Metropolitan Wage Differentials: Can Cleveland Still Compete?

Introduction Labor costs are often cited as one of the primary reasons for the economic hardships plaguing many older industrial cities, such as Cleveland. Of course, other factors, such as local taxes, proximity to markets, product cycles, and energy costs may also contribute to the area's diminished ability to compete with other regions in attracting and retaining businesses. Nonetheless, since labor costs represent an important part of total production costs, the initial presence of significant wage differentials among metropolitan areas may have been a major factor in the economic expansion of Sunbelt cities and the relative decline of Snowbelt cities. In turn, divergent patterns of growth resulting partly born firms relocating in low-wage areas may have caused wage levels to converge. With respect to the effect of differential labor costs on firm location and on regional employment growth, two aspects of labor costs must be considered. First, there is more to examining labor cost differentials across regions than simply looking at regional differences in wage rates. Firms consider not only the amount they pay workers, but also the productivity of their workers. Stated simply, an employer is willing to pay a worker in Cleveland a higher hourly wage than a worker in Atlanta, for example, if the Cleveland worker is more productive than the Atlanta worker. Therefore, a comparison of regional wage differentials is much more meaningful when these wages are adjusted for differences in worker skills. Randall W. Eberts is Associate ProThis paper was prepared for the fessor, and Joe A. Stone is W. E. Labor Conference, which was held Miner Professor, Department of EcoMay 2, 1985, in Cleveland, and which nornics, University of Oregon. was sponsored jointly by the Regional Economic Issues (REI) Program and the Federal ReSe~e Bank of Cleveland. The study was supported in part by a grant from the Cleveland Foundation through the RE1 Program. The authors would like to acknowledge useful comments by Michael Fogarty, director of the RE1 Program, and by Joseph Antos. In addition, the authws are especially indebted to Ralph Day for technical cornouter assistance.