Trump's Fractured Coalition

Executive summary

Trump's coalition is fracturing in real time — and tariffs are the fault line. A new survey of 595 respondents finds that nearly half of people who know a 2024 Trump voter believe that person would not vote for Trump again today, while the president's signature trade policy simultaneously ranks as both the most supported and a leading source of opposition among the same sample.

Four findings stand out. First, the Epstein files are the single most-opposed policy in the study — drawing more disapproval than tariffs, Iran, or deportations — even among respondents who otherwise back Trump's agenda. Second, 44.7% of respondents say a Trump 2024 voter they know would defect, abstain, or go third-party if an election were held tomorrow. Third, tariffs generate a paradox: 23.9% of policy-support selections went to tariffs, yet 16.9% of not-support selections did too — and even self-identified Trump job-performance approvers are more likely to oppose his tariff handling specifically. Fourth, respondents with higher psychological openness scores are measurably less likely to approve of Trump's job performance or support any of his featured policies.

Context

This study was fielded as Trump's second term moved past its early phase and into contested policy territory — a moment when initial executive energy collides with real-world consequences for household budgets, foreign entanglements, and unresolved controversies like the Epstein files. The 595-respondent sample answered eight questions spanning job approval, policy support and opposition, economic performance, voter retention, and open-ended economic impact, giving researchers a layered picture of sentiment rather than a simple thumbs-up/thumbs-down on the president.

The sample skews toward politically engaged respondents — people who have opinions on Venezuelan geopolitics, Iran airstrikes, and trade policy — which matters for interpretation. Results likely reflect a more informed and opinionated slice of the public than a random national sample would capture. That means support figures may understate casual Trump approval and opposition figures may reflect more ideologically committed critics.

Still, the study's value lies not in its raw approval numbers but in the internal architecture of opinion: which specific policies generate split verdicts even among those who broadly like what Trump is doing, and which psychological traits predict where that split falls. The Retrospect analysis layered behavioral correlations on top of the survey responses, identifying personality trait signals — particularly OCEAN Openness and Prism Resilience — that predict policy opposition with statistically meaningful effect sizes.

The backdrop is a national polling environment showing Trump's approval at 35–37% in May 2026, driven by dissatisfaction with grocery prices, housing costs, and energy bills. The Federal Reserve has documented a real, gradual 8.5% price increase on Chinese-origin goods by December 2025 — not a spike, but a slow climb that compounds each month at checkout. That economic context makes the tariff paradox finding especially consequential: it is not just an abstract policy disagreement but a pocketbook grievance playing out in real time.

Findings

The Tariff Paradox: Loved and Hated by the Same People

No policy in this study generates more split verdicts than tariffs. Enacting Global Tariffs tied for first place in the support rankings at 23.9% of all policy-support selections — yet it also captured 16.9% of not-support selections, making it simultaneously a rallying point and a grievance. That double presence is unusual: most policies cluster clearly on one side or the other.

What makes this finding sharper is what the Retrospect behavioral analysis found underneath it. Respondents who said Trump is doing a good job overall were still more likely to select tariffs as a policy they do NOT support — a moderate-confidence cross-tab that suggests leadership loyalty and policy-specific dissatisfaction are operating independently. In other words, you can think Trump is a strong president and still believe his trade moves are hurting you.

External data reinforces this split. Federal Reserve research tracking 200,000 household retail transactions found that tariff costs on Chinese-origin goods were passed through to consumers at a rate of at least 30%, producing an 8.5% year-over-year price increase by December 2025. That wasn't a one-time shock — it was a slow climb that continued each month. Meanwhile, only half of Trump's own voters believe his China tariffs will benefit U.S. companies, and roughly one in four say the tariffs are actually damaging America's negotiating leverage.

Epstein Files: The Issue That Cuts Across Every Coalition

If tariffs are the most contested policy, the Epstein files are the most simply opposed. Handling of the Epstein Files was the top not-support selection at 25.2% of all opposition picks — while simultaneously being the lowest-ranked support item at just 12.2%. No other policy in the study has a wider gap between opposition and support.

The cross-cutting nature of this disapproval is what makes it politically significant. The Retrospect analysis found that respondents who supported all four other Trump policies tested — tariffs, Iran, deportations, and the Maduro capture — were dramatically more likely to also select Epstein files as a policy they do NOT support. The sample for that specific cross-tab is small (n=10), so the effect size should be read cautiously, but the directional signal is consistent with national polling: 80% of Americans support releasing all Epstein files, including 67% of Republicans. Trump's net approval on his handling of the investigation sits at -26 in external national polling.

