A Response to Emerson and Van Buren

Abstract

We applaud the effort by Emerson and Van Buren (1992) to replicate our model using LISREL techniques and we concede that their approach is more thorough in assessing measurement error. Emerson and Van Buren's analysis strategy offers the advantage of estimating the strength of effects after allowing for the impact of measurement error (Andrews, Abbey & Halman 1991; J6reskog & Sorbom 1989). One weakness of their presentation is that they do not provide measures of the overall goodness of fit of the models. Substantively, we take exception to their conclusion that our model is more adequate at explaining the concept of social rights than welfare support. Emerson and Van Buren contend that the reason that our model is not supported is because we have incorrectly operationalized government intervention and welfare support. They question the distinction between our concepts of government intervention and social rights on the basis of two empirical findings: a high beta for government intervention in predicting social rights and a principal components analysis of the items that comprise these two measures. While we may not have the best indicators of government intervention or social rights, we maintain that a theoretical argument can be made for making a distinction between the principle that government has a responsibility to intervene on an aggregate level to rectify social inequalities vs. the principle that individuals have a right to a basic standard of living. Moreover, we constructed our model with a hierarchy in mind. The endogenous variables are increasingly more specific. Work ethic and government intervention represent general principles, social rights bring the principles to an individual level, and finally welfare support represents fiscal support for implementing some of these principles. One cannot a priori conclude that there is necessarily a consistency among these principles at varying levels of specificity. For example, one can agree to a particular principle but not be willing to endorse a policy that seeks to implement the principle. Much of the literature on racial attitudes points to the gap between willingness to endorse positive racial attitudes and a commitment to specific affirmative action policies (e.g., Schuman, Steeh & Bobo 1985).