Explaining Changes in Women's Earnings and Employment from 1970 to 2010: A Quantitative Analysis of Discrimination and Labor Force Hypotheses

Abstract

My dissertation examines three factors that influence women's workplace inequality: the Pregnancy Discrimination Act, the U.S. Supreme Court ruling in Meritor Saving Bank v. Vinson, and women "opting out" (i.e., voluntarily leaving) of work to take care of their families. Each of the three essays attempts to resolve one or more fundamental questions in the literature on women's workplace inequality. Additionally, each of these factors is described in its own essay, and these essays are presented in chronological order of the events they discuss. The first substantive chapter, Chapter Two, examines the association between the 1978 Pregnancy Discrimination Act (PDA) and women's employment rates, men's employment rates, and women's disadvantage in employment rates. The PDA declared pregnancy discrimination a type of sex-based discrimination, making it illegal under Title VII. Among other provisions, the PDA extended temporary disability benefits (TDB) to pregnant workers who were already covered by TDB for non-pregnancy-related conditions. This extension was designed to increase the likelihood of women's returning to work after giving birth. There is a debate within the literature on whether the PDA positively affected women, hurt women, or had no effect on women. This paper uses difference-in-difference modeling with Current Population Survey data from 1968 to 2010 to weigh in on this debate. There is a positive association between the PDA and the difference between men's and women's employment because it narrows the gap between women's and men's employment rates. However, the PDA had no effect on women's employment rates and a negative effect on men's employment rates. Chapter Three examines the association between wages and the 1986 U.S. Supreme Court decision in Meritor Savings Bank v. Vinson, which declared workplace sexual harassment illegal under Title VII of the 1964 Civil Rights Act. I test two hypotheses, a "naïve hypothesis," which argues that anti-sexual harassment policies increase women's wages, and a "critical hypothesis," which contends that these policies hurt or had no effect on women's wages. Using a difference-in-difference model with Current Population Survey data from 1977 to 2010, I find that there is a negative association between Meritor and women's wages. Furthermore, I find no association between Meritor and either men's wages or the gender wage ratio. Finally, Chapter Four examines the association between women opting out and socioeconomic status (SES). I explore two contrasting storylines within the opting out literature. The first storyline focuses on the lives of high-SES women and the likelihood that they opt out. The other storyline focuses on the lives of low-SES women and the likelihood that they opt out. It is important to note that neither storyline denies the existence of the other, but research that focuses solely on one class of women may miss a significant part of the story. Using the Current Population Survey from 1980 to 2009, I first describe the trends of opting out over this 30-year span. I then test six SES-related variables (education of spouse, education of respondent, household income, below poverty line, receive welfare income, and receive food stamps). I find that low-SES women are opting out at similar rates between 1980 and 2009, but the rates for high-SES women have declined. Additionally, I find overwhelming evidence that low-SES women are more likely to opt out.