State Business Cycles and Their Relationship to the National Cycle: Structural and Institutional Determinants

Abstract

During business cycles state and regional incomes fluctuate in the same direction as national income. In some states and regions income fluctuates more than the national average; in other states and regions it fluctuates less. Although no general explanation for differences in business cycles across states exists, an influential set of explanations focuses on differences in price flexibility. Gordon [2; 3] has discussed the employment implications of sticky prices and made some international comparisons of unemployment rates. His results imply that decreased price flexibility leads to increased variability of income and employment. Olson [6; 7] and Colander and Olson [1] have suggested that differences in state unemployment rates arise from differences in price flexibility. Differences in price flexibility in turn originate in differences in structural characteristics and institutions [2]. If a state's structure and institutions allow prices to be flexible, most of the adjustment to changes in aggregate demand will be made by prices rather than output. If a state's institutions tend to enforce price rigidity, most of the adjustment to changes in aggregate demand will be made by output rather than prices. The absence of acceptable state price indices prevents direct tests of the hypothesis that differences in price flexibility explain differences in income variability across states. The inability to measure price flexibility by state turns out not to be a serious problem, however, because we are primarily interested in what determines the magnitude of fluctuations in state income. Institutional and structural differences across states lead to the differences in price flexibility that cause differences in the business cycle. Price flexibility is merely the transmission mechanism. Moreover, it may not be the only transmission mechanism. Structural and institutional factors may affect state business cycles through a host of channels. We will therefore focus directly on the relationship between structures and institutions and state business cycles. By adopting such a procedure, we may miss a particular transmission variable, but we will be more likely to identify the total effect of structure and institutions on performance. Institutional and structural variables will be compared across states rather than regions, an attractive alternative. We used states rather than regions because states embody the political