FLIPR 2026 Midterm Election Forecast Model

The FLIPR Model: Silver Bulletin's 2026 Midterm Forecast

Nate Silver and collaborator Eli McKown-Dawson have launched FLIPR, a comprehensive forecasting model covering every competitive Senate, House, and gubernatorial race heading into the 2026 midterm elections. The model — whose name evokes the possibility of chamber flips — is updated continuously as new polling data arrives, running roughly 40,000 simulations per update cycle to generate probabilistic outcomes. The joint modeling approach is one of its distinguishing features: rather than forecasting Senate, House, and governor's races in isolation, FLIPR treats them as correlated events, allowing the system to calculate the likelihood of split outcomes such as Republicans holding the Senate while simultaneously losing the House majority.

The forecast draws on a broad polling infrastructure, pulling most race-level election surveys from The New York Times poll database while independently collecting generic congressional ballot and congressional approval data. Silver Bulletin's polling averages incorporate several layers of adjustment before feeding into the model. Polls are weighted by reliability, corrected for house effects — the systematic tendencies of individual pollsters to lean toward one party — and adjusted based on whether respondents are registered voters or likely voters, a distinction that historically matters significantly in midterm turnout environments. A timeline adjustment also updates older polls to reflect movement in the national generic ballot, preventing stale data from anchoring projections in a fast-moving political environment.

The Political Landscape: Democrats Gaining Ground, With Regional Asterisks

As of the late August update, the competitive terrain favors Democrats broadly but unevenly. A tour of the 48 most competitive House races published in the same cycle identifies where the party's gains are most likely to materialize — and where structural factors complicate their path. The clearest obstacle is mid-decade redistricting in the South, where Republican-controlled legislatures have redrawn maps in ways that provide meaningful insulation to incumbent GOP members and reduce the number of genuinely contestable seats in that region.

This is a significant development in the 2026 cycle. Redistricting typically happens once a decade following the census, but several Southern states have pushed through additional map changes outside that conventional window — a practice that has faced legal challenges but has nonetheless reshaped the battlefield. For Democrats, who are counting on a favorable political environment driven by voter response to the second Trump administration's policies, the geographic tilting of the map adds a structural tax on their vote efficiency. Winning a larger share of the national popular vote does not automatically translate into a proportional gain in House seats when the underlying district lines are drawn to pack Democratic voters into fewer constituencies.

The generic congressional ballot — arguably the single most important macro indicator for House forecasting — is being tracked with a likely-voter adjustment that Silver Bulletin applies to generate a more electorally realistic snapshot. Historically, the party out of the White House performs better among likely voters in midterm cycles, and that dynamic is embedded in the model's assumptions. Donald Trump's approval rating dashboard runs in parallel, serving as another key input: presidential approval has been among the strongest predictors of midterm House outcomes in modern political history, and the direction and magnitude of Trump's numbers will shape how many marginal seats become genuinely competitive.

Methodology and What's New for 2026

Silver Bulletin describes the FLIPR methodology as substantially updated from prior cycles, with the detailed technical documentation suggesting meaningful changes to how the model handles uncertainty, correlated errors across races, and the treatment of sparse polling environments — conditions that apply to the majority of House districts, which receive little or no individual polling. In those cases, the model relies heavily on fundamentals: district partisanship, incumbency, candidate quality signals, and the national environment as proxied by the generic ballot and presidential approval.

The joint modeling of chambers is worth dwelling on. In prior forecasting approaches — including Silver's earlier work at FiveThirtyEight — Senate, House, and governor's races were often treated with significant independence. Modeling them jointly captures the reality that a wave large enough to flip the House almost certainly affects Senate races too, and vice versa, because both reflect the same underlying national mood. The 40,000-simulation architecture allows the model to map out the full distribution of possible outcomes, including tail scenarios: a near-sweep by Democrats, a surprisingly resilient Republican performance, or the kind of split-decision result that produces divided chambers with distinct partisan majorities.

Why 2026 Matters Beyond the Midterm Norm

Midterms are structurally important — control of the House and Senate determines whether a president's legislative agenda can advance, stall, or be actively reversed — but 2026 carries additional weight given the breadth of policy changes enacted or attempted in the second Trump term. Immigration enforcement, executive authority, federal spending, and regulatory rollbacks have all generated significant political energy on both sides, and the midterms will serve as the first major national referendum on that agenda.

The Senate map in 2026 presents its own complexities, with a mix of Democratic and Republican incumbents defending seats in states that voted differently in 2024 than in earlier cycles. Gubernatorial races, meanwhile, will shape redistricting authority heading into the post-2030 census map-drawing process — making some of those contests consequential well beyond their immediate policy implications.

FLIPR's probabilistic outputs will shift considerably between now and Election Day in November 2026, and Silver Bulletin has built the model to update continuously as polling accumulates. The August baseline reflects a landscape that is fluid but directionally legible: Democrats are competitive for House control, the Senate remains a harder climb, and regional redistricting is functioning as a meaningful drag on what might otherwise be a strong Democratic environment.