← Back to simulator

Election Environment

Drag right for a Republican wave, left for a Democratic wave.

District Map

Share of districts that are packed safe-D seats. Slide right to gerrymander more aggressively in Democrats' favour.
Share of districts that are packed safe-R seats. Slide right to gerrymander more aggressively in Republicans' favour.

Candidates

Higher → Democrats run more moderate candidates across all districts.
Higher → Republicans run more moderate candidates across all districts.
Higher → Democrats run more moderate candidates in swing districts.
Higher → Republicans run more moderate candidates in swing districts.
Higher → Democrats run more moderate candidates in red districts.
Higher → Republicans run more moderate candidates in blue districts.

Voters

Higher → voters punish ideologically extreme candidates more heavily, relative to how partisan the district is.
Lower → election outcomes track district partisan lean and candidate ideology more deterministically. Higher → more random per-district swings.

Historical House Seats and Election Margins, 1992–2024

Pick a US House election cycle from the thumbnails below and three comparison rows appear at the top: the cycle's actual result (top), the same sliders run on that cycle's real districts (middle), and the same sliders run on the simulator's synthetic chamber (bottom). Cell colour = party of victory + margin of victory. The lean axis is a Cook-style PVI — a weighted average of recent presidential margins minus the same weighted national popular vote. Presidential years use a 4-to-1 weighting of the current pres election and the prior one (current × 4 + prior × 1); midterm years use a 2-to-1 weighting of the two prior pres elections (more recent × 2 + the one before × 1).

Pick a cycle to compare

Loading data…

Sources. House results 1998–2024: FiveThirtyEight election-results. House results 1992–1996: MIT Election Data + Science Lab (1976–2018 constituency returns). Presidential margin 2008–2024: The Downballot / Daily Kos Elections per-CD calculations (2024 in 2024-cycle districts; 2008/2012/2016/2020 in 2020-cycle districts; 2000/2004/2008 in 2002-cycle districts). Presidential margin 1992/1996/2000 in 1992-cycle districts: Atlas of U.S. Presidential Election Results by Congressional District (Kiernan Park-Egan / WWU, 2022), election data from the Almanac of American Politics (Barone & Ujifusa) and Dave Leip's Atlas.

Coverage. Every cycle matches the full 435 D/R seats against a partisan-lean source. Louisiana's federal elections (jungle primary + occasional December runoff) and the two court-ordered Texas mid-decade redraws (TX 1996, TX 2006) are absorbed by falling back from general to runoff to jungle primary on a per-district basis. Fusion-ticket candidates (mostly New York Republicans cross-endorsed on Conservative / Independence lines, and Democrats on Working Families) are tallied per-candidate so all of one person's ballot lines count together. Three Independents who held a seat and consistently caucused with a party are coloured by their caucus: Bernie Sanders (VT-AL, 1992–2006) → D, Jo Ann Emerson (MO-08, 1996) → R, Virgil Goode (VA-05, 2000) → R.

Caveats. 1994's PVI bin is single-cycle (just 1992) — 1988 was held in 1980s district lines and can't be projected cleanly into 1992-cycle boundaries. 2002's PVI is single-cycle for the same reason (1996 lives in 1992-cycle lines, 2000 in 2002-cycle). For 2008–2016 the join between House (contemporary districts) and presidential margin (2020-cycle districts) is approximate — districts redrawn after the 2010 census don't match exactly, so a few seats per cycle inherit a slightly-different neighbour's lean rather than their own.