Does Industry Mix Matter in Regional Business Cycles?

Abstract

This paper tackles two issues : (1) lead and lag relationship among regions and the role of the industry mix effect to this phenomenon are explored ; (2) concurrent and lagged effects of the industry mix on the regional economic fluctuations are measured explicitly with the national shock identified from the principal components method. The empirical analysis focuses on five Midwest states. The findings reveal that, the business cycles of Michigan, Ohio, Indiana and Wisconsin coincide with the national cycle while the cycle of Illinois lags the national cycle by 3 to 4 months. This phenomenon turns out to be generated from the differences in industry structure since the manufacturing sector reacts promptly to the national shock while the services sector respond in a few months. As a result, relatively service-oriented Illinois lags other neighboring states. Analysis reveals that the industry mix effects explain more than 60 percent of the variance of the state coincident index and around 40 percent of the variation of state total non-farm employment. In addition, the simulation of VAR model demonstrates that the transmission mechanism and autoregressive property of economic activity expand the time differences in the business cycles among regions caused by the industry mix effects.