Gender-based pay discrimination has been illegal in the United States for over half a century, but significant wage gaps between male and female workers persist. There are several, complex factors driving the gender pay gap, such as occupational segregation (differences in industries, with more men in higher- paid industries), vertical segregation (fewer women in senior, better paying positions) and differences in years of experience, and finally, the “motherhood penalty.”
It is also true that women are often paid a lower compensation than their male counterparts, even when they are in the same industry and hold the same job/position. This gap is known as the adjusted gender wage gap, where variables such as age, education, experience, occupation, industry, and location are controlled, but the wage gap still persists, albeit at a lower rate than the unadjusted gender wage gap.
How are socioeconomic variables like race, age, and educational attainment related to the gender pay gap? How does the gender pay gap differ by state, country, or continent? What are some jobs in the U.S with the widest gender pay gap? In order to better understand the extent and nature of the gender pay gap in the U.S and beyond, I performed descriptive analytics on three datasets using R.
It is important to note that most of the analysis was performed using macro data on the unadjusted gender wage gap, which doesn’t control for the variables mentioned above. However, one dataset I used did control for variables like industry and occupation, which will be further discussed in my analysis below.
Across the three datasets, I used one, unified metric to measure the gender pay gap: women’s earnings as a percentage of men’s. As such, if women’s earnings as a percentage of men’s is low, it indicates a wide gender pay gap — a high difference in wages earned by female and male workers. To be more precise, for instance, if women’s earnings are, on average, 81% of men’s, that means the gender pay gap is about 19%. This might be potentially confusing, so I will try my best to clarify this point throughout my analysis.
- U.S Bureau of Labor Statistics: Highlights of Women’s Earnings in 2019
- U.S Bureau of Labor Statistics: Annual table, Earnings by detailed occupation and sex (2020)
- OECD: Decile Ratio of Gross Earnings (2019)
- Wikipedia: Table of presidential elections by states since 1972
- Country and Continent Codes List
— — — — — — — — — — — — — — — — — — — — — — — — — —
I. Analysis conducted using <U.S Bureau of Labor Statistics: Highlights of Women’s Earnings in 2019>
Women’s earnings as a percentage of men’s, from 1979 to 2019
How has the gender pay gap changed in the U.S in the last 40 years? Let’s find out. Here, we are only focusing on full-time wage and salary workers.
Women’s earnings as a percentage of men’s have increased from 1979 to 2019, which shows that the gender pay gap has been narrowing. In 1979, women were paid 62.3 cents for every dollar men were paid, while in 2019, women were paid 81.5 cents for every dollar men were paid. Again, this dataset shows the unadjusted gender wage gap, meaning that variables like occupation, seniorities, and age have not been controlled for. It is worth noting that the gender pay gap has narrowed rapidly from 1979 to 2009, but progress has more or less stalled from 2010 to present.
Analyzing the gender wage gap at the state level
First, I began by plotting a map of the U.S that shows the gender wage gap at the state level.
The median usual weekly earnings of full-time wage and salary workers in 2019 (data source: Bureau of Labor Statistics) was used to create this plot. This map shows that states like California have high women’s earning as a percentage of men’s, which means that their gender pay gap is narrow. In contrast, women’s earnings as a percentage of men’s in states like Wyoming and Indiana are low, which indicates a wide gender pay gap.
While this map is very useful, it doesn’t provide a detailed overview of exactly which states have the widest/narrowest gender wage gap. Using the same data that was used to create the map above, I outlined 7 states with the widest gender wage gap (i.e. lowest women’s earnings as a % of men’s) and 7 states with the narrowest gender wage gap (i.e. highest women’s earnings as a % of men’s.)
I noticed that the states with the widest gender wage gap are all states that are traditionally considered red states, and the states with the narrowest gender wage gap are all states that are typically considered blue states. I wanted to further explore this relationship between a state’s political party association and its gender pay gap. To do so, I merged the gender wage gap data with another dataset (source: Wikipedia) that showed which candidate each state supported in U.S presidential elections from 1972 to 2020.
I first converted each state’s voting history from 1972 to 2020 into numeric terms; if the state voted Republican every single election during that period, it had a value of 0, and if the state voted Democrat every single election during that period, it had a value of 1. Then, I plotted the relationship between a state’s political party association and its gender wage gap.
While more statistical analysis must be performed to determine whether the two have a statistically significant relationship, the two variables did appear to be moderately correlated. When I ran a simple linear regression model using the states’ political party association as the predictor, it yielded a R square value of 0.42 and a p-value of 3.04e-07.
Again, this plot confirmed that blue states tend to have a narrower gender wage gap than red states. While more research must be conducted to understand what is primarily driving this difference, one might suspect it may be due to Democratic state legislators being more willing to pass equal pay legislations within a state than Republican state legislators.
