Can elections be predicted like hurricanes or sporting events? American University professor Allan Lichtman believes so, and he has the track record to prove it. Using his unique “13 Keys to the White House” system, Lichtman has, according to him, correctly predicted the outcome of 10 out of the last 10 presidential elections.
In this episode of the WhoWhatWhy podcast, we talk with Lichtman to unpack his groundbreaking method, which goes beyond fleeting polling numbers to consider the complex interplay of 13 immutable forces that, he says, determine who will occupy the Oval Office. From economic conditions and social unrest to foreign policy successes and failures, Lichtman’s keys offer a tantalizing glimpse into the underlying dynamics that drive American politics and human behavior in general.
As the 2024 election looms on the horizon, with a Biden vs.Trump rematch, Lichtman shares his insights on how his keys are already aligning. While he’s not ready to make his final prediction just yet, he hints that the pieces are falling into place for a showdown that could once again defy conventional wisdom.
Apple Podcasts
Google Podcasts
RSS
Full Text Transcript: (As a service to our readers, we provide transcripts with our podcasts. We try to ensure that these transcripts do not include errors. However, due to a constraint of resources, we are not always able to proofread them as closely as we would like and hope that you will excuse any errors that slipped through.)
Jeff: Welcome to the WhoWhatWhy podcast. I’m your host, Jeff Schechtman. It seems that every day new polls try to tell us about the outcome of the 2024 election, but polls are at best a snapshot of a particular moment in time. They’re not predictors of what might happen weeks, months, or a half a year from now. For that, there is no algorithm, no AI, no mathematical formula. What there is is history and experience.
My guest, American University history professor Allan Lichtman, brings both to this exercise. He has correctly predicted the results of nine of the past 10 presidential elections. Using his unique formula of 13 keys to the White House. Professor Lichtman is gearing up for perhaps his greatest challenge yet, forecasting the outcome of Biden versus Trump round two. But he is not just playing soothsayer. His predictions are based on a carefully constructed formula that looks at factors like the state of the economy, foreign policy success and failures, social unrest, and the candidate’s charisma.
While he hasn’t locked in his final prediction for 2024 yet, the compelling evidence is already leading him in some very clear directions. We’re going to dive into his history of election forecasting, examining his 13 keys, and get a take on how the 2024 election is shaping up. It is my pleasure to welcome Professor Allan Lichtman back here to the WhoWhatWhy podcast. Allan, thanks so much for joining us.
Allan: My pleasure.
Jeff: Well, it’s great to have you here. Talk a little bit first about how this method began to evolve. The idea, not of polling, but of actually using specific things to begin to predict how elections might turn out.
Allan: I’d love to tell you I developed my prediction system, the keys to the White House, by ruining my eyes in the archives and deep contemplation. But if I were to say that to quote the late, not-so-great Richard Nixon, that would be wrong. I developed the keys serendipitously like so many other discoveries in 1981 when I was a distinguished visiting professor at the California Institute of Technology in Southern California.
And there I met the world’s leading authority on earthquake prediction, Vladimir Keilis-Borok. And he said, “You and I are going to collaborate.” And I said, being brilliant and foresightful, “No we’re not. Earthquakes may be a big deal here in Southern California, but I have to go back to Washington, D.C. where I teach at American University. Nobody cares about earthquakes there.” He said, “Oh no, I solved earthquakes.”
But get this, in 1963, Keilis-Borok was a member of the Soviet Scientific Delegation that came to Washington and negotiated by far the most important treaty in the history of the world. It’s the one that saved us from poisoning our atmosphere, our oceans, and our soil, the Nuclear Test-Ban Treaty. And he said, “In Washington, I fell in love with politics and always wanted to use the methods of earthquake prediction to predict elections. But look, I live in the Soviet Union. Elections? Forget it, it’s [the] supreme leader or off with your head, but you are an expert in American history and the presidency.”
So we became the odd couple of political research. And like any other breakthrough, we began by reconceptualizing presidential elections. Remember, this is 1981, not as Carter versus Reagan, not as Republican versus Democrat, not as liberal versus conservative, but in earthquake terms. As stability, the White House party keeps the White House, earthquake, the White House party is turned out. We then looked at every American presidential election from the horse and buggy days of politics in 1860 when Abraham Lincoln was elected right up to the election of Ronald Reagan in 1980.
