Evaluating the Evidence on Union Employment and Wages

Abstract

Using Current Population Survey data for the years 1973–86, this study tests the hypothesis that the steep decline of union employment shares since 1973 is related to increases in union wages relative to nonunion wages. The authors show that although the aggregate union wage premium has changed little since 1973, some industry-specific premiums have risen substantially, and the industries with the rising premiums are those experiencing declines in union employment shares. The increase in the premium explains from 21% to 64% of the decline in union shares depending on the equation specification. The paper also presents a detailed set of union wage premiums and employment shares for the sample period.