The leading alternative theories that have been offered to explain organizational changes are domination theory, market power, technology, life cycle, pecuniary economies, and strategic behavior. 1 shall consider them seriatim.
1. Domination Theory
Domination theory focuses on human actors. There are those who possess economic power and those who do not. The organization of economic activity is under the control of those who possess power. The reason why one mode is chosen over another is that it permits those who are in control to extend and perfect their power.
This theory of organizational innovation presumably applies to the rela- tions both between capitalists and workers and among capitalists themselves. The thesis that work is hierarchically organized so as to prevent Workers from gaining power is examined in Chapter 9. Consider, therefore, whether power theory explains confrontations in intermediate product markets between cap- italists. Porter and Livesay report that during the “first two centuries after the initial English settlement on the North American continent, urban merchants dominated” (1971, p. 5). Those “urban merchant capitalists . . . were the wealthiest, best informed, and most powerful segment of early American society” (p. 6). The all-purpose merchant nevertheless gave way to the spe- cialized merchant early in the nineteenth century, which merchants then became “the most important men in the economy” (p. 8). The specialized merchant in tum found his functions sharply cut back by the rise late in the 1800s of the integrated manufacturer: “The long reign of the merchant had finally come to a close. In many industries the manufacturer of goods had also become their distributor. A new economy dominated by the modem, integrated manufacturing enterprise had arisen” (1971, p. 12).
Power theory must confront two troublesome facts in explaining those changes. First, why would the general purpose and later the special purpose merchants ever permit economic activity to be organized in ways that removed power from their control? Second, why did power leak out selectively—with the merchant role being appropriated extensively by some types of manufacturers but not by others? As developed above, the transaction cost approach explains both in terms of efficiency.
To be sure, this does not preclude the possibility that power is also operative. For example, entrenched interests may sometimes be able to delay organizational transformations. Power enthusiasts have not, however, demon- strated that significant organizational innovations—those in which large transaction cost sayings are in prospect—are regularly defeated by established interests. There is abundant evidence to the contrary. Within the economic arena,14 therefore, if not more generally, I submit that organizational innova-tions for which nontrivial efficiency gains can be projected will find a way to subdue (or otherwise will be accommodated by) opposed interests. Power is relegated to a secondary role in such a scheme of things.
2. Market Power
Market power arguments can be brought to bear on organizational innovation in two ways. One is that possessors of market power simply prefer certain organizational arrangements. The second is that organization is used strate- gically as an impediment to rivals.
The latter is considered in subsection 7.6, below. Porter and Livesay appear to appeal to the former in explaining why manufacturers integrated into distribution in some industries and not in others. Thus they observe that the “incidence of oligopoly and large size was much less frequent” among manufacturers that did not integrate forward than among those that did (1971, p. 214). It is noteworthy, however, that a number of large firm/concentrated industry groups are included among those nonintegrators: breakfast cereals, hand soaps, soup, and razor blades, to name a few. Those industries would presumably be prime candidates for forward integration if oligopolistic prefer- ences rather than efficiency were driving the organizational outcomes.
3. Technology
The argument that technological imperatives explain organizational outcomes is an old one. As explained earlier, however, the common ownership of two stages that are operating in a cheek-by-jowl association with each other is to be understood as a solution to a troublesome bilateral bargaining relation. Steady state thermal economies can always be realized by placing autonomous blast furnaces and rolling mills alongside one another, whatever the ownership structure. Choice among structures thus turns on how to mediate the interface in response to disturbances. This is a transaction cost issue.
Although forward integration into distribution is an anomaly if addressed in terms of physical or technical aspects, Chandler has come forward with an alternative technological explanation for that condition. While acknowledging that successful organizational innovations serve, among other things,.to ccon-omize on transaction costs (1977, p. 256), the main factor on which he relies in explaining forward integration is what he refers to as “economies of speed” (pp. 281,298; Chandler and Daems, 1979, pp. 30-31). According to Chandler and Daems, such economies
. . . could only be realized … if a managerial hierarchy carefully scheduled the flows. . . . Therefore, when and where a new technology permitted mass production and when or where new markets permitted mass distribution, such administrative coordination turned out to be more efficient than when the movement of goods between units was a result of a multitude of market transactions. [1979, p. 31]
Although economies of speed remain unspecified, appeal to an intuitive notion of such economies leads to a number of anomalous results. Why didn’t manufacturers comprehensively integrate into distribution for the sale of ciga- rettes, beer, and branded packaged goods? Why were small, standardized producer durables sold through independent distributors while manufacturers sold and serviced large, unique producer durables themselves? I submit that fungible human assets were employed for the retail sale and service of cigarettes, other packaged goods, and standardized producer durables, while that was not the case for large, unique producer durables. It is this (together with the economies of scope available for the former set of products and not for the latter plus the diseconomies of bureaucracy that attend forward integration) rather than “economies of speed” differentials that explain the pattern.
