Trump’s gamble just might be worth it, but since we’re seeing bigger shifts
(back) toward Democrats in Obama-Trump congressional districts as compared to Romney-Clinton or consistent partisan districts, it’s unclear that it will. It’s probably a fool’s errand to try to figure that out now, with just under 34 hours to go until election results start rolling in.
Democrats are also hoping that they can market the president’s morals as repulsive enough to spur both turnout and persuasion among suburban, college-educated whites — and especially women. They looked poised to do so
, at least to some extent.
Mainly for these reasons, but also others — congressman Steve King of Iowa’s 4th congressional district is being chastised for racist comments, and Chris Collins and Duncan Hunter could be headed for defeat because of corruption scandals — Democrats look poised to gain enough House seats to form a majority. Their path(s) to 218 seats run mainly through these issues and demographic patterns outlined above. But just how likely is that? And what about the Senate? Governorships? State legislatures?
The discussion above sets the stage for the 2018 midterm elections. Great — but what actually transpires on that stage is what really matters. Across the country, Democrats will be challenging Republicans for control of various levels of government. Our House forecasting model at The Economist
thinks that one of the patterns above — or an unforeseen one — will deliver the House to Democratic hands 86%
of the time — about 5 trials out of 6, or the chance that you roll a die and get a 1.
At the Senate level, Nate Silver reckons that Democrats have a 14%
, or 1-in-7, chance of winning control of the chamber. Why? You’re likely familiar with this story by now; Democrats are defending a high number of seats they won in red — sometimes very, very red — states in the 2012 midterm elections. Back then, the political environments in the various states were much more favorable for candidates with a (D) next to their name. Since North Dakota’s Heidi Heitkamp has trailed in 8 of 10 polls taken in the state, according to 538
, it’s likely that she’ll miss out on another term. This means Democrats have to pick up an extra seat to win control of the upper legislative house—a total of three
pickups instead of two (they’re currently at 49 seats, and only getting to 50 would leave Vice President Mike Pence with a tie-breaking, reliably Republican vote, so they need 51 for a majority). Democrats could win in the vulnerable currently Republican states of Arizona and Nevada and still need victories for either Beto O'Rourke in Texas or Phil Bredesen in Tennessee — events which only have 1-in-5 or roughly 20% chances of happening individually.
At the state level, Democrats look poised to flip 7-8
governors mansions, according to me and my colleague’s work, and ~500
state legislative seats (or 7 chambers), according to political scientist Carl Klarner. This could position them well to influence crucial policies like congressional redistricting, since most governors will have vetos over proposed map redraws after the 2020 census and subsequent reapportionment. Democratic victories would mean fewer gerrymanders.
But as an SNL skit
made clear on Saturday night, Democrats aren’t taking this year’s forecasts for granted. It might be a happy consequence of the failure of many media types to properly evaluate our quantitative predictions in 2016 that they might better understand uncertainty this time around.
Different outcomes and their pre-postmortems
Let’s discuss the most likely outcomes and the reasons why they might come true.
Scenario 1) Democrats hold the House because Obama-Trump voters came home.
In this outcome, Republicans lose the ten seats that they’re currently forecast
to lose in the Midwest because voters who switched from Obama to Trump have reverted back to blue, likely as a result of Republican efforts to deprive them of health benefits in mid-2017—efforts that were later revealed to be a first step to the party’s full eventual repeal of Medicaid in America. The average voter in these districts was also slightly white, more educated, and female than they were in 2016, helping Democrats make up ground that they lost when “working class whites” (voters primed with racial animus) came out in droves for Mr Trump. This reversion to 2012 doesn’t help Senator Heitkamp enough to win in North Dakota, and Democrats win 49 or 50 seats in the Senate.
Scenario 2) Suburban White women really drive a Whole Foods revolt against POTUS in suburban America
. In this case, the famous “Year of the Women” comes to fruition among the mass public (as well as among Democratic candidates) and causes the blue wave to wash over just shy of two dozen
House districts in suburban America. Women, asked by exit pollsters why they (A) voted (B) for Democrats cite president Trump’s moral impropriety as the main reason. In well-educated red seats, a nontrivial number of white Republican male #NeverTrumpers are swayed by their wives and vote Democratic, handing Democrats a few extra pickups and the House majority. However, the inverse trend was observed in rural America; white men — especially those without college degrees — voted at higher than normal rates in rural America, claiming Heidi Heitkamp and Claire McCaskill as casualties of a continual realignment of the Senate to presidential voting patterns.
Scenario 3) Nonwhites fail to materialize in the South and West, non-college educated whites are motivated by late-breaking identity politics. In this scenario, Democrats win in most of the suburban House districts in which they are favored but fail to have received a necessary swell in turnout among liberal whites and low-propensity turnout Hispanics in California and Florida, landing them at either a bare minority or majority. They might have power, but either way, have little more room for affecting Washington policy than they used to — a (moral) loss in the House is compounded by losses in Arizona and Nevada actually expands the GOP majority in the Senate.
