Labor Studies Program Meeting

October 23, 2009
David Card, Organizer

Laura Giuliano, University of Miami
Effects of the 1996 Federal Minimum Wage Law on Employment, Substitution, and the Quality of Teenage Labor Supply: Evidence from Personnel Data

Using personnel data from a large U.S. retail firm with more than 700 stores nationwide, Guiliano examines the firm's response to the 1996 federal minimum wage increase. First, increases in average wages had negative, but statistically insignificant effects on overall employment. Second, however, increases in the relative wages of teenagers led to significant increases in the relative employment of teenagers, and especially of more productive teenagers from affluent ZIP codes. This second result is consistent with models that link labor demand to labor market participation, and in particular suggests informational asymmetries may be important in the teenage labor market.


Barry Hirsch, Georgia State University; and Christopher R. Bollinger, University of Kentucky
Wage Gap Estimation with Proxies and Nonresponse

Earnings nonresponse in the Current Population Survey (CPS) is about 30 percent in the monthly surveys and 20 percent in the annual March surveys. Half of CPS earnings records rely on "proxy" respondents, among whom nonresponse is particularly high. Even if nonresponse is random, severe bias attaches to wage equation coefficient estimates on non-match (and some imperfectly matched) imputation attributes. If nonresponse is ignorable (that is, conditional missing at random), then unbiased estimates can be achieved by omitting imputed earners. Bollinger and Hirsch use selection models and longitudinal analysis to examine whether CPS nonresponse is ignorable and how proxy responses affect reported earnings. Based on reasonable instruments to identify selection, they conclude there is negative selection into response for men and, to a far lesser extent, women. Wage equation slope coefficients are affected little by selection but because of intercept shifts, wages for men and to a lesser extent women are understated, as are gender wage gaps. Longitudinal results reinforce the qualitative conclusion that imputation understates earnings, but gender differences are less clear-cut. Cross-sectional estimates of proxy effects on reported earnings suggest large differences in the effects of spouse and non-spouse proxies. These results are driven by heterogeneity, with panel analysis suggesting that both spouse and non-spouse proxy respondents report about 2 percent less than do self respondents. For most wage equation analyses, response bias and proxy reports are of second order importance and the simple exclusion of imputed earners provides a reasonable first-order approach.

Cecilia Machado, Columbia University
Selection, Heterogeneity and the Gender Wage Gap

Observed measures of the gender wage gap may be biased by selection because of differences in the characteristics of the workforce relative to the overall population. This problem is particularly severe for women, who have non-participation and missing information on wages at still very high levels. Existing approaches generally either impose a unique parametric correction procedure or assume the sign of average selection is known. In the case of female selection, however, both positive and negative rules likely co-exist, rendering the sign of average selection unknown. Machado proposes an alternative estimator that recovers a local measure of the gap in models with unobserved heterogeneity in selection. The local measure applies to the subpopulation of "always employed," who are the women that remain employed in the presence of a young child. This is a relevant subpopulation for measuring the gender wage gap as the "always employed" women are similar to men in labor force attachment. In CPS data from 1976 to 2005, Machado shows that the gender wage gap has narrowed substantially for this group, a more than twofold reduction from a -.573 to -.267 gap points (female minus male log wages).


Jennifer Hunt, McGill University and NBER
Why Do Women Leave Science and Engineering

Hunt uses the 1993 and 2003 National Surveys of College Graduates to examine the higher exit rate of women compared to men from science and engineering relative to other fields. She finds that the higher relative exit rate is driven by engineering rather than science, and shows that 60 percent of the gap can be explained by the relatively greater exit rate from engineering of women dissatisfied with pay and promotion opportunities. Contrary to the existing literature, Hunt shows that family-related constraints and dissatisfaction with working conditions are only secondary factors. Her results differ because she uses non-science and engineering fields as a comparison group.


Kerwin Charles, University of Chicago and NBER; Jonathan Guryan, University of Chicago and NBER; and Jessica Pan, University of Chicago, GSB
Sexism and Women's Labor Market Outcomes

Charles, Guryan, and Pan examine the extent to which cross-market differences in women's relative labor market outcomes are determined by differences across markets in sexism - defined as views about the appropriate role women should play in society. Using data from the GSS to measure sexism, they show that selection-corrected gender wage gaps and relative employment rates are significantly related to the degree of sexist views held by the median male, but not with male sexism at the 10th or 90th percentile. Consistent with a standard labor supply model in which sexism lowers women's offered wage, the researchers find lower relative employment of women in more sexist markets is concentrated among women who would have worked few hours in sexism's absence. Finally, they show that the patterns described for male sexism are not apparent for female responses to the GSS questions. The results are robust to a variety of extensions, including alternative strategies for correcting for gender skill differences, and selection. They argue that these results are consistent with a taste-based model of discrimination (Becker (1957)), and are especially striking in light of results from Charles and Guryan (2008) who find that racial wage differences are related to the left tail of the racial prejudice distribution, rather than the median or right tail - exactly as the prejudice model predicts for a group whose prevalence in the labor market much less than that for women. The results suggest that sexism has important implications for the workings of labor markets for men and women.