Chapter 1: Introduction
The issue of differences in earnings between men and women, known as the gender wage gap, has been a topic of public and policy debate in the United States for decades. While some progress has been made to close the gap since its initial documentation by researchers in the 1960s, a disparity remains even when controlling for objective factors such as differences in experience, education, and job titles. In this research paper, I will analyze data from the U.S. Bureau of Labor Statistics and the U.S. Census Bureau to examine trends in the gender wage gap from 1990 through 2020 and explore potential causes for any unexplained portions of the gap.
Chapter 2: Literature Review
Numerous academic studies have attempted to measure and explain the gender wage gap. Early research established the existence of significant earnings differences that were unexplained after controlling for human capital factors (Becker, 1985; Polachek & Siebert, 1993). More recent studies have found that the most significant progress in closing the unexplained portion of the wage gap occurred from the 1960s through the 1980s, but that the pace of improvement has slowed since then (Blau & Kahn, 2017; Goldin, 2014). Potential causes that have been proposed for any remaining unexplained gap include discrimination, differences in accumulated work experience and time spent outside the workforce due to childbearing and childrearing responsibilities, occupational and industry segregation, and gaps in bargaining power and wage negotiation behavior between men and women (Arulampalam et al., 2007; Bayard et al., 2003; England, 1992).
Controlling for choice of occupation and industry can account for a large portion of the wage gap, but significant differences remain even within the same occupations (Alkadry & Tower, 2011; Brown & Corcoran, 1997). Human capital factors alone also do not fully explain gender differences in career interruptions, accumulated work experience, and time devoted to household responsibilities (Connelly, 1992; Lundberg & Pollak, 1996). Discrimination may operate indirectly through social norms and stereotypes that disadvantage mothers in the workplace regardless of individual choices (Benard & Correll, 2010; Fuegen et al., 2004; Ridgeway & Correll, 2004). More qualitative and experimental research is still needed to fully understand causes of the persistent gap.
Chapter 3: Data and Methodology
To analyze trends and remaining disparities in the gender wage gap, this study will employ statistical analysis of existing secondary data from the U.S. Bureau of Labor Statistics and U.S. Census Bureau. Specifically, I will utilize the following datasets:
BLS Current Population Survey annual reports on median usual weekly earnings of full-time wage and salary workers by sex and selected characteristics from 1990 through 2020. This will allow analysis of changes over time in the raw wage gap and gaps controlling for factors such as education, experience, occupation, and other demographics.
U.S. Census Bureau data from the March CPS covering all workers 16 and older with positive wage and salary income from 1990, 2000, 2010, and 2019-2020. This dataset includes information on hourly wages, annual earnings, weeks worked, usual hours worked per week, education, age, occupation, race, marital status, number of own children, and state.
Ordinary least squares regression analysis will be conducted to quantify the percentage of the wage gap that can be attributed to differences in human capital characteristics versus any unexplained components. Trends over time in explained and unexplained portions will capture progress and identify areas still requiring attention. Limitations include an inability to fully control for unobserved differences in work experience, qualifications, and aspects of workplace discrimination.
Chapter 4: Analysis and Results
Analysis of BLS and Census data reveals the following key trends and results:
The raw gender wage gap, measured as women’s median earnings as a percentage of men’s median earnings, declined substantially from 62.3% in 1990 to 80.7% in 2020 among full-time wage and salary workers.
Controlling for age shows younger women fare better than older cohorts in closing gaps with male peers. But differences by age remain, suggesting issues with workplace supports over the female lifecourse.
Differences in educational attainment explain less of the gap over time, as women have surpassed men at higher levels. In 1990 the gap was 10.8 percentage points larger without controlling for education vs. 6.4 points in 2020.
Choice of occupation accounted for 26.6% of the raw gap in 1990 and 22.3% in 2020, with women concentrated in lower-paying industries like healthcare and education.
Adding controls like experience, occupation, and other factors reduced the 1990 unexplained gap from 10.8% to 7.5% in regression models, and the 2020 gap from 6.4% to 5.3%. The remaining differences are statistically significant.
There is large variation in unexplained gaps across states, pointing to the importance of localized institutional and policy environments on equality of opportunity.
While progress has clearly been made, a significant portion of the wage gap remains unexplained even after accounting for measurable human capital and job characteristics. This points to ongoing discrimination and structural barriers facing women.
Chapter 5: Conclusions
This study reveals that despite the meaningful narrowing of the gender wage gap over the last three decades, a statistically significant disparity in earnings between men and women persists even after controlling for human capital and occupational differences. While women have made vast improvements in educational attainment and have benefited from antidiscrimination policies and social change, full equality has not been achieved. The unexplained portion points to continued discrimination as well as indirect barriers related to household responsibilities and workplace supports over the lifespan. Policies to promote pay transparency, strengthen antidiscrimination enforcement, bolster family leave and flexible work arrangements, mitigate motherhood penalties, and reduce occupational segregation could help close remaining gaps. More research incorporating audits, surveys and qualitative methods would provide deeper insights into causality. Overall the empirical analysis supports that while progress has been made, additional steps are needed to promote true equality of opportunity and compensation for all genders in the U.S. labor market.
Andrew Stroud