This is an archived article and the information in the article may be outdated. Please look at the time stamp on the story to see when it was last updated.

On Thursday, Mitt Romney made a bold prediction.

“I think President Trump will be re-nominated by my party easily, and I think he’ll be reelected solidly,” Romney told big-dollar donors at a retreat in Deer Valley, Utah. “I think that not just because of the strong economy and because people are increasingly seeing rising wages, but I think it’s also true because I think our Democrat friends are likely to nominate someone who is really out of the mainstream of American thought and will make it easier for a president who is presiding over a growing economy.”

Yes, it is somewhat ironic that Romney is predicting future success for Trump given that the party’s 2012 presidential nominee gave a speech almost two years ago exactly in which he said this: “Donald Trump is a phony, a fraud. His promises are as worthless as a degree from Trump University.”

But, that flip-flop is easily explained by the fact that Romney wasn’t running for anything in 2016. Now he’s running for the Senate — and he doesn’t want to antagonize the Trump base, which remains extremely loyal to the President. Politicians do political things. Dog bites man.

What I’m much more interested in is not why Romney made the 2020 prediction about Trump but rather, based on history and what we know about Trump as of today, whether he’s right. Is Trump a lock to be the GOP nominee — and a solid favorite to win a second term?

The first question is easier to answer than the second. And the answer is: Yes. History suggests that incumbent presidents in the modern era, even when they do face a serious primary challenge, always win their primaries.

That doesn’t mean there haven’t been close calls. In 1980, Ted Kennedy started as the frontrunner in the primary contest against President Jimmy Carter but wound up losing — largely the result of an inability to articulate why he actually wanted to be president. Four years prior, Ronald Reagan ran an ideological challenge to President Gerald Ford’s right but came up just short. The Reagan challenge damaged Ford and fueled Reagan. The former lost his reelection bid to Carter, the latter parlayed his increased profile into the 1980 GOP nomination and, eventually, two terms as president.

There’s little polling evidence currently available that suggests Trump would face a serious challenge from within his own party in 2020. He remains extremely popular within his own party — despite his middling approval numbers in the general electorate; in CNN’s May national poll, 86% of Republicans approved of the job Trump was doing while just 9% disapproved.

That doesn’t mean that Trump will avoid a GOP primary challenge — Ohio Gov. John Kasich and Arizona Sen. Jeff Flake both seem serious about challenging the incumbent. It just means that he isn’t likely to lose to them or, really, any Republican.

Now for the tougher part of Romney’s prediction: That Trump will win the 2020 general election “solidly.”

Much of Romney’s calculus is based on his belief that Democrats will ultimately choose “someone who is really out of the mainstream of American thought.” And they might! But the 2020 Democratic field is totally in flux at the moment and it’s literally impossible to anticipate who might eventually be the nominee and why.

Which leaves us with looking back at the history of past presidents at this time in their tenures — and comparing how Trump stacks up. There are a few measures I took a look at to see if any/all of them were predictive.

1. Electoral votes in first presidential election: There’s no real evidence that how many electoral votes you take in your first go around indicates anything about your chances of winning a reelection race. George W. Bush famously won only 271 electoral votes in 2000 but won a second term four years later. His father, George H.W. Bush, claimed 426 electoral votes in 1988 only to lose to Bill Clinton four years later. Just for kicks: Trump’s 304 electoral votes in 2016 is the third lowest total in a winning presidential bid dating back to Jimmy Carter. (Bush’s 271 is the lowest, followed by Carter’s 297 in 1976.)

2. Job approval numbers: Trump’s job approval in the latest Gallup poll is 41%. That’s lower than any of the past six presidents at the 500-day mark of their tenures. Here’s where each of them stood in Gallup: Obama (47%), George W. Bush (74%), Clinton (46%), George H.W. Bush (67%), Ronald Reagan (45%) and Carter (44%).

Throw the W. Bush number out because his approval ratings were hugely inflated in the aftermath of Sept. 11, 2001. But even with that aside, there’s no obvious trend — George H.W. Bush was the most popular president of the last six at this point in his term and he went on to lose. Reagan and Carter had very similar approval numbers at this point; the former carried 49 states in his reelection race, the latter lost a bid for a second term.

3. Unemployment rate: In May, the unemployment rate dipped to 3.8% — the lowest it’s been since 1969. In fact, the current unemployment rate is half the 9.3% it was in Obama’s second year in office. And it’s nowhere near the 10.8% rate for Reagan in 1982. Of course, Reagan got re-elected. And Carter, who had a 6% unemployment rate in 1978, lost. What’s more important, traditionally, when it comes to the unemployment rate is the trendline. If the unemployment rate is headed downward, presidents usually benefit. If it’s going up, it’s usually bad news.

The simple fact is that there is no hard-and-fast — or even close — indicator that, two and a half years out from a reelection race, can tell us whether a president will win. Presidencies are heavily affected by circumstances out of their control — unforeseen events that shape history and a president’s role in it.

So, Mitt may be right. Or wrong. We just won’t know which for another few years.