This paper uses the 2011 Phoenix Area Social Survey to evaluate the plausibility of the assumptions made with pure characteristics sorting models to rationalize incomplete stratification of households across local communities by income. The analysis with the New Ecological Paradigm, a well-recognized index of environmental attitudes, confirms the correlations in equilibrium outcomes implied by these models. As a result, it provides the first empirical support for the role of differences in the tastes for public goods as one explanation that provides the rationale for the commonly observed sorting outcomes.