This article differentiates a social etiology model focused on identifying the social antecedents of one particular mental disorder from a social consequences model concerned with the overall mental health consequences of various social arrangements. In the social etiology model, people with disorders other than the one particular disorder singled out for investigation are implicitly classified as "well." This disorder-specific model is inappropriate for the more general sociological task of identifying the consequences of various social arrangements, such as concentrated poverty, racial segregation, and gender stratification. It is problematic because these consequences are typically nonspecific, not limited to one particular disorder. From this perspective, persons classified as "well" in the disorder-specific model who have a different disorder are misclassified. Consequently, the impact of social arrangements is underestimated, and estimates of causal effects are biased. To address these problems, the full range of theoretically derived mental health outcomes needs to be simultaneously analyzed. (Contains 2 notes.)