4. Life Cycle
George Stigler (1951) has advanced a theory of vertical integration in which life cycle features are prominent. Extensive integration is favored at both the early and late stages of an industry’s development and less integration occurs at the intermediate stages. Integration in the textile industry is held to be consistent with the hypothesis (Stigler, 1951).
Both Porter and Livesay (1971, p. 132) and Chandler (1977, p. 490) read the evidence differently. Specifically, Porter and Livesay contend that “while large firms may pass through the three stages described by Stigler, they frequently engage in reintegration as a result of rising, not declining, demand” (Porter and Livesay, 1971, p. 132). The predicted stage two reduction in vertical integration is not borne out in the aluminum industry either (Stuckey, 1983, pp. 26-46).
I submit that life cycle analysis needs to be joined with transaction costs in order for observed patterns of vertical integration to be explained. More interesting, moreover, than the disputed demand features referred to above is the following life cycle phenomenon: As customers and independent middlemen become more knowledgeable of the technology and as reliability of an item increases (so that service requirements decrease), the transaction cost incentive to maintain a forward market presence by a manufacturer decreases. Accordingly, items that were once marketed by an integrated sales and service organization can often be returned to the market in the later stages of a product’s life cycle.
That has numerous ramifications, among which is the viability of discount houses for selling mature products. Also, public policy toward forward integration ought to make allowances for life cycle features. The likelihood that forward integration is justified by transaction cost considerations is much greater for products that are sold before maturity sets in. The possibility that such integration is continued at mature stages because it serves strategic entry impeding purposes likewise deserves consideration.
Those issues have a bearing on the differential performance of firms whose complex products are technologically on a parity but which follow different early stage marketing strategies. The success of IBM in relation to Sperry-Rand (and, later, RCA and GE) may well turn on the intensive sales and support that IBM offered for its relatively unfamiliar but complex product in the 1950s, which were critical formative years in the computer industry’s history.
5. Pecuniary Economies
Vertical integration may be adopted as a device by which to avoid excise taxes (Coase, 1937; Stigler, 1951). Whether this has been an important factor in explaining vertical integration in the United States has never been established. I conjecture that it has been of minor significance in relation to the real transaction cost savings reported above.
Corporate taxes (and tax credits) have been a factor for some mergers, but probably more for conglomerate than for vertical mergers in the United States—especially in the period following World War II (although this remains to be investigated). Plainly tax considerations played a principal role in the early postwar acquisitions of Royal Little (Sobel, 1974, p. 356). And they continue to influence conglomerate acquisitions to this day, the attempted takeover of the Mead Corporation by Occidental Petroleum being a recent example.49 Whether such assets, once acquired, will be effectively managed is an organization form issue (see Chapter 11). Whatever the immediate incentives to integrate, therefore, transaction cost issues still must be addressed. (Those conglomerates that adopted a holding company rather than an M-form structure would presumably be less well suited to deal with complexity and adversity and would be shakeout candidates as events progressed.)
6. Strategic Behavior
Strategic behavior has reference to efforts by dominant firms to take up and maintain advance or preemptive positions and/or to respond punitively to rivals. The object in both instances is to deter rivalry. An example of the first kind would be forward integration into distribution where the resulting transaction cost savings are negligible. Punitive strategic behavior is illustrated by predatory pricing. These troublesome issues are discussed more fully in an antitrust context in Chapter 14. Suffice it to observe here that strategic behavior mainly has relevance in dominant firm or tightly oligopolistic industries. Since most of the organizational change reported above occurred in nondominant firm industries, appeal to strategic considerations is obviously of limited assistance in explaining the reorganization of American industry over the past 150 years.
None of the six alternative theories of organizational structure/innovation treated here makes more than a piecemeal contribution to an understanding of the reshaping of the American economy, and some are plainly misconceived. Transaction cost economizing, by contrast, not only applies broadly—to the changing governance of intermediate product markets, labor markets, corporate governance, and regulation—but also helps to explain many of the micro- analytic details and some of the general movements of vertical integration.
Source: Williamson Oliver E. (1998), The Economic Institutions of Capitalism, Free Press; Illustrated edition.