The Epstein issue appears to function as a trust signal rather than a partisan one. Respondents who are otherwise aligned with Trump's agenda are not exempting his Epstein handling from scrutiny — they are actively singling it out as the one area where he has lost them.

Voter Retention: Nearly Half Would Defect

The study's most politically consequential number: only 55.3% of respondents believe someone they personally know who voted for Trump in 2024 would still vote for Trump if an election were held tomorrow. The other 44.7% split into three groups — 19.3% say that person would now vote Democrat, 14.3% say they'd sit out entirely, and 11.2% say they'd go third party.

This is a proxy measure, not a direct self-report of vote intent, and it reflects respondents' perceptions of people they know rather than a representative sample of 2024 Trump voters. But the direction aligns with external evidence. Navigator Research found that 20% of 2024 Trump voters explicitly regret their vote, with regretters skewing younger, more moderate, and deeply pessimistic about the economy. Critically, those regretters remain unfavorable toward the Democratic Party (66% unfavorable) — meaning they are not simply switching teams. They are drifting toward abstention or third-party options, which is exactly what this study's distribution shows.

Trump's net approval on the economy among his own 2024 voters fell from +75 to +66 in just two weeks in April 2025 polling. That 9-point drop in a fortnight is the kind of erosion that compounds.

Iran and Venezuela: Foreign Policy Fault Lines

The Iran conflict ranks second in not-support selections at 23.8% while appearing fourth in support at 18.8% — a net-negative gap that makes it the second most contested policy after Epstein. External polling confirms the asymmetry: 45% of Americans oppose the Iran airstrikes versus 38% who support them. By May 2026, opposition had hardened further, with 64% opposing U.S. military action in Iran altogether.

Venezuela tells a different story. The capture of Venezuelan President Maduro is the most net-positive policy in the study — third in support at 21.2% and last in opposition at just 16.7%. That makes it the least polarizing of the five issues tested. But nationally, the picture is more divided: 36% support versus 39% oppose U.S. military force in Venezuela, and 51% say Trump should have sought congressional authorization first. Study respondents appear meaningfully more favorable toward the Maduro operation than the general public, likely reflecting the sample's engaged, politically aware composition.

Who Opposes Trump: Personality Predicts More Than Demographics

Beyond the specific policies, the study's trait analysis surfaces a consistent pattern: psychological openness is the strongest single predictor of both Trump job disapproval and policy opposition in this sample. Respondents with higher OCEAN Openness scores were less likely to think Trump is doing a good job (r = -0.222, p = 0.0073, n = 145) and less likely to support any of his featured policies (r = -0.267, p = 0.0072, n = 40). These are high-confidence findings from the Retrospect analysis.

This is consistent with a large research base. A meta-analysis of 73 studies covering nearly 72,000 people found that Openness correlates negatively with political conservatism at r = -0.18 — and that correlation strengthens in low-threat conditions. Importantly, nine-wave longitudinal data show no evidence that Openness causes changes in political ideology; the relationship reflects stable co-occurrence, not a personality-to-politics pipeline.

Also notable: respondents with higher Prism Resilience scores (r = 0.258, p = 0.0029, n = 66) and higher Prism Persistence scores (r = 0.218) were more likely to indicate opposition to Trump's policies — and that opposition was spread across multiple policy areas rather than concentrated in any single issue. Resilient, persistent individuals are not simply reactionaries to one policy; they appear to hold a generalized, durable disaffection.

Conclusion

The picture this study paints is not one of a collapsing coalition but of a coalition under sustained internal stress. Trump retains a core of durable supporters — but even within that core, specific policies are generating friction that loyalty alone isn't smoothing over. Tariffs are the clearest example: they are both a rallying cry and a household cost. The Epstein files are a trust issue that cuts across partisan lines. And the voter retention signal, however indirect, points toward an electorate that is drifting rather than deciding.

Watch three things in the months ahead. First, whether tariff-driven price increases continue their gradual climb — the Federal Reserve's 30% pass-through estimate on Chinese goods was measured through December 2025, and further escalation would widen the gap between Trump's trade rhetoric and kitchen-table reality. Second, whether the Epstein file handling remains a persistent drag or fades as a salience issue; bipartisan 80% majorities demanding transparency are not a number that typically resolves quietly. Third, whether voter regret among Trump's 2024 coalition crystallizes into organized opposition or disperses into abstention — the third-party and no-vote signals in this study suggest the latter is the more likely near-term path, which matters enormously for turnout math in future cycles.