| Vertical grants are an integral part of fiscal relations in federalist countries (e.g., Boadway and Shah 2009). Following Silkman and Young (1982), a number of studies analyzed their impact on the efficiency of local public service production. Kalb (2010) finds negative effects on cost efficiency caused by the German local government fiscal equalization grants (for the German state of Baden-Wu¨rttemberg). Balaguer-Coll, Prior, and TortosaAusina (2007) and Balguer-Coll and Prior (2009) report a significantly negative effect of current grants from higher levels of government on the technical efficiency of Spanish (Valencian) municipalities. The same holds for De Borger and Kerstens (1996) and the cost efficiency effect of generalpurpose grants on Belgian municipalities. Loikkanen and Susiluoto (2005) find (at least for some years) a negative impact of state block grants on technical efficiency in Finland. In contrast, according to Worthington (2000), the general-purpose grants to the Australian local governments have no effect. Grossman, Panagiotis, and Wassmer (1999) analyze US central cities and find no effect of state grants and a positive effect of federal grants on the technical (output-)efficiency. Finally, Geys and Moesen (2009) report a significantly positive impact on cost efficiency for Flemish municipalities. In this article, we argue that this mixed evidence may result from the fact that the studies differ in the degree to which they control for an important inference-disturbing factor. The starting point of our argument is the fact that vertical grants in many countries are designed to reduce the differences in fiscal capacity between municipalities. If a grant scheme discriminates in favor of municipalities with below-average fiscal capacity but preserves the initial ordering in fiscal capacities among municipalities, high per capita grants coincide with low fiscal capacity even after fiscal equalization and thus less leeway for slack and inefficiency. Thus, municipalities that receive high per capita grants face—on average—stricter fiscal constraints and thus are (forced to be) technically more efficient. This does not imply any causal relationship between high vertical grants and high levels of efficiency. Instead, a third variable—the municipality’s fiscal capacity before grants—drives both the amount of per capita grants and the fiscal capacity after grants and thereby the degree of efficiency. In this article, we show that this effect is highly relevant. In the second section, we present a simple model to illustrate the role of grants and income for public-sector efficiency. Based on our model, we show how fiscal equalization systems that reduce fiscal stress influence the correlation between vertical grants and efficiency and how cross-sectional studies that do not control for the inference-disturbing effect of the vertical grants system may lead to the wrong conclusions regarding the causal impact of vertical grants on efficiency. |