Essays on Empirical Fiscal Federalism

This thesis comprises three essays (each corresponds to one chapter) on empirical fiscal federalism. Using spatial econometric methods, they investigate how space, politics and economics affect the distribution of public expenditure in Mexican municipalities in the new millennium.

Chapter 1 investigates the presence of public expenditure spillovers among municipalities in Mexico from 2001 to 2021. Using dynamic spatial panel data models with common factors, it estimates the spatial and temporal interdependence in municipal spending across total expenditure and expenditure categories: use of services, gross capital formation, and subsidies and social benefits. Results show that municipal governments exhibit incremental spending behaviour and that public goods are complementary across space, with significant direct and indirect effects driven by employment rates, household income, federal transfers and population density.

Chapter 2 examines the political dimensions of fiscal mimicking in Mexican municipalities using regime-switching spatial Durbin models. Electoral concentration triggers swing voter dynamics, with politically vulnerable mayors mimicking more intensely to appease pivotal voters. Vertical alignment with higher levels of government triggers core voter dynamics, with politically dominant mayors mimicking to retain their base of supporters. The interaction of both dimensions reveals an asymmetry. National alignment combined with concentration reinforces the core voter pattern, regional alignment combined with concentration reverses it, and the joint combination of both vertical alignments restores the swing voter pattern.

Chapter 3 examines the flypaper effect (the tendency of intergovernmental transfers to stimulate local public spending by more than an equivalent increase in private income) across the expenditure categories analysed in Chapter 1. A positive and significant flypaper effect is present in total expenditure, use of services, and subsidies and social benefits, while gross capital formation shows no flypaper effect. The heterogeneous analysis across the moderator distributions shows that the effect declines with political concentration and with own resource taxation for total expenditure and use of services, while for subsidies it rises with political concentration and falls with own resource taxation. Gross capital formation remains zero across the entire range of both moderators.