The Effect of Economic Conditions on the Success of Equal Employment Opportunity Laws: An Application to the Sex Differential in Earnings
实证分析1970年代经济衰退如何削弱了1964年民权法案第七章对女性收入和性别收入差距的改善效果，发现执法效果与失业率负相关，对政策评估有参考价值。
W ORSENING economic conditions curtailed the effectiveness of equal employment opportunity (EEO) laws during the early 1970s. But for the recession of the 1970s, Title VII of the 1964 Civil Rights Act would have increased female earnings and narrowed the sex differential in earnings by more than it actually did. This paper empirically distinguishes the effects of Title VII itself from those induced by the economic environment.I Overall between 1968 and 1974, enforcement of Title VII increased female earnings and narrowed the sex differential; but, despite overall gains, enforcement occurring prior to the law's March 1972 amendments had reduced the earnings of both sexes by 1974 (Beller, 1979). Although negative long-run effects of EEO laws are consistent with certain microeconomic predictions (Heckman and Wolpin, 1976),2 the coincidence of much of this six-year period with a cyclical decline in economic activity suggests that those effects might have been caused by macroeconomic conditions.3 A recession can erode gains and impede progress toward EEO goals by increasing the monetary and psychic costs that firms complying with EEO laws face. If such costs vary inversely with cyclical changes in economic activity, then we should expect to find a pro-cyclical pattern in firms' compliance. Compliance by firms also depends directly upon the expected costs of violating the law. Firms' costs vary with their (1) probability of getting caught violating the law and (2) probability of paying a penalty once caught. The 1972 amendments to Title VII increased these expected costs; allowing a discrete change in the effects of these probabilities in 1972 can capture the amendments' impact.4 A continuous change in the effects of these probabilities over time can capture the interaction between compliance costs and the business cycle. Upon this framework, we can build an econometric model. Its estimates show the effect of Title VII on earnings in 1968 through 1974 to be inversely related to the aggregate unemployment rate. These results confirm that macroeconomic conditions can alter the effects of social programs, a fact which analysts must take into account.