Is Trump's Approval Rating the House GOP's 2018 Ceiling?
Will Democrats take back the House in 2018?
There are a lot of ways to attack this question, and of course no way to be certain of the answer. It’s clear that a wide variety of factors (e.g. how the electoral map is drawn, the national political environment, which incumbents choose to retire, etc.) all matter. But most analysts would argue that President Trump’s approval rating will play a central role -- a strong rating would help Republicans while a low rating would aid Democrats. So it’s worth trying to figure out how much presidential approval matters and how election watchers should interpret these polls.
I used a large dataset of polls (generously provided by Joe Bafumi, Christopher Wlezien and Robert Erikson) to get a better handle on these questions. And in recent elections, presidential approval has provided a ceiling of sorts for the incumbent party, making it difficult for that party to outpoll the president. This pattern might not hold for Trump (see the final section on extrapolating from these results), but if it does, these patterns will be useful for interpreting the polls in the coming months.
Presidential Approval Can Act Like a Ceiling for the Party in Power
In recent midterms, the incumbent party has struggled to outpoll its president. The 2010 election is a good example.
It might be tough to understand this graph at first glance, but it’s actually simple. Each point represents the difference between a generic poll from Bafumi et al.’s dataset and President Obama’s presidential approval rating (as tracked by the Roper Center) heading into Election Day 2010. So if a point’s vertical location is at zero on a given day, that means there was a poll where support for House Democrats equaled Obama’s final approval rating. The trendline is a local regression with a significant amount of smoothing.
That trendline tells most of the story here. While a few polls had Democrats outperforming Obama’s rating, most of the surveys showed Democrats getting as much or a little less support than the president. That became a problem for them in 2010, when Republicans won in a wave and took back the lower chamber.
Others have noted that Democrats had trouble outpacing Obama in the midterms, but below we see that the same pattern appeared in 2006 as well. As the Iraq War and the aftermath of Hurricane Katrina dragged on, President Bush’s approval ratings dropped. Republicans had trouble outpacing him, and they lost the House in a historic wave.
The polling was sparser in 1994, but Bill Clinton’s Democrats suffered a similar fate. Individual surveys varied (due to randomness as well as defensible choices made by pollsters), but polls showed that the Democratic Party failed to outpace Clinton and lost the House majority for the first time in four decades.
Finally, Democrats had trouble outpacing Obama’s approval in 2014. This graphic isn’t as clear as the other two because the final poll heading into that election was on the low side -- it showed Obama with a 41 percent approval rating, while the RCP average had him at 42 percent. Either way, the dynamic is similar.
While Democrats performed slightly better than Obama, according to this poll, the results were basically consistent with the other three cycles shown. Support for the president’s party was roughly equal to his approval rating.
Not every election conforms to this pattern, however. In the wake of the 9/11 terrorist attacks, the country rallied around President Bush and gave him very high marks. But in November 2002, Republicans underperformed these sky-high ratings, as Democrats who approved of Bush nonetheless stuck with their party on the congressional level. In 1998, Bill Clinton benefited from GOP overreach related to Monica Lewinsky, but Democrats underperformed his approval rating significantly.
History has also kept the sample size low here. For most of the second half of the 20th century, Democrats dominated House elections and presidents enjoyed high approval ratings if the country seemed safe and the economy was humming along. Only recently -- in the era of political polarization -- did these patterns start to emerge.
But under normal circumstances (two-party competition, no out-of-the-ordinary events drastically changing opinions of the president), the incumbent party may find it difficult to outperform the president’s approval rating.
The Two Big Ifs: Events and Empiricism
It’s not hard to imagine how someone would conclude, based on the analysis above, that Democrats have a big advantage in the coming midterms. Trump’s approval rating is low, and if Republicans can’t get above it, they will likely have midterm losses. But there are two important caveats regarding that conclusion.
First, Trump’s approval rating could improve before Election Day 2018. There’s still a year and a half to go, and anything could happen between now and then. Moreover, Republicans have some advantages (a large number of incumbents, a favorable map due to geography and gerrymandering) that might allow them to hold the House without winning the House popular vote. In other words, Trump wouldn’t necessarily have to get to a 50 percent approval rating for Republicans to keep the majority.
(The flip side of this caveat, of course, is that Trump’s political situation could become worse. His approval could drop, and some of the GOP’s built-in advantages could disappear -- e.g. incumbents in competitive districts could retire if they foresee a tough re-election fight.)
Second, it’s possible that the patterns of the past don’t apply to Trump. In 2016, the Republicans won the House popular vote while Trump lost the presidential popular vote. And it’s possible that some voters are unhappy with Trump but, because of stronger distaste for the Democratic Party, they will still vote for Republican House candidates.
There’s no guarantee that this scenario will play out either. Voters often punish the party in power for presidential overreach, and as the GOP passes Trump’s agenda (or at least attempts to) they may become more closely tied to him. And right now, the generic ballot polls show Republicans with the same level of support as Trump -- roughly 40 percent.
That being said, many aspects of the Trump phenomenon have defied previous patterns and we’ve only seen four midterm elections in the last 23 years where the president’s rating served as a ceiling for his party. So it’s worth noting these results while also being cautious about extrapolating them to 2018.