And we used the methods of earthquake prediction, of pattern recognition to see what patterns in the political environment were associated with stability and earthquake, following my insight that presidential elections are primarily votes up or down on the strength and performance of the White House party. And that historical analysis led us through the pattern recognition to the development of the 13 keys you outlined and a simple decision rule. The keys are always: they’re true/false questions always phrased so an answer of true favors stability, an answer of false favors earthquakes.
And if six or more of the keys go against the White House party, you have a predicted earthquake, fewer than six, you have predicted stability. And I’ve since used that key ever since I predicted Ronald Reagan’s reelection in April 1982, more than two-and-a-half years ahead of time. Now you may say, “Oh, Ronald Reagan’s reelection everyone knew that.” Yes, in retrospect. But in 1982, America was in the worst recession since the Great Depression to that point and Ronald Reagan’s approval ratings were no better than Jimmy Carter’s at a comparable point in his presidency, and everyone was talking about a one-term president.
Jeff: Talk a little bit about how the landscape has changed, though. We’re looking at a period now where the media landscape has changed, the way people get information has changed, and we’re looking at an election essentially which some would argue of two incumbents. Talk about how that impacts what it is, and how you look at this.
Allan: None of that impacts anything at all in terms of the keys. The keys have been like the North Star of politics. They have been unchanged since Keilis-Borok and I developed them in 1981. Look, retrospectively, the keys go back to 1860. There have been enormous changes in our economy, in our society, in our demography, in our politics. Back then we didn’t even have automobiles, much less jet planes or computers or polls. Women couldn’t vote, most Blacks couldn’t vote.
My ancestors from Eastern Europe aren’t even here yet. We were an agricultural society. So if the keys have held up since 1860, they certainly would hold through the kinds of changes you are talking about, which are less monumental. Look, every four years, someone tells me [you’ve] got to change the keys. We have an African American running now, never had that. We have a woman running now, never had that. We had social media, never had that.
Of course, in 1860, we didn’t even have radio and television. And my answer is always the same. You cannot change a model on the fly. That is a recipe for making errors. And that’s why I don’t change my keys in response to what people say are the unique circumstances of every election. Yes, every election is unique, but every election has the same underlying structure as modeled by the keys.
Jeff: Talk about how voting methods impact this. The concern today about how people vote, the different methods by which they vote, early voting, the way so many things in that regard have changed.
Allan: Doesn’t affect the keys at all. Look, back in 1860, we didn’t have the secret ballot. People voted openly on either ballots color-coded by the parties, or actually through voice voting. And again, my key to go all the way back to that period. And the changes from 1860 to 1980 again are far greater than any of the changes in recent years. That’s why the keys have been so successful.
And I have to make one point. I’ve predicted 10 out of 10, not nine out of 10. [The year] 2000 was a stolen election. The wrong man was elected president. My prediction was right. When the Supreme Court stopped the voting in Florida that gave Bush the electoral college, he was ahead by 537 votes. But in fact, Al Gore should have won that election going away by tens of thousands of votes based on the intention of the voters. As I proved in my report to the United States Commission on Civil Rights, the wrong man was elected president because one out of every nine to 10 ballots actually cast by African Americans were rejected by election officials as compared to just one out of 50 ballots cast by whites.
If the rejection rate for Blacks was not zero, but just equal to that of whites, Al Gore would’ve won by more than 40,000 votes. And most of those rejected Black ballots were not the hanging chads or the dimples. They were so-called overvotes where Blacks were told [to] punch in Gore, but just to make sure write in Gore and all those ballots were invalidated. And subsequent studies by political scientists have validated my point. So I would say 10 out of 10, not nine out of 10.
Jeff: Talk about the role of the economy. We hear over and over again how important economic conditions are. How do they play within the context of these 13 keys?
Allan: Well, the economy counts for two keys. The short-term economy is the economy in recession during the election year, and the long-term economy [is] the real per capita growth during the term compared to the previous two terms. So two of my 13 keys relate to the economy. We’re all caught in the same error that Bill Clinton led us to. “It’s the economy stupid.” Not always. The economy is important. It’s two of my keys, but it’s not the only thing.
