Industry Trade Balance and Domestic Merger Policy: Empirical Evidence from U.S. Merger Policy for Manufacturing Sectors

Article excerpt

I. INTRODUCTION

A voluminous literature (e.g., Williamson, 1968; Fisher, 1987; White, 1987; Farrell and Shapiro, 1990) exists concerning the welfare effect of mergers involving only domestic competitive implications; that is, domestic merger activity within a closed economy. Globalization, however, is fast fictionalizing the notion that merger policy can be completely embedded within one nation (Melamed, 2000). Accordingly, the intersection between trade and competition policy--a previously neglected topic--has recently received a great deal of scholarly attention (e.g., Richardson, 1999; Vandenbussche, 2000; Horn and Levinsohn, 2001). Further, and closer to the topic at hand, a growing dialogue exists on the design of merger policy in an open-economy setting (e.g., Barros and Cabral, 1994; Levinsohn, 1997; Head and Ries, 1997; Neven and Roller, 2000b, 2003; Mavrodis and Neven, 2001; Zhang and Chen, 2002). Such debate is particularly healthy, as the globalization of mergers and acquisitions suggests that larger nations with commitments to antitrust (e.g., the United States) will increasingly face the dilemmas previously experienced by midsize nations with commitments to antitrust (e.g., the United Kingdom and Germany).

The question motivating this work is how does the industry trade balance impact the tenor of merger policy for a specific industry sector. Within the above dialogue on merger policy in an open-economy setting exists a subliterature that specifically considers the impact of industry trade balance on optimal domestic merger policy. Invoking various oligopolistic scenarios, Barros and Cabral (1994), Head and Ries (1997), Levinsohn (1997), Sorgard (1997), Yano (2001), and Zhang and Chen (2002)--all find positive (negative) trade balances to conditionally favor more lenient (strict) domestic merger policy. Explicit in all these works is that antitrust agencies face a national welfare criterion of which foreign producers and consumers are not part; yet consumer surplus (not national welfare) is the stated criterion for many antitrust agencies. Moreover, Landes and Posner (1981), Ghosal (2002), and others make the intuitive argument that imports ameliorate anticompetitive behavior in domestic markets, thus suggesting the converse of above: Positive (negative) trade balances conditionally favor strict (lenient) domestic merger policy. Such discord is made more problematic once one recognizes the political economic nature of antitrust, the dangers of using "strategic" merger policy, and the bureaucratic discretion held by most antitrust authorities.

It remains difficult then to theoretically predict how trade balance impacts the tenor of domestic merger policy. Solving how--and whether--an industry trade profile alters domestic merger policy is ultimately an empirical question. Consequently I test here the impact of trade balance on merger policy outcomes in the U.S. manufacturing industrial sectors. A comprehensive panel dataset--covering U.S. merger policy at the two-digit standard industry classification (SIC) code level over the 1982-2001 period--allows testing whether trade balance acts to promote strict or lenient merger policy. The empirical results support higher trade balances leading to stricter merger policy.

The article is organized as follows to support the analysis. Section II reviews the competing theories on how trade balance might impact the nature of domestic merger policy and formulates two competing propositions that motivate the empirical testing. Section III presents the results of the empirical tests. Section IV concludes.

II. THEORIES AND COMPETING PROPOSITION FORMULATION

Barros and Cabral (1994) sparked the subliterature on trade balance and the optimal tenor of domestic merger policy in an open-economy setting with their extension of Farrell and Shapiro's (1990) pioneering approach to mergers in a closed economy. The Barros and Cabral setup consequently conforms rather closely to that of Farrell and Shapiro in a number of dimensions. …