This chapter shows that company (or firm) effects significantly influence business-line profitability with industry effects controlled. A fundamental question underlying this book is whether (privately) efficient diversification could be because of firm-specific advantages that are not totally resolvable into industry-level effects. The chapters of Parts II and III taken together imply that firm effects in profitability and R&D exist and that differences in the types of purposive diversification chosen by competing manufacturers with different firm-specific capabilities may underlie the differences in profitability and in R&D effort and performance across firms.

Introduction

The mutual interdependence recognized by sellers in a concentrated market protected from entry implies behavior that differs from that predicted by the model of pure competition which takes as given large numbers of sellers or free entry. For example, in such a concentrated market the equilibrium price is expected to exceed marginal cost, and resource allocation is therefore expected to be inefficient. Yet empirical verification of the a priori behavioral significance of variance in markets' structures is controversial. This chapter cannot resolve all controversy, but within the general linear model, it provides insights about the importance of traditional models of structure and profitability, firm effects, and the effect of capital intensity on oligopolistic coordination. Purposive diversification provides a potential explanation for the unexplained systematic variance that is documented.

This chapter explores how purposive diversification of R&D affects R&D behavior and productivity in U.S. manufacturing. The findings support the predictions developed in Chapter 8.

Purposive diversification

The chapter shows that R&D diversification in large U.S. manufacturing firms is purposive, exploiting complementarities of various research activities and forming groups of related industry categories. The purposively diversified firms behave differently from randomly diversified or undiversified firms. One aspect of the behavioral differences between purposively diversified firms and others is that the former allocate relatively more R&D funds to industry categories where the appropriation of the returns from R&D is easier and relatively fewer funds to those categories where returns are more difficult to appropriate. Finally, R&D expenditure and productivity are more closely linked at the group level than at the industry-category level, suggesting that spillovers of knowledge across industry categories are important. The chapter's findings provide a key explanation for the firm effects found in Part II, because firms in the same industry category typically engage in very different forms of purposive R&D diversification and some do not diversify at all.

In Section 9.2, I use my methodology to distinguish nonpurposive or random diversification from any purposive diversification into a set of related, or close, R&D activities. “Nonpurposive” diversification here is discerned as diversification dissimilar to patterns found in significant clusters of firms.

In June 1990, the U.S. House of Representatives passed a bill (U.S. House, 1990) designed to extend to joint ventures in production the protection offered cooperative research and development (R&D) efforts by the National Cooperative Research Act of 1984 (NCRA). By the following summer, both the U.S. House and the U.S. Senate hammered out and passed similar bills. The historical context of the NCRA is important; the act was passed during a dramatic redirection of U.S. government policy toward business combinations. Uncritical pronouncements about the efficacy of such combinations abounded. Recounting in this chapter the history of those pronouncements, from which the proposed laws on production joint ventures have evolved, and questioning their economic validity, suggests first that the laws promoting cooperation are not likely to be a panacea for lagging U.S. competitiveness and second that they may actually do great harm.

Introduction

The Reagan administration expected much good from business combinations. The declining international competitiveness of U.S. industrial products added importance to the expectations. The claims of advocates of change in antitrust laws led to new laws and proposals for still more new laws championing combinations among U.S. industrial competitors as ways to promote efficiency and meet the challenges of international competition.

That historical context of the NCRA, reviewed in Section 13.2, implies that the act would have passed even if it were not sound economic policy. Section 13.3 reviews economic theory explaining why the NCRA may not promote desirable R&D behavior.

This chapter will depart from the standard theoretical formulation in which all of the firms in the research and development (R&D) race anticipate the same single prize. Chapter 8 developed the standard formulation to explain that when, from society's perspective, firms underinvest in R&D because they do not appropriate all of the social value of their innovations, more competition can increase R&D investment toward the socially optimal amount. This chapter treats research possibilities as diverse among firms, because the point of the chapter is that another potential gain from competition is more diversity in research.

To develop the idea theoretically, I shall use the Chamberlin/Osborne type of model discussed in Chapter 2. As explained there, we could also develop the idea using the theory of multiperiod games; however, my point can be made most simply by using what Chapter 2 called the Chamberlinian shortcut. To explore diversity empirically, I shall use the variance within and across industries in the systems orientation of patent portfolios. Industry effects will be used to control for the differences across industries in the applicability of systems research, while firm effects will capture the idea that industry traits do not determine completely the systems orientation of firms. Firms can employ different strategies even when in the same industry category. The intra-industry diversity associated with the firm effects that I estimate could of course be associated with things other than the rivalry mechanism that I shall posit.

