Voters often oppose welfare-enhancing policies because they fail to anticipate how others will adjust their behavior. We show that policy design can align political support with efficiency even under biased beliefs. We study priority policies, a remedy whose off-equilibrium incentives make reform attractive even to biased voters. In a theory-guided experiment, participants vote between an inefficient status quo and a treatment-specific welfare-improving reform. A Pigouvian-like tax attracts only 27.5 percent support; adding priority incentives raises it to 41.7 percent, and a pure-priority policy raises it to 70 percent. Treatment effects are mostly driven by biased participants.