A Small Town's Census Figure Is Not From the 1-Year Survey — It Cannot Be, Because 1-Year Estimates Stop at 65,000 People
The Census Bureau's own guidance limits ACS 1-year estimates to "areas with populations of 65,000+", and a 5-year estimate is 60 months of collected data, not a single year. As of 2026-09-25.
There is a population floor, and it is written down
A town of 30,000 has no ACS 1-year estimate. Not a noisy one — none. The Census Bureau's guidance lists the 1-year product as "Data for areas with populations of 65,000+".[1]
So a figure quoted for a small place came from the 5-year product, and the Bureau says as much: the 5-year estimates are for "Examining tracts and other smaller geographies because 1-year estimates are not available".[1]
"2020-2024" is not 2024
A 5-year estimate is a period, not a year. The page defines it as "60 months of collected data" and gives "2020-2024 ACS 5-year estimates" as the example. The 1-year product is "12 months of collected data".[1]
That is the trade the Bureau states outright. The 5-year product is for when "Precision is more important than currency". A more reliable number is also an older one.[1]
Two comparisons the Bureau tells you not to make
First: "Only compare 1-year estimates with 1-year estimates, and 5-year estimates with 5-year estimates." Mixing the two products is ruled out, not merely discouraged.[2]
Second: "Do not compare overlapping datasets" — with the Bureau's own example, "do not compare 2005-2009 ACS 5-year estimates to 2006-2010 ACS 5-year estimates". Consecutive 5-year releases share four years of data, so the difference between them is not a change over one year.[2]
There is also a year to set aside: "Data users should not compare 2020 ACS 1-year experimental estimates" with other estimates. For 2020 the Bureau released experimental estimates instead of the standard 1-year products.[2]
If the question is "how many people", this is the wrong survey
The Bureau puts it plainly: "The strength of the American Community Survey (ACS) is in estimating characteristic distributions. If you are looking for population totals, we recommend the 2020 Census or Population Estimates Program."[2]
And one product no longer exists going forward — "ACS 3-year estimates have been discontinued", with "no new 3-year estimates will be produced", though the datasets already published remain available. A citation to a 3-year estimate is therefore a citation to a closed series.[1]
What this page has not established
This page rests on one institution — the Census Bureau's own guidance pages — with no independent cross-check. It reports the rules the Bureau publishes, not how well the estimates perform.[1][2]
Not established: the margin of error for any particular estimate, what the 1-year supplemental product covers, and which specific tables changed between vintages. The comparison pages route those questions to year-by-year guidance this page did not open. As of 2026-09-25.[2]
Verified facts
Cross-checked against 2+ independent sourcesThis section contains facts cross-checked against multiple sources.
The U.S. Census Bureau's guidance page on choosing American Community Survey datasets states that 1-year estimates provide "Data for areas with populations of 65,000+", and that a 5-year estimate is built from "60 months of collected data", giving "2020-2024 ACS 5-year estimates" as its example.[1] 1 sources
Reported, not confirmed
Not cross-checked — do not read as factFrom here on: claims and speculation that are not cross-checked.
The page describes the 5-year product as the choice when "Precision is more important than currency" and for "Examining tracts and other smaller geographies because 1-year estimates are not available".[1] single-source ×1 · U.S. Census Bureau, Using 1-Year or 5-Year ACS Data
A footnote states that "ACS 3-year estimates have been discontinued" and that "no new 3-year estimates will be produced", while the previously published 3-year datasets remain available.[1] single-source ×1 · U.S. Census Bureau, Using 1-Year or 5-Year ACS Data
"Only compare 1-year estimates with 1-year estimates, and 5-year estimates with 5-year estimates."[2] single-source ×1 · U.S. Census Bureau, Comparing ACS Data
"Do not compare overlapping datasets" — the page's own example is "do not compare 2005-2009 ACS 5-year estimates to 2006-2010 ACS 5-year estimates".[2] single-source ×1 · U.S. Census Bureau, Comparing ACS Data
"The strength of the American Community Survey (ACS) is in estimating characteristic distributions. If you are looking for population totals, we recommend the 2020 Census or Population Estimates Program."[2] single-source ×1 · U.S. Census Bureau, Comparing ACS Data
"Data users should not compare 2020 ACS 1-year experimental estimates" with other estimates; the Bureau released experimental estimates instead of the standard 1-year products for 2020.[2] single-source ×1 · U.S. Census Bureau, Comparing ACS Data
Timeline
- Model
- claude-opus-5
- Time
- 09/25/2026, 05:14
- Body characters
- 2,394
- Sources
- 2 sources adopted
- Model
- claude-opus-5 (both Census Bureau guidance pages opened and searched for each quoted sentence before it was used; the generator aborts on any miss)
- Time
- 09/25/2026, 05:14
- Verdict
- Passed
Show revision history (2)
| 09/25/2026, 09:00 | First authored (claude-opus-5) | Created |
| 09/25/2026 | First version. Both guidance pages were opened and every quoted sentence searched for in the live page before the file was written; the generator aborts if any quote is absent. | Updated |
Frequently asked
Why is there no 1-year ACS estimate for my town?
The Census Bureau's guidance lists 1-year estimates as covering "areas with populations of 65,000+". Smaller geographies are covered by the 5-year estimates.[1]
Does a 2020-2024 5-year estimate describe 2024?
No. The Bureau defines the 5-year product as "60 months of collected data" and uses "2020-2024 ACS 5-year estimates" as its example of that.[1]
Can I compare a 1-year estimate to a 5-year estimate?
The Bureau says to "Only compare 1-year estimates with 1-year estimates, and 5-year estimates with 5-year estimates."[2]
Can I compare 2005-2009 with 2006-2010?
No. "Do not compare overlapping datasets" — that pair is the Bureau's own example of what not to do.[2]
Should I use ACS for population totals?
The Bureau recommends the 2020 Census or the Population Estimates Program for totals, and says the ACS's strength is estimating characteristic distributions.[2]
What happened to the 3-year estimates?
They have been discontinued and no new ones will be produced, though the previously published datasets remain available.[1]
What about 2020?
The Bureau states that data users should not compare 2020 ACS 1-year experimental estimates with other estimates; experimental estimates were released instead of the standard 1-year products.[2]
Official links
Sources
- [1] Using 1-Year or 5-Year American Community Survey Data primary
- [2] Comparing ACS Data primary