Evaluation of the Gender Wage Gap Using Linked Census and Administrative Records
The gender wage gap using linked census and administrative records, focusing on measuring and decomposing the wage gap in detailed occupation categories. The data includes employer-reported earnings, work histories, and occupational characteristics to provide substantive contributions to the literature on the wage gap. The goal is to understand how factors like work history and occupational sorting influence the wage differential between men and women.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
An Evaluation of the Gender Wage Gap using Linked Census and Administrative Records November 2, 2021 Brad Foster and Marta Murray-Close U.S. Census Bureau Christin Landivar and Mark deWolf U.S. Department of Labor 1
Disclaimer Any opinions and conclusions expressed herein are those of the authors and do not necessarily reflect the views of the U.S. Census Bureau or the U.S. Department of Labor. The underlying data used throughout this presentation are protected by Titles 13 and 26 of the U.S. Code. Several steps have been taken to avoid unwarranted and unauthorized disclosure. The Census Bureau s Disclosure Review Board has approved all statistics and estimates presented today for public release under approval numbers CBDRB-FY2019-CES005-002, CBDRB-FY2019- CES010-002, and CBDRB-FY2019-461 2
Context 60,000 Men $52,146 55,000 50,000 Median Yearly Earnings ($) 45,000 $41,977 40,000 Women 35,000 30,000 25,000 20,000 1962 1967 1972 1977 1982 1987 1992 1997 2002 2007 2012 2017 Source: Fontenot, Semega, and Kollar. 2018. Income and Poverty in the United States: 2017. U.S. Census Bureau. Issued September 2018. Data from the CPS ASEC. 3
Motivation and Goals 1. Measurement of gender wage gap (commissioned by Congress) Employer-reported earnings (SSA-DER and W-2s) and reconstructed work histories Measurement and decomposition of the wage gap in 316 detailed occupation categories Census benefits related to survey-based income measurement 2. Substantive contributions to the wage gap literature Whether/how length of work history influences decompositions Account for work history and occupational sorting simultaneously Relationship between occupational characteristics and occupation-specific residual gaps (Goldin 2014) 4
Data CPS DER Links CPS-ASEC (2004-2013) to SSA-DER (1978-2012) 25-year work histories ACS W-2 Links ACS (2015 and 2016) to W-2s (2005-2016) 10-year work histories 316 Detailed occupation categories O*NET occupational characteristics (Goldin 2014) Time Pressure, Competition, Hazards, Autonomy, and Communication/Teamwork Restrict to full-time, year-round workers ages 25 to 54 in civilian sector occupations Adjust for non-random variation in PIK assignment 5
Measuring Wages We discuss the wage gap in terms of hourly wages, which we construct as follows: Sum all earnings and deferred compensation reported by employers for a given year (from SSA-DER and IRS W-2s) Divide by the number of weeks worked in the past year (from CPS and ACS) Divide by the number of usual hours worked per week (from CPS and ACS) Adjust to 2017 dollars using Consumer Pricing Index inflation factors and top- code wages at $500/hr 6
Measuring Work Histories Rule: An individual worked in a given year if their wages were greater than those that would be earned by someone working 520 hours at the federal minimum wage (Goldin and Mitchell 2017) Improvement over relatively standard approximation of potential experience (Age Years of Education 6) Potential experience overestimates actual experience from work histories by about 3.6 years for women and 2.9 years for men, on average. In some of the most populous occupations, potential experience can overestimate actual experience by 4 years among men and nearly 7 years among women. 7
Other Variables Age, Age2 Race/Ethnicity (5-category) Marital status, Presence/age of children* Educational attainment (5-category) Weeks worked last year, Usual hours worked per week Industry (13-category) Region (4-category) Survey year *These variables were used for summary statistics, but were left out of final decomposition models due to their multicollinearity with other variables, like work history, weeks worked last year, and usual hours worked per week. 8
Oaxaca-Blinder Decomposition ?? ??= Total gap Effects of Observed Differences between Men and Women ?? ???? Explained + ???? ?? Unexplained Residual Gaps + ?? ?? Interaction ?? ?? 9
Raw and Residual Gender Wage Gaps In ACS-IRS data (2015-2016), the median women-to-men hourly wage ratio is $20.31/$24.82, or 0.82. Around 30% of this gap is explained , leaving a residual gap of about 12.5%. In CPS-DER data (2004-2013), the median women-to-men hourly wage ratio is $19.75/$24.84, or 0.80. Using 25 years of work history, around 40% of this gap is explained , leaving a residual gap of about 12%. Source: American Community Survey (2015-2016) linked with IRS W-2s (2005-2016); Current Population Survey (2004-2013) linked with SSA-DER (1978-2012). Median wages calculated using administrative earnings. Sample is limited to full-time, year-round civilian sector workers age 25 to 54. CBDRB-FY2019-CES005-002 and CBDRB-FY2019-461. 10
