A Trump flag at a rally for President Trump in El Paso last month. (Adria Malcolm/Bloomberg News)
Data analyst and political columnist

We live in an era of horribly misinformed political decision-making. In 2015, the Trump-skeptical Republican elite believed that Trump would naturally fade over the course of the primary, so they remained split between mostly implausible presidential candidates (looking at you, Jeb), and Trump won. That same year, major Democrats cleared the field for Hillary Clinton, who turned out to be the second most disliked presidential candidate ever and one of the few Democrats capable of losing an open race to Donald Trump. In Alabama’s special Senate election in 2017, Trump helped scandal-ridden Republican Sen. Luther Strange make it to a primary runoff with Republican Roy Moore, not realizing that Moore would win that runoff, subject America to a long race where the Republican was an (alleged) pedophile and ultimately lose the general.

Most of these mistakes were based on bad political calculations. People overestimated Clinton’s strength, the demographic resiliency of the Democratic coalition or Moore’s weakness, and their plans fell apart. And both Republicans and Democrats are in danger of making new mistakes by underestimating Trump’s strength in 2020.

The conventional wisdom on Trump’s reelection odds are much too bearish. Trump has a better chance of winning again than many pundits seem to think. And both Democrats and Republicans need to take Trump seriously and recalibrate their decision process or risk facing serious consequences.

Getting the odds right

I think Trump has roughly 50-50 odds of winning reelection at this point. The other wonks who are more bearish on Trump aren’t crazy — over half of the country disapproves of his job performance — but I think my view is reasonable.

The economy is the most promising indicator for Trump.

Emory University professor Alan Abramowitz put together a simple statistical model that uses economic growth, the president’s approval rating and whether the president’s party has held the White House for one term or more to predict the results of presidential elections. The basic logic of the model is simple. A solid economy (which Trump has) and high approval ratings (which he doesn’t) are good signs for a president’s reelection chances. If you feed the model current economic and poll numbers (which obviously can change by the time we get to Election Day) the results suggest that Trump would get about 51.4 percent of the two-party presidential vote.

Models like this don’t have pinpoint accuracy and they don’t take the electoral college into account. But they’re useful for ballparking the results. And the data suggests that a good economy, a not-amazing Democratic opponent (models such as this often implicitly assume that the parties nominate candidates who have roughly even political skills) and the advantages of incumbency might be enough to get Trump close or carry him over the finish line.

Trump has some big, obvious weaknesses, too. His approval rating has been low for almost his entire first term, and a Democrat could easily win by snagging people who disapprove of the president. Moreover, Trump’s approval rating has been unusually stable — he’s been bumping around somewhere between the high 30s to low-to-mid 40s for much of the last two years.

Trump’s approval rating could remain steady until the election — which could (especially if economic numbers dip) lead to a Carter-esque defeat against the Democratic Ronald Reagan. (Kamala D. Harris? Bernie Sanders?)

But I think the conventional wisdom is discounting the possibility that Trump’s approval improves enough for him to win and overestimating the (still very real) likelihood of total collapse.

Trump’s approval rating has been sitting in the low 40s for a while, but he doesn’t need to do much to improve his numbers. Trump’s approval rating tends to slide upward when he’s not actively pushing it down by pushing an unpopular health-care law or shutting down the government in pursuit of a border wall. If Trump were to calm down a little bit (which I know is a big ask), maybe work out some very basic happy-sounding compromises with House Democrats (or at least make a big show of trying), it’s not impossible to imagine him getting into the mid-to-high 40 percent range. If Trump is in that range, he has a solid chance — he was able to win over some voters who disliked him in 2016, and a Democratic nominee who runs on a far-left platform might drive some softer, Trump-disapproving Republicans back into the fold.

Trump may choose (as he has for most of his presidency) not to do any of this and stay stuck in the low 40s. But I think people are underrating how easy it would be for Trump to improve some and get into a range where he has a solid chance of winning. Approval ratings change — historically speaking, a president’s approval rating roughly 770 days from inauguration (which is where we were about a week ago) doesn’t have a strong relationship with their eventual reelection vote percentage.


