In the final days of the campaign, Donald Trump is experiencing an unprecedented decline, transforming what initially seemed a sure victory for the former president into a likely win for Kamala Harris, according to Thomas Miller, a data scientist at Northwestern University who has a track record of accurate election forecasts.
Just under two weeks before Election Day, Trump was leading significantly after a remarkable comeback from a substantial deficit, appearing poised for a resounding victory. However, Harris countered by shifting focus from Trump’s policies to highlighting his “unstable” personality and “obsession with revenge.” Despite this, Trump’s ascent continued, driven by voter dissatisfaction with the Biden administration, particularly concerning economic issues.
Miller notes that while growth and employment figures seem positive, voters are more concerned with the higher costs of groceries, housing, and loans, alongside their lack of savings and need for multiple jobs. Additionally, many Americans are increasingly worried about the country’s involvement in foreign conflicts, such as those in Israel and Ukraine. Trump’s quasi-isolationist stance is resonating with some voters.
With limited time before the election, Miller speculated that only a significant event favoring the Democratic ticket could alter the race’s trajectory.
The turning point occurred at a Trump rally at Madison Square Garden on October 27. Planned as a pinnacle event, it featured controversial figures who reiterated Trump’s extreme views, including Rudy Giuliani, Stephen Miller, and Elon Musk. The event led to several missteps, notably comedian Tony Hinchcliffe’s offensive remarks and businessman Grant Cardone’s derogatory comments about Harris.
These incidents did not convey unity but instead seemed to alienate certain voter groups, as noted by Republican Nikki Haley. Following the rally, angry feedback from women, especially educated suburban women, further highlighted the negative reception.
In the aftermath, Harris gained substantial ground in electoral votes, according to Miller’s model based on data from PredictIt, a political wagering platform. Trump’s lead drastically reduced over several days, with Harris eventually overtaking him.
Miller’s model, which does not rely on polls but rather prediction markets, is calibrated to account for potential Republican bias in these markets, noting that they might overestimate GOP chances due to demographics of the bettors who tend to be risk-taking males.
With Harris now edging ahead, Miller suggests that her current advantage may be understated. In light of prediction markets’ leanings and recent events, Miller predicts a possible Harris-Walz victory, highlighting her move towards the center against Trump’s increasingly extreme position.
Miller’s real-time tracking “tinyticker” detected the rapid shift following the MSG rally, capturing its immediate effects, contrasting with traditional polls that may delay capturing such dynamics. What seemed an insurmountable lead for Trump just a week prior has swiftly diminished, leading to a tight race with Election Day approaching.
