Uninsured Rate Under 65 (Census SAHIE)

Share of Georgians under 65 without health insurance, statewide and for every county, yearly 2008-2024 - Census model-based estimates (SAHIE).

Source: U.S. Census Bureau, Small Area Health Insurance Estimates (SAHIE)
Year coverage: 2008–2024
Updated: annual
Version: 1.0.0
Schema hash: 1d8825bcb02e

Detail levels available

StateCounty · default

Overview

Health insurance coverage of people under 65 from the U.S. Census Bureau's Small Area Health Insurance Estimates (SAHIE), for Georgia and each of its 159 counties, 2008-2024: the share of people without health insurance (the headline uninsured rate), the number uninsured and insured, and the group's size, each with a 90 percent margin of error, by sex, age band and family income relative to the poverty line, and - for the state only - by race/ethnicity. SAHIE estimates are model-based: they combine the American Community Survey with administrative records and population estimates, and are the only single-year uninsured estimates for every county.

Row grain: One row per year, county_fips, district_code, tract_geoid, demographic, sex, age_bracket, income_to_poverty.

Key metric

uninsured_rate — Uninsured Rate (Under 65) (proportion)

The single headline value most answers want.

JSON schema →Data contract (YAML) →

Variables

Every column in the fact table. The code is the exact field name the API returns.

