All posts

Where US county poverty is most concentrated

The county-level picture of poverty in the United States is far from even. Across 3,144 counties on the US mainland, the median county poverty rate sits at 13.2%. The highest-poverty counties run two to four times that rate, and the geography is unusually concentrated.

The household figures here come from the American Community Survey, the Census Bureau's annual household data release that covers every county in the country. We pulled the county-level poverty rate (population 5,000 and above) and grouped the top thirty.

Four regions hold most of the top thirty

Of the thirty US mainland counties with the highest poverty rates, 25 sit inside one of four named regions:

  • Tribal lands on the Northern Plains. Oglala Lakota County, South Dakota, Todd County, South Dakota, and Dewey County, South Dakota.
  • The South Texas border. Dimmit County, Texas, Zapata County, Texas, Presidio County, Texas, Zavala County, Texas, and Starr County, Texas.
  • Central Appalachia. Wolfe County, Kentucky, Knox County, Kentucky, McCreary County, Kentucky, Cumberland County, Kentucky, Magoffin County, Kentucky, and Calhoun County, West Virginia.
  • The Mississippi Delta. Holmes County, Mississippi, Madison Parish, Louisiana, Coahoma County, Mississippi, East Carroll Parish, Louisiana, Claiborne Parish, Louisiana, and Concordia Parish, Louisiana.

These are not statistical noise. They are places with deep, persistent, multi-generational poverty, often a long way from a major metro. The same counties have appeared on poverty rankings for as long as county-level statistics have been published, and the story is more about structural geography (isolated reservations, an extractive coal economy, plantation-era agriculture, a remote southern border) than about year-to-year economic cycles.

Just how far above the national median these counties sit

The mainland US county median for household poverty rate is 13.2%. The highest-poverty counties run far above that:

  1. Oglala Lakota County, South Dakota: 52.8% poverty rate, median household income $34,769.
  2. Dimmit County, Texas: 44.8% poverty rate, median household income $33,409.
  3. Wolfe County, Kentucky: 38.1% poverty rate, median household income $29,052.
  4. Holmes County, Mississippi: 38.0% poverty rate, median household income $29,434.
  5. Madison Parish, Louisiana: 36.7% poverty rate, median household income $37,267.
  6. Zapata County, Texas: 37.4% poverty rate, median household income $36,527.

Income context

Poverty rate is one window. Median household income widens it. The national median across mainland counties is $63,690. The counties at the top of each cluster sit roughly half that level:

  • Oglala Lakota County, South Dakota: $34,769.
  • Dimmit County, Texas: $33,409.
  • Wolfe County, Kentucky: $29,052.
  • Holmes County, Mississippi: $29,434.

The compound effect is what separates these counties from a typical low-income suburb. The same households face higher poverty rates and lower median earnings together, in places where the local labour market has been thin for decades.

What this means for a campaign brief

If a national campaign assumes a US household at around the national median, that brief leaves these counties out by construction. The household it pictures is not common in any of these regions at scale. A national average sits between two very different populations, and a campaign written to the average is, in practice, a campaign written for the suburbs.

The reverse case is just as common. A campaign that explicitly targets lower-income households often ends up planning around state-level averages, which paper over how concentrated the lowest incomes actually are. The four regions in this post hold a disproportionate share of the country's lowest-income households, and they reach those households through very different media, retail, and trust channels.

Public data like this is the best way to check a brief against the country it is meant to reach. Cambium AI builds synthetic populations from these same public data files, so a marketing or research team can see where a brief holds and where it does not, county by county, before the budget is set. The four regions in this post are a sensible place to start.

Data source: U.S. Census Bureau, American Community Survey 2023 (ACS) 5-Year Estimates

Ask AI about this post:
ChatGPT