Empirical Strategies for Incorporating Weak Complementarity into Continuous Demand System Models

Abstract

This paper conceptually and empirically compares alternative strategies for incorporating weak complementary into continuous demand system models. The repackaging, integrating back, and discontinuity approaches are evaluated in terms of their behavioral implications and potential usefulness for applied research. The conceptual review suggests that the repackaging approach offers the most flexibility and tractability. The empirical comparison suggests that qualitatively similar policy inference arises from previously employed repackaging approaches. These estimates can be similar to use-related welfare estimates derived from non-weakly complementary models using a decomposition approach suggested by Herriges, Kling, and Phaneuf (2004), although the latter are sensitive to arbitrary assumptions about how to decompose use and nonuse values.