Schumpeterian dynamics and metropolitan-scale productivity

This study seeks to make a contribution to our understanding of economic growth and development by exploring the relationship between Schumpeterian dynamics and the productivity performances of U.S. metro-scale regional economies. It presents the results of a sequence of models whereby job reallocation—the total volume of job entry and exit—along with several other theory-relevant covariates is used to explain variation in productivity growth across U.S. metro-regions during the 1986–99 study period. Descriptive statistics are used to characterize the individual dimensions of the Schumpeterian dynamics being used as predictors. The insights they provide are then used to develop and test explanatory models using a variety of econometric methods. A final multivariate Error Correction Model (ECM) estimated by Feasible Generalized Least Squares (FGLS) is used to investigate how the short-term Schumpeterian dynamics affect metro-scale productivity in a longer-run historical context. The robustness of the basic model is also explored by examining this relationship across a variety of metro-region classifications. In addition, a Recursive Regression Model (RRM) is also employed to locate more precisely when in time Schumpeterian dynamic effects actually registered their effects.; The study finds that despite being systematically obscured by conventional economic analyses, massive job reallocation within U.S. metro-regions has been the norm, rather than the exception. Moreover, both job creation and job destruction dynamics separately tend to follow the contours of the business cycle, while job reallocation remains relatively more stable. When job reallocation is introduced in the productivity function along with other predictors, including several varieties of industrial restructuring, scale economies, and human capital endowments, it is found to register large and significant positive effects on metro-scale productivity growth. Schumpeterian dynamics do not exert a uniform influence on metro-regional productivity performances over time. Recursive regression results indicate that before 1996, the effects of the Schumpeterian dynamics were muted into insignificance by job creation and destruction dynamics that often canceled each other out as metro-regions underwent continued industrial restructuring. Since 1996, however, the positive effects on metro-region productivity growth have been consistently strong.