COMPETING AT HOME TO WIN ABROAD: EVIDENCE FROM JAPANESE INDUSTRY.

Publication: Review of Economics and Statistics

Publication Date: 01-MAY-01

Bookmark this article Print this article Link to this article Email this article Digg It! Add to del.icio.us RSS

COPYRIGHT 2001 MIT Press Journals

I. Introduction

This study explores the influence of domestic competition on international trade performance, using data from a broad sample of Japanese industries. The role of domestic rivalry in international competitiveness is highly salient at a time when the globalization of markets has raised questions about the appropriateness of domestic antitrust policy and other forms of domestic policy towards competition. Japan is a particularly interesting setting in which to explore this question because it is frequently argued that competition within Japan is limited. Many Western observers, for example, have asserted that Japanese industry is internationally competitive because of cooperation among Japanese rivals, sheltering from international competition, and selective intervention in competition orchestrated by the Ministry of International Trade and Industry (MITI) and other governmental agencies.(1) Related to this, a persistent theme in explaining Japanese competitiveness has been the role of benign antitrust laws and weak antitrust enforcement, under which even cartels are legal in many circumstances. It is argued that this fosters efficiency and avoids the "destructive" competition ascribed to Western industries.(2)

We explore three competing hypotheses about the effect of domestic competition on international market performance. The first, consistent with much literature on Japan, is that local collusion and limits on domestic competition enhance international competitiveness; here, the relation between the intensity of local rivalry and international competition is negative. Support for this hypothesis can be drawn from the literatures of trade and innovation. When there are impediments to trade such as tariffs, transport costs, and transaction costs that allow domestic firms to price-discriminate between domestic and foreign markets, and domestic and foreign products are perfect substitutes, firms face a more elastic demand for exports than for domestic sales. The domestic price will be higher than the export price. Domestic collusion, then, will be associated with higher exports (Caves & Jones, 1973; Brander, 1981; Brander & Krugman, 1983).

Consistent with this view is a long line of argument suggesting that monopolists will be more innovative than firms facing competition. Schumpeter (1943) argued that monopolies are a necessary evil for expanding R&D, because there are economies of scale in R&D, market power allows monopolists to fund more R&D, and large firms are willing to take greater risks. Concentrated domestic industries, then, would be associated with more innovation and enhanced international competitiveness.

A second hypothesis is that the intensity of domestic competition will have little or no association with international competitiveness because the distinction between domestic and international competition has been rendered unimportant by the widespread internationalization of markets. Reich (1991), for example, argued that the national identity of corporations and their homebase location has become increasingly irrelevant in international competition. The intensity of rivalry internationally is what is important, not domestic competition in any one country. This same school of thought leads to the conclusion that domestic antitrust policy in international industries need not concern itself with conditions in the local market.

A third and final hypothesis is that there should be a strong positive association between domestic rivalry and international competitiveness. This hypothesis draws on Porter (1990), who argues that domestic rivalry pressures firms to innovate and upgrade while fostering positive static and dynamic externalities in the local business environment (such as, supplier availability, easier access to technology and market information, and specialized human resource development).(3) In this theory, rivalry among domestically based firms offers greater benefits to competitive upgrading than either imports or foreign companies with minimal investment in the nation. Local rivalry not only gives rise to positive externalities, but it creates stronger competitive incentives together with greater pressures to upgrade productivity, because local rivals neutralize advantages due to input costs and other local business conditions.(4) The literature on the economics of innovation contains theoretical support for the idea that competition is associated with greater rates of innovation under the assumption of strong ex post appropriability (Arrow, 1962; Loury, 1979; Lee & Wilde, 1980). In the same vein, Scherer (1980) argued that insulation from competitive pressures breeds bureaucratic inertia and discourages innovation. Porter's theory also builds on a tradition going back to another argument advanced by Schumpeter (1934, 1943), which sees competition as a dynamic process of creating new products and process. Here, dynamic domestic competition would spur innovation and productivity improvement and, in turn, international competitiveness.

This study explores these three competing hypotheses by examining the effect of domestic competition on the trade performance using new data on a broad sample of Japanese industries. We employ market-share instability in the Japanese market to measure domestic rivalry directly, rather than using structural variables such as seller concentration to proxy the intensity of competition. We also investigate the relative influence of market-share instability, import penetration, and domestic protection as influences on trade performance. Section II reviews the relevant literature. Section III introduces the statistical model and variables. The data are described in section IV. Section V presents our empirical results, and section VI offers some conclusions.