One of the reasons so many pundits got 2016 wrong, it was a strong economy, and based on the economy, Hillary Clinton should have won going away. One of the strongest economies we’ve ever had was the economy of the 1960s. Based on the economy, Hubert Humphrey should have beat Richard Nixon by a landslide in 1968. So the economy counts, but it’s not the only thing.
That’s why the keys are so successful because as you outline in the beginning, they look at a broad range of indicators, third parties, incumbency, internal party contest, social unrest, scandal, policy change, foreign policy successes and failures. And the only keys that apply to the candidates at all are very high threshold keys, asking whether the incumbent party candidate is one of those once-in-a-generation, inspirational candidates. And of course, asking whether the challenging party candidate is not one of those truly once-in-a-generation, Ronald Reagan, Franklin Roosevelt inspirational candidates.
Jeff: Are the keys weighted differently in different elections?
Allan: Great question. Another one of the secrets to the success of the keys as opposed to all the other models, all of which create errors, is the keys are unweighted. The problem with weighting is obviously the weights have to be based on past elections. But for a future election where you’re predicting but don’t know the answers, the weights can change unpredictably leading to errors.
But another secret to the success of the keys raised by your question about the weights is called trigger effect. That is, if one key is strong enough, it can trigger other keys. Some people say to me about the economy, “Are you telling me the Great Depression of 1932 when FDR beat Herbert Hoover, only counted for two keys?” And my answer is, directly, but it triggered other keys. It triggered social unrest, the bad economy. It triggered incumbent party losses in the midterm elections of 1930. It triggered a charismatic candidate, Franklin Roosevelt, coming out to run four years before he thought he would, or the Vietnam War in the 1960s triggered social unrest. It triggered Lyndon Johnson, the incumbent president, to withdraw from the election, losing the incumbency key, and it triggered a big contest for the incumbent party’s nomination losing another key.
Jeff: Talk about how these keys lined up when you were looking at them in 2016.
Allan: The critical factor, the crucial sixth key that led me to turn the election against Hillary Clinton, and obviously by default for Donald Trump. Nothing to do with Donald Trump. Any generic Republican would’ve been a predicted winner. The critical key was Bernie Sanders’s contest against Hillary Clinton, which turned the sixth and fatal key, internal party contest.
Other keys, of course, that the Democrats had already lost were incumbency since it was an open seat, mandate since they lost the midterm elections of 2014, they didn’t have a big foreign policy success in the second term to follow up the killing of bin Laden in the first term, or a big domestic policy achievement, policy change to follow up the Affordable Care Act in the first term.
Jeff: To what extent is incumbency critical now? I know you said the keys aren’t weighted, but how different is it when there’s an incumbency versus an open seat?
Allan: Incumbency is very important. Again, no key is weighted, but typically when you have an open seat, obviously that loses the incumbency key. It also typically triggers a loss of the contest key, as we saw in 2016 or 2008 for the Republicans. That is, if there isn’t an heir apparent, you’re going to get a contest for the incumbent party nomination. So that’s two keys down.
By the way, this is very relevant to what’s going on now. You heard all the pundits saying, Biden’s too old. He should step down. There should be someone younger, as if there’s some Democratic chieftain then up in the sky who could dismiss Biden and pick some younger candidate. I have no idea who it would be, but in fact, like so much else of punditry, that leads us down to the primrose path of error.
Biden running, as we saw, secures the incumbency key and the contest key. That means of the remaining 11 keys, six more would have to fall to predict a Democratic defeat. Biden doesn’t run, they lose the incumbency key, and of course, it would trigger a loss of the contest key since there’s no heir apparent. That’s now two keys lost off the top, only four keys would have to fall to predict a Democratic defeat.
Jeff: Are we looking at this race, this current race 2024 differently with respect to incumbency? Because we basically have two presidents running.
Allan: Well, not really. Incumbency only counts for one key. Donald Trump does not win an incumbent’s key. There’s no winning keys for the challenger. All the keys are counted for and against the White House party. Now, where being a former president could in fact influence the election if as a result of being a former president, it turned the charisma key, key 13, against the White House party because Donald Trump had so much luster as a former president that he emerged as a broadly inspirational candidate. That hasn’t happened.
