In economic models of behavior consumers are assumed to value the goods and services they purchase based on stable preferences over externally identifiable attributes such as quality. These models predict that consumers will respond to changes in price in a way that is independent of the source of the price change. Yet research in the behavioral sciences indicates that consumers that are emotionally attached to a consumption good or other behavior might respond with resistance when policies threaten their consumption or behavior. Moreover, policies that in fact validate some emotional attachments can stir a stronger preference for the good or behavior. Reviewing both survey and experimental data from the literature, we demonstrate how such emotional responses can create hidden costs to policy implementation that could not be detected using standard welfare economic techniques. Building upon Rabin’s work on fairness in games, we propose a partial equilibrium model of emotional response to policy whereby preferences are endogenous to policy choices. In accordance with evidence both from our own analysis and the field, we propose that confrontational policies (such as a sin tax) increase the marginal utility for a good, and that validating policies (such as a subsidy) also increases the marginal utility for a good. A social planner that ignores potential emotional responses to policy changes may unwittingly induce significant dead weight loss. Using our model, we propose a feasible method to determine if emotional deadweight costs exist, and to place a lower bound on the size of these costs.