A changing tide?

By Ruy Teixeira
12 April 2018
Looking Ahead to the 2018 midterm elections
Created with Sketch.

The 2018 election in the United States will be the first national vote since Donald Trump’s stunning populist triumph in November, 2016. As such, it will tell us a great deal about the popularity (or lack thereof) of Trump’s version of the Republican party and the Democrats’ chances of mounting a robust comeback after the 2016 debacle.

What’s at stake?

All 435 seats of the House of Representatives will be decided. Democrats need a net gain of 23 seats to take back control of that body. The Senate has 35 of 100 seats up for election this November and the Democrats need a gain of 2 seats to gain control.

Notably, the particular seats up for takes in the Senate present a daunting challenge. The overwhelming majority of these seats – 26 of 35 – are already held by the Democrats and thus will have to be defended from Republican challenge. Moreover, 10 of these seats are in states that Trump carried in 2016, compared to just one seat the Republicans have to defend in a state won by Hillary Clinton.

The federal House and Senate elections get the most attention from observers and pundits. And they are certainly very important. But let’s face it, 2018, even with very favorable results in the House and the Senate, is not going to be the start of a new progressive era in the United States. No, that is really a 2020s thing when President Trump is (hopefully) defeated and Democrats have enough strength in the states to dominate the next round of redistricting, thereby allowing them to push back against Republican gerrymandering and translate their underlying political support into actual political victories.

That’s why the most significant results of the 2018 election may well be those for state, not federal, offices. Here’s what’s at stake in state elections:

  • 36 governors’ races, 26 of which are currently in Republican hands. And of the 26 Republican-held seats, 13 are in states that Obama wonin at least one of his presidential victories.
  • At least half of state Senate seats in 42 states (in 15 of these states, the entire Senate is up).
  • Every state House seat in the overwhelming majority of states.

These results will set the playing field for state elections in 2020 and the redistricting thereafter. Procedures in states vary but the typical setup is for the state legislature to be in charge of the actual redistricting with the governor having veto power.

In 34 states, the governor who will be in office for the upcoming redistricting will be elected this year (two were elected last year, which the Democrats bagged) and in 30 states half or more of state senators who will preside over the process will be elected this year.

Of course, 2020 will be important too, but the revolution, so to speak, starts this year.

Lay of the Land

But will it really start this year? There are a number of favorable indicators that say it just might.

Start with President Trump’s job approval rating. Currently it stands at about 40 per cent approval vs. 54 per cent disapproval. This is bad. Very bad. No president at this point in his term has been this poorly rated – or even close – in the history of polling.

There is no indicator more consequential for the political environment than the approval rating of the incumbent President. Countless political science studies show a strong link between how poor this rating is and losses by the incumbent party at every level in a midterm election (typically a difficult election for the incumbent party anyway).   This means every GOP candidate at every level is fighting the serious headwind of historically high Presidential disapproval.

Of course, if Trump’s approval rating goes up significantly between now and November, Republican prospects will improve. But this seems unlikely. Trump’s approval ratings have been remarkably stable over the last year, sometimes a point or two higher than 40 per cent and frequently a point or three below. Nothing in the current trajectory of his approval ratings – or of the apparent popularity of his latest tweets or legislative moves – suggests he is breaking out of this box.

Nor is he likely to get a spike in approval from the relatively good economy between now and November. He hasn’t yet – in fact, Trump’s ratings have dramatically underperformed relative to the health of the economy throughout his Presidency.

So, in all likelihood, the GOP is stuck with a very unpopular President as we move toward the 2018 elections. Other indicators are also ominous.

  • The generic congressional ballot, where respondents declare whether they favor Republicans or Democrats in the upcoming congressional elections, has been strongly in favor of the Democrats since about last May and currently stands at about a 9 point Democratic advantage. Models indicate that Democratic margins on the Congressional ballot, both now and more generally, predict seat gains sufficient to take back the House.
  • In special (i.e., not part of the regular cycle) federal and state legislative elections since 2016, the Democrats have outperformed their expected vote in these districts–based on underlying partisanship and previous election results—by an average of 13 points. Most stunningly, in a recent special election to fill the Pennsylvania 18th Congressional District House seat, Democrat Conor Lamb was victorious in a district that voted for Donald Trump by almost 20 points. These special election results also very predictive of outcomes in the next regular election—and at all levels.
  • In the regularly-scheduled non-federal 2017 elections, the Democrats also over-performed. The contest for governor of Virginia—which was supposed to be close—was won easily (54-45) by Democrat Ralph Northan over Republican Ed Gillespie, who had attempted to emulate Trump by running an anti-immigrant scare campaign. And downballot in the Virginia House of Delegates—the lower house of the Virginia legislature—the Democrats flipped 15 seats—going from a lopsided 66-34 disadvantage to a near-tie (51-49). The latter result has considerable predictive power for the outcome of the 2018 elections.
  • To add to Republican problems, recent anti-gerrymandering court decisions have given the Democrats a leg up in some states. Most notably, in Pennsylvania, the state supreme court mandated a new congressional district map that gives Democrats a good chance of picking up one quarter of the seats they need to win back the House in just one state.
  • In addition, indicators of party mobilisation like voter enthusiasm, relative turnout and candidate recruitment are all heavily on the Democrats’ side.

All this suggests that the Democrats have an excellent chance of taking back the House and making big gains in state governors and legislatures. They even have an outside shot of taking back the Senate, despite the very unfavorable map cited above.


So it should be a very good election for the Democrats in 2018. But how good? There are always things that could go wrong – a sudden change in the political environment, a spike in enthusiasm among Republican voters, a stronger-than-expected advantage for incumbents. But arguably the key variable lies elsewhere in Democratic strategy itself.

One might radically simplify this strategic question as conversion vs. mobilisation. There are some voices within the Democratic party who insist that the party should concentrate on mobilising its base in minority and educated suburban districts and essentially forget about getting the votes of anyone who voted for Trump in the 2016 election. Others argue that, while base mobilisation is indeed hugely important, a really big wave election depends on reaching into less favorable voter groups and picking up enough support to flip seats in a wide variety of districts.

The data would appear to support the latter position. The Democrats’ electoral progress since 2016 is not explicable without factoring in an ability to swing white non-college-educated and 2016 Trump voters. Case in point: The recent special election in Pennsylvania’s 18th Congressional District was narrowly won by Democrat Conor Lamb in a district that was around 60 per cent white non-college and voted by almost 20 points for Donald Trump. Post-election data analyses show definitively that Lamb’s victory was attributable both to his ability to mobilise the Democratic base and persuade a significant number of white non-college-educated 2016 Trump voters to support him.

It would thus be foolish for Democrats to concentrate on only certain kinds of districts and ignore others. In reality, the Democrats have reasonable chances in districts with a wide range of demographics. The only real mistake they can make is not to cast their net widely enough to take advantage of these openings.

So: how high the wave? We shall find out in November.

Image credit: Michael F. Hiatt / Shutterstock.com