Donald Trump appeals to a very narrow base and to turn the charisma key against the White House party, you have to be broadly appealing. Like Ronald Reagan, who won two landslide elections and created a lot of Reagan Democrats, Trump lost two elections by a combined 10 million votes. Essentially a record number and I’ve yet to find [chuckles] more than a tiny handful of Trump Democrats. So being a former president hasn’t done Trump any good in turning key 13 against the White House party.
Jeff: To what extent does it work its way into these keys when we look at things like political realignment and demographic shifts?
Allan: Not at all. There have been enormous demographic shifts since 1860. As I said, people from Eastern Europe weren’t here yet. People from Latin America and Asia weren’t here yet. That is a massive, enormous demographic shift. Not to mention the shift from an agricultural then to an industrial, then to a post-industrial economy. Not to mention shifting from no women voting to women’s suffrage or virtually no African Americans voting to African American suffrage. And the keys have held through all those changes.
Jeff: Talk a little bit about the way 2024 is lining up relative to the keys as you see it.
Allan: Right now Biden is only definitively down two keys, the mandate key based on losses in the house elections of 2022, and the incumbent charisma key because he’s no Franklin Roosevelt. There are only four other keys that are shaky. Third-party, and let me explain that. To turn that key, the third-party candidate has to get at least five percent of the popular vote, but I don’t know, obviously, until after the election. So predictively, it’s the only time I have to rely on polls, but I don’t take them as face value.
To turn that key, the third-party candidate has to be consistently polling as we get closer to the election at 10 percent or more. It’s the Lichtman rule of halves or the wasted vote syndrome. I love you Ross Perot, but you can’t win, so I’m not going to vote for you. I don’t think RFK Jr. is going to consistently poll at 10 percent or more. I think he only goes down as we get closer to the election, but I haven’t called that key yet.
I thought social unrest was safe until last week when these campus protests emerged. They’re nowhere near the anti-Vietnam War protests in the ’60s. The biggest enemy to the Democrats is not Donald Trump, it’s not Marjorie Taylor Greene, it’s Bernie Sanders. It was Bernie Sanders who cost the Democrats the re-election in 2016.
Now it’s Bernie Sanders who is utterly misrepresenting these campus demonstrations, comparing them to the anti-Vietnam War demonstrations when millions upon millions of people were out in the street, not just a few thousand demonstrations that looked like they are going to fizzle. I don’t understand Bernie Sanders, why he seems constantly intent on torpedoing the Democrats as best he can. But that key has not been called yet, but suddenly it’s shaky. And obviously, the foreign policy success and failure keys are shaky depending on what happens in the Middle East and Ukraine.
Jeff: And at what point does this solidify? How far out from the election can you comfortably make these kinds of predictions?
Allan: There’s no set answer. Sometimes the keys fall into place early. In the very hard-to-call 2012 election, when a lot of pollsters and pundits were calling it very late for Romney, I called it for Obama in 2010 because I saw the keys already falling into place in favor of his re-election.
I then got out of the blue, a 30-page attack from none other than the compiler of polls, Nate Silver, saying, “No, no, you can’t predict this early.” Of course, being a professor, I wrote a 30-page response, which boiled down to, “I can because my system is based on how elections really work. You can’t because you’re based on polls, and polls are meaningless this far out.” Even closer, they’re meaningless.
Ultimately, close to the election, Silver came around to my point of view. I wrote him an email saying, “Let’s do a joint article on how two authorities with fundamentally different methods came to the same answer.” Never heard from him. Needless to say, he’s not my favorite person, and I did not shed a single tear when, quite rightly, he got bounced out of FiveThirtyEight and is no longer apparently doing election work.
Sometimes it falls into place early, sometimes very late. I didn’t call 2016 until September, and I didn’t call 2020 until August, and I don’t expect to call 2024 until August.
Jeff: What role, if any, do polls play in your analysis of the things that get you to these keys, whether it’s polling on the economy, polling on social unrest, polling on various issues that may be different than just polling with respect to the candidates?
