Our most recent book is The Comeback: The 2024 Elections and American Politics. It includes a chapter on congressional and state elections.
Over the past decade, polling problems—especially the plunge in response rates—have made election forecasting harder. So instead of a single prediction, here are three plausible scenarios: rare, medium, and well-done.
Rare. In this scenario, the polls have systematically underestimated the GOP vote, and the Democratic advantage in early and mail ballots abruptly vanishes as Republicans surge on Election Day.
- House: Democrats pick up about a dozen seats, enough to win a majority but below expectations.
- Senate: Democrats lose Michigan and win North Carolina, for no net change. Republicans retain control of the chamber.
Medium. The polls and early-vote estimates are roughly on track, and 2026 is a good Democratic year.
- House: Democrats gain about 20 seats, giving the new majority leadership some cushion on close roll-call votes.
- Senate: Democrats win Maine, Michigan, North Carolina, Ohio, Texas, and either Kansas or Alaska. A net gain of five gives them a 52-seat Fetterman-proof majority.
Well-Done. Surges among Black, Hispanic, and young voters lead to Democratic overperformance.
- House: Democrats gain 25 seats or more.
- Senate: Democrats win Alaska, Kansas, Iowa, Maine, Michigan, North Carolina, Ohio, South Carolina. They also score a huge upset in Mississippi or Louisiana. The new Democratic Senate majority is 55 seats. Independent surprise victories in Nebraska or Idaho bring the GOP total down to 44.
My guess is that "well-done" is more likely than "rare," but uncertainty is the watchword.
