What is this site?
It brings county-level demographic, health, and economic data together from public federal sources — Census ACS, CDC, HUD, CMS and more — so every measure is comparable across the same counties, states, and the US.
Where does the data come from?
All of it is public government data. Each source, its vintage, and its release schedule are listed on the Sources & Dashboards page, and every measure in the dataset records its exact source.
What is ACS-5?
The Census Bureau's American Community Survey 5-year estimates — five years of survey responses pooled together so that even small areas have reliable numbers. It is the backbone of this site, and the latest release covers 2020–2024.
What areas does it cover?
All U.S. counties, with state and national figures for comparison. Some measures go down to the census tract level; a few are only available at the state or national level.
Does this include Americans living overseas?
No. Like the Census, this site counts the resident population — people living in the United States. U.S. citizens living abroad — roughly 1% of the population, an estimated several million people — are not in the ACS or decennial counts. Because there is no official list of them, the best estimate comes from the Federal Voting Assistance Program's Overseas Citizen Population Analysis.
How current is the data, and how often is it updated?
It varies by source. ACS-5 updates once a year (usually each December); CDC, HUD, CMS and others follow their own schedules. Each source's typical release month and lag are shown on the Sources & Dashboards page.
How much history is there?
The curated reports — topic, county, provisional — show the current year and the prior year. That is deliberate: the priority is timeliness. Government data is already about a year old when it is published, and the goal here is to load it within about ten days of release, so two years is enough to show the direction of travel while staying current.
ACS is roughly 60–70% of the measures on this site, and for those the history runs deeper: the ACS5 Data Explorer carries every ACS field from 2020 to the current year, down to census tracts. One caution — definitions, geographies, and methods change over time, and a trend line drawn across those breaks can mislead. ACS-1 & 5 Changes Since 2020 flags every table dropped, added, or relabeled, so check a table there before you trend it. For history deeper than 2020, use IPUMS. Two exceptions — Life Expectancy and Homelessness — are having their history extended, because current county-level figures for them do not exist anywhere else.
Where does Life Expectancy come from?
This one is the exception, and the difference matters. Life expectancy (and premature death) is calculated here by the author from CDC WONDER mortality files — a 5-year pooled figure matched to the ACS 5-year window. It is not an official CDC-published statistic, and you will not find these exact numbers on CDC's site: CDC WONDER is the underlying mortality data and should be credited as such, but the life-expectancy calculation itself is original work. Few sources publish county-level life expectancy, and none have the current numbers — which is why it is computed here.
Where do the county homeless numbers come from?
Also derived, not published. HUD counts homelessness through its annual Point-in-Time count but reports it by Continuum of Care (CoC), not by county — and CoC boundaries do not line up with counties (some cover a city and a county together, many rural "Balance of State" CoCs span dozens of counties, and HUD publishes no official county crosswalk). The county figures here rely on a CoC-to-county crosswalk built by the author, assembled from HUD geography, state service-area maps, planning-region maps, an older published crosswalk, and manual corrections. Confidence varies by county, and a few counties split across more than one CoC — treat these as careful approximations, not official HUD statistics. Every CoC, listed by state, is on the HUD CoCs page.
Can I download the data?
Yes, and it's free. The reports, charts, maps, and analyses throughout the site have a download button that exports the data you are looking at — and, where it makes sense, the full measure list behind it. Everything here is built from public, open government data, so you are welcome to download and use it for non-commercial purposes. Please credit the original source (listed on the Sources & Dashboards page), and check that source for official figures.
Who maintains this site?
Generally one person, working part-time. New data is loaded within roughly ten days of release, and the sources are reviewed about once a month. The roadmap below is a set of intentions, not dated commitments — things land when the numbers are confident, not before.
What is being worked on?
A few things, in rough priority: Life expectancy — full period life tables giving county life expectancy at birth in 5-year spans (the largest piece in the queue; the crude death-rate series 1979–2024 is already up on the Life Expectancy page as a first step). Completing the HUD CoC-to-county crosswalk by confirming assignments directly with the CoCs, then publishing it as a downloadable file that does not exist anywhere else. Higher-frequency data — bringing in more current releases (monthly BLS county unemployment, provisional mortality) with the fresher-but-provisional trade-off made explicit. More sources under review — candidates like nonfatal injury (NEISS-AIP), education (NAEP), and childhood immunization (NIS), each judged on whether it reaches county level and fills a real gap. A ready-to-use county data section for CHA/CHIP reports, with sourced, benchmarked figures that drop straight into a Community Health Assessment. And short training videos on the concepts behind the data, once the site settles.
Is this an official government website?
No. This is an independent, non-commercial site that uses public data only. Always check the original source for official figures.
Found an error, or want to get in touch?
Please say so. Corrections from people who work with this data locally — a bad number, a broken link, a source that has moved, a CoC assignment that does not match what you know on the ground — are the most valuable feedback there is, and they are acted on. [enable JavaScript to see the address]