American television is carved into 210 Designated Market Areas — Nielsen's county-by-county map of who watches which local stations. Nobody drew them around states, districts, or voters. Yet they decide whose news you see and where campaigns can buy an ad. Half the country lives in a market that straddles a state line — and half of all congressional districts are split across two or more of them.
Markets were drawn by signal reach and viewing habit, not by any political authority. Four consequences fall out of the geography itself.
A market takes its name from the city its stations broadcast from. For millions of people that city sits in a different state — so the local news covers another state's governor, another state's legislature, and campaigns must buy ads across a border to reach them.
| State | Share in an out-of-state market | People | Anchored in |
|---|
Buy these and you have reached half the country. Click a row to pin it on the map.
| # | Market | Population | Counties | States | 2024 margin |
|---|
A campaign cannot buy a congressional district. It buys a media market, and pays for everyone in it. Turning the 2026 district lines against the market map gives the exchange rate. The New York market is in a class of its own — it holds the 25 costliest districts in the country — so the table below shows at most three districts per market to make the rest visible.
Three cycles, one unmoved constraint
Between 2022 and 2026 the House was redrawn twice and 181 districts changed shape. The number split across two or more markets went from 209 to 213. Mapmakers optimize for partisanship; the media map is simply not a constraint they are working against — though underneath the flat total, — individual districts did change which markets they depend on.
Costliest districts to advertise in
| District | 2024 | Main market | You pay for |
|---|
Districts split across the most markets
| District | Markets | Largest share | Main market |
|---|
Boundaries. Nielsen DMA polygons (206 lower-48 markets; Alaska's three and Honolulu are not in this file, so the national count of 210 shows here as 206). Nielsen builds every market out of whole counties, so the borders you see are county borders.
The crosswalk is derived, not downloaded. No public county→DMA table was available, so each of the 3,107 lower-48 counties was matched to a market by point-in-polygon on its representative point. Every lower-48 county matched.
Population is the Census Bureau's 2024 county estimate (Vintage 2024). Connecticut replaced counties with planning regions in 2022, so its eight legacy counties carry 2020 estimates scaled to the state's 2024 total. Votes are 2024 county-level presidential returns.
Congressional districts are the enacted lines for 2022, 2024 and 2026 (432 of 435 each cycle; Alaska and Hawaii sit outside the market file), the 2026 set including the mid-decade redraws in ten states. Use the year buttons beside the Districts toggle to switch cycles; every district figure on this page follows that choice. To split each district's population across markets, county populations and district populations were reconciled by iterative proportional fitting over district×county area overlaps — so both totals are respected rather than area alone, which badly misreads dense urban districts. Shares, not absolute counts, drive every figure here.
One gap. Nielsen splits Riverside County, CA between Los Angeles and Palm Springs. A whole-county method cannot, so Riverside lands in Los Angeles and Palm Springs is shown as no-data. It is the only market affected.