Kansas, Georgia, South Dakota. We have cycled through a lot of wild-card red states over the last month in surveying the landscape of competitive Senate races. But the most likely wild card might be the one we always expected: Alaska.
Recent polls suggest that Mark Begich, the Democratic senator, has made gains over recent weeks. The new polls, the first to show Mr. Begich ahead in more than a month, add considerable uncertainty to a race in a state that has a history of inaccurate polling. But even if Mr. Begich does not lead, he is most likely within striking distance, especially given the unusually robust Democratic turnout effort that is taking place.
The state fell off the radar over the last few weeks because just about every unsponsored survey was showing Dan Sullivan, the Republican, in the lead. But over the last few days, two Alaska-based pollsters have shown Mr. Begich with a substantial lead.
On Friday, Hellenthal and Associates, a Republican-leaning Alaska-based firm, showed Mr. Begich ahead by a 10-point margin. Mr. Sullivan led by 4.5 points in the last Hellenthal poll, conducted in mid-September.
Equally striking was that the rest of the Hellenthal poll seemed eminently reasonable. Party registration seemed about right, with 33 percent saying they were registered Republicans and 18 percent saying they were Democrats. In 2010, 30 percent were registered Republicans and 15 percent were registered Democrats. The state’s embattled Republican governor, Sean Parnell, led by two points — no worse than other polls, if not better than most. Voters who said they “usually” supported Republicans outnumbered Democrats by 40 to 23.
Then, on Monday evening, Ivan Moore Research showed Mr. Begich ahead by a modest margin. The poll offered two results, one with a loose and one with a tight likely voter screen. Mr. Begich led by seven points among fairly likely voters, and led by eight points among the most likely voters.
The poll, released on Facebook, offered few details about the composition of the sample or likely voters. It did not release results for the state’s competitive governor’s contest, either. But there was evidence that the poll might have a somewhat unrepresentative, Democratic-leaning sample: The poll showed the state’s longtime Republican House Representative, Don Young, leading by just two points among fairly likely voters and trailing by five points among the most likely voters. Mr. Young is not thought to be especially vulnerable this year; the Hellenthal poll, in contrast, showed Mr. Young leading by 15 points.
On the other hand, the possibility that Mr. Young might find himself in a somewhat tight race isn’t entirely out of the question. He has found himself in tight races before. And last week in a speech with students and teachers, he came under attack for blaming insufficient support from friends and family for the suicide of an Alaskan high school student. The comments were the latest in a potentially damaging string of incidents, including his comparison of same-sex marriage to bull sex in front of the same audience of high schoolers and a verbal confrontation with his Democratic opponent.
I doubt Mr. Young is locked in a two-point race, but it would not be surprising if his numbers were weaker than they ought to be. It thus might not be bad news for Mr. Begich if he were only running 10 points better among fairly likely voters than the Democratic congressional candidate, Forrest Dunbar, as the poll suggested.
Despite these two polls, it would be hard to argue that Mr. Begich has a decisive lead. That’s because Harstad Research, a Democratic firm conducting research for the Democratic-aligned Senate Majority PAC, made an ill-timed decision to release a survey showing Mr. Begich and Mr. Sullivan tied at 44-44, just a few hours before the Hellenthal poll showed Mr. Begich ahead by a large margin.
It is not usually a good sign for a candidate when his own partisan polling doesn’t show him ahead. But at the time it was released, the poll was the best result for Mr. Begich since the last time the Democrats released an Alaska poll, more than a month ago. The implication might be that it’s also the first time that Democrats found themselves in a tie in a month — perhaps another indication of gains for Senator Begich.
The only other new data comes from the New York Times/CBS News/YouGov panel of this past weekend, which showed Mr. Sullivan up by three points. But as I’ve emphasized in every release of YouGov data, I’m not sure it’s worth focusing too much on anything from Alaska. The panel is extremely small, with just 561 respondents this time around. That might be fine for a normal probability survey, which attempts to take a random sample of adults, but it’s a big headache for a survey that doesn’t rely on random sampling.
YouGov uses a technique known as matched sampling, in which representative respondents are selected out of an unrepresentative panel. The underlying assumption is that there is a fairly representative sample of America’s population among YouGov’s 1 million panelists; the challenge is to select it.
Finding 1,000 representative panelists out of 1 million is easy to imagine. Finding 500 representative panelists out of 500 panelists just doesn’t happen for a nonprobability, online Internet panel. But that’s basically what we’re working with in Alaska, where YouGov is interviewing almost its entire panel and weighting it heavily. The modeled margin of error, based on simulations repeating YouGov’s methodology, is plus or minus 8.6 percentage points.
Alaska Election Preview »
Before YouGov, Moore, Harstad and Hellenthal, the last polls in Alaska were conducted more than three weeks ago by Rasmussen, Fox News and CNN. All three showed Mr. Sullivan in the lead, but all three showed him up by between just five and nine points among white voters — the type of margin that could be erased by nonwhite voters if they represented about 20 to 25 percent of the electorate and voted for Mr. Begich by 20 to 25 points — as expected. Apparently, none found Mr. Begich so far ahead among nonwhite Alaskans.
It is possible that Mr. Sullivan was overperforming among nonwhite voters. It is also possible that the polls were struggling to sample the state’s Democratic-leaning population of Alaska Natives, who represent 13 percent of the voting-eligible population, and sometimes do not have home addresses, let alone landline telephones.
Over all, there’s considerable uncertainty about the state of the race in Alaska. But between the history of polling errors in Alaska and the signs that Mr. Begich might be somewhat better positioned than he was a few weeks ago, there’s a very good case that Mr. Begich is close enough that he could pull off an upset with a superior turnout effort.
If any Democrat should have a superior turnout effort, it’s Mr. Begich. The Democrats have invested near the point of diminishing returns.
My colleague Derek Willis reported last month that Democratic campaigns and aligned groups were outspending Republicans by an 8-to-1 margin on field operations in Alaska. Democrats there were spending nearly nine times as much per capita as Democrats in North Carolina. Sasha Issenberg reported in The New Republic that Democrats would have 130 staff workers in Alaska, or one for every 3,000 Alaskans who voted in 2012. The most recent New York Times/CBS News/YouGov poll found that a staggering 43 percent of registered voters said they had been contacted by the Begich campaign.
Because Alaska’s population is no larger than the average congressional district, a startlingly small numbers of voters are enough to swing control of the Senate. Just 7,500 voters would be enough to overcome a three-point deficit.
