The paper presents a Bayesian model for estimating ideological ambiguity of political parties from survey data. In the model, policy positions are defined as probability distributions over a policy space and survey-based party placements are treated as random draws from those distributions. A...