Raw Gender Wage Ratio by Occupation (ACS-IRS) Source: American Community Survey (2015-2016) linked with IRS W-2s (2005-2016). Sample is limited to full-time, year- round civilian sector workers age 25 to 54. CBDRB-FY2019-CES005-002. 11
Experience and the Wage Gap (CPS-DER) PERCENTAGE OF WAGE GAP EXPLAINED 0.0% 1.0% 2.0% 3.0% 4.0% 5.0% 6.0% 7.0% 8.0% 9.0% 10.0% 5 years 0.1% Years of Work Experience included in Decomposition 10 years 2.7% 15 years 5.1% 9.3% 25 years Source: Current Population Survey (2004-2013) linked with SSA-DER (1978-2012). Sample is limited to full-time, year-round civilian sector workers age 25 to 54. CBDRB-FY2019-461. Note: Decomposition also includes education, hours worked, age, metro status, region, race/ethnicity, industry, and occupation. 12
Experience and the Wage Gap (CPS-DER) PERCENTAGE OF WAGE GAP EXPLAINED -10.0% -5.0% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% 0.1% 2.7% Years of Work Experience included in Decomposition Work history 5.1% 9.3% -5.7% -5.7% -5.7% -5.7% Education 5 years 10 years -1.1% -1.4% -1.5% -1.7% 15 years Usual weekly hours 25 years 14.3% 14.1% 14.1% 14.1% Occupation 21.7% 21.8% 22.0% 22.0% Industry Source: Current Population Survey (2004-2013) linked with SSA-DER (1978-2012). Sample is limited to full-time, year-round civilian sector workers age 25 to 54. CBDRB-FY2019-461. Note: Decomposition also includes Age, Race/Ethnicity, Region, and Metro Status. 13
O*NET Characteristics and the Wage Gap (ACS-IRS) PERCENTAGE OF WAGE GAP EXPLAINED -10.0% -5.0% 0.0% 5.0% 10.0% 15.0% 20.0% 25.0% Communication -5.1% Autonomy -0.9% Time pressure -0.4% Returns to Overtime 2.2% Occupational hazards 2.2% Proportion Female 9.7% Competition 14.9% All Occupation Characteristics Combined 22.5% Source: American Community Survey (2015-2016) linked with IRS W-2s (2005-2016). Sample is limited to full-time, year- round civilian sector workers age 25 to 54. CBDRB-FY2019-CES010-002. Note: Decomposition also includes Work History, Age, Race/Ethnicity, Education, Usual Weekly Hours, Industry, Region, and Metro Status. 14
Residual wage gaps by occupation To better understand gaps within occupations, we run 316 occupation-specific decompositions and estimate each occupation s residual gap We find wide variation in residual wage gaps In some occupations, all or nearly all of the gap is explained In others, residual gaps in logged wages as large as 0.63 (or $1.88/hour in favor of men) To what extent are O*NET characteristics associated with these residual gaps? Predict residual gaps as a function of occupation characteristics 15
Occupation Characteristics and Residual Gaps Bivariate regression -0.046** Multivariate regression -0.027 Occupational characteristic Proportion female O*NET characteristics Time pressure Competition Occupational hazards Autonomy Communication Returns to overtime 41-49 hours 50+ hours N R2 0.004 -0.003 0.011** -0.012** -0.021*** -0.002 -0.006 0.012** -0.006 -0.013* 0.387** 0.281*** 316 -- 0.227 0.355*** 316 0.137 Source: American Community Survey (2015-2016) linked with IRS W-2s (2005-2016). Sample is limited to full-time, year- round civilian sector workers age 25 to 54. CBDRB-FY2019-CES005-002. Note: Residuals are obtained from 316 occupation- specific decompositions. Occupation characteristics are standardized to Mean = 0 and SD = 1. 16
Summary & Conclusions Current wage gap is around 18%-20% at the median of the hourly wage distribution Survey- and AR-derived hourly wage gaps are not significantly different from one another Work history results show that we can explain more of the overall gap as we include longer histories, but the effects of other variables are robust When facing trade-offs in publicly-available data, err on the side of more detailed industry/occupation categories 17
Summary & Conclusions Gendered sorting into industries and occupations accounts for approximately one-third of the overall gap a much larger share than work history Wage gap varies significantly by occupation While at parity in some occupations, others see gaps as large as 45% Gap is larger than average in 19% of occupations, but smaller than average in 51% of occupations Finance occupations are common among those with the largest wage gaps Occupation-specific residual gaps vary widely In some occupations, all or nearly all of the gap is explained In others, residual gaps as large as 0.63 (or $1.88/hour in favor of men) exist 18
Summary & Conclusions To some extent, the occupation effect is attributable to characteristics of occupations and not just sorting/segregation Competition and hazards increase the gap Autonomy and communication reduce the gap If one can isolate the features of occupations that have high and low residual differences by gender, one can figure out what factors make for more equal pay. (Goldin 2014: 1102) Residual gaps tend to be smaller in occupations requiring more communication and teamwork Residual gaps tend to be larger in more hazardous occupations, and in occupations with disproportionate returns to overtime 19
Thank you! Questions/Comments? thomas.b.foster@census.gov Working Paper link: https://www2.census.gov/ces/wp/2020/CES- WP-20-34.pdf 20
