Industrial diversification and manufacturing employment stability: A study of the state of Connecticut

Studying economic fluctuations of regions has been an important area of inquiry in urban and regional economics. To gain an understanding of this problem, some researchers have sought to establish a relationship between regional cyclical behavior and industrial composition. While this relationship has been questioned on conceptual grounds, empirical evidence has shown that industrial composition is partly responsible for, and often quite significantly related to regional cyclical behavior.;Despite the empirical support there is still a lack of consensus on the merits of diversification as a policy to promote regional stability. This is due, in part, to the variety of approaches taken to examine this relationship. Differing definitions of diversification have been employed to measure the industrial structure of a cross-section of variously defined regions. Moreover, the level of economic detail on which the research is based has been usually highly aggregated (1-digit and 2-digit SIC).;The ability of three traditional measures of diversification to track manufacturing employment stability in Connecticut between 1964 and 1983 is examined. Two recent measures are also considered. One, the portfolio variance, is found to provide an improvement in measuring instability as it accounts for interindustry covariation.;The analysis of employment is based on an essentially 3-digit disaggregation. This permits a comparative analysis of the measures at the 2-digit and 3-digit levels. It is discovered that their relative explanatory ability is affected by the level of disaggregation and the choice of detrending technique employed.;An industrial recruiting model is developed to aid regional policymakers in recruiting strategy. Based on a disaggregated variance/covariance matrix, it determines sector levels that will minimize the variance in overall employment. Interindustry production linkages are considered by incorporating an input-output (I-O) table into the model. While sector variances, covariances and size influence the results, I-O linkages and final demand also impact the minimum variance solutions. Simulating an employment increase in relatively stable defense sectors shows how instability is spread through sectors related to defense.