The purpose of this paper is to compare theories of firms and mergers offered by Oliver Williamson and Thorstein Veblen. Williamson elaborated Coase's theory of the finn into a transaction cost framework that resembles Veblen's approach in some respects, in that both place emphasis on transactions in explaining the emergence and expansion of large firms. Yet this apparent similarity is misleading and the differences revealing. I will illustrate these differences with a brief interpretation of firm strategies from the early business history of the automobile industry and then conclude that the "new institutionalism" and "new industrial organization" derived from the ideas of Williamson have ignored Veblen to their peril.
In The Theory of Business Enterprise, first published in 1904, Veblen wrote:
Business consolidation . . . [eliminates] the pecuniary element from
the interstices of the system as far as may be. . . . with the result
that there is a saving of work and an avoidance of that systematic
mutual hindrance that characterizes the competitive management
of industry . . . The heroic role of the captain of industry is that of a
deliverer from an excess of business management. It is a casting
out of business men by the chief of business men [Veblen 1932, 29].
Although this passage has some definite transaction cost overtones, Veblen precedes it with a clearer version of what appears to be early transaction cost logic:
[t]he amount of "business" that has to be transacted per unit of
product is much greater when the various related industrial processes
are managed in severalty than where several of them are
brought under one business management. A pecuniary discretion
has to be exercised at every point of contact or transition, where the
process or its product touches or passes the boundary between different
spheres of ownership. . . . THe greater the parcelment in
point of ownership [of the industrial processes], the greater the
amount of business work that has to be done in connection with a
given output of goods or services, and the slower, less facile, and
less accurate, on the whole, is the work [Veblen 1932, 28].
It was three decades later when Coase posed his now famous question: if the price mechanism is the most efficient mechanism for allocating resources in a market economy, why do firms exist? The answer was that economic agents incurred transaction costs when using the price mechanism. The greater the number, and the complexity, of transactions, the greater the costs involved in transacting [Coase 1937). Williamson, following in Coase's footsteps, elaborated on the reasons that transactions were costly, noting that uncertainty, idiosyncraticity, complexity, informational asymmetry, and opportunism were inherent to transactions, which made it difficult to coordinate highly interdependent production and distribution processes through the market mechanism alone [Williamson 1985, 1990]. So firms emerged, and merged, to reduce the costliness of transactions with other firms by bringing more activities within one governance structure--or, as Veblen put it, to reduce the "severalty" of business managements.
Veblen, Coase, and Williamson were all interested in how interactions between firms shaped firm and industry structure: for Veblen, mergers were, in part, a response to threats by other firms to interfere at business interstices, which could obstruct seamless industrial processes; mergers were efforts to gain control of these interstices for "guidance and coordination of industrial processes with a view to economies of production" [Veblen 1990, 299]. In other words, firms merged primarily to minimize interstitial adjustments. For Williamson, all interactions between firms were best understood by examining the costs of both routine and idiosyncratic transactions. Mergers occurred when transaction costs were both high and reducible by combining two or more firms under the same management. In other words, firms merged primarily to reduce the costs of making transactions [Williamson 1985, 129; 1990, 235].
The similarities end there. …