The Cutting Room Floor – May 1, 2026

Stuff that didn’t make the cut but was too good to toss

Friday, May 1, 2026 | 2 min read

It’s Friday. Here’s some stuff we found interesting this week but couldn’t fit into a full post. Take what’s useful, ignore what’s not, have a good weekend.

Mocking ICE’s “NICE” rebrand drops message effectiveness by 5.7 points.The jokes write themselves. The frame doesn’t work. Leave the comedy to Kimmel.

62% say a “strong leader who doesn’t bother with parliament and elections” sounds like a good idea. Then 88% say it would be harmful if the president actually ignored Congress.Voters love decisiveness as an abstraction. They reject every specific version of it. This gap is one of the more underrated tensions in American public opinion right now. → Authoritarianism tests well as a vibe. It tests badly as a policy.

97% say the U.S. spends too much on foreign aid. 0% say too little. Near-unanimity is rare in survey research. This one stopped being a policy question a while ago. It’s something closer to a psychological consensus.

79% favor the death penalty for murder. 44% support the DOJ’s actual expansion decision.The gap between principle and mechanics is doing a lot of work here. The unsure rate on the specific expansion — 22% — was the highest of any policy question we tested this week. Voters know where they stand on capital punishment. They’re much less sure when you hand them the specifics of firing squads and new drug protocols.

67% want AI regulated now, even if it slows innovation. 7% favor hands-off. One of the most lopsided findings in any survey this month. Rare bipartisan consensus. Nobody’s running on it yet. Who knew.

84% favor raising taxes on millionaires. Strong opposition: literally 0%. The rarest political species: a policy position with no intense enemies. This ran on the same survey where 74% say elites have rigged the system. The two findings are not unrelated. → The economic populism window is open. It’s been open.

If any of these raised more questions than they answered — good. That's where we'd start too. The full findings are available on request.

Methodology: All data from Tavern Research Breaking News surveys, online panel, fielded April 23–28, 2026. Sample sizes range from n=511 to n=534 per survey. Margin of error ±6.4–7.5%. AI-assisted drafting, human-verified analysis. Powered by the same tools we build for our clients.