Donald Trump has a 91 percent chance of winning a second term in the US presidential elections scheduled to be held in November this year, according to an American political science professor who has rightly predicted five of the six elections since 1996.
In a detailed blog, Professor Helmut Norpoth explains his Primary Model, wherein Trump is predicted to win 362 electoral votes in a possible match-up against his Democratic rival Joe Biden, who, the model predicted, would get 176 votes.
Norpoth makes the surprising prediction despite recent polls showing Trump in trouble while surveys showing diminished public optimism. Many surveys have also reflected that voters feel the US president is poorly managing the surging virus and the languishing economy.
According to a Newsweek report, Norpoth, however, said the COVID-19 pandemic that has gripped the US and the world and led to a near-economic meltdown, could swing the election to Biden if Trump's approval rating drops significantly.
According to the report, "while the Hill/HarrisX national poll, released on Wednesday, has Biden leading by 4 percent over Trump, an Economist/YouGov survey gives Biden a lead at 49 percent to Trump's 40 percent".
The election prediction model is based on primary performances in New Hampshire and South Carolina, while taking into account the first-term electoral benefit that the party in power has enjoyed in the last 200 years of American election history, professor Norpoth says in the blog.
Interestingly, in a possible match-up with Bernie Sanders in November, Trump's chance of winning would rise to 95 percent, he says.
In that scenario, Trump would get 390 electoral votes, Sanders 148, he adds.
"On the Democratic side, Biden and Sanders split the primaries in New Hampshire and South Carolina while Trump handily won the Republican Primary in New Hampshire," the professor writes in the blog.
"Winning the early primaries is a major key for electoral victory in November," he adds.
The prediction model uses presidential elections results from 1912, the year presidential primaries were introduced, to estimate the weight of primary performance.
An analysis of elections from 1912 to 2016, it said, showed that "the primary picks the winner, albeit retroactively, every time except in 1960 and 2000".
According to Mediaite, the two elections that the primary model failed to predict were the 1960 election of John F Kennedy and the 2000 election of George W Bush.
Another factor goes in favour of a second Trump presidency, the Stony Brook professor professor explained, is that the "the incumbent party is favored to win re-election" after one term in the White House. The professor showed results from 1960 onwards in his blog to back this claim.
It should be noted that the model used all primaries for prediction of elections held before 1952, but after that only the New Hampshire Primary has been used as a rule . South Carolina, which has a large and most loyal Democratic constituency and African-Americans, was added to the model in 2008. But the GOP primary was cancelled in South Carolina this year.