Where the Stranded Live

The same map, seen from opposite sides

The pair of maps below shows, county by county, what share of each party's local members are stranded — living in a county their party lost in that cycle's presidential vote. Each map covers the 579 counties with at least 50 pooled CES partisan respondents; smaller cells are suppressed and drawn in beige. Because strandedness is decided by the county result, most counties sit at one extreme or the other; the in-between shades are counties that flipped between 2020 and 2024, leaving the same residents stranded in one cycle but not the other.

Read the Democratic map and your eye goes to the South and the interior: a broad scatter of Sun Belt suburbs, mid-size metros, and small cities where Democrats live in numbers but lose anyway. Read the Republican map and the stranded population snaps to the coasts and the big-metro cores — Los Angeles, the Bay Area, New York, Chicago, Seattle, Denver, Washington. These are not mirror images of one another. They are different geographies of political exile.

A Southern story and a big-city story

Put numbers on the asymmetry. Among all Democrats, 36.4% live in the census South. Among stranded Democrats, it is 47.6% — nearly half. The West runs the other way: it houses 23.9% of all Democrats but just 13.3% of the stranded ones, because a Western Democrat usually lives in a county the party carries. The single biggest concentrations of outvoted Democrats are Florida (9.6% of all stranded Democrats), Texas (8.2%), Pennsylvania (6.3%), and Ohio (5.3%) — states where Democrats are numerous but blue counties are few.

Stranded Republicans are the inverted image. California alone houses 12.6% of them — more than any other state — and California, New York, and Illinois together hold 24.0%. Even Texas (9.7%) and Florida (8.1%) contribute their own blue-metro Republicans, conservatives living in Houston, Dallas, Austin, Miami, and Orlando counties that their party lost. The stranded Republican is, overwhelmingly, a city or inner-suburb dweller in a state or metro the party has written off.

2024 flipped who lives behind enemy lines

For four straight presidential cycles, Republicans were clearly the more stranded party. In the 2008 cycle, 50.5% of Republicans lived in a county Obama carried, against 31.9% of Democrats in McCain counties. The gap stayed Republican-heavy through 2012 (44.6% vs 34.5%) and 2016 (39.8% vs 35.8%), then blew out in 2020: 44.6% of Republicans stranded versus just 28.4% of Democrats — a 16.2-point gap, the deepest Democratic insulation in the series. The median Democrat that cycle lived in a county Biden carried by about 15.2 points.

Then 2024. For the first time since the 2004 (Bush) cycle, the gap inverted: 40.3% of Democrats now live in a county Trump won, versus 33.5% of Republicans in Harris counties. The result is robust to counting partisan leaners (41.3% vs 35.1%), and the median Democrat's county cushion thinned from D+15.2 to D+9.1.

Nobody had to move for this to happen. Between 2020 and 2024, 86 counties flipped from Democratic to Republican and exactly zero flipped the other way. Decompose 2024's stranded Democrats against the 2020 map: 76.1% of them live in counties that were already red in 2020 — the long-standing Southern and small-metro stranded population — but 23.9% were stranded by a county that flipped under their feet. That second group is 9.6% of all Democrats: roughly one in ten woke up newly outvoted at home without changing their address, in places like Miami-Dade and a belt of Sun Belt and border counties. The same tide un-stranded 9.0% of Republicans.

This is the load-bearing caveat of the whole exercise: strandedness is tide-sensitive. A red-shifting map raises Democratic strandedness mechanically, with zero migration. Three-quarters of the 2024 stranded-Democrat population is structural; one quarter is the 2024 tide itself. Whether the inversion persists depends on whether those 86 flips do.

More landslide counties than ever — barely more landslide people

Strandedness is the flip side of the Big Sort, Bill Bishop's famous observation that Americans increasingly live in landslide counties. The full county record bears him out, with a twist. In 1976, 1,150 of 3,104 counties (37.0%) were decided by twenty-plus two-party points, and they held 26.4% of the two-party vote. By 2020, landslide counties held 58.1% of the vote; in 2024 the count of landslide counties hit an all-time record — 2,493 of 3,113, or 80.1% — while the share of votes cast in them slipped slightly, to 57.2%. Four-fifths of counties are now landslides; they contain a bit more than half the voters.

The person-level CES series tells the same story: the share of American adults living in a landslide county climbed from 46.4% in the 2008 cycle to 58.0% in 2020, and eased only to 57.4% in 2024 — a dip of under one point on the vote-weighted series, not a reversal. The sorting plateaued at a high altitude. What changed in 2024 is the partisan composition: blue-metro margins compressed while rural red margins deepened, which is exactly the mechanism that stranded those new Democrats.