The Concept and Measurement of Regional Industrial Diversification

Abstract

The relationship between the economic structure of subnational regions and their relative responsiveness to fluctuations in national economic activity has been a persistent vein of inquiry within regional economics for nearly thirty years. The pursuit of diversified local economic structures in response to potentially unstable dependence upon single industries or groups of industries is a thoroughly ensconced dimension of virtually every contemporary local development effort. The professional research upon which such planning orientation is presumably based, however, has been less than satisfying in both its theoretical and empirical dimensions. In fact, a distinct pessimism permeates the literature with respect to the relationship between diversification of industrial structure and stability of local economic activity [see Isard, 10, 276; Richardson, 15, 148]. In this paper we shall note that such pessimism may be premature. Previous attempts to define and measure industrial diversity in economically meaningful ways have suffered from significant conceptual shortcomings; and empirical attempts to relate such definitions to historical stability have, almost without exception, been such as to understate the potential of diversification efforts to reduce instability in regional economic activity. The conceptual difficulties of the four principal measures of regional industrial diversity or diversification are here demonstrated, and the limitations of empirical studies to date are noted. A new conceptual approach is offered and is shown to lead naturally to a new measure of industrial diversification which appears to alleviate several of the problems encountered in previous measures. The results of empirical analyses of the relative ability of the new measure to explain statistically the historical instability of a cross-section of 52 U.S. metropolitan regions are also reported. The results provide some evidence not only that the new measure seems to explain variation in instability considerably better than its precursors but also that it appears to do so sufficiently well in absolute terms to raise new in erest in the policy goal of intra-regional diversification.