor, industrial and national accounts surveys validly determine production input and output behavior, but do not adequately explain the structural or intertemporal processes embedded within the economic organization of the transactions themselves. It will take a different kind of analytical framework such as TCE along with its own data set to accomplish the latter. Dependent on externally derived data, traditional microeconomics struggles at high economic and financial costs to reliably generate the deep knowledge that must be acquired to understand the adaptive ingenuity that goes on within the interfaces of inter-firm transactions across markets. This topic of “deep knowledge” is discussed in more detail in the Stylized Case Study: Derivation of Economic Governance listed in the Case Study Archive.

What is deep knowledge? 

Deep knowledge is the vast body of knowledge that often lies fallow and unknown in risk assessments; hence, numerous risks are not priced out and remain obscured in the price as a risk premium. Risk premiums, however, are not benign numbers. A part of deep knowledge is that farsighted traders capitalize on the information that is embedded within them. Risk premiums obscure and act to disadvantage the counterparty, which in PPP infrastructure is frequently government.

Traditional analytical foci are not able to access deep knowledge because it centers on the institutional environment of the polity. This is the big picture aggregate side of economic exchange where hazards are known and generally addressed. Hazards, however, are unknown, obscure, and ubiquitous and particularly so across inter-firm market interfaces between autonomous parties. Transaction hazards are precursors to transaction costs. Part of the art of TCE is to know if and when and what to do when a hazard is ready to become a cost. The twin characteristics of the decision makers who are involved is fundamental to this challenge. The second reason is that traditional analytical lens is ill-suited to the phenomena under study, so an entire area of economic phenomena that has production cost generating implications escapes analysis and documentation.[viii]

Deep knowledge has something to do with the explanations of all of the discovered regularities of human interaction that occur within the economic exchange interface of a valued asset between a buyer and a seller who are subjected to legal and regulatory environmental constraints. Well known regularities already exist such as participation, ownership, moral hazard, gaming, and impossibility of selective intervention, free riding, group dynamics, bilateral dependency, comfort levels, and many others. Regularities have a unique feature, which is why they occur regularly, and unbundling them – “bilateral dependency” for example, unravels the deeper knowledge having to do with the incentives of the actors involved in an economic exchange of a special kind. Knowing what the precursors are to human incentives that occur at a specific intertemporal time frame in contracting and within a particular trading environment is valuable knowledge for the analyst who is designing the economic organization of that exchange and others going forward.

The above regularities of human interaction came to our consciousness either through the casual sifting of large numbers of interactions over a long period of time, or through the empirical research of organizational theorists. The problem is it has taken 8000 years since the dawn of writing to identify only a handful of these regularities while technological change races along seemingly at exponential speeds. Mining this deep knowledge should be a high priority for analysts.

Like the many layers of the onion, the topic of “deep knowledge” is unfathomable. There is no telling what lies in the next layer. (Once again, refer to the case study about how transaction complexity overtook the institutional capacities of the state listed in the Case Study Archive)

Accessing “deep knowledge” and improving deal-making

To access deep knowledge requires posing the problem of economic organization as a problem of contracting. Contracting greatly facilitates the economic analysis of organizations.  “[T]he firm serves as a focus where conflicting objectives of individuals are brought into equilibrium within a framework of contractual relations.” (Jensen, Meckling, 1976, pp. 311-12). “A contract is defined as an agreement between buyers and sellers in which a triple PKS (price, asset specificity, and safeguards) defines the terms of the contract assuming quantity, quality and duration are specified.” Operationalizing this requires the knowledge of law, economics, and organization theory. TCE is an alliance of these disciplines. (Williamson, 1996)

What makes Williamson different from his predecessors is that he has been studying the social and physical phenomena of economic exchange for over forty years in a slow, molecular fashion. With the exception of Ronald Coase, John Commons and Chester Barnard, for most of the pre and post World War II period, there were few empirical studies; now there are over one thousand attesting to the numerous regularities that mark the connection between the human factors of economic exchange and the structural incentive apparatus of economic governance. (Klein and Shelanski, 1995; Macher and Richman, 2006; Geyskens, Steencamp, and Kumar, 2006, p. 531) CITE LATEST 2008 LITERATURE SURVEY ON TCE



  1. I knowingly say “a” lesson because the recent financial crisis is certainly more complex than the absence of deep knowledge.  Although, it is my contention that the absence of deep knowledge significantly hindered the decision-making of CEOs and financial analysts on Wall Street and Main Street.
  2. Lots of statistical data on production inputs and outputs and highly trained economists and financial analysts in mathematical analysis proved difficult for the economics profession to ignore at the time.
  3. I say this more out of conjecture than out of knowledge.  It simply makes more sense to me that Williamson would have considered the difference between financial (i.e., cash) and economic (i.e., social) costs and opted for the combination of the two and not one over the other. The western economy is too complex to consider otherwise. 
  4. To make sure I get it right, here are Williamson’s own words on organization. “Organizations are susceptible to analysis by (1) making the transaction the unit of analysis and contract the framework for operationalizing economic exchange; (2) emphasizing workably realistic behavioral assumptions of bounded rationality [i.e., ignorance] and opportunism [self interest with guile]; (3) employing discrete structural analysis rather than marginal analysis and (4) adopting remediableness by which comparisons [and benchmarking] to hypothetical ideals are avoided. (1985). Distribution of gains is indeterminate, even if it is in the interest of autonomous traders to fill gaps and correct errors to realign incomplete contracts in long-term bilaterally dependent relations; where relations among traders are emphasized as costly bargaining is involved.” Williamson, 1996
  5. The commons versus the enclosures is one example; and democracies versus socialism, and market economy versus command economy are two other examples that come to mind.
  6. So far public suffering has not reached the level of the Great Depression even though it has been over a year since the sub-prime market started to tank. This does not mean it won’t as things could get worse.  
  7. My contact with Professor Williamson began five years ago. I first contacted him in 2004 while I was in Manila, Philippines serving as the Trade and Investment Policy Advisor for USAID Philippines. I read most of his publications by that time but I can’t attest to having gained an acceptable level of literacy until several years later. At that time he agreed to serve as an advisor to me.  Three years later, he responded to my request for the assistance of a Ph.D candidate in his program at U.C. Berkeley to serve as a peer reviewer of several papers I had written on applying TCE to my work over the years. He responded and the PhD candidate Sharon Proczter helped immensely in getting my papers prepared for uploading to my web site: www.
  8. Although outside the scope of this essay, I view current modes of analysis, e.g., financial engineering, as having reached or are close to their plateau in terms of providing new understandings about diversifying risk. I am probably wrong on this, as mathematical permutations are seemingly unlimited.

This article is filed in other-articles. Start or participate in a discussion below or click here to see more articles in this category.


Please share your thoughts and comments about the article. We look forward to hearing from you. Fields marked with "*" are required: