Economic research has focused intensely in recent years on the role played by financial markets for real economic activity. Based on ideas tracing back at least to Schumpeter (1912), and inspired by the early contributions of Goldsmith (1969), Gurley and Shaw (1955), and McKinnon (1973), the work of King and Levine (1993 a,b), Demirguc-Kunt and Maksimovic (1998), Levine and Zervos (1998), Rajan and Zingales (1998), Levine, Loayza and Beck (2000), among others, has provided robust empirical evidence that broader, deeper financial markets are strongly associated, causally, with better prospects for future economic growth.
Having established this basic finding, the research effort is now focused on the analysis of the mechanisms through which finance affects real economic activity. What are the specific characteristics of financial markets that seem to affect firms and industries in non-financial sectors of production? For example, does it matter whether banks are privately or government owned (La Porta, Lopez-de-Silanes and Shleifer, 2001), or whether there is higher or lower protection for financial contracts (Levine, 1999), or whether banks are in a more or less competitive environment (Jayaratne and Strahan, 1996, Cetorelli and Gambera, 2001)? And, what specific characteristics of firms and industries are especially affected by finance so that it eventually translates into higher economic activity?
This paper contributes to this line of research by investigating the role of well-defined characteristics of banking markets on equally well defined industry characteristics in production sectors. More precisely, we investigate the impact of bank concentration and bank deregulation on measures of industry structure in non-financial sectors. We ask whether concentration of market power in banking has an effect on entry in a given sector, on the average size of existing firms in a sector, and on the overall firm size distribution within a sector.
Using data on U.S. local markets for banking and non-financial sectors, we find that more vigorous banking competition that is, lower concentration and looser restrictions on geographical expansion is associated with more firms in operation and with a smaller average firm size. In fact, the whole firm size distribution shifts to the left as our measures of banking competition increase. Because we exploit data at the industry level, we are able to control for alternative (omitted) variables that may drive market structure both within and outside banking by exploiting differential usage of external finance across industrial sectors.
Whether bank competition is “good or bad” for economic activity has been and continues to be a lively topic of research and policy analysis. In addition to the conventional argument that concentration of market power in banking means lower equilibrium amounts of credit, it has also been claimed that banking market power is actually needed for banks in order to establish valuable lending relationships. Hence, whether more or less competition in banking is socially desirable is still under discussion. This paper thus contributes to expand our understanding of the economic role of bank concentration and competition.
The number of competitors in a sector, the average firm size and the composition between small and large firms are all important factors having a bearing on conduct and market performance. They are therefore important determinants of the sector’s capital accumulation and growth and consequently of the sector’s contribution to the overall level of economic activity. Various related streams of literature have focused on determinants of product market competition (e.g., Brander and Lewis (1986), Chevalier (1995), Kovenock and Phillips (1995, 1997), Maksimovic (1988)), on firm size (e.g., Kumar, Rajan and Zingales (2001), Campbell and Hopenhayn (2002)) and on firm-size distribution and more general industry dynamics (e.g., Lucas (1978), Jovanovic, (1982), Evans (1987), Hopenhayn (1992)). This paper relates to these parallel lines of research and makes a contribution bridging them together.
Our evidence is consistent with that documented in several recent papers focusing on banking concentration and competition policies across countries. Cetorelli (2001) provides evidence of larger average firm size in countries with more concentrated banking. Along similar lines, Cetorelli (2003a) finds that enhanced bank competition following passage of the Second European Banking Directive brought a reduction in average firm size. Matching data on job creation and destruction in US manufacturing sectors with banking data across US markets, Cetorelli (2003b) shows that more bank concentration implies less entry and thriving of younger firms and also delayed exit of older firms. Again based on cross-country data, Beck, DemirgucKunt and Maksimovic (2003) find that higher bank concentration is associated with more financing obstacles, especially for smaller firms.
Our study is an important addition to this literature because we are able to measure banking structure at the local level rather than at the country level. Thus, our data offer a distinct advantage because much of the research on bank market power suggests that the relevant geographical market for banking services, especially for small firms or potential entrepreneurs, is local (e.g. county, city or, perhaps, state) rather than national or international (see, for example, Berger, Demsetz and Strahan, 1999). Moreover, our data allow us to explore not only how average firm size responds to banking competition, but how the whole size distribution responds. By doing so, we are able to test more directly whether more or less bank competition is beneficial for all firms in a sector, or whether instead the effect may be different for firms in distinct size classes.
In the remainder of the paper we first flesh out the theoretical links between banking concentration and industrial structure in order to motivate our empirical tests (Section II). In Section III, we present the data set and the main variables used in the analysis. Section IV documents the empirical results, and Section V concludes.
