The purpose of this paper is to empirically examine the robustness of the government failure theory. A core feature of the government failure theory is demand heterogeneity. Previous studies have brought into question the robustness of the government failure theory, due to inconsistent results concerning the explanatory power of demand heterogeneity. Therefore, in this paper we revisit this important research agenda using US state level panel data. We find the two-way fixed effects model a suitable model for testing the government failure theory's robustness and present findings which indicate that observable demand heterogeneity has a positive effect on the size of the nonprofit sector. This paper also empirically examines the relevance of the complementary financing hypothesis in terms of the cooperative nature of the governmental and nonprofit sector relationship; that is where governments delegate the production of quasi-public goods to the nonprofit sector. Copyright CIRIEC, 2004.