An Empirical Evaluation of Profit Cycle Theory

This study tests key hypotheses of Markusen's profit cycle theory of regional growth by examining the recent time path of industrial development for eight U.S. sectors. The analysis uses annual data for 1954–1990 and incorporates the influence of national economic trends and random effects in identifying stages of the profit cycle. Based upon those stages, the hypothesized changes in the behavior of other aspects of the growth cycle are assessed in simple statistical models. The results of the empirical tests are mixed. Employment follows a general pattern of early rapid growth followed by slow growth and eventually decline, but the pattern is not symmetric. The value-added curves are similarly asymmetric, and some even contradict the expected monotonicity of the growth and decline process. The lack of spatial dispersion is probably the most striking result of the study. The stability over time of the employment percentage in top states is inconsistent with the spatial dispersion implications of the profit cycle model.