Movie producers and exhibitors make various decisions requiring an understanding of moviegoer’s preferences at the local level. Two examples of such decisions are exhibitors’ allocation of screens to movies and producers’ allocation of advertising across different regions of the country. This study presents a predictive model of local demand for movies with two unique features. First, arguing that consumers’ political tendencies have an unutilized predictive power for marketing models, we allow consumers’ heterogeneity to depend on their voting tendencies. Second, instead of relying on the commonly used genre classifications to characterize movies, we estimate latent movie attributes. These attributes are not determined a priori by industry professionals but rather reflect consumers’ perceptions, as revealed by their moviegoing behavior. Box-office data over five years from 25 counties in the U.S. Midwest provide support for this model. First, consumers’ preferences are related to their political tendencies. For example, we find that counties that voted for congressional Republicans prefer movies starring young, Caucasian, female actors over those starring African American, male actors. Second, perceived attributes provide new insights into consumers’ preferences. For example, one of these attributes is the movie’s degree of seriousness. Finally, and most importantly, the two improvements proposed here have a meaningful impact on forecasting error, decreasing it by 12.6%.