When we last took an overview of Senate races in December, Republicans appeared to be slight favorites to take control from Democrats, with a net gain of four to five seats representing the most likely outcome.
Since then, however, Republican fortunes have diminished somewhat because of problems with the quality of some candidates and key retirements. Although Republicans are most likely to gain seats on balance because Democrats have considerably more incumbents up for re-election, the question of whether the Republicans will win enough to gain control now appears to be closer to a tossup. In fact, the outcome may depend on who wins the presidential election, as well as whether an independent Senate candidate who is favored in Maine will caucus with the Democrats.
The table below contains the FiveThirtyEight estimates of the probability of victory for each party in the 33 Senate seats to be contested this November. These forecasts, although they are informed by polling and other objective measures, are ultimately reflections of my best judgments and also account for “intangible” factors like candidate quality and the partisan orientation of the state. (We traditionally switch over to purely tangible factors in the summer, once polling becomes more robust and we post the official forecast model for the Senate.)
Although these estimates run on a scale from 0 percent to 100 percent, we also classify the races into different categories like “Lean Democrat” and “Likely Republican” for ease of interpretation. Nine races have switched categories since our previous update: six in favor of Democrats, two in favor of Republicans, and one in favor of an independent candidate. These changes do not reflect shifts in the overall political environment, which remains fairly evenhanded, but rather circumstances particular to each race.
Shifts in Favor of Democrats
We had previously categorized the Senate race in California as Likely Democrat. The Democratic incumbent up for re-election there, Dianne Feinstein, has only average approval ratings and at age 78 was a plausible risk for retirement. However, Mrs. Feinstein is running in the state’s June 5 primary and, between California’s Democratic politics and its expensive airwaves, it would take a strong and well-financed Republican challenger to defeat her. Instead, the Republican field in the state has been disorganized, with few candidates capable of running a full-fledged campaign. We now classify the state as Safe Democrat.
In Indiana, Democrats have a pickup opportunity because of the defeat of the longtime Republican incumbent, Richard G. Lugar, in the primary last week. Although the candidate who defeated him, State Treasurer Richard Mourdock, should be capable of running a competent race, Democrats have a strong candidate in Representative Joe Donnelly, who survived the Republicans’ electoral onslaught in 2010. The only poll of the race so far showed a tie between Mr. Mourdock and Mr. Donnelly. We give Mr. Mourdock a slight edge because of Indiana’s overall partisan orientation and classify the race as Lean Republican. But the race is much more competitive than it would have been if Mr. Lugar had survived the primary.
In Michigan, the Democratic incumbent Debbie Stabenow gained considerable ground in the polls after a questionable advertisement by one of her potential Republican opponents, Peter Hoekstra. Mrs. Stabenow may also be helped by the state’s improved economy, as Michigan’s unemployment rate has dropped substantially, to 8.5 percent from a peak of 14.1 percent. We now classify the race as Likely Democrat rather than Lean Democrat. Republican chances to win it may depend in part on how much time and money Mitt Romney invests there in the fall.
Republicans are still favored to pick up a seat in Nebraska, where Ben Nelson, an unpopular Democrat, is retiring. However, Democrats have the better-credentialed candidate in Bob Kerrey, the former senator and governor, whereas Republican primary voters on Tuesday elected an inexperienced candidate, State Senator Deb Fisher, in an upset. Mrs. Fisher should still be the favorite because of Nebraska’s strong Republican orientation, and she led Mr. Kerrey in polls in March. Still, the candidate-quality gap is substantial and could translate into her running an inferior campaign, and candidate quality is a variable that we have found to have predictive power in the past. We now classify the race as Lean Republican rather than Likely Republican.
The hardest Senate race to call may be in North Dakota, where Democrats have another tough defensive assignment after the retirement of Kent Conrad. Like in Nebraska, the partisan climate in the state leans toward Republicans. However, Democrats are somewhat more unified behind their candidate, Heidi Heitkamp, the former state attorney general. The only recent poll of the race showed Mrs. Heitkamp slightly ahead, but it was commissioned by the North Dakota Democratic Party and such partisan polls often exaggerate their candidate’s standing. For now, we classify the race as Lean Republican, rather than Likely Republican as before, but the race could revert to Likely Republican or move into the tossup category pending the release of nonpartisan polling. One measurable factor working in Republicans’ favor is that their candidates have combined to raise more money than Mrs. Heitkamp has.