Column	Type	Unit	Range	Null?	Description
`year`Year	integer	—	—	no	Estimate year: the calendar year SAHIE models, from that year's single-year American Community Survey responses combined with administrative records and population estimates (2024 = 2024). An annual series: each year is its own estimate, not a pooled period. Served 2008-2024, the ACS-based SAHIE series; the 2006-2007 estimates (an earlier method based on the Current Population Survey, not comparable) are excluded. Full description▾ Estimate year: the calendar year SAHIE models, from that year's single-year American Community Survey responses combined with administrative records and population estimates (2024 = 2024). An annual series: each year is its own estimate, not a pooled period. Served 2008-2024, the ACS-based SAHIE series; the 2006-2007 estimates (an earlier method based on the Current Population Survey, not comparable) are excluded.
`county_fips`County FIPS	string	—	—	yes	5-digit county FIPS code (state 13 + county). Populated only on county-level rows; NULL on the state rows. Full description▾ 5-digit county FIPS code (state 13 + county). Populated only on county-level rows; NULL on the state rows. Joins counties; adds its label columns.
`district_code`District Code	string	—	—	yes	Always NULL: SAHIE publishes no school-district geography. Kept for the census domain's standard key shape. Full description▾ Always NULL: SAHIE publishes no school-district geography. Kept for the census domain's standard key shape. Joins districts; adds its label columns.
`tract_geoid`Tract GEOID	string	—	—	yes	Always NULL: SAHIE publishes no census-tract geography. Kept for the census domain's standard key shape. Full description▾ Always NULL: SAHIE publishes no census-tract geography. Kept for the census domain's standard key shape. Joins tracts; adds its label columns.
`demographic`Demographic	string	—	—	no	Race/ethnicity (state rows only); 'all' is everyone under 65. Full description▾ Race/ethnicity (FK to the demographics dimension), published by SAHIE for the STATE only: county rows are always 'all'. Mutually exclusive groups: 'hispanic' is Hispanic or Latino of any race; 'white', 'black', 'native_american' (American Indian and Alaska Native), 'asian', 'pacific_islander' (Native Hawaiian and Other Pacific Islander) and 'multiracial' (two or more races) are the not-Hispanic groups; 'all' is everyone. They split Hispanic origin the same way as census/acs/population_by_race but follow the Census Bureau population estimates' race categories, which have no 'some other race' group (those people are counted in a specific race), so they are close to, but not identical with, population_by_race's groups (which include 'other'): do not mix counts from the two topics. 2008-2020 publish only white, black and hispanic (they do not sum to 'all'); from 2021 all seven are published and sum exactly to 'all'. Never published for age_bracket 'under_19'. Sex is on the sex column, not here. 8 allowed values▾ `all` `white` `black` `hispanic` `native_american` `asian` `pacific_islander` `multiracial` Joins demographics; adds its label columns.
`sex`Sex	string	—	—	no	Sex: all, male or female. Full description▾ Sex: 'male', 'female', or 'all' (both sexes). A partition: male + female = all for every count (exactly from 2013; within 1 person in 2008-2012 by rounding). Published at state and county level, but not for age_bracket 'under_19' (both sexes only). 3 allowed values▾ `all — total / aggregate` `male` `female`
`age_bracket`Age Bracket	string	—	—	no	Age band; 'all' is everyone under 65. Bands overlap - never add them. Full description▾ Age band. 'all' is everyone SAHIE covers: people UNDER 65 (SAHIE publishes nothing for 65 and older). The other bands nest and overlap rather than partition: under_19 (ages 0-18), 18_64, 21_64 (from 2014), 21_29 (from 2024), 40_64, and 50_64 (county rows from 2010; state rows from 2008). 18_64 contains 21_64, 40_64, 50_64 and 21_29, and under_19 and 18_64 both include 18-year-olds, so counts must never be added across bands. under_19 is published for both sexes and all races only. 7 allowed values▾ `all — total / aggregate` `under_19` `18_64` `21_64` `21_29` `40_64` `50_64`
`income_to_poverty`Income to Poverty	string	—	—	no	Family income as a percent of the poverty line; 'all' is every income. Ranges overlap - never add them. Full description▾ Family income as a percent of the federal poverty threshold (the income-to-poverty ratio). 'all' is every income. The others are cumulative or overlapping ranges, not a partition: at_or_below_138_pct, at_or_below_200_pct, at_or_below_250_pct, at_or_below_400_pct, and 138_to_400_pct (above 138% up to 400%, from 2012). Each 'at or below' range contains the lower ones, so counts must never be added across ranges. 138% of poverty is the Affordable Care Act's Medicaid-expansion income limit. 6 allowed values▾ `all — total / aggregate` `at_or_below_138_pct` `at_or_below_200_pct` `at_or_below_250_pct` `at_or_below_400_pct` `138_to_400_pct`
`uninsured_rate`Uninsured Rate (Under 65)	number	proportion	—	yes	Share of people under 65 without health insurance (model-based estimate). Full description▾ Share of the group without health insurance coverage (0-1): SAHIE's published percent uninsured divided by 100, served as published and never recomputed. A model-based estimate for people under 65. From 2013 it equals num_uninsured / num_population within rounding (0.0005); in 2008-2012 the published percent can differ from the counts' ratio by up to 0.0086 (0.86 points), a property of the source.
`uninsured_rate_moe`Uninsured Rate Margin of Error	number	proportion	—	yes	90 percent margin of error for uninsured_rate (0-1 scale; the published percent margin / 100): the 90 percent confidence interval is uninsured_rate plus or minus this value (the source's published bounds equal exactly that and are not served). Always positive. Never the headline metric. Full description▾ 90 percent margin of error for uninsured_rate (0-1 scale; the published percent margin / 100): the 90 percent confidence interval is uninsured_rate plus or minus this value (the source's published bounds equal exactly that and are not served). Always positive. Never the headline metric.
`num_uninsured`Number Uninsured	integer	count	—	yes	People in the group without health insurance coverage, as published (model-based estimate) - the rate's numerator. Full description▾ People in the group without health insurance coverage, as published (model-based estimate) - the rate's numerator.
`num_uninsured_moe`Number Uninsured Moe	integer	count	—	yes	90 percent margin of error for num_uninsured, as published. Always positive. Full description▾ 90 percent margin of error for num_uninsured, as published. Always positive.
`num_population`Number in Group (Under 65)	integer	count	—	yes	People in the group (the age x sex x race group within the income range), as published - the universe and the rate's denominator. On income_to_poverty 'all' rows it is the Census Bureau's population estimate for the group; on income-range rows it is itself a model-based estimate. Full description▾ People in the group (the age x sex x race group within the income range), as published - the universe and the rate's denominator. On income_to_poverty 'all' rows it is the Census Bureau's population estimate for the group; on income-range rows it is itself a model-based estimate.
`num_population_moe`Number Population Moe	integer	count	—	yes	90 percent margin of error for num_population, as published. 0 on every income_to_poverty 'all' row: there the group size is a population estimate SAHIE treats as fixed (a control), so the source publishes no model error for it. That 0 does not mean the count is exact, and it does not mean the rate has no uncertainty (uninsured_rate_moe carries that). Positive on every income-range row. Full description▾ 90 percent margin of error for num_population, as published. 0 on every income_to_poverty 'all' row: there the group size is a population estimate SAHIE treats as fixed (a control), so the source publishes no model error for it. That 0 does not mean the count is exact, and it does not mean the rate has no uncertainty (uninsured_rate_moe carries that). Positive on every income-range row.
`num_insured`Number Insured	integer	count	—	yes	People in the group with health insurance coverage, as published (model-based estimate). num_uninsured + num_insured = num_population within 1 person (independent rounding). Full description▾ People in the group with health insurance coverage, as published (model-based estimate). num_uninsured + num_insured = num_population within 1 person (independent rounding).
`num_insured_moe`Number Insured Moe	integer	count	—	yes	90 percent margin of error for num_insured, as published. Equal to num_uninsured_moe on income_to_poverty 'all' rows, where the group size is fixed; generally different on income-range rows. Always positive. Full description▾ 90 percent margin of error for num_insured, as published. Equal to num_uninsured_moe on income_to_poverty 'all' rows, where the group size is fixed; generally different on income-range rows. Always positive.

