Ready-to-use precinct shapefiles for Mosaic, with population, race/age demographics, and recent election results pre-joined.
Each <State>_DRA_Mosaic.zip contains a single TIGER/Line VTD shapefile with 14 attributes:
| Column | Type | Meaning |
|---|---|---|
GEOID |
string | 11-char VTD GEOID20 (state FIPS + county FIPS + VTD) |
CTY |
string | 3-char county FIPS |
POP_TOTAL |
int | 2020 total population |
VAP_TOTAL |
int | 2020 voting-age population |
VAP_WHITE |
int | 2020 VAP, White alone, not Hispanic |
VAP_BLACK |
int | 2020 VAP, Black any-part (incl. Hispanic) |
VAP_LATINO |
int | 2020 VAP, all Hispanic regardless of race |
VAP_ASIAN |
int | 2020 VAP, Asian any-part (incl. Hispanic) |
TRUMP_16 / CLINTON_16 |
int | 2016 presidential R / D votes |
TRUMP_20 / BIDEN_20 |
int | 2020 presidential R / D votes |
TRUMP_24 / HARRIS_24 |
int | 2024 presidential R / D votes |
Full schema notes and definitional choices: see DATA_LICENSE.md.
TRUMP_24 / HARRIS_24 for DT_EST_24 / KH_EST_24 / EST_24_FLG. The estimates come from applying each county’s reported 2020 → 2024 swing uniformly across its precincts, so county and state sums match true 2024 totals exactly. Per-precinct values are modeled — treat with caution.These shapefiles are joins of two upstream sources:
Because they incorporate DRA-packaged data, the joined shapefiles inherit DRA’s license:
Mosaic itself is MIT-licensed; this data is not. Use accordingly.
See DATA_LICENSE.md for the full attribution table, upstream provider list, and schema choices baked into the join.