County Clusters
3,104 U.S. counties grouped into 12 clusters using machine learning. Based on 32 demographic and political features including population, race, education, income, and 2020 presidential voting patterns.
Powered by pgvector embeddings
K-means clustering (K=12)
HNSW similarity index
0
124 counties
Strongly Dem, Urban, College-Educated
1,017,486 avg pop
D+37.1% avg margin
44.3% college+
$98K median income
View all counties →
1
593 counties
Strongly Rep, Rural, White
36,560 avg pop
R+54.3% avg margin
21.7% college+
$68K median income
View all counties →
2
274 counties
Strongly Rep, Suburban
153,428 avg pop
R+26.5% avg margin
34.7% college+
$89K median income
View all counties →
3
161 counties
Lean Dem, Rural, High-Black
32,872 avg pop
D+11.8% avg margin
16.9% college+
$46K median income
View all counties →
4
325 counties
Strongly Rep, Rural
52,286 avg pop
R+38.5% avg margin
17.9% college+
$59K median income
View all counties →
5
238 counties
Lean Dem, Suburban, College-Educated
279,358 avg pop
D+13.0% avg margin
39.0% college+
$87K median income
View all counties →
6
43 counties
Lean Rep, Rural, High-Hispanic
210,724 avg pop
D+7.1% avg margin
18.3% college+
$53K median income
View all counties →
7
475 counties
Strongly Rep, Rural, White
21,608 avg pop
R+56.1% avg margin
16.0% college+
$55K median income
View all counties →
8
444 counties
Strongly Rep, Rural, White
46,890 avg pop
R+23.0% avg margin
24.4% college+
$68K median income
View all counties →
9
114 counties
Lean Rep, Suburban, College-Educated
117,142 avg pop
R+4.6% avg margin
37.5% college+
$65K median income
View all counties →
10
49 counties
Competitive, Rural, High-Native
17,501 avg pop
D+5.6% avg margin
17.8% college+
$61K median income
View all counties →
11
303 counties
Strongly Rep, Rural, White
6,230 avg pop
R+64.0% avg margin
24.8% college+
$65K median income
View all counties →
Clustering Methodology
Counties are embedded into 32-dimensional space using demographic data (population, race, income, education) and political data (2020 presidential vote margins). K-means clustering (K=12) groups similar counties together.
The HNSW (Hierarchical Navigable Small World) index enables fast similarity search, powering the "Similar Counties" feature on county profile pages.
Data sources: U.S. Census Bureau ACS 2024, PL 94-171 (2020), MIT Election Lab