November 8, 2026 · By FutureBallot
Turnout, partisan reactions and policy concerns help explain the electorate. A separate calculation asks how a different House majority could have emerged from the district returns.
Democrats won 235 House seats in 2018; Republicans won 199, with one seat unfilled. A majority of the 435-seat House requires 218. The official election returns establish the result. Surveys help explain the people behind it, without assigning a single cause to every vote.
What mattered to voters?
In KFF’s May 2018 survey, the economy and jobs, health care, and gun policy led the topics voters wanted candidates to discuss. After the election, Pew’s open-ended question about vote choice found that support for one’s own side and opposition to the other side appeared more often than particular policies. Opposition to Trump featured in Democratic voters’ answers. Those findings address different questions; neither proves health care was irrelevant.
The Census Bureau estimated that 53% of voting-age citizens voted, the highest midterm rate in four decades in that survey. This is self-reported participation among citizens, not the percentage of all residents who voted.
Pew’s later study matched respondents to turnout records. Among 2018 voters who had not voted in the preceding presidential election, 68% reported a Democratic House vote and 29% a Republican vote. Records helped verify participation; candidate choice still came from respondents. The finding shows why participation belongs in the explanation alongside changing preferences.
How close was control under a vote-switch model?
In Utah’s 4th district, Ben McAdams received 134,964 votes and Mia Love 134,270: a gap of 694. Moving one vote directly from McAdams to Love reduces that gap by two. A switch of 347 produces a tie at 134,617 each; 348 puts Love ahead, 134,618 to 134,616.
Republicans would need nineteen additional seats to reach 218. Among 199 eligible Democratic-won districts with both a Democratic and Republican candidate in the archived House Clerk returns, the nineteen lowest switch requirements sum to 59,699. The table shows every selected district and names the Democratic winner and Republican opponent for each one.
| District | Winner | Opponent | Winner votes | Opponent votes | Switches | Running total |
|---|---|---|---|---|---|---|
| UT-4 | Ben McAdams | Mia B. Love | 134,964 | 134,270 | 348 | 348 |
| CA-21 | TJ Cox | David G. Valadao | 57,239 | 56,377 | 432 | 780 |
| GA-6 | Lucy McBath | Karen C. Handel | 160,139 | 156,875 | 1,633 | 2,413 |
| OK-5 | Kendra Horn | Steve Russell | 121,149 | 117,811 | 1,670 | 4,083 |
| ME-2 | Jared F. Golden | Bruce Poliquin | 142,440 | 138,931 | 1,755 | 5,838 |
| NM-2 | Xochitl Torres Small | Yvette Herrell | 101,489 | 97,767 | 1,862 | 7,700 |
| NJ-3 | Andy Kim | Thomas MacArthur | 153,473 | 149,500 | 1,987 | 9,687 |
| SC-1 | Joe Cunningham | Katie Arrington | 145,455 | 141,473 | 1,992 | 11,679 |
| FL-26 | Debbie Mucarsel-Powell | Carlos Curbelo | 119,797 | 115,678 | 2,060 | 13,739 |
| NY-22 | Anthony J. Brindisi | Claudia Tenney | 127,715 (four party lines) | 123,242 (three party lines) | 2,237 | 15,976 |
| VA-2 | Elaine G. Luria | Scott Taylor | 139,571 | 133,458 | 3,057 | 19,033 |
| VA-7 | Abigail A. Spanberger | Dave Brat | 176,079 | 169,295 | 3,393 | 22,426 |
| CA-39 | Gil Cisneros | Young Kim | 126,002 | 118,391 | 3,806 | 26,232 |
| IA-3 | Cindy Axne | David Young | 175,642 | 167,933 | 3,855 | 30,087 |
| CA-10 | Josh Harder | Jeff Denham | 115,945 | 105,955 | 4,996 | 35,083 |
| MN-7 | Collin C. Peterson | Dave Hughes | 146,672 | 134,668 | 6,003 | 41,086 |
| TX-7 | Lizzie Pannill Fletcher | John Abney Culberson | 127,959 | 115,642 | 6,159 | 47,245 |
| NY-11 | Max N. Rose | Daniel M. Donovan, Jr. | 101,823 (three party lines) | 89,441 (four party lines) | 6,192 | 53,437 |
| CA-45 | Katie Porter | Mimi Walters | 158,906 | 146,383 | 6,262 | 59,699 |
What the number means
For each eligible district, switches equal the whole-number part of half the winner–opponent gap, plus one. We sort those requirements and sum the lowest nineteen. Each selected opponent must also finish ahead of every other candidate and of official-return categories such as blank votes, scattering, void and write-in totals. Other districts, turnout and the unfilled seat stay fixed.
This is a minimum within the vote-switch model, not an estimate of persuadable people. Maine’s 2nd district uses the Clerk’s reported final-round totals from its ranked-choice election. We treat switches there as transfers between the two finalists; we do not reconstruct ballot rankings or replay earlier eliminations. Additional voters, different turnout or a uniform national swing would require different calculations.
The survey findings describe the electorate that produced the result. The arithmetic measures one precisely defined way to change control. For another election with a different political setting, see 2010: One election, different kinds of power.
Get an email when FutureBallot publishes a new explainer or video, or follow the Analysis page.