News & Politics

He Predicted Both Trump’s Election and Impeachment. What Else Does He Know?

A conversation with Allan Lichtman, who's correctly predicted every election since 1984.

(Photo by Allan Lichtman)

UPDATE: On Aug. 5, Lichtman revealed his prediction for the 2020 election, saying in a video that Trump will lose to Biden, according to his prediction system.


The kingdom of political punditry has always been divided between its two unruly houses: The tweed jacket set, holding forth on history and the political mood from vaunted column inches, and the self-styled data gurus, who claim to prefer spreadsheets to soliloquies.

Straddling this breach is Allan Lichtman, a history professor at American University, who occupies a more uncertain dominion. Lichtman has predicted the correct outcome of every presidential election between 1984—where he made his first prognostication (a Reagan landslide) in the pages of Washingtonian—to 2016, offering the very lonely opinion that Donald Trump would narrowly vanquish Hillary Clinton.

To make these predictions, Lichtman uses a quasi-scientific method he calls the “thirteen keys”—thirteen factors, like the health of the economy and “social unrest”—that add up to determine whether the incumbent can hold onto the White House.

Licthman is not a political scientist—he’s a historian—so he doesn’t harmonize with the number-crunchers of the predictions world. But neither is Lichtman an academic’s pundit, who would typically steer away from bold declarations about the future. He has a flare for showmanship that can be off-putting to critics, and has come in for grief for inviting liberal bias into his scholarship. A misfit in either camp, Lichtman is still largely ignored by both—his primary suitors today are media personalities, while political wonks and academics still tend to cast his work as “suspicious” and “hyperbole.”

And yet, doubtless to the annoyance of many, Lichtman continues to get it right. Plus, his winning streak now includes a new mark: His prediction that Donald Trump would be impeached, which he made in 2016.

In a conversation edited for length and clarity, Lichtman discussed the Trump era, defended his use of history as a guide for prediction, and turned his oracular powers to the future.

Here we are, with Trump on the verge of impeachment. I admit that I actually dismissed your prediction as attention-seeking. Has anyone else sent letters of contrition?

No one’s coming to me with mea culpas for doubting my prediction that this president would be impeached, which I actually made at the same time I predicted his election, in September of 2016. So I simultaneously predicted his victory and his impeachment. And believe it or not, after the election, I did get this note on the Washington Post interview where I predicted Trump’s election. It said, “Congrats, Professor—Good Call.” And in big sharpie letters was the name Donald J. Trump.

Where did he send it?

My home address. But of course, since the president doesn’t read, he probably didn’t read far enough into my Post interview to see the next big prediction.

It’s like a Greek tragedy.

It has the elements of that. I think it’s more tragic than anything else. But mostly what people have come to me saying is, Wow, how did you ever predict this impeachment more than three years ago? I don’t have a formula for predicting impeachment. But I had studied his entire career, and his campaign, and I concluded this was someone who had no respect for the law, no respect for the truth, who had spent his entire life enriching himself, and promoting himself, often at the expense of others. This was someone who had never been held accountable for anything he had done, not for discriminating against minorities, not for hiring illegal immigrants, not for declaring multiple bankruptcies, not for lying about President Obama’s actual place of birth—nothing. And I saw the ways in which he had cozied up to the Russians in the 2016 campaign.

And I concluded that all of these traits were exactly the kinds of characteristics that would lead a president to be impeached, that he wouldn’t change any of his ways once he became president.

Your decision to predict Trump’s impeachment must have seemed like a leap of faith—

Not faith. It wasn’t faith.

Well, a leap of faith in the sense that, for a historian, it’s quite a big risk. In September 2016, you had no idea who would control Congress in four months, much less 2018.

If you’re going make your name as a predictor, you’ve got to take risks. Otherwise, if you’re only making safe predictions, you’re worthless as a predictor. For example, in 1988, when Michael Dukakis was 18 points ahead of George H.W. Bush, I predicted that Bush would beat Dukakis. And I took a load of grief for that prediction—people writing to the president [of the university], saying, Why’d you hire this guy? People asking, Where did you get your degree Dr. Lichtman, mail order? So I’m used to making these risky kinds of predictions.

And by the way, it was contingent on a lot of factors. I did believe back then that the the Democrats would retake the Congress in the midterms, because midterm elections in particular in recent years have been very negative for the party holding the White House. So yes, there were a lot of moving parts, but it was a very risky prediction. No risk, no gain.

Your book, “The Case for Impeachment,” makes an empirical argument against Trump and treats that as the basis for impeachment. And you got some pushback, I’m thinking particularly in the New York Review of Books, because legal scholars said that a lot of your arguments just weren’t plausible reasons to impeach someone.

Yes, we had a nice debate.

But it’s true that none of the reasons you offered are actually the reasons Trump is being impeached. I mean, I’m sure there are people who wish he was impeached for the Paris Accords. But he’s not. So would it be fair to say that you were right, but for the wrong reasons?

The most overlooked and serious case for impeachment is that Trump is putting the survival of humanity at risk, by not just halting, but throttling back efforts to combat catastrophic climate change. At a time when every scientific analysis is warning that we are at the brink of true existential disaster for all of humanity. I think that is a crime against humanity which he should be impeached for. And indeed, the International Criminal Court, we’re not subject to its jurisdiction, but it makes crime against environment one of the crimes against humanity. But just as Richard Nixon was not impeached for his worst crime, which was the illegal war that killed more than one hundred thousand people in Cambodia, I don’t think Trump will be impeached for his worst crime.

