Pollsters are picking up the pace after a slow start in this midterm election season. Sunday featured the release of a trio of NBC News/Marist polls of the Senate races in Arkansas, Kentucky and Colorado, while the online polling firm YouGov released polls of almost every Senate race in conjunction with the New York Times and CBS News.
The bottom line is not much has changed. The FiveThirtyEight forecast model gives Republicans a 65.1 percent chance of winning the Senate with the new polling added, similar to the 63.5 percent chance that our previous forecast gave them on Friday.
But the path to a Republican majority is becoming a little clearer — and the problem for Democrats is that it runs through six deeply red states.
Republicans have long been favored to win back the seats in Montana, West Virginia and South Dakota, three states where Democratic incumbents have retired. Montana and West Virginia are virtual locks for Republicans. South Dakota is slightly less certain only because of the presence of a third-party candidate, the former Republican Senator Larry Pressler. However, Pressler was at only 6 percent of the vote in the YouGov poll, lower than in other recent surveys.
The GOP’s next two easiest targets are in Louisiana and Arkansas, where the Democratic incumbents Mary Landrieu and Mark Pryor are struggling with electorates that have turned much redder since their tenure in office began.
Both the NBC/Marist and YouGov polls put Pryor behind among likely voters in Arkansas. Each poll also showed Pryor in a tie with his challenger, Republican Rep. Tom Cotton, among registered voters, but that’s not entirely good for news for Democrats; several earlier polls of the state had shown Pryor ahead among registered voters, leaving open the prospect that he could win the election with a strong turnout. Pryor retains a chance to keep his seat — about a 30 percent chance, according to our forecast — but he may need both a strong turnout and a strong close to his campaign.
The new polling did not affect our forecast in Louisiana, since the YouGov poll was a survey of the state’s Nov. 4 open primary and not its probable Dec. 6 runoff between the top two candidates, Landrieu and the Republican Bill Cassidy. (The FiveThirtyEight forecast is projecting the runoff results in Louisiana and not the primary; it appears unlikely that either Landrieu or Cassidy will get 50 percent of the vote on Nov. 4, which would avert the runoff.) The poll was nevertheless bad news for Landrieu, since it showed her with just 36 percent of the vote against 48 percent for the top two Republicans, Cassidy and Rob Maness, combined. Based on previous runoff polling, Republicans have about a 70 percent chance of winning the seat, about the same as in Arkansas.
The sixth red state, which could be decisive in giving Republicans a majority, is Alaska. There, the YouGov poll gave the Republican Dan Sullivan a 6-point lead over the Democratic incumbent Mark Begich, reversing a July YouGov poll that had put Begich 12 points ahead.
It could be easy to make too much of the YouGov poll. It had a small sample size — about 400 likely voters — in a notoriously hard-to-poll state. Nonetheless, it’s among the only recent polls of Alaska and shows a result more in line with how other red-state Democratic incumbents like Pryor and Landrieu are polling.
The FiveThirtyEight model evaluates a series of “fundamentals” factors in each state in addition to the polling. Ordinarily, these receive relatively little weight in the model, but they get more in states like Alaska where the polling is thin. In Alaska, the fundamentals model puts Begich 8 points behind Sullivan, principally because his left-of-center politics are further removed from Alaska’s median voter than Sullivan’s and because Begich only narrowly won his election in 2008. Public fundraising is another factor in the fundamentals estimate; Sullivan has raised nearly as much in individual contributions as Begich, in contrast to other states where Republican challengers face larger fundraising gaps against Democratic incumbents.
Because of this, the FiveThirtyEight model already had Begich as a slight underdog before the new poll was released — although he’s a heavier underdog now, with a 30 percent chance of keeping his seat, similar to Landrieu and Pryor.
But we shouldn’t lose sight of the big picture. Republicans can win the Senate solely by winning Alaska, Arkansas, Louisiana, Montana, South Dakota and West Virginia, states which voted for Mitt Romney over Barack Obama by an average of 19 percentage points in 2012.
Another potentially vulnerable Democratic incumbent got some better news. In Colorado, both the NBC/Marist and YouGov polls put the Democratic incumbent Mark Udall ahead of his Republican challenger, Rep. Cory Gardner — as did a Rasmussen Reports poll conducted late last week. Udall is now a 63 percent favorite in Colorado, improved from a very slight underdog when the FiveThirtyEight model launched last week.
However, if Begich holds on in Alaska or Pryor does in Arkansas, Republicans retain some backup options. Iowa and North Carolina are tossups, according to the FiveThirtyEight model. Democrats remain favored in Michigan — this weekend’s YouGov poll put the Republican Terri Lynn Land slightly ahead, but the consensus of surveys show Democrat Gary Peters with the lead — but it remains a highly competitive race. Republicans face longer odds in New Hampshire and Minnesota, but they aren’t entirely safe for Democrats either.
Democrats could greatly complicate Republicans’ math by picking off one or more Republican-held seats. Kentucky, which once appeared to be Democrats’ best option, may be fading as an alternative. The NBC/Marist and YouGov polls were the latest in a long string of polls to show Republican minority leader Mitch McConnell ahead of the Democratic challenger Alison Lundergan Grimes; McConnell’s chances of keeping his seat are now up to 84 percent, according to the FiveThirtyEight forecast. The Democrats have slightly better prospects in Georgia and Kansas, although the latter state remains exceptionally difficult to forecast after the Democratic candidate, Chad Taylor, ended his campaign last week. (YouGov polled Kansas but it did not include the center-left independent candidate Greg Orman in its survey, who could caucus with Democrats if he wins.)
A mild piece of good news for the Democrats is that the turnout gap may not be as large as it was in 2010. Both YouGov and NBC/Marist released results among both registered voters and likely voters in each of the states they polled — and on average, their likely-voter models showed the Republican candidate doing a net of about 2.5 percentage points better than in the registered-voter version of their surveys. That’s in line with the historical average gap between registered and likely voters in midterm years — rather than the 6-point gap that persisted throughout 2010.
The FiveThirtyEight model automatically shifts registered-voter surveys toward Republicans to make them equivalent to likely-voter polls; this is one of the reasons our forecast is slightly more favorable to Republicans than others you might encounter. It’s likely that other models will show a shift toward Republicans and come to more closely match ours as more polling firms begin to release likely voter results.