With the US presidential election in just a few weeks, several pollsters are busy conducting opinion polls and analysing them. Political polls, however, are not necessarily reliable indicators of election results, especially in a close contest.
On the other hand, Allan Lichtman, a history professor at American University in Washington, DC, has made a straightforward yet almost Nostradamus-like prediction about the US presidential elections that has, thus far, had a 90% accuracy rate.
Well, how many pollsters can boast such a coveted prediction record? Using an unconventional approach he and his geophysicist friend Vladimir Keilis-Borok developed in 1981, Lichtman disregarded opinion polls and pundits in favour of examining the fundamental issues that, according to them, have shaped every presidential election since 1860.
Since 1984, his prediction system, “The Keys to the White House,” has been evaluating the strength and performance of the incumbent party and predicting the outcomes of US presidential elections using 13 historical factors, or “keys” (four political, seven performance, and two personality).
The Gore miss
He has been right in all except one of his predictions, which was George W. Bush’s contested triumph against Al Gore in 2000. Lichtman, however, contends that even that stain is unfair, pointing out that thousands of votes that were disallowed came from people who had honestly attempted to support Gore but unintentionally spoilt their ballots. Thirteen broad, true-false questions that gauge the effectiveness of the White House party are the keys.
The party mandate, White House scandal, incumbency, primary contest, third party, shortand long-term economies, policy changes, incumbent and challenger charisma, foreign policy success and failure, and social unrest are the keys. Lichtman’s model predicts that the incumbent party’s candidate will lose if six or more keys are against them.
And, in early September, Lichtman predicted Kamala Harris would win the 2024 presidential election this time around; eight of the 13 keys supported her. Lichtman argues that the Democrats gained House seats even though they lost seats, so the midterm gains key is false for Harris. The incumbent key is also false because Biden withdrew from the contest.
It is to be noted that Lichtman had previously angrily objected to replacing Biden as the Democratic nominee following his appalling performance in the June debate and rejected the veracity of opinion polls suggesting that Biden’s chances of winning had been harmed.
Harris’s chances
According to Lichtman, Harris benefitted from the lack of a formidable third party due to the withdrawal of Robert F. Kennedy Jr.’s independent campaign, favourable short- and long-term economic indicators, significant legislative accomplishments carried out by the Biden administration, and the absence of social unrest or scandal surrounding the White House.
In addition, she had the advantage of not having to fight for the party’s nomination to replace Biden, as she gained the support of other contenders before the Democratic National Convention last month. Even if two still unresolved keys on foreign policy achievements or failures fell in Trump’s favour, they would not be enough for him to win the election, according to Lichtman’s formula. Lichtman, however, ignores polling trends, campaign tactics, and events.
Even an event like Taylor Swift’s endorsement has little bearing on Lichtman’s 13-point model. However, that can shape public perception and even voting turnout. “None of these ephemeral events, not the debate, not the purported attempted assassination, not JD Vance saying he made up a story about immigrants eating cats and dogs... None of that changes the fundamentals of the election,” Lichtman said. But is America really ready to elect a woman of colour to the presidency and create history? Is it really that simple this time?
“My prediction is that Kamala Harris will win in 2024, but everyone still needs to VOTE,” Lichtman stated recently. It should be mentioned that despite his success, Lichtman’s methodologies have frequently been questioned by political scientists, pollsters, and other historians who claim that his model is unscientific and too simplistic.
And, to be honest, assessing some of the keys is quite subjective. And one wonders, amid the waves of opinion polls, if similar “keys” may be created for the Indian elections as well – possibly by sifting through historical election data?
Given the complex, heterogeneous, and multifaceted nature of our elections in a truly multi-party system where various strong regional parties and various types of pre- and post-poll alliances exist and undergo constant evolution, that may be a more daunting and subjective endeavour. Can any political pundit, though, attempt to formulate such a model? That might add more fun to election prediction, for sure.
The writer is Professor of Statistics, Indian Statistical Institute, Kolkata