Paradoxes of Heterogeneous Returns: The Negative Selection Syndrome in Stratification Research

This paper examines the paradoxes of heterogeneous socioeconomic returns associated with common models of attainment in stratification research. Specifically, the paradoxes stem from the pervasive presence of negative selection across diverse substantive contexts—what we call the negative selection syndrome (NSS), including Brand and Xie’s (2000) finding that economic returns to college are the greatest for those who are least likely to attend. We develop a formal model of the paradoxes, specifying the common structure and crucial parameters of the phenomena, conducting simulation to demonstrate the regularities underlying the paradoxes and deriving testable predictions of the model. The model also explains why NSS is prone to be observed when propensity score matching method is used to produce counterfactual causal estimation of treatment effect on the treated and the treatment effect on the untreated. The model also connects NSS with other paradoxes in stratification research, such as the relative incompetence syndrome.