This visualization explores the shape of the United States labor market, examining differences in income between professions, gender, race / ethnicity, age, and education to better understand various pay gaps.
Below is the gap between the pay for males and females (using census gender definitions). The vertical line represents overall median pay. On the left, you can hover over (or click) to see that the median female makes 7% less than the overall median while the median male make about 9% more.
However, when talking about differences in pay, how you frame the question is really important. If we expand our understanding to look per occupation, food preparation and serving sees a fairly small gap compared to the disparity in legal occupations. As we will see shortly, there may be reasons to explain this dynamic but consider what factors might cause these gaps.
Adding more nuance, checking the group size option shows how many people are in each group by gender. Here, for example, "healthcare support" sees a smaller gap but a large disparity in participation by gender.
What about other types of gaps? Switch to "race and ethnicity". Notice that the "Hispanic" and "Black" populations see lower wages than the "White" population across many occupations.
Click the "metrics" checkbox. Note that "Pay" refers to median hourly pay (salaried employees' hourly equivalent). Meanwhile, Gini refers to the Gini Index which is a measure of inequality (higher being more inequitable). This in mind, pay for "Healthcare practitioner and technical" (like doctors) exceeds that of the "Healthcare support" professions but it also sees higher racial / ethnic disparity.
Of course, going back to potential explanations, consider that gaps may reflect differences in educational attainment. Click on edit filters and uncheck everything but "Advanced" and "College" under education. Then, click close filters. Note that, despite filtering down just to just college graduates, similar disparities between those identifying as "White" versus those identifying as "Black" or "Hispanic" still emerge.
However, income isn't the only place gaps arise. Select unemployment with the filters still applied. In many (though not all) professions, those identifying as black see higher unemployment.
For example, while architecture and engineering sees only 1.5% unemployment, those identifying as black within that profession see unemployment at about a point higher than their white counterparts despite having applied filters for only those with college degrees.
What other trends can you find? For example, consider filtering by hours worked (by default, only those working at least 35 hrs included). When you find something interesting, there's a share button where you can send others a link to your results.