The Electorate Decides: Measuring the Gap Between Forecast and Reality in 2024
Inside the slimmest of margins: what the data says about who really showed up.
I. The Verdict Is In — But the Questions Remain
In the closing days of October 2024, Quantus Insights released its final projection: Donald Trump 49.1%, Kamala Harris 48.3%. As close as close can be and an accurate snapshot of a divided nation.
Final Results for 2024 Presidential Election
Trump: 49.8
Harris: 48.3
Now, with the benefit of Catalist’s exhaustive voter file analysis in What Happened in 2024, we compare the prediction to the post-election data to see who showed up, who defected, and where expectations matched or missed the mark.
II. What Catalist Found: A Coalition Redrawn
Catalist’s 2024 analysis confirms what many suspected: Trump expanded his coalition, not just in size but in diversity. While Harris held ground with regular voters, especially older women and college-educated whites, her campaign bled support from the margins of the Democratic base:
Young voters, especially young men, shifted rightward.
Latinos, once a Democratic stronghold, moved to near parity.
Black men, particularly younger ones, also peeled away from the party.
Urban areas saw reduced turnout and erosion in Democratic support.
Trump’s strategy relied less on persuading moderates and more on energizing less engaged voters who feel alienated by elite institutions and it worked. His support gains came disproportionately from voters Catalist calls “irregulars” — those who don’t show up every cycle but do when Trump’s name is on the ballot.
III. Comparing the Quantus Model to Reality
Quantus’ final crosstab estimates were largely in line with Catalist benchmarks, notable feat in a volatile cycle. Here's how their key group-level estimates stacked up (see table):
White voters: Dead-on for Trump (57%), just 1 point under for Harris.
Black voters: Slightly underestimated Harris support (+4 pts), with Trump support slightly overstated (3 pts).
Hispanic voters: Quantus underestimated Trump support by 2 points and slightly overestimated Harris by 1 point — very close to actuals.
Education split: Within 1–2 points across college vs. non-college and racial subgroups.
Age: The biggest underestimates for Harris were among 18–29-year-olds (–2 pts) and 30–44-year-olds (–1 pt), matching Catalist’s findings of youth defection.
Net takeaway: Quantus got the big picture right. We called the final margin with near surgical precision and tracked the major demographic shifts that shaped the 2024 race. If there was any gap, it was minor—slightly underestimating turnout among younger, lower-propensity voters, many of whom leaned toward Trump. But even that margin was narrow, and the direction of movement was correctly forecast.
These voters aren’t easy to model. They don’t vote like clockwork, don’t respond to pollsters, and don’t follow the script. They're younger, often male, less formally educated, and more skeptical of institutions, including the media and political parties. But in 2024, they showed up, and they went for Trump.
Catalist’s voter file confirmed what Quantus picked up: Harris held her base, but Trump expanded his reach. Not by persuasion alone, but by mobilizing a slice of the electorate that isn’t always counted on—and often isn’t counted at all.
Pollsters often miss because they get the vote preference right, but the composition wrong. Quantus didn’t. We saw the realignment coming. In a cycle full of noise and surprises, we came closer than most to reading the country as it actually voted.
IV. The Broader Picture — And A Word of Caution
If there’s one lesson from this cycle, it’s this: the old assumptions are dying. Democrats can no longer count on young voters and communities of color to show up or vote blue at historical margins. Republicans, long dismissed as unable to grow their base, have found in Trump a candidate who brings new voters into the fold: irregular, disaffected, nontraditional.
Still, the margin of error remains narrow. While Trump claimed the popular vote plurality, the battle was closer in the battlegrounds. Small shifts in turnout or loyalty could have flipped the map.
V. Final Thought — In a New Century
The tectonic plates of American politics are shifting. The New Deal coalition is a relic. The Reagan Democrats are back — or their sons are — wearing Carhartt, carrying debt, and voting Trump. What the Quantus/Catalist comparison shows is not just polling accuracy or error. It reveals a deeper story: a realignment in motion, where cultural identity, distrust of institutions, and economic insecurity drive voter behavior more than party labels or policy plans.
It’s not about which pollster is best. It’s about who understands who’s going to show up and what’s moving them.