The dissertation offers an analysis of the role of budgetary institutions on the determination of fiscal outcomes. In the second chapter I provide a theoretical model that rationalizes differences in fiscal outcomes of two countries that are supposed to obey the same set of numerical constraints on the budget. I argue that these differences arise from heterogeneity in the degree of budgetary transparency that make these rules more or less binding. Moreover, the model is able to accommodate not only long run results, where stronger institutions will always cause more constrained fiscal outcomes, but also short run implications, where countries with relatively stronger institutions can be paired with relatively unconstrained outcomes. The main lesson of the chapter is that, in a democratic environment, transparency of the budgetary process is the main ingredient responsible for the good behavior of the government, and that numeric constraints will have very different effects depending on the level of transparency.

In the third chapter I conduct an empirical investigation across a set of countries, of the effects of budgetary institutions on fiscal outcomes. I exploit a new dataset on budgetary practices across countries, to construct several measures of the three recognized budgetary institutions: numerical rules, procedural rules, and budgetary transparency. The main finding of the chapter is that among budgetary institutions, transparency is the only one that is consistently associated with more fiscal discipline, a finding that goes in hand with the results of the model in the previous chapter.

The fourth chapter provides an empirical investigation of the effects of budgetary transparency on fiscal outcomes in the American States. I construct a transparency measure across time from the mid 1980s that allows me, not only to look at the evolution of transparency in the American States, but to take account of possible fixed effects in the estimations. My results essentially corroborate those obtained elsewhere in the literature, that greater fiscal transparency among the American States is associated with larger size of government, but I show that this effect is less robust and economically relevant than previously thought.