Finally, in Pennsylvania, Republicans were in need of a strong candidate to seriously challenge the Democratic incumbent, Bob Casey, who is fairly popular. Instead, their field failed to coalesce and they picked an inexperienced nominee, the businessman Tom Smith, in their primary in April. We classify the race as Safe Democrat, rather than Likely Democrat as before.
Shifts in Favor of Republicans
Two races have shifted in favor of Republicans. In New Jersey, Robert Menendez, a Democrat, has middling approval ratings and is below 50 percent in the polls against the most likely Republican nominee, State Senator Joseph Kyrillos. Mr. Kyrillos would need to catch some breaks to win the race — his fund-raising has lagged Mr. Menendez’s so far, and he might need Mr. Romney to be more competitive in New Jersey than Republican presidential candidates have been in recent years. Still, there is a plausible path to an upset, and so we now classify the race as Likely Democrat rather than Safe Democrat.
We had classified Texas, where Kay Bailey Hutchison, a Republican, is retiring, as Likely Republican, hedging our bets somewhat as open-seat races are inherently somewhat competitive by default. But Democrats have failed to recruit a good candidate, while Republicans have several well-qualified choices. A significant candidate-quality advantage, and Texas’ substantially Republican partisan orientation, is enough for us to classify the race as Safe Republican.
Shift in Favor of an Independent Candidate
The Senate race in Maine has already been upended twice: first, when Olympia Snowe, a Republican, announced her retirement, and when former Gov. Angus King, an independent, announced that he would run. Mr. King was popular as governor and now has a large lead in the polls. He is formidable enough, in fact, to have deterred several major-party candidates from running, particularly Democrats. Although independent candidates inherently face some difficulties in winning elections — they cannot receive money from party committees, for instance — Maine has a history of treating independents favorably, as Mr. King’s two terms as governor demonstrated. Thus, we classify the race as Likely Independent. If there is an upset, it could be by the Republican candidate in the event that votes on the center and the left split between Mr. King and the Democrat — although Democrats could avert that outcome by essentially surrendering the race to Mr. King.
Instead, the most interesting question in Maine might be which party Mr. King would caucus with if he wins. The Democrats are a much more likely possibility than the Republicans; Mr. King’s politics are center-left, and he endorsed Representative Chellie Pingree, a Democrat, in 2010. Still, Mr. King has declined to reveal his intentions if he wins, suggesting at times that he might not caucus with either party or that his decision could depend on the overall partisan environment after the November elections.
The ambiguity over Mr. King’s status complicates the overall forecast. Currently, we project the most likely outcome to be Republicans winning 50 seats, Democrats 49, and Mr. King the seat in Maine. Under those circumstances, the Democrats would retain control of the Senate if Mr. King caucused with them and President Obama won re-election, making Vice President Joseph R. Biden Jr. the tiebreaking vote. Otherwise, Republicans would control the chamber.
However, it should be remembered that this is just one potential outcome out of a fairly wide range of possibilities. In 2006 and 2008, Democrats won all or almost all of the races that would have been classified as tossups or as leaning toward one party at this stage. Republicans did the same in 2010.
This year, if Republicans won all the lean and tossup races, they would pick up a net of nine seats from Democrats. That would give them control of 56 seats — a sizable majority whether or not Mr. King aligned with them.
Conversely, if Democrats won all the lean and tossup races, they would actually gain a net of three seats from Republicans, or four if Mr. King caucused with them. That would give them 55 or 56 seats over all.
What would produce such a shift in the overall partisan environment is unclear, but a substantial economic upturn or downturn, a major foreign policy success or failure, a significant gap in turnout among Democratic and Republican voters, a partisan row in Congress like last year’s debt ceiling debate or a major scandal involving Mr. Obama or Mr. Romney are all possibilities. Put more simply, the possibility for major news events is a “known unknown,” and there is plenty of time between now and November.
Still, if the elections were held today, the Senate races would be more localized than they were in 2006, 2008 or 2010, and Mr. King’s decision and the outcome of the presidential race could swing the partisan balance.