| Abstract | Earmarked grants are usual policy tools to correct local fiscal distortion, while the policy evaluations to such programs are controversial. This paper argues that the effectiveness of earmarked grants depends on the rules to distribute grants, such as the rule of assisting weak or rewarding excellence, and the stage of economic development. As economy grows, the rule of rewarding excellence will be superior to the rule of assisting weak, for the former rule can constrain moral hazard of local government and result in more balanced public expenditure structure. Empirically, taking the rural compulsory education expenditure guarantee mechanism reform in 2006 as background, we test the interrelationship between earmarked education grant and local education expenditure using nonrecursive structural equation models with a sample of 423 counties in central China. The empirical results give support to the hypothesis of validity of earmarked grants and refuse the moral hazard hypothesis. Such evidence indicates prevalence of the rule of rewarding excellence, which leads to more local spending on education. This finding is of importance to build an equitable and efficient earmarked grants system. |