Donald Trump and Kamala Harris appear on track to be on presidential ballots come Election Day. Daily polls suggest that this will be a close race. Yet Allan Lichtman, the noted historian and election prognosticator, is unimpressed with polls as predictive tools (he is mostly right). Instead, he relies on 13 keys to predict the winner of the election.
Lichtman (a historian and long-standing Democrat) has been vociferous in criticizing the Democratic Party for abandoning President Biden as their candidate, given that it flips one additional key to false (incumbency), providing additional headwinds for Democrats to retain the White House.
Is he correct? Certainly, based on the definitions of the 13 keys. If five or fewer keys are false, then the incumbent party wins the election. With Biden gone, the Democratic Party self-inflicted a false key.
Using a baseball “money ball” metaphor, giving away a key is like a sacrifice bunt or trying to steal a base. Actions that give the opposing team an out is not smart analytics.
But what is being gained by giving away a false key?
With Biden gone, voters have what they have asked for, namely, an alternative to a Trump-Biden rematch. Will this suppress interest in Robert F. Kennedy Jr., ensuring that the third-party key is firmly true?
All the other keys appear unaffected with Biden off the ticket.
The 13 keys have been enormously successful in forecasting presidential elections. The approach uses pattern recognition to match what has occurred in the past with what is occurring in the present to infer how the incumbent party should fare in the election. In today’s vernacular, this is an application of artificial intelligence, with past elections serving as data by which the pattern recognition model learns.
In reviewing the 13 keys, seven have likely already been determined (1: Party Mandate, which is false; 2: No primary contest, which appears to be true based on Harris’ rapid and uncontested rise to the top of ticket; 3: Incumbency, which is now false with Biden dropping out of the race; 5: Economy not in recession, which is true; 6: Long-term economic growth greater than previous terms, which is true; 12: Charismatic incumbent, which is false; 13: Uncharismatic challenger, which is true).
Republicans may disagree with me about keys 5, 6, and 13, but those involve more political posturing than objective facts.
That leaves six keys, of which at most two can be false for the Democrats to retain the White House.
Though the keys have been remarkably predictive, the original paper, published by Lichtman and Keilis-Borok in the Proceedings of the National Academy of Science in 1981, gives two disclaimers.
“We neither claim that other parameters cannot be used for the same purpose nor suggest methods for predicting future elections.” The first point is that there may be other parameters that may also align with historical outcomes of presidential elections. The second point is that the keys were not originally designed to be predictive.
This means that other factors may contribute to winning the White House that are not included in the 13 Keys. For example, whether a candidate has been convicted of a felony has never been incorporated into the keys, with Trump holding this distinction.
Presidents who have been impeached have never been elected to a second term (Andrew Johnson, who could not even earn the Democrat nomination in 1868, Bill Clinton, who was impeached in his second term, and Trump, who was impeached twice). The age of a candidate has also not been considered as factor, with Trump now topping the list.
Any such other factors could either replace or more likely supplement the existing keys.
Note that the original paper contained 12 keys, not the 13 used today. Lichtman has co-authored several books since then, refining his approach and giving political pundits plenty of ammunition to debate about the current race.
Whether times have changed sufficiently to suggest that the 13 keys have lost their luster is impossible to assess. With Trump and Biden, they both possess serious flaws that alienate different factions of voters, forcing some to either vote against the candidate they like the least, or simply staying away and not voting. With Biden off the ticket, Harris is now an alternative that did not exist just a few weeks ago.
The 13 keys are simple, direct and easy to understand. A superior alternative has yet to be discovered and demonstrated. Polling is a sampling nightmare, trying to align those polled with those who will actually vote. The shortcomings of polling were most recently on full display in 2016.
History suggests that the 13 keys work, even if they do not fully capture all the factors that may go into predicting who will win the White House, as Lichtman himself acknowledges in his original paper. How this year’s election, and the candidates, measure up by the voters remains anyone’s guess. Yet taking into account all the yet-to-be-called keys suggests that Harris’ lucky number may very well be 47.
Sheldon H. Jacobson, Ph.D., is a professor of computer science at the University of Illinois Urbana-Champaign.