Scenario 4) The left tail of the distribution happens. In this scenario, polls have been underestimating enthusiasm for Democratic candidates among nonwhites and young voters across the nation. Finally, it seems, millennials are turning out — to the dismay of many pollsters’ turnout models that failed to account for this possibility. Democrats sweep suburban and urban districts, many with sizable university populations, and in AZ, NV, ND, and TX to capture control of the Senate. This is the “nobody saw this coming!” event, although both scenarios had about a 1-in-10 chance of happening.
Scenario 5) See section below, titled “Is the future the past? Polarization and negative partisanship.”
Scenario(s) 6+) Something else? The mixture of intertwined — sometimes more than others — Demographic and political covariation hurts our ability to really imagine what some electoral scenarios look like. It’s distinctly possible that the actual outcome isn’t yet on our radar.
Is the future the past? Polarization and negative partisanship
Finally, I fear that I have not discussed enough the reasons for why we got here, why the midterms are competitive how and where the are. Let me remedy this issue presently. A fifth explanation for the midterms.
Primarily, I believe that the most profound trend in American politics today — that of political polarization, largely because of a rise in negative feelings, or affect, for the other party — dubbed negative partisanship
in political science literature, explains well why we see the midterms electorate that we see today. Because voters increasingly view the opposing party negatively, they are less and less likely to vote for Republicans if they’re a Democrat (or Democrats if they’re a Republican). This has decreased the rate of split-ticket voting, explaining an increase in the ability
to predict House and Senate elections with presidential vote shares for the corresponding geographies. These problems are only compounded by the tendency for voters to stay in media echo chambers, associate mostly with members of their own party (or go bowling alone
This means that we should expect, over time, that the the 25 House districts that voted for Hillary Clinton and a Republican House representative to switch hands in coming years. The reverse is true as well; we should expect the 12 Trump-Democratic districts to flip in that direction over time. Of course, there are many other factors at hand in deciding this year’s elections — I’m just explaining a tangible effect of negative partisanship and polarization. However, it’s no coincidence that of these districts, 2 Trump-Democratic districts are projected to flip to Republicans on Tuesday and 18 Clinton-Republican districts are favored to change hands.
It is worth noting how this is accomplished. As Americans increasingly sort into their own partisan bubbles, decreasing the amount of the time the cross the aisle, margins of victory are going to get both slimmer and more stable as elections take place closer to the political equilibrium. However, the closeness to that equilibrium depends on voter engagement and turnout. If one party is more motivated than the other to turnout to vote, the party on top is going to see a boost in vote share that is bigger today than it was in, say, 1990, just because more of their voters are locked in. And because true independent voters make up a relatively small share of voters, these disparities in turnout, a big driver of the purported “blue wave” in 2018, will also slowly become big prizes for electoral campaigns in America. As strategists turn toward efforts to turn out their base, a lack of persuasion campaigns will likely only exacerbate the issue, reinforcing the broader trend of polarization. (Tangentially, I must note that this trend is complicated by the fact that Democrats also have a built-in advantage
if more voters turnout overall, due to there being more Democratic than Republican voters in America.)
I highlight this partisanship and polarization argument as both (A) a prospective explainer for a trend we’re seeing this year and (B) a theory for what we might come to see in politics in the future. I look forward to testing this theory over time, but it’s worth pointing out that this year’s midterm elections are being well-forecast by historic “fundamental” indicators in US politics — all of which have been influencing election outcomes prior to our current state of hyper polarization — contrary to what some people may argue.
Where to next?
It is most likely that America wakes up Wednesday morning with a Democratic House and Republican Senate. In this case, Democrats will likely immediately invest committee staff in issuing subpoenas and initializing investigations to provide much-needed oversight of the Trump Administration’s massive, currently under-checked federal government. They will be forced to work on moderate policy proposals that have a chance of making it through the Senate and past the president’s veto. However, a group of progressive Democrats demand attention for liberal causes, and continually introduce bills to expand Medicare for all Americans, decrease military funding, raise taxes on the super-wealthy and reign in corporate greed.
However, this is by no means the only plausible outcome. It takes no stretch of the mind to imagine a November 7th in which Trump and Republicans are emboldened by a surprise win in the House and the subsequent disappearance of many moderate members of their party. Conservative lawmaking can now begin again at a pace accelerated by the absence of maverick troublemaker John McCain. Indeed, now that Kevin Cramer is the Senator from North Dakota, Republicans can afford defections from the Liberal Left’s shadow GOP lawmakers, Senators Murkowski and Collins. A new conservative movement is born in America. This is a distinct possibility, however impractical.
Still, there’s the under-discussed fact that a Republican victory in the House is virtually assured to arise out of a Democratic victory in the popular vote. This would mark an unbridled assault on small-“d” democracy in America, the cementing of minority rule at every level of the federal government. The greatest liberal experiment in the world would be cast into a political crisis unmatched in modern US history.