Wording Doesn’t Move Morality Polls

A large randomized Pew experiment (May 2025, n≈8,937) tested whether asking about 'homosexuality' versus 'homosexual behavior' changes respondents' moral judgments. The two phrasings produced nearly identical shares calling the subject 'morally wrong' (39% vs. 41%), with no statistically significant differences across major demographic and religious subgroups. — If common moral‑opinion questions are robust to small wording shifts, media and policymakers should be less quick to treat framing differences as explanations for apparent opinion change or disagreement.