The Covariability of Industrial Concentration and Employment Fluctuations

Abstract

The United States' recent experience with high unemployment' and inflation has stimulated interest in reexamining the other-than-macroeconomic determinants of these policy relevant variables. There are, for example, several theories which suggest that seller concentration plays a role in determining employment fluctuations. However, the various theories which explain the supposed link between the two variables offer conflicting conclusions as to the direction of the impact.2 Some suggest that industrial concentration aggravates employment fluctuations. Others conclude that concentration dampens employment fluctuations. Additional research on the subject is important because of This paper analyzes the effect of industrial concentration on cyclical employment fluctuations using a covariance-based measure of employment fluctuations. This employment variable measures the extent to which employment in each sample industry fluctuates over the business cycle with movements in the rest of the economy. The test results show that the cyclical stability of employment in an industry is not adversely affected by its concentration ratio. The nature of each industry's output is the significant determinant of cyclical employment fluctuations. Adjusting for product characteristics, the tests show that cyclical employment is, if anything, more stable in concentrated industries. * We have enjoyed the financial support of the General Electric Foundation and the Research Foundation of the City University of New York. We wish to thank the anonymous referee for his or her helpful suggestions. The conclusions of this study are solely ours and do not reflect the views of the U.S. General Accounting Office or the City University of New York. 1. Clarkson and Meiners (1977) present the hypothesis that official unemployment rates are overstated by approximately two percentage points for the years since 1974. The reason given is the introduction in 1974 of various work registration requirements for welfare recipients. These requirements increased unemployment statistics because they added to the official labor force a large new body of nonworking and, for the most part, unemployable individuals. 2. E.g., see Means (1936) and Blair (1956) on administered prices; Williamson (1964) on organizational slack; and Wachter (1970) and Ross and Wachter (1973) on wage premiums.