Toss-Up, Lean, Likely, Safe: How Race Ratings Work

Three organizations dominate professional race ratings in the United States. Their assessments move money, recruit candidates and determine where campaigns spend resources.

Cook Political Report, founded by Charlie Cook in 1984, is the industry standard. Cook analysts rate every competitive House, Senate and gubernatorial race using a combination of polling, fundraising data, partisan lean, candidate quality and historical performance. Cook also developed the Cook Partisan Voting Index (PVI), the most widely used measure of district partisanship. Cook is subscription-based and primarily serves political professionals.

Sabato's Crystal Ball, run by University of Virginia professor Larry Sabato and his team, is the most academically prominent forecasting outlet. It focuses on Senate and House elections with detailed race-by-race analysis. Sabato's overall accuracy record is excellent — in 2020 and 2022, his final Senate ratings had the correct winner in every race.

Inside Elections (formerly known as the Rothenberg Political Report, founded by Stuart Rothenberg in 1984) provides similar ratings with particular strength in House analysis. Nathan Gonzales, its editor and publisher, is a prominent TV analyst. Inside Elections uses slightly different language — it adds half-steps like "Lean Republican (Toss-up)" to add granularity.

All three organizations update their ratings continuously throughout an election cycle. Major rating changes — particularly when a race moves from Likely to Lean or from Lean to Toss-up — generate significant media coverage and can trigger fundraising surges or candidate recruitment.

Note on Inside Elections: Inside Elections uses additional half-step categories — "Lean Democrat (Toss-up)" or "Lean Republican (Toss-up)" — to signal races at the boundary between categories. This provides more granularity than Cook and Sabato's seven-category system.

Race ratings are not just descriptive — they actively shape the race they describe. When Cook or Sabato moves a race to Toss-up, it triggers a cascade of downstream effects that make the race more competitive, partly fulfilling the forecast.

Fundraising: Major donors, party committees (DCCC, NRCC) and PACs use ratings to prioritize spending. A race that moves from Likely Republican to Lean Republican will immediately attract Democratic donor attention. A candidate in a Toss-up district can fundraise off the rating itself — "Cook rates this race a Toss-up, every dollar matters." Safe seats attract almost no outside money regardless of how strong the candidate is.

Media coverage: Political journalists focus disproportionately on Toss-up and Lean races. A Toss-up Senate race in a swing state will receive 10x the national coverage of a Likely Republican race in a safely red state, even if the underlying policy stakes are comparable. Candidate earned media — free press coverage — correlates strongly with race competitiveness.

Candidate recruitment: Party committees use ratings to recruit strong candidates into competitive races. High-quality candidates (experienced elected officials, strong fundraisers, well-known local figures) are much more likely to run when ratings suggest a realistic path to victory. Toss-up or Lean races in favorable party environments can attract dramatically better candidates than Likely or Safe races.

Voter mobilization: Ground game investments — field offices, canvassing, voter contact — are concentrated in competitive districts. A rating change from Safe to Likely can trigger both parties to begin turnout operations in a district they previously ignored.

The track record of major forecasters is strong for the outer categories and appropriately uncertain at the center. The key insight is that a Toss-up rating does not mean the forecaster has no idea — it means the race genuinely could go either way, and over many cycles, roughly half of Toss-up races go to each party.

In 2022, Cook Political Report's final House ratings correctly predicted the winner in approximately 430 of 435 races. Every Safe race went as predicted. Every Likely race went as predicted. Most Lean races went as predicted. The five misses were primarily in Toss-up or late-moving Lean races — precisely where uncertainty was acknowledged.

Notable historical misses illustrate where ratings fail. In 2012, several Senate races rated Lean Republican (Missouri, Indiana) flipped to Democrats after Republican candidates made damaging public statements late in the campaign. The structural rating was correct — those states leaned Republican — but the ratings couldn't forecast candidate-specific implosions.

In 2020, most forecasters rated Maine, Iowa and North Carolina Senate seats as Toss-ups. All three went Republican in an environment where Democrats were expected to overperform. The structural data pointed toward potential Democratic wins; the actual results reflected Republican-leaning electorates asserting themselves on Election Day. This is within the expected uncertainty range for Toss-up ratings.

The 2026 midterm cycle is in early stages as of April 2026, and ratings will shift significantly as primaries conclude, polling accumulates and the national environment develops. The following is a snapshot of early-cycle assessments from major forecasters.

House: Republicans hold a 220-213 majority. Most forecasters currently rate the House as Lean Democratic — reflecting historical midterm patterns and Trump's below-50% approval — but with the caveat that the margin needed is small (5 seats) and the gerrymandered map limits Democratic opportunity. The competitive battlefield is concentrated in roughly 30-40 seats, primarily Republican-held districts in New York's suburbs, California, Pennsylvania, Virginia and Arizona that have competitive PVI scores (R+1 to D+3).

Senate: Democrats face a structurally difficult Senate map in 2026. Republicans are defending relatively few seats in competitive states, while Democrats must defend seats in states like Georgia, Michigan and Pennsylvania — competitive states where Trump performed well in 2024. Most forecasters rate the net Senate environment as Lean Republican for 2026, meaning Republicans are more likely than not to hold or expand their majority.

Key Senate races drawing early attention: Georgia (Democratic Sen. Jon Ossoff in a state Trump won by 2%), Michigan (Democratic Sen. Gary Peters retiring, open seat) and New Hampshire (Democratic Sen. Jeanne Shaheen retiring, open seat in a state Harris carried narrowly). On the Republican side, Maine (Sen. Susan Collins) and Nevada (open seat in a state Trump narrowly won) are potential Democratic pickups.