The organizational failures framework applies to the study of a variety of exchange relations that are not prima facie associated and applies symmet- rically to market and nonmarket forms of organization alike. It attempts to identify ultimate, as contrasted with immediate, sources of transactional friction. As compared with previous efforts to explain market and organi- zational failures, it relies on a relatively small set of explanatory factors.
The emphasis on transactions favored here bears a similarity, which is sometimes quite close, to John R. Commons’ study of institutional economics in the 1930’s. Thus I, like Commons, regard the transaction as “the ultimate unit of economic investigation” (1934, p. 6). Also, Commons addressed himself to what he referred to as the “going concern” and the “working rules” that appertain thereto ( 1934, p. 69), with special emphasis on contracting for future events. In the language favored here, Commons was concerned with the development of efficient modes for completing transactions which involved recurrent contracting under conditions of uncertainty. The significant differences between the working rules (which include the incentive and control instruments and associated constitutional powers) of markets and internal organizations are plainly germane for comparative institutional choice-making purposes. Thus, although there is no simple correspondence between the rather elaborate quasi-judicial vocabulary developed by Commons for evaluating transactions and the language of the organizational failures framework, his general approach and many of the institutional phenomena which were of concern to Commons are the same as those dealt with here.
The organizational failures framework consists of a set of human factors on the one hand, and a related set of environmental factors on the other, which together explain the efficacy of contracting. The environmental factors are uncertainty/complexity and small numbers, which are familiar to economists from the prominent role they play in the market failure literature. The human factors are bounded rationality and opportunism, which are somewhat less familiar. Not that discussions of these human attributes have never appeared previously in the economics literature, but an effort to identify these as key human factors and to relate these to the environmental factors mentioned above has not previously been attempted. In addition, there is a derived condition, referred to as information impacted- ness, and a systems related influence, referred to as atmosphere, that completes the framework. My remarks here focus on these less familiar categories, with emphasis on the way in which each is affected by internal organization.
Bounded rationality can take either computational or language forms. The pairing of limited computational capacity with uncertainty is what makes a complete set of contingent claims contracts infeasible or very costly to write. In circumstances where incomplete contracting exposes the agents to the negotiations to considerable risk of opportunistic behavior, alternatives to market forms of organization may be favored instead. Internal organization can thus arise on this account.
Internal organization not only serves to curb such opportunism (see below), thereby relaxing the need for complete contracts, but hierarchy also has advantages in computational and planning respects, in that it facilitates the specialization of decision-making and economizes on communication expense. Economies of communication are realized in part through specialization of information collection and dissemination, partly through idiosyncratic coding economies, and partly by promoting the convergence of expectations [in the sense of Malmgren (1961) ].
As indicated, bounded rationality is manifested not merely as a com- putational constraint but is also the result of language limitations. This latter, when paired with an information impactedness condition, can also occasion internal organization. Here the bounded rationality problem is that participants to a transaction sometimes lack the ability to communicate successfully about the nature of the transaction through the use of words or symbols that are contractually meaningful. The requisite language may not exist or the individuals concerned may not have access to it — possibly because they are intrinsically limited in their ability to acquire it.
But while the individuals in question may be language-limited in con- tractual respects, they may nevertheless be able in other respects successfully to communicate about the phenomena in question. It is merely the failure, or prohibitive cost, of formal language that impedes the exchange. A picture, according to a Chinese proverb, is said to be worth ten thousand words. A demonstration may be worth ten million. Apprenticeship systems, which involve learning- and teaching-by-doing, are often the most efficient means of communication for this reason.
The assumption that individuals behave in a self-interested way is so commonplace to economics that it would seem scarcely to warrant separate attention. Opportunism, however, is more than simple self-interest seeking. It is self-interest seeking with guile: agents who are skilled at dissembling realize transactional advantages. Economic man, assessed with respect to his transactional characteristics, is thus a more subtle and devious creature than the usual self-interest seeking assumption reveals.