H. Empirical Evidence on the Link between Domestic Rivalry and Export Performance

Although a good deal of empirical literature examines the connection between international trade and domestic market structure, it has mainly focused on the role of international trade in influencing industry profitability. Leading examples include Pugel (1980), Yamawaki (1986), and Patterson and Abbott (1994).

Empirical studies of the relationship between domestic rivalry and export performance have been few, and they report conflicting findings. On the one hand, Pagoulatos' and Sorensen's 1976 analysis of 88 three-digit SITC industries found that the U.S. seller-concentration ratio was positively related to U.S. industry exports as a percentage of OECD exports. They argue that, when barriers to trade are present, firms possessing market power in their domestic market will have greater exports because they can engage in dumping and because control over the domestic market may allow more-aggressive pursuit of export opportunities.

Yamawaki and Audretsch (1988), on the other hand, found that four-firm producer concentration had a negative and significant effect on the Japanese share of all exports to the U.S. markets in a small sample of 24 three-digit SIC Japanese industries. In a highly concentrated domestic industry in which the interdependence among firms is well recognized, firms will have lower level of pre-trade output and higher prices, reducing export competitiveness. Yamawaki and Audretsch also found that high concentration of the Japanese relative to the U.S. industry had a negative and significant effect on Japanese exports to the United States, consistent with the above argument. Note that Yamawaki and Audretsch's dependent variable is a measure of bilateral trade and not overall trade performance.

Audretsch and Yamawaki (1988), in a study of 213 four-digit SIC industries, also found that domestic seller concentration had a negative influence on U.S.-Japan bilateral trade performance, and that legal cartelization status had no significant effect on bilateral trade. They also found that relative R&D intensity between the United States and Japan had a positive and significant impact on the trade balance. If more-intense domestic competition leads to higher relative R&D intensity, these results would be consistent.(5)

III. Specification

In these and virtually all studies of the relationship between competition and trade, the extent of competition is proxied by the industry concentration ratio. Our approach is to measure rivalry (market conduct) directly using market-share instability and relate it to trade performance, controlling for factor endowments. We also explore the role of import protection and other variables on trade performance.

Although our argument is that market-share instability reflects the intensity of competition, another possibility is that share instability is the result of exogenous disturbances. We thus estimate two models: a market-share instability model and a trade-performance model. By including exogenous shocks among the explanatory variables in the first model, we evaluate their importance in explaining share instability. The second trade-performance model is our primary interest. We also test the possibility that market-share instability and trade performance are simultaneously determined by employing two-stage least-square (2SLS) estimation.

A. Model 1

(1) MSINST = a + [b.sub.1]R&D + [b.sub.2]ADV + [b.sub.3]C4 + [b.sub.4]C4SQ + [b.sub.5]PCAP + [b.sub.6]SHIPGRW + [b.sub.7]SHIPVAR + [b.sub.8]JIS + [b.sub.9]BARRIERS72 + [b.sub.10]CARTEL + [b.sub.11]EXCARTEL + [Epsilon]

Dependent variable

Market-Share Instability: Following Caves and Porter (1978), we calculate market-share instability from the sum of individual market-share fluctuations of leading firms between 1973 and 1990. Measures of market structure, such as the number of firms, the four-firm industry concentration ratio, and the Herfindahl index only indirectly measure market conduct. In contrast, there are strong theoretical reasons that instability in market positions is a sign of active competition, whereas stable market shares will be associated with oligopolistic collusion. Stigler (1964) noted that fixing market shares is probably the most efficient of all methods of combating secret price reductions, and Shepherd (1970) observed that successful collusion will tend to hold market shares virtually constant. Allen (1976) also noted that, under conditions of imperfect collusion, temporarily stable market shares are probably the best that can be attained. Although constant market shares can also stem from vigorous but stalemated competition, such a running standoff is relatively improbable. The greater the stability of shares, then, the higher the likelihood that overt cooperation or strong recognition of mutual dependence is present; conversely, churning among market shares, especially that of leading firms, should be associated with active competition whatever the level of concentration.(6)

We calculate market-share instability in two ways (appendix A provides the details). One is absolute instability, or the sum of the absolute value of the...

Read the full article for free courtesy of your local library.