For the economy to work well, resources should flow freely from one industry to another in response to changing demands and costs. The capital market can allocate resources by means of the entry and exit of firms. But entry can be by new firms or existing ones through diversification. Chapter 1 observes that a multimarket firm may use its own internal organization to allocate resources more efficiently than the arm's-length market mechanism could do. For example, Williamson (1970) and Weston (1970) stress advantages of internal capital transfers over the market. Because of such advantages, Gort (1962, p. 4) and Rumelt (1974, p. 2) state that multimarket operation of firms will speed redeployment of resources in response to profitable opportunities. They would be right if multimarket operation were not coincident with multimarket contact.

A priori impact of multimarket grouping

While diversified companies may have advantages that would facilitate the movement of capital, as explained in Chapter 2 they have enhanced opportunity for coordination if they meet in several markets. Multimarket groups are groups of diversified firms whose activities span the same markets to a significant extent. Multimarket grouping of sellers could reduce the flow of resources, thereby inhibiting a socially desirable competitive process, if it proceeded until the diversified sellers recognized their mutual dependence and coordinated a reduction in competition, tacitly or otherwise.

The findings about industrial diversification described in previous chapters suggest some broad lessons about what it takes for a country's manufacturers to be competitive in international markets. Since the results for U.S. manufacturers are broadly consistent with and complementary to findings in the literature about the manufacturers of other countries, I shall venture some conclusions about the archetypal successful international competitor. Those conclusions also suggest new policies; recall that the declining effectiveness of U.S. manufacturers in international trade provided important motivation for the NCRA policy toward cooperative research and development (R&D) and for the R&D tax credit. Further, the policy initiatives suggested by the research in this book that uses U.S. data are potentially applicable to the policies of many countries because the U.S. approach to antitrust policy is increasingly being applied throughout the world. For example, one prominent antitrust policy publication reports that “For the first time in the history of modern Japan, antitrust enforcement officials in 1991 displayed visible clout and achieved visible results, and they expect to increase their role in the economy during 1992”. Antitrust enforcement activity is not only increasing in Japan, but in Canada, the European Community, the United Kingdom, Germany, France, and Italy as well. Further, with grants from the Agency for International Development, the U.S. Department of Justice and the Federal Trade Commission are assisting the nations of Eastern Europe to develop antitrust enforcement programs.

Chapter 2 explained that one way conglomerate mergers could create market power is by increasing multimarket contact and symmetry among a market's firms. Chapter 2 showed that multimarket contact can theoretically increase the stability of cooperative-like behavior among the rivalrous firms in any given market in which the firms meet. Further, as emphasized in Chapter 1, if the various motives for diversification rest on industry-specific properties, then the extension of diversification will tend to increase multimarket contact with the potential for increasing cooperative-like behavior (whether or not bolstering such behavior was primary among the firm's objectives). This chapter tests the hypothesis that mergers serve to increase multimarket contact and symmetry.

Overview

Section 3.2 measures the contact and symmetry created by two large conglomerate mergers just after the passage of the Celler–Kefauver Act and illustrates the change in market structure expected if a merger were designed to increase market power. Of course, the diversification and ensuing symmetry could be a response to any sort of synergy based on industry characteristics. As explained in Chapter 2, gains in potential for cooperative-like behavior from increased multimarket contact could in principle induce diversification, but as Chapter 1 has explained, there are strong “innocent” explanations for diversification that can make it impossible to show that much diversification is on balance because of such potential.

As we have seen in Chapter 2, current antitrust policy ignores multimarket contact. Yet, as Chapter 2 explains, multimarket contact among firms may increase their ability to exercise market power (i.e., to raise price above competitive levels) without engaging in collusion that would violate U.S. antitrust laws. Multimarket contact can increase market power because such contact may allow oligopolists to effect a noncompetitive price with decisions made individually.

The hypothesis

One could even argue that multimarket contact may at times be necessary if a profitable oligopolistic consensus among diversified firms is to be reached. Chapter 2 explains that disagreements of the battle-of-thesexes sort might otherwise intrude. As Section 5.2 details below, Bain's classical observations about market structure and performance were conditioned on high multimarket contact among the leading firms in each of the industries that he sampled. Thus, the classical work linking monopolistic pricing with oligopoly is consistent with the view that multimarket contact may be necessary for consensus among diversified oligopolists. This chapter analyzes Bain's sample, then tests the hypothesis that multimarket contact among diversified oligopolists is necessary for their market power because without the contact they will be unable tacitly to agree on a supracompetitive price.

Bain's sample

A different interpretation of Bain's result: Sellers' recognition of mutual dependence within markets is the typical focus of empirical models of oliopolistic market power; conjectural variations and strategic games reflect sellers' interdependence.