Race, Gender, and Geography

Deep-dive analysis of voter turnout across racial, gender, and geographic dimensions using Census CPS microdata. Unlike the CIRCLE published estimates, this analysis lets us slice the data any way the sample size supports.

Race × gender breakdown

Within-race gender gaps

For each racial group, how different are men's and women's turnout rates?

Race × region

Where does each racial group vote at the highest rates?

Cells with fewer than 30 respondents are suppressed to avoid unreliable estimates.

Race × education

Does the education turnout gap look different across racial groups?

The education gradient — higher education = higher turnout — appears in every racial group, but the slope varies. Groups where the slope is steeper may see larger turnout gains from interventions that increase educational attainment or that target non-college populations.

Full breakdown table

Every race × gender × region combination with reliable sample size:

Change from a comparison year

Sample size matters

All cells in this analysis use at minimum 30 respondents. Cells below that threshold are suppressed because the margin of error becomes too wide for meaningful interpretation. For context, the standard error for a proportion with n=30 is roughly ±9 percentage points — large enough that small differences are not interpretable.

If you need state-level breakdowns by race, the CPS sample is generally too thin except for the largest states (CA, TX, NY, FL). The State-by-State page uses CIRCLE's aggregated voter file methodology which handles this better for overall youth turnout, though it does not publish reliable race × state crosstabs.