Allan: I never poll on issues. That, again, leads you down to the primrose path of error. When Ronald Reagan won nearly every state in 60 percent of the vote in 1984, he was pretty well tied on the issues with Walter Mondale. Hillary Clinton was ahead on most issues over Donald Trump. Issues cannot predict, just as horse race polls cannot predict. By the way, you absolutely are right in saying polls are just snapshots and are abused as predictors, but it’s worse than that.
They tell you the typical error in polls is a random plus and minus three percent. Wrong. That’s [a] statistical error. That’s the error you would get if you had a huge jar of red and green balls and you pulled out a sample and you estimated the percentage of red and green balls in the jar. But human beings are not red and green balls. The election is not held today. That’s meaningless. They may not have focus. They may change their mind. They may lie. Plus, no one’s voted, so they’ve got to guess who the actual voters are going to be. They call it likely voters. It’s a pure guess.
And this other error is not random, it’s directional. In 2016, the error underestimated Republicans, and recently it underestimated Democrats, based on the 2023 off-year elections and the special elections. We had two of them in New York, one for the seat out of which George Santos got bounced. A poll just a couple of days before the election had it dead even, no daylight. The Democrat won by eight points.
Just had another special congressional election in New York this week, and the latest polls had the Democrat ahead by nine percent. He won by 36 percent. So not only do the polls have much bigger errors than they tell you, the error is not random. So yes, you can use polls to some extent for things like responses to what’s going on in Gaza and Ukraine. That’s not determinative, but at least that can give you some insight into foreign policy, success and failure, along with the substance of what’s happening. [But it] can’t be just polls.
Jeff: Do these keys apply to other races other than presidential races?
Allan: I tried that. I developed a set of keys, one set for all 50 states, keys to the Senate. I was right about 85 percent of the time. But what do you think everyone focused on? The 15 percent [I got wrong]. So I gave up. I decided I did not want to tarnish my perfect record in presidential prediction. But maybe someone else could try, but I’m out of that business. It’s very hard to have one set of keys to 50 states, and certainly not to 435 congressional elections.
Jeff: Given that events happen at greater speed today, and given that voting is getting earlier and earlier when the process of voting starts, does that matter?
Allan: No. I can’t stress this too strongly. None of these things matter. Nothing matters unless it affects a key. Now, let me then back off of that a little bit. I’m not psychic Jeane Dixon. I’m not a crystal ball. I’m not Mike Johnson, the Speaker of the House who claims the Almighty speaks to him. My keys are based on history, and as I’ve explained, they’re incredibly robust, going back retrospectively to 1860 and prospectively to 1984.
But it’s possible that there could be such an unprecedented cataclysmic change in historical patterns that it could affect an election outside the keys. It hasn’t yet, but we do have something going on right now. Donald Trump sitting on the dock in a criminal trial, charged with 34 felonies. And the media, again, has this all wrong. It’s not a hush-money trial. Donald Trump is not on trial for paying hush money. It’s an election fraud charge. He’s on trial for falsifying his business records to conceal the payment from the American people and the tax collectors. That’s the crime.
And the media is also off base and stressing, “This is the first time a former president has been on trial criminally.” It’s much bigger than that. It’s the first time any major party candidate, former president or not, has been on trial for a criminal offense. So assume, hypothetically, Donald Trump is convicted of 34 felonies. That may be an event so unprecedented, so cataclysmic, that it could affect the election. It could also affect a key. For example, it might discourage voting for a third-party candidate because people think, “Well, maybe that’s going to help this convicted felon.”
Jeff: Well, I suppose we’ll revisit with you in August when it comes down to your final predictions. American University professor Allan Lichtman, I thank you so much for spending time with us today here on the WhoWhatWhy podcast.
Allan: Thank you for an extraordinary insightful interview. I appreciate it.
Jeff: Thank you. And thank you for listening and joining us here on the WhoWhatWhy podcast. I hope you join us next week for another radio WhoWhatWhy podcast. I’m Jeff Schechtman. If you like this podcast, please feel free to share and help others find it by rating and reviewing it on iTunes. You can also support this podcast and all the work we do by going to whowhatwhy.org/donate.
