All the Crystal Balls Say Democrats Will Gain in Midterms

(Bloomberg Opinion) -- Here’s the bottom line on how the battle for the House of Representatives looks as we approach the fall campaign season. Every way of forecasting the outcomes now concludes that Democrats are favored to reach at least 218 seats and therefore gain a House majority — but there’s still a substantial chance that Republicans, while losing seats, could retain a slim majority. The plausible range remains very wide. In fact, I wouldn’t be shocked by anything from a dud of a 12-seat gain for the Democrats all the way up to a larger-than-2010 landslide of 70 seats.

There are basically three ways to forecast House elections. One is by looking at polls, using both national polls and individual district polls. That corresponds with the “lite” version of the Nate Silver forecast at FiveThirtyEight, and shows Democrats with an estimated 69 percent chance of reaching 218, with an average gain of 34 seats and an 80 percent range of Democrats gaining from nine to 61 seats. See also G. Elliot Morris’s fairly similar results.

Another is by “fundamentals” models which look at regularities between such things as presidential approval, the economy and exposure — that is, how many seats a party currently holds and how many incumbents are running. Seth Masket ran a simple model of that type at Mischiefs of Faction back in February, and it pointed to Democrats picking up 45 to 50 seats. Silver also incorporates something similar in his more sophisticated models (as he explains in a highly recommended overview of what his forecast is saying).

And then there are district-by-district reported forecasts at sites including Inside Elections, the Crystal Ball and DailyKos, which take into account a variety of factors to assess the chances of what will happen in each House seat. Those don’t give us quantitative results, but we can still use them to get a good sense of what’s going on. For example, currently the Cook Report has 202 seats classified as safe, likely or lean Democratic (with only nine of those in that weaker “lean” group). If Democrats win all of those and both of the currently Democrat-held toss-ups, they need only take 14 of the 27 Republican-held toss-ups to get to 218. And that still leaves Democrats plenty of targets — not only the remaining Republican-held toss-ups, but also the 26 lean Republican and even the 25 likely Republican seats. Again, the same thing holds here as with the numerical forecasts: It’s far easier to see paths for Democrats to win, and perhaps win big, but it’s still quite possible that Republicans will keep their losses relatively small and hold on to their majority. 

If you want to combine all of these methods, use Silver’s “deluxe” model. But the truth is that the differences between any of them are smaller than the likely errors involved, so it doesn’t really matter a lot what you look at, since they’re all basically saying the same thing: Democrats are favored, Republicans have a solid chance, and there’s a wide range of plausible outcomes. 

While it’s always possible, I don’t expect much that really matters about this to change until the last few weeks of the campaign. The polls will ebb and flow, but even if those changes are real — and there are large margins of error in these types of surveys — the odds are that even real shifts may quickly dissipate. That is, good or bad news about Trump might change his approval numbers, and Republican polling, by a few percentage points, but it will burn off in a couple of weeks as the news cycle moves on. What could make an oversize difference, however, is any big events in the last few weeks of the campaign, when the effects don’t have enough time to dissipate. 

So I won’t tell you to not to obsess over a handful of changes in Cook ratings or not to keep an open tab on Silver’s forecast and keep hitting refresh on it. After all, I can’t stop myself from doing so. But odds are we’re still where we are now in mid-October, and in fact we very well might be in the exact same place until the polls close on Nov. 6.

1. Michael Tesler at the Monkey Cage has the numbers on Trump’s approval ratings among African Americans.

2. Elaine Kamarck and Alexander R. Podkul have some data that suggests the Republican Party may not be quite as Trump-centric as some assume. 

3. Vox’s Sean Illing talks with political scientist Suzanne Mettler about why many Americans don’t connect their dislike of government programs with the government programs they like and benefit from.

4. Historian Julian Zelizer on Trump’s enemies list.

5. Heather Hurlburt on one potential consequence of Trump’s foreign policy: Germany getting the bomb

6. Stan Collender on the demise of Trump’s military parade

8. Catherine Rampell calls on free speech conservatives to stand up when it matters.

9. Greg Sargent on Trump and loyalty. 

10. Jennifer Bendery has the latest on the Blake Farenthold saga

11. And my Bloomberg Opinion colleague Noah Feldman is absolutely correct that a free press is not (necessarily) a neutral press — and that James Madison and the other Founders would never have thought so given that media neutrality, whatever its faults and virtues, wasn’t an 18th century idea at all.

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Jonathan Bernstein is a Bloomberg Opinion columnist covering politics and policy. He taught political science at the University of Texas at San Antonio and DePauw University and wrote A Plain Blog About Politics.

©2018 Bloomberg L.P.