How to Read a County Unemployment Rate (and Why Last Month Is the Wrong Comparison)

Harris County, TX had an unemployment rate of 5.0% in August 2026, a preliminary figure. That one number comes from two published counts and one division, and reading it well means knowing what the counts measure, why the month before it is the wrong comparison, and how much the figure can move for reasons that have nothing to do with the local economy.

What the rate is made of

The U.S. Bureau of Labor Statistics' Local Area Unemployment Statistics (LAUS) program publishes four figures for every county each month: the number of people employed, the number unemployed, the two added together as the civilian labor force, and the unemployment rate. The rate is the unemployed count divided by the labor force, times 100, rounded to one decimal. For Harris County in August 2026:

126,999 unemployed ÷ 2,548,851 in the labor force × 100 = 4.98%, which BLS publishes as 5.0%.

“Unemployed” has a precise meaning. A person is counted only if they had no job during the reference week, were available to work, and had actively looked for work in the previous four weeks. Someone who has given up looking, is in school full time, has retired, or is caring for family is not unemployed and not in the labor force either, so they appear in neither count. A county's rate can fall because people found work or because people stopped looking, and the rate alone does not say which. The labor force count beside it does: if it shrank while the rate fell, read the fall with care.

Why last month is the wrong comparison

BLS seasonally adjusts the national rate and the state rates, removing the regular pattern that repeats every year: school staff off in summer, farm work in harvest months, retail hiring before the holidays, construction slowing in winter. It does not seasonally adjust county figures. So a county's published rate carries the whole calendar in it. Over the 13 months the Harris County page shows, the rate ran from 4.3% in December 2025 to 5.2% in June 2026, and much of that range is season, not change.

The same month a year earlier has the same season in it. Harris County was 5.0% in August 2025 and 5.0% in August 2026, so the part that actually changed over the year is 0.0 points. That is the comparison every county page on this site makes, and the reason none of them prints a month-over-month change: the number would be real, and it would mislead.

The state and the nation sit beside the county on the same basis. Texas, not seasonally adjusted, was 4.6% in August 2026; the United States, from the Current Population Survey and also not seasonally adjusted, was 4.3%. The national figure you hear in the news is the seasonally adjusted one and will differ by a few tenths; comparing a county with it would mix two bases.

Preliminary means it will change

The newest month BLS publishes carries a preliminary flag. When the following month is released, the preliminary figure is revised, usually by a tenth or two of a point, and the flag moves to the new month. Every spring BLS also re-estimates the previous years for every area, which can move the year-earlier figure on a page. A county page here marks the preliminary month in words, and the snapshot behind it is refreshed each month so the revision replaces the preliminary figure rather than sitting beside it.

Where county figures come from

Nobody surveys each county's households every month. BLS estimates each state with a statistical model that combines three inputs: the Current Population Survey of households, the Current Employment Statistics survey of employers, and counts of unemployment insurance claims. County figures are built from the same inputs and constrained so the counties add up to the state and the states to the nation. They are the official statistics for the county, and they are estimates, with error BLS does not publish at the county level.

Size matters more than anything else in reading them. In a county with 4,000 people in the labor force, 40 people changing status moves the rate by a full point; a plant closing or a seasonal workforce arriving shows up as a swing that would be invisible in a large county. That is why each page prints the counts, and why this site publishes no list of the highest- and lowest-unemployment counties: an order over 3,000 estimates with different error and different seasonal patterns would mostly be ordering noise.

One more thing the figures count: residents, not jobs. LAUS measures where people live. A county whose residents commute to work elsewhere shows their employment, not the health of its own employers.

Reading a page in order

Start with the rate and the same month a year earlier, together. Then the labor force count, to see how big the base is and whether it moved. Then the state and the nation on the same basis, for scale. The 13-month series shows the season the county lives in. The county unemployment rate pages on this site are laid out in that order, with the division shown so the rate can be checked by hand, and with the month BLS left unpublished during the 2025 lapse in appropriations marked rather than filled in.

Sources

U.S. Bureau of Labor Statistics, Local Area Unemployment Statistics, county and state series, not seasonally adjusted, read from the public flat files on 2026-10-08; figures for Harris County, TX (FIPS 48201) and Texas, August 2026 (preliminary) and August 2025. U.S. Bureau of Labor Statistics, Current Population Survey, national not-seasonally-adjusted unemployment rate, series LNU04000000. Informational only, not economic or career advice.

Informational only, not professional or financial advice.