About the data

Sources

• Election results by ward: Wisconsin Legislative Technology Services Bureau (LTSB), which collects results from the Wisconsin Elections Commission and splits each reporting unit's votes among its wards by population.
• Official totals and candidate names: the Wisconsin Elections Commission canvass via OpenElections; MIT Election Data + Science Lab.
• National generic ballot: the U.S. House popular vote, from the FEC's Federal Elections reports (2002–2022) and the Clerk of the House (2024).
• Population: U.S. Census Bureau 2010 and 2020 Census, annual population estimates, and the American Community Survey's citizen voting-age population tables.
• District maps: LTSB (2002, 2011, 2022 and 2024 plans).

One map for every year

Ward lines change almost every year. To compare places over time, every election is shown on today's (2025) ward boundaries: older results are carried onto current wards in proportion to where people live, using 2020 Census blocks. Tested against a case with a known answer, about 0.5% of votes end up in a different ward, and totals for counties and the state are unchanged. Ward-level numbers are estimates; county and statewide numbers match official results closely.

District views and district pages add up wards inside each district. Statewide races are shown on one district map across all years, so you can see how today's (or an older) district voted in every election. A district's own races are official results, shown only for years its map was in force. Candidate names for 2004 and 2024 legislative and U.S. House races aren't in the sources yet, so those rows say “Democratic candidate” and so on.

Measures

Margin

Democratic minus Republican share of the two-party vote. D+5 means the Democrat won 52.5% to 47.5%.

Lean vs. state

A place's margin minus Wisconsin's in the same race. D+10 lean = ten points more Democratic than the state.

Shift

The change in a place's margin between two elections.

Shift vs. state / vs. nation

The shift minus Wisconsin's shift, or minus the national generic ballot's shift. A ward that moved D+10 while the state moved D+8 and the nation D+6 is +2 vs. state and +4 vs. nation.

Trend score

The average shift per election cycle over 2010–2024 beyond Wisconsin's own swing and beyond what population change (growth and changing racial makeup) predicts. "Trending blue" or "red" needs at least 1.5 points per cycle and 5 of the last 6 election pairs pointing the same way.

Turnout

Votes in the race divided by the estimated voting-age population that year. A few small places exceed 100% where population estimates lag behind growth.

Beyond demographics

Each place's shift relative to the state is compared with its demographics from the Census Bureau's American Community Survey (2019–23 and the change since 2008–12): college degrees, household income, age, race and ethnicity, home ownership and density, plus how far it leaned at the start. A weighted regression across places gives the shift places like it made; "beyond demographics" is the rest. The map measure averages it over 2012→16, 2016→20 and 2020→24, per cycle. Survey estimates for small towns have wide margins of error, so single small places matter less than patterns. Details on the Trends page and in the project's analysis notes.

Turnout (registered)

Votes divided by registered voters, including Election Day registrations, as each municipality reported them to the U.S. Election Assistance Commission's Election Administration and Voting Survey (2016–2024). Wards and districts (2022 on) and the county parts of municipalities that span counties are split using the Wisconsin Elections Commission's ward lists, or by adult population where a place isn't on them.

Community type

Municipalities by 2020 population density: urban (2,500+ per square mile), suburban (750–2,500), small town & exurban (100–750), rural (under 100).

Efficiency gap

Wasted votes (all votes for losers, plus winners' votes beyond 50%) for one party minus the other, as a share of all votes. Uncontested seats are counted as 75–25.

The projection tool

Wisconsin's margin is set to the national generic ballot plus Wisconsin's offset from the nation (by default the 2016–2024 average). Each ward keeps its recent lean (averaged over 2022 and 2024, the method that predicted past elections best), and turnout re-weights the wards. It's a baseline for "what would it take?" questions, not a forecast: it knows nothing about candidates, and the national environment itself is the biggest unknown.

Known limits

• 2004 results run about 0.5% short of official statewide totals (gaps in the source data).
• 2002–2010 legislative district totals from ward data are approximate; district tables use official counts where available.
• The 2004 State Senate data is unreliable and is left out of the legislative analysis.
• Candidate names for 2024 legislative races outside Waukesha County are not yet available.
• Midterm comparisons with the nation set the Governor's race against the national House vote, so strong candidates show up as Wisconsin "outperforming".