Donald Trump is historically unpopular, and Republicans are underperforming his margins in special elections across the country. Some might be tempted to look at these numbers and conclude that Trumps’ re-election effort (which is already underway) is doomed.
But I don’t think that’s the right conclusion.
This table shows the approval ratings of each incumbent president since Truman 400 days after their inauguration (data is from FiveThirtyEight and David Leip’s Election Atlas) along with their popular vote margin when they were up for re-election.
There’s no real pattern in that table.
Low approval ratings don’t forecast doom or victory. Ronald Reagan was dealing with a recession in early 1982, and Republicans lost 26 House seats in that year’s midterm elections. But Reagan recovered by 1984, winning every state except for Washington, D.C., and Minnesota (the home state of his opponent, Walter Mondale). Jimmy Carter, on the other hand, had a 47 percent approval rating in 1978, floundered for significant portions of his term and went on to lose to Reagan by about 10 points in 1980. Ford’s approval rating more closely resembles Trump’s, but it’s harder to compare the two since Ford took over partway through Richard Nixon’s second term and Trump was elected on the normal schedule.
Trump could recover by 2020. The economy is good, and if he manages to take advantage of that (provided there’s no downturn) it’s possible to imagine him winning re-election. President Obama wasn’t particularly popular in early 2010 and Democrats lost the 2010 midterms in a landslide. But Obama recovered and won re-election by a decent margin in 2012.
Conditions could also get worse for the president. His approval rating on the economy is often higher than his overall approval rating—suggesting that an economic downturn could damage him in the polls. Trump could then struggle on policy, drive numerous persuadable voters toward the Democrats and lose re-election in Carter-esque blowout. Or we could see a scenario somewhere between the middle of the two—e.g. a close loss, another popular vote loss coupled with an electoral college win, etc.
The point here is that we’re a long way away from 2020, and any prediction based on Trump’s approval rating is really just a guess.
And there’s not much outside of approval ratings that would allow us to confidently predict the 2020 results. After controlling for presidential approval and economic conditions, Emory University Professor Alan Abramowitz found that first-term incumbents often perform better than non-incumbents. Abramowitz’s model has received some fair criticism, but the overall point is that incumbent presidents—including four of the last five, dating back to Ronald Reagan—often win re-election.
Beyond that, there’s not much that would help us predict 2020. The political and economic forces that shape elections are weird and unpredictable.
Trump’s current approval rating does matter in one regard: 2018. Presidential approval polls are our most direct measurement of how the public feels about the president, and they’ve been linked to how well the president’s party performs in midterm elections. If President Trump remains unpopular, Republicans will likely have trouble holding the House in 2018. Additionally, if Trump’s approval remains low, some congressional Republicans may distance themselves from the president and refuse to back his legislative pushes. That is why you should watch his approval ratings.