Analyzing how socioeconomic variables like race, educational attainment, and age group impact the gender wage gap in the U.S
(1) Race and Ethnicity
It appears that Asian and white women, on average, actually experience the widest gender pay gaps relative to Asian and white men, respectively. However, it must be noted that this dataset only presents the gender pay gap within each race group, meaning that it doesn’t provide information on gender pay gaps across race groups (for instance, black women’s earnings as a % of white men’s.) As such, the fact that the gender pay gap is the widest for Asian and White Americans may be due, in part, to the fact that Asian and white men make much more than black or Hispanic men. In fact, the Economic Policy Institute’s 2015 research article demonstrated that Black and Hispanic women experience the widest pay gaps, more so than White and Asian women, when women’s median hourly wages were expressed as a share of white men’s.
(2) Educational Attainment
In the 70s and 80s, the gender wage gap was the narrowest for individuals with bachelor’s degrees and higher. Over time, however, the gender wage gap for individuals with higher educational attainment widened, and starting from the late 2000s, individuals with bachelor’s degrees and higher actually face a wider gender pay gap than other groups do.
On average, it appears that the gender wage gap is the widest for older age groups.
II. Analysis conducted using <U.S Bureau of Labor Statistics: Annual table, Earnings by detailed occupation and sex (2020)>
This dataset is particularly valuable, as it shows both (1) the number of male and female workers and(2) the median weekly earnings for male and female workers for each occupation surveyed. As such, this dataset allowed me to gather useful insights not only about the gender pay gap but also about the share of female workers in each occupation. While this dataset doesn’t control for variables like seniorities and age, it does control for industry and occupation choice.
To start, I wanted to understand whether there is a relationship between the share of female workers as a % of total workers and the gender wage gap across various occupations.
According to this scatterplot, there doesn’t seem to be a clear relationship between the two variables.
A high-level analysis of (1) proportion of female workers and (2) the gender pay gap by occupation type
In addition to providing information about each occupation, this dataset also grouped certain occupations into ‘occupation types’ that we could analyze at a high-level.
Occupation types like “healthcare support,” “community and social service”, and “office and administrative support” had high shares of female workers in its workforce and a relatively narrow gender pay gap. On the other hand, despite having a relatively high proportion of females in the workforce, legal occupations had a particularly wide gender pay gap.
Comparing the gender pay gap by occupation
After analyzing each occupation type, I then analyzed the gender pay gap for each individual occupation. This allowed me to observe which occupations have the narrowest/widest gender wage gap.
While occupations like “producers and directors” and “office and administrative support workers” had the narrowest gender wage gap, occupations like “medical scientists” and “financial managers” had the widest gender wage gap.
Comparing the proportion of female workers by occupation
While occupations like “personal appearance workers,” “nurse midwives,” and “healthcare diagnosing or treating practitioners” had the highest proportion of women in their workforce, occupations like “sales engineers,” “molders and molding machine operators,” and “cost estimators,” as well as several engineering occupations, had the lowest proportion of female workers.
III. Analysis conducted using <OECD’s Decile Ratio of Gross Earnings (2019)>
Finally, I wanted to understand how the United State’s gender pay gap compares to other OECD countries. This dataset allowed me to observe the extent of the gender pay gap on a global scale and where the US stands relatively.
Median gender wage gap for OECD countries, 2014–2019
Compared to the average gender wage gap across all OECD countries (as highlighted above), the United States has a wider gender wage gap. In fact, of all the OECD countries, the United States has the 7th widest gender wage gap.
While Korea and Japan have the widest gender wage gaps, countries like Costa Rica, Romania, and Croatia have the narrowest gender wage gaps. I noticed how Asian countries appeared to have wider gender wage gaps, and I wanted to compare the gender wage gap at the continent-level as well. To do so, I merged the OECD dataset with another dataset that matched individual countries with the continents they belonged in.
Comparing the gender wage gap for OECD countries by continent
Corroborating my earlier observation, countries in Asia did appear to have the widest gender wage gap, followed by North America, Europe, Oceania, and South America. It is worth noting a key limitation to this analysis: most OECD countries are in Europe, and there are only 2 countries in South America and 5 countries in Asia that are members of the OECD, so extrapolating beyond the relevant group, OECD countries, may be inaccurate.
Some critics say the Bureau of Labor Statistic’s data doesn’t compare apples to apples, as variables like seniorities, age, and years of experience aren’t controlled for. Yes, that is true, but nonetheless, advocates for gender pay equality consider the agency’s wage gap data to be a critical measure of how women are faring in the workplace.
“Pay equity is a critical part of being a truly inclusive workplace,” Julie Ann Overcash, vice president of human resources and global director of compensation and benefits at Intel says, and it “creates more-productive, high-performing organizations.” In order to close the gender pay gap, company leaders and government officials must continue the push toward paying workers equally for similar work.
According to this New York Times article, if current trends continue, the gender wage gap is expected to close in around 38 years. Yet, for Black and Hispanic women, “the deadline is a whole century away.” Dr. C. Nicole Mason, president of the Institute for Women’s Policy Research, says, “if we do nothing, my daughter, and daughter’s daughter, will not see pay equity in their lives.”
Until then, we must continue to educate ourselves on the root cause and possible solutions to the gender pay inequality, both at the institutional and individual level. Analyzing underlying nuances, like the relationship between the gender pay gap and various socioeconomic variables, how the gender pay gap differs by occupation, as well as the geographical differences, whether it be by state, country, or continent, is critical to furthering our understanding of the issue.