Opportunism can involve either data distortion or the making of self- disbelieved promises. Data distortion is possible, of course, only if an in- formation impactedness condition obtains. Symmetrically, information impactedness by itself would pose no transactional problem if individuals could be relied on candidly to reveal the information to which they have selective (low cost) access. It is the proclivity of individuals to distort to their advantage the data to which they have preferred access which poses the contractual problem. Ceteris paribus, forms of organization which serve to check these distortion tendencies, possibly by reducing the information disparities between the parties through auditing, will be favored. Internal organization may obtain for this reason.
Atmosphere is a concept which noneconomists are apt to find less objec- tionable than economists. Indeed, much of what economists do is to rationalize how human behavior is to be understood in familiar net gain terms, and to debunk such relatively loose concepts as atmosphere. Scratch an economist and you find a deeply committed rationalist.
That the rationality approach has powerful advantages is beyond dispute. But there is a risk that it will be employed with excessive zeal. There is a tendency to focus on the parts, which often lend themselves to rationalization, sometimes at the expense of the whole.11 Reference to atmosphere has the advantage that it expressly raises systems considerations. What are the prevailing attitudes, how are these influenced by the nature of the transactions, and with what spillover consequences?
The usual assumption in economics, implicit if not explicit, is that individuals regard their relation to the system in a strictly neutral, dispas- sionate manner. Subject only to the condition that transactions are techno- logically separable, each transaction can be priced separately and metered independently. I submit, however, that technological separability is merely a necessary and not a sufficient condition for transactions to be regarded independently. Attitudinal separability must also be established. To process transactions which are technologically separable but attitudinally interrelated, in either individual or group respects,183 as though they were independent necessarily misses spillover effects in the pricing and metering of these transactions. Suboptimality will result if the fiction of independence is maintained when in fact attitudinal spillover exists at the margin.
The problem in all of this is to identify when such attitudinal consider- ations operate strongly and when they can be safely neglected. I conjecture that transactions which affect conceptions of self-esteem and/or perceptions of collective well-being are those for which attitudinal considerations are especially important. The employment relation is the leading instance of a market exchange where the influence of metering intensity on work attitudes needs to be assessed with care. More specifically, efforts to divide the employment relation into parts and assess each separately in strictly calculative, instrumental terms can have, for some individuals at least, counterproductive consequences.
It is interesting to speculate why individuals do not regard all market transactions neutrally. Though I do not pretend that it is a complete expla- nation, I submit that one of the reasons for this is that individuals are generally not schizophrenic with respect to their economic and noneconomic identities. Individuals who experience nonmetered externalities in noneconomic contexts and reach nonmarket accommodations thereto bring the attitudes and experiences which evolve in these nonmarket circumstances to the workplace as well. Rather than regard transactions in strictly quid pro quo terms, with each account to be settled separately, they look instead for a favorable balance among a related set of transactions.13
The above discussion of the affirmative ways in which hierarchy affects each of the factors that appears in the organizational failures framework may be summarized compactly as follows:
Bounded rationality. Hierarchy extends the bounds on rationality by permitting the specialization of decision-making and economizing on communication expense.
Opportunism. Hierarchy permits additional incentive and control techniques to be brought to bear in a more selective manner, thereby serving to curb small-numbers opportunism.
Uncertainty. Hierarchy permits interdependent units to adapt to unforeseen contingencies in a coordinated way and furthermore serves to “absorb” uncertainty.
Small numbers. Hierarchy permits small-numbers bargaining indeterminacies to be resolved by fiat.
Information impactedness. Hierarchy extends the constitutional powers to perform an audit, thereby narrowing ( prospectively at least) the information gap that obtains between autonomous agents.
Atmosphere. As compared with market modes of exchange, hierarchy provides, for some purposes at least, a less calculative exchange atmosphere.
These are the advantages of hierarchy. In circumstances where market exchange would occur between small numbers of traders under conditions of uncertainty, hierarchical exchange is apt to be favored. But two caveats should be noted. First, hierarchy experiences both size and transactional limits — for the reasons given in Chapter 7. Second, the advantages indicated are merely potential. For this potential to be realized requires that appropriate principles for decomposing hierarchies be followed and the requisite incentive and control apparatus supplied. This is the organization form issue, as developed in Chapters 8 and 9.
Source: Williamson Oliver E. (1975), Markets and hierarchies: Analysis and antitrust implications, A Study in the Economics of Internal Organization, The Free Press.