Single-race population estimates, national, 2000-2024 (backfill)
Source:R/data-pop_singlerace_full.R
pop_singlerace_full.RdAnnual US resident population by age (5-year bins, 18 groups), sex, single-race
group (the six OMB-1997 alone categories), and Hispanic origin, from the Census
Bureau Population Estimates Program (PEP). This is the 2000-2024 BACKFILL of the
single-race series: it extends pop_singlerace (2020-2024 only)
back to 2000 by stitching three PEP vintages, each contributing a disjoint,
non-overlapping calendar-year range.
Format
A data frame with 10800 rows and 9 columns
- year
year of observation (2000-2024)
- age
starting age for the 5-year age bin (0, 5, ..., 85 = 85+)
- sex
"male"or"female"- race
single-race group:
white_only,black_only,american_indian_only,asian_only,nhopi_only,multiracial- hispanic_origin
"non_hispanic"or"hispanic"- pop
population count
- vintage
PEP vintage for that row:
"int2000","int2010", or"V2024"- scheme
race scheme (
"single")- source
data source (
"census_pep")
Details
Use it with single-race death counts (remap_race()/categorize_race()
codes 101-106, coded 2020 and later). It is a single, consistent taxonomy across
all 25 years (RACE 1-6), NOT comparable to the bridged pop_bridged
or the legacy single-race-alone pop_est. The 2020-2024 slice is
identical (to the person) to the frozen pop_singlerace. Only the
finest cells are stored; "total"/"both"/"all" are
synthesized on demand, never stored, so aggregation never double-counts.
The vintage column records each row's provenance:
"int2000" (2000-2009, the 2000-2010 intercensal estimates),
"int2010" (2010-2019, the rebased 2010-2020 intercensal estimates – NOT
the postcensal V2020 series, which was never re-based to the 2020 census), and
"V2024" (2020-2024, the Vintage-2024 estimates).
State- and county-level single-race backfill denominators are distributed
separately (too large to bundle) as pop_singlerace_state_full /
pop_singlerace_county_full and fetched via
download_pop_data(scheme = "single") / get_pop_county().