How proportional representation would have changed the U.S. House, 2016–2024
For each recent U.S. House election, this takes the actual statewide vote and asks what the seat split would have been if every state allocated its seats proportionally (using the Sainte-Laguë method), with no polling or projection involved. Comparing that to the seats each party actually won isolates the distortion built into the winner-take-all district map.
| Cycle | Actual result | Under PR | Shift under PR |
|---|---|---|---|
| 2016 | D 194 / R 241 | D 214 / R 221 | +20 toward Democrats |
| 2018 | D 235 / R 200 | D 233 / R 202 | +2 toward Republicans |
| 2020 | D 222 / R 213 | D 223 / R 212 | +1 toward Democrats |
| 2022 | D 213 / R 222 | D 213 / R 222 | even |
| 2024 | D 215 / R 220 | D 213 / R 222 | +2 toward Republicans |
Across the last five cycles the gap between the proportional result and the actual one is usually small — about two seats — with one striking exception: 2016.
Why 2016 is the outlier
In 2016, Democrats won about 49% of the House vote nationwide but only 45% of the seats. Proportional representation would have closed almost all of that gap, adding roughly 20 Democratic seats. And it wasn’t one rogue state. A dozen states with Republican-drawn maps (Texas, Pennsylvania, Ohio, North Carolina, Michigan, Georgia, and more) each turned a near-even statewide vote into a lopsided delegation. Those were the most aggressive maps of the decade, drawn after the 2010 census and not yet thrown out by courts.
It also reflects something the map keeps showing: Democratic voters cluster in cities, so the party piles up big margins in a few districts and “wastes” a lot of votes. By 2018 Pennsylvania’s map had been struck down and redrawn, and a strong Democratic year turned votes into seats more efficiently. The maps behind 2022 and 2024 are close to nationally balanced: independent-commission and court-ordered redraws (Michigan, North Carolina) landed alongside offsetting new gerrymanders — Democratic in Illinois, Republican in Florida and Texas — so they roughly cancel out. That’s why proportional representation changes far fewer seats now — the small gaps you see in the other cycles.
Explore the retrospectives on the interactive map →
Prior-cycle vote totals come from the MIT Election Lab (U.S. House 1976–2024); 2024 from the U.S. House Clerk. See the methodology, the most distorted delegations, or the national map. Data is CC BY 4.0; cite as “The Proportional House (proportionalhouse.org), accessed [date].”