Filters

Query parameters the fact endpoint accepts. Comma-separated values are OR within a parameter; multiple parameters AND together.

Parameter	Kind	Allowed values	Notes
`year`	year_exact	any value	—
`year_min`	year_range	any value	—
`year_max`	year_range	any value	—
`detail`	detail	`countiesstates`	enum-enforced, default: counties
`district_code`	foreign_key	any value	multi-value
`county_fips`	foreign_key	any value	multi-value
`tract_geoid`	foreign_key	any value	multi-value
`demographic`	foreign_key	8 allowed values▾ `all` `white` `black` `hispanic` `native_american` `asian` `pacific_islander` `multiracial`	multi-value, enum-enforced
`sex`	categorical	`allmalefemale`	multi-value, enum-enforced
`age_bracket`	categorical	7 allowed values▾ `all` `under_19` `18_64` `21_64` `21_29` `40_64` `50_64`	multi-value, enum-enforced
`income_to_poverty`	categorical	`allat_or_below_138_pctat_or_below_200_pctat_or_below_250_pctat_or_below_400_pct138_to_400_pct`	multi-value, enum-enforced
`district_type`	dimension_attribute	7 allowed values▾ `commission_charter` `resa` `standard` `state_agency` `state_charter` `state_school` `state_special`	multi-value, enum-enforced
`demographic_category`	dimension_attribute	12 allowed values▾ `aggregate` `disability` `economic_status` `esol` `foster_care` `gender` `grade` `homeless_status` `migrant_status` `military` `race` `sped`	multi-value, enum-enforced

Example requests

Run these against the base URL — they return live data.

All rows (first page), counties level

curl "https://georgiacivicdata.org/api/v1/census/uninsured"

counties data for 2024

curl "https://georgiacivicdata.org/api/v1/census/uninsured?year=2024"

Filtered to sex = male for 2024

curl "https://georgiacivicdata.org/api/v1/census/uninsured?year=2024&sex=male"

One county (FIPS 13121), 2024

curl "https://georgiacivicdata.org/api/v1/census/uninsured?county_fips=13121&year=2024&detail=counties"

Download 2024 as CSV

curl "https://georgiacivicdata.org/api/v1/census/uninsured?year=2024&format=csv"

Notes & limitations

NULL handling

Zero is a real, reported value.

Limitations

People UNDER 65 only: SAHIE publishes nothing for ages 65 and older, so uninsured_rate is the under-65 rate (age_bracket 'all' means under 65), not an all-ages rate. Every figure is a Census Bureau model-based estimate, not a direct count. age_bracket and income_to_poverty values overlap: never add across them. Race/ethnicity (demographic) is state-only, never for under_19; native_american, asian, pacific_islander and multiracial start in 2021. Before 2021 only white, black and hispanic are published, so the groups do not sum to 'all' until 2021. The model combines American Community Survey responses with administrative records and population estimates. Not comparable with census/acs/health_insurance_by_citizenship, which covers all ages over a pooled multi-year survey period. year is a single estimate year, an annual series rather than a pooled period: year-to-year change can be read, but judge it against both years' 90 percent margins of error (the *_moe columns). 2008-2024 are served; the 2006-2007 estimates, made with an earlier method based on the Current Population Survey, are not comparable and are excluded. Coverage grows: county 50_64 from 2010, 138_to_400_pct from 2012, 21_64 from 2014, 21_29 from 2024. uninsured_rate is the Census's published percent / 100, never recomputed: from 2013 it equals num_uninsured / num_population within rounding, but in 2008-2012 the published percent can differ from the counts' ratio by up to 0.86 points. County estimates are controlled to the state's, so the 159 counties sum to the state row (exactly from 2013, within 16 people before) and male + female = all. num_population_moe is 0 on every all-incomes row because the group size there is a population estimate SAHIE treats as fixed; that is not a claim the count is exact, and the rate keeps its own margin. SAHIE publishes every cell: there is no suppression and no NULL. The published percent insured (exactly 1 - uninsured_rate, same margin) and the 90 percent confidence bounds (estimate plus or minus the margin) are not served. State and county geography only; there is no school-district or tract geography.