The Concentration-Earnings Hypothesis: Reconciling Individual and Industry Data in U.S. Studies.

The authors provide at least a partial reconciliation of industry and individual level examinations of the influence of concentration on wages. They find that neither aggregation per se nor use of different wage measures accounts for the general failure of industry level studies to confirm a correlation. Instead, the failure results from serious misspecification forced by an incongruence by constructing an industry level data set that eliminates it and by showing that concentration emerges as significant, a result in harmony with most individual level studies. As a consequence, the results of previous industry level studies should be discounted. Copyright 1990 by Blackwell Publishing Ltd