The special election to fill Montana’s U.S. House seat took a weird turn on Wednesday night when the Republican candidate, Greg Gianforte, reportedly body-slammed Guardian reporter Ben Jacobs. An audio tape of the incident seemed to indicate that Gianforte snapped after Jacobs asked him about the Republican health care bill. What effect all this will have on today’s special election is … unclear. Suffice to say that we don’t have much precedent for election-eve body-slamming.
Gianforte had been favored over Democrat Rob Quist, but not overwhelmingly so. Importantly, roughly two-thirds of the vote (or more) may have already been submitted before Wednesday in early voting. But even if Gianforte holds on to win, this might be another special election in which the Republican candidate vastly underperforms Trump’s showing in the 2016 presidential election, cranking up the anxiety level of GOP officials all over the country.
Most pollsters gave Gianforte the lead. Gravis Marketing had him up 49 percent to 35 percent. A Change Research poll put Quist closer, with Gianforte ahead 49 percent to 44 percent. Two Google Consumer Surveys put into the field over the last three weeks actually had Quist up by 7 and 14 percentage points, respectively. In other words, there’s a wide spread in the publicly available data.
The problem with public polls for special House elections is they aren’t all that accurate. In special House elections dating back to 2004, the true margin of error for any poll is a little bit more than +/- 13 percentage points. In this race, caution is extra warranted. Gravis Marketing’s only previous foray into a special House election, back in 2015, resulted in them missing the final result by 23 percentage points. Change Research has never produced a public horserace poll besides this one. Finally, there are some methodological issues with how Google Consumer Surveys conducts its horserace polls.
The consensus among insiders is that Gianforte has a small advantage. Internal Republican polls have been reported to show a race in which Gianforte is up 2 to 4 percentage points. Those internal surveys — while possibly leaked in an attempt to skew the expectations game — do track with assessments from the Cook Political Report and Inside Elections, which give Gianforte only a slight edge.
As always, we’re interested in who wins — a Republican win in Montana gives House Speaker Paul Ryan another member in his majority — but we’re also interested in what the Montana result tells us about the national political environment. And when we’re judging the latter, we need to look at the margin of victory, not just who wins and loses.
Specifically, we’ll want to compare the margin in Montana tonight to previous presidential results in the state. In a neutral political environment, we wouldn’t expect the Montana House special election to be at all close. Montana is about 21 percentage points more Republican than the nation as a whole, according to weighted average of the 2012 and 2016 presidential results.1 (That is, if there were a tie in the national popular vote for president, a Republican would be expected to win Montana by 21 points.)
The state hasn’t elected a Democrat to the House since Pat Williams won re-election in 1994, and Zinke won re-election by 16 points in 2016. Indeed, there are 120 other Republican-held House districts that lean more Democratic than Montana. Even a close Gianforte win (say, by 10 points or less) would be consistent with a national environment that heavily favors Democrats. A Gianforte loss would likely set off panic among Republicans and signal to them that being associated with Trump (as Gianforte has tried to be) is toxic. Trump’s approval rating in ruby red Montana is probably under 50 percent, and it’s even worse nationally.
If Gianforte wins by only a small margin or loses, it would be consistent with the three previous special election results so far this year. While Republicans haven’t lost a race that a House Republican won in 2016, the Democratic candidates have, on average, outperformed expectations by 16 points.
|DISTRICT||PRIMARY||DEM MARGIN IN SPECIAL ELECTION||DEM LEAN IN LAST TWO PRESIDENTIAL ELECTIONS||DIFF.|
|California 34||April 4||+87||+69||+18|
|Kansas 4||April 11||-6||-29||+23|
|Georgia 6||April 18||-2||-9||+7|
The consistency here is key. Any single House special election is susceptible to district-specific factors. (There’s the body-slamming incident, to take just one example. And even before that, Gianforte seemed a bit out of place in Montana. He was a billionaire tech entrepreneur from New Jersey who was last seen losing Montana’s 2016 gubernatorial race, even as Trump was cruising in the state.) But special elections as a group have done a decent job of forecasting the following midterm’s House results. When a party vastly underperforms the past presidential vote consistently, it tends to do poorly in the following midterm. If the average House Republican candidate has underperformed nationally by 16 points once all the special elections occur, it would be on par with 2006, when Democrats took back the House.
In other words, if the GOP candidate puts in an underwhelming performance in Kansas, and in Georgia, and in Montana, it’s probably safe to conclude that there’s something going on nationally rather than in just those three states specifically. We already heard Republicans try to explain away Ron Estes’s relatively poor performance in Kansas by pointing to Gov. Sam Brownback’s unpopularity. And while Gianforte’s flaws probably give Quist (who himself has issues) a better chance than normal in such a red state, Montanans typically don’t send a non-incumbent Democrat to Congress unless the national mood is poor for Republicans. The state’s Democratic Sen. Jon Tester originally won office in 2006 when he beat then-incumbent Sen. Conrad Burns. That year, Tester benefited from President George W. Bush’s 47 percent approval rating in the state. Williams originally won the state’s only congressional seat when President George H.W. Bush was so unpopular that he lost Montana in the 1992 presidential election.
So how will we know if Gianforte is vastly underperforming Trump as the results come in tonight? Here’s the weighted past presidential vote by county, relative to the nation as a whole. I’ve also included how well Gianforte did compared with his statewide performance in the 2016 gubernatorial election.
Where winning will be easiest and hardest for Rob Quist
|DEMOCRATIC LEAN IN…|
||LAST GUBERNATORIAL ELECTION▲▼
||LAST TWO PRESIDENTIAL ELECTIONS▲▼
|Lewis & Clark||+19||+13|
Both the past gubernatorial and presidential results suggest that Gianforte should do well in populous Flathead County in the northwest, while Quist should do well to the south of Flathead in Missoula County. Nearby Lake County has been the most consistent bellwether in the state. The truth is, though, that we may need to need to wait for the results from the most populous county: Yellowstone. Gianforte barely won it in 2016 as he lost by just 4 percentage points statewide. Meanwhile, the past presidential vote suggests Gianforte should win it by about 7 points to be on track to win overall. If Quist is within single digits in Yellowstone, Gianforte is — at a minimum — vastly underperforming Trump in the state.
In any case, we’ll be live-blogging the election here on FiveThirtyEight, starting at around 9 p.m. Eastern, and hope that you’ll join us to see whether or not Gianforte holds his own even after body-slamming a reporter, underperforms Trump but still wins, or sets off the national GOP fire alarm by losing to Quist.