And while of course I could not have foreseen the particular events in the Ukraine, the characteristics that I pointed out—the concern only for his own interests, disregard for the law, disregard for the truth—the things I highlighted in my book, are exactly the traits that led him quite gratuitously to bring this on himself through his Ukrainian scandal. Which, by the way—people are not aware of this—goes far beyond just shaking down, and I think extorting and bribing, the Ukrainian president. It involves the setting up of a whole rogue private foreign policy operation, accountable to none of the regular lines of authority, and that’s operating independently, directly contrary to American interests, and apparently also operating for the financial interests of those involved, notably the guy heading it up, Rudy Giuliani and his associates.

Last point—who benefits from declining a relationship, a troubled relationship, between the U.S. and the Ukraine? The Russians. And by the way, who benefits from Trump’s policy in Syria? The Russians. So I think, the things I pointed out in my book, while not specific to these events, certainly explain them. Including my chapter on the Russia connection.

How did all this start? What was it about predictions that tugged on your sleeve?

First of all, historians don’t make predictions, typically, historians write about dead people, so they don’t have to take any risks. I didn’t think about being a predictor when I first got my Ph.D at Harvard in 1973. I’d love to tell you I came across my prediction system through years of ruining my eyes in libraries, archives, year of deep contemplation. I became a predictor, totally by accident, in 1981. I met another visiting scholar, Vladimir Keilis-Borok, who was the world’s leading authority on earthquake prediction. We became the odd couple of political predictions. We developed thirteen keys to the White House by looking back from 1860 to 1980, using the methods of earthquake prediction, which were pattern recognition, to see what patterns were associated with victory and defeat to the incumbent party. In other words, the key to the system is reconceptualizing presidential election in geophysical terms. Stability is the party holding the White House wins. The earthquake is they lose. Our system of 13 keys is not random, but based on the proposition that American presidential elections are essentially a vote up or down on the strength and performance of the party and power. And virtually all our keys gauge those factors.

What I want to ask is how that deals with the nature of chance. After 2016, there was this big reckoning about what percentages really mean. If Trump has a 33 percent chance of winning, what does that mean? And the same idea with Jane Mayer’s work [on the scholar Kathleen Hall Jamieson], which gauged the chances that Russian interference made a difference in the election.

My systems is totally different from what 538 does, Nate Silver, the guys at Princeton, their consortium. What they do is just compile polls, and polls aren’t predictors. Polls are just snapshots. They don’t predict anything. Because polls always change, and they keep changing. They just compile polls. That’s why they were so wrong in 2016. My system is based on a theory of how presidential elections actually work. Not just a mindless compilation of polling data. Secondly, unlike any other system, I outline in each of the thirteen keys what has to happen during the course of a presidential term. Third, my system suggests a new way of understanding campaigning. It says campaigning as usual doesn’t matter. The negative ads, the cheap soundbites, the attack strategy, the campaign tricks.

Really? What about all the analysis that suggests Comey—




—that Comey basically swung the outcome of the election with his last-minute announcement about Clinton’s emails. And that research suggests that if this chance thing hadn’t happened, the election would have gone differently.

When you looked at flawed data, you reach flawed conclusions. The fact that I was able to predict it before James Comey should tell you everything you should need to know. All of these retrospective analyses are  worthless. I’ll give you a good example. After Mitt Romney lost in 2012—when a lot of people said he was going to win after he creamed Obama in that first debate—there were all of these post-election analyses, half of which show Romney lost because he was too conservative, that he should have been more to the middle. The other half said data showed Romney lost because he wasn’t conservative enough, he didn’t fire up the conservative base. These post-hoc analyses are worthless, unless you have a track record of predicting elections.

Sorry, but I just have to push back on this. There’s no chance event during a campaign that could change outcomes? Nothing at all? Someone gets assassinated?

Look, I’m not God. I don’t’ have a pipeline to the almighty. I’m not a psychic genius with a crystal ball. I only base my predictions on history. All I can tell you is that this is an incredibly robust system. It’s held up retrospectively for 220 years, and prospectively for 36 years. Of course some out-of-the-blue, black swan event like an assassination could happen. But short of that, the system is robust and it’s held up every year.

Have you made a prediction about 2020 yet?

No, I have not.

What about impeachment? Do you think that will affect the outcome in 2020?

Absolutely—the “scandal” key. I wrote several articles saying that the one thing Democrats could do to help themselves, and in fact it’s probably essential—necessary, if not sufficient—would be to turn the scandal key, by making Donald J. Trump only the third president in U.S. history to be impeached by the full House. We’re looking at foreign policy failure, perhaps in Syria. And then looking to the future: Is the economy going to turn sour, and turn another key? If Trump is battered enough in the impeachment process, will there be a real challenge—a third party? Will a charismatic challenger like Barack Obama emerge in the primary process? These are the keys to look for. And they’re all laid out. All you have to do is read my book. The polls won’t tell you what to look for. Nate Silver won’t tell you.

Benjamin Wofford
Staff Writer

Benjamin Wofford is a contributing editor at Washingtonian.