Current job approval doesn't tell us much (David Byler/The Washington Post)

Obviously, there’s another side to this situation — Trump has a lot of catastrophe potential. A recession could hit, and his numbers could take a truly intense dive. Special counsel Robert S. Mueller III could come up with something truly damning. One of the other many investigations into Trump could also blow a hole in his support.

But I would urge caution here. Nobody knows exactly when the next recession will hit, how deep it will be or whether the economy will be able rebound by election time. It’s possible to imagine cases in which a recession hits early enough for Trump to take credit for the rebound. And I don’t think anyone truly knows how the Mueller probe will end. Most of the time, people underestimate the probability of catastrophic events, but the punditry seems to have the opposite tendency with Trump. The End of Trump has been right around the corner since the summer of 2015, yet he’s still here.

To me, this situation adds up to Trump having roughly 50-50 odds of winning reelection. There are a lot of ways that things could go very wrong for Trump, but there are also a lot of ways things could improve enough to put him over the top. Other people might come with somewhat different math, but, regardless, it’s a risk to assume that Trump is a one-termer walking.

If Democrats assume Trump is toast, they might make some huge mistakes

Democrats have just started the process of picking their presidential nominee. And if they underrate Trump too much, they may make some big strategic mistakes.

Democrats could underestimate Trump and incorrectly think they have more room to move to the left than they really do. I want to be clear here — this isn’t a “Back To The Center, Democrats” take. I think a Democratic Party that adopts a people-versus-the-powerful ethos, heads left on economics and pushes on issues where Trump has been especially un-populist could work. But not every left-leaning position is as popular as marriage equality. Some 2020 Democrats (most notably Kamala Harris) have signaled that they’re open to some form of reparations for slavery. Whatever you think of the merits of that policy, it polls badly.

Underrating Trump could also lead Democrats to make bad decisions about the map. It would be easy for Democratic politicians to look at Trump’s low approval numbers, the growing number of Asian American and Latino voters, and conclude that they should de-emphasize the Midwest (or take the region for granted) and run hard in long-term targets such as Georgia, Arizona and Texas so they can run up the electoral college score.

Democrats shouldn’t do that. They should try to play on a broad map that includes Midwestern swing states as well as suburban, diversifying America. It’s smart for Democrats to try to get Republicans to spend money and effort on Georgia and Arizona, but Pennsylvania, Michigan and Wisconsin are still arguably the lowest hanging fruit and the best route to 270 electoral votes.

Republicans need a real plan

Republicans probably can’t plan too much for what would happen if Trump loses. Whether Trump exits in 2025, 2021 or sooner, the party seems destined for a post-Trump identity crisis. Instead, Republicans should make sure they have a plan for what to do if Trump wins.

Trump seemed to come into office with clear policy instincts but with a less clear ideology. He was an immigration hard-liner who wanted to undo President Barack Obama’s legacy and avoid committing himself to a Paul Ryan style of fiscal conservatism. But when he took office, he basically spent his political capital on less-than-popular parts of the Ryan agenda (i.e., traditional GOP tax cuts and attempting to repeal and replace the Affordable Care Act). If Trump gets a second term, he might, depending on the results in the House and Senate, have the chance to try to create a more coherent policy vision. A competent president would try for that.

Maybe more importantly, a second Trump term would shape the long-term trajectory of the GOP. No matter what happens in 2020, the GOP will have to figure out which elements of Trump-ism they should keep (maybe an openness to violating orthodox libertarian economics) and which parts need to be memory-holed fast (a “both sides” approach to events such as the white nationalist rally in Charlottesville). But if Trump wins twice, it will be harder for the rest of the GOP to argue with the Trumpists. In that case, the GOP will have a serious planning challenge — trying to create a coherent policy platform that works for an increasingly diverse country while Trump’s face is stamped on your party. The GOP should take this possibility just as seriously as the chance that he loses in a landslide and start planning for it. An unexpected victory can turn contain the seeds of future defeats.

Read more:

Jennifer Rubin: How to lose in 2020 if you are a Republican

Catherine Rampell: Democrats in 2020 are at risk of turning into Republicans in 2016

Jennifer Rubin: Should Democrats take the Green New Deal literally or seriously?

Megan McArdle: ‘We’re nuts!’ isn’t a great pitch for a Green New Deal

Paul Waldman: Will the 2020 Democrat be progressive, or electable against Trump? Probably both!