Essays on fiscal consolidation and monetary policy

This dissertation consists of three essays regarding fiscal and monetary problems that typically affect developing economies. Chapter one presents new evidence that the composition of fiscal adjustments (i.e. the type of expenditures that are cut and/or revenues that are raised) matters for their persistence. Using detailed revenue and expenditure reports for 73 countries between 1990 and 2005, including OECD and non-OECD members, two different methodologies are considered. The first one is an adaptation of Alesina and Perotti (1995, 1996) that uses a descriptive analysis to separate the fiscal adjustment episodes into "successful" and "unsuccessful" groups according to some persistence criterion, and then compares the differences in composition among them. The second one is based on a formal regression framework and uses survival analysis, where the probability of ending a fiscal adjustment is modeled as a function of its composition and control variables. The results show that cuts in the government wage bill, non-wage consumption and transfers (in this order of importance) are particularly more associated with a longer duration of fiscal adjustments. On the revenue side, the results show that raising taxes on households and social contributions are also satisfactory alternatives in terms of contributing for a more persistent adjustment, while raising taxes on business and indirect taxes are clearly inferior alternatives.;The second chapter analyzes the effects of jurisdictional uncertainty and capital controls on cross-country real interest rate differentials. The inspiration comes from an article by Arida, Bacha, and Lara-Resende (2005), which argues specifically for the Brazilian case that, risks associated with the jurisdiction and currency inconvertibility are relevant determinants of the level of short-term real interest rates. A methodology based on their definition of jurisdiction uncertainty is formulated, and then a set of institutional variables is used together with an index of capital controls to test their conjecture and variants of it. The results are by and large unfavorable not only to Arida, Bacha, and Lara-Resende (2005)'s conjecture, but also to variants of their argument. The results further indicate that traditional monetary and fiscal factors are far more relevant to explain the level of short-term real interest rates than the binomial jurisdictional uncertainty/currency inconvertibility is.;The third chapter investigates the relationship between the high share of zero-duration floating rate bonds in the domestic public debt and monetary policy in Brazil. Two possible channels of this relationship are tested. First, a higher share of floating rate debt could be weakening the impact of changes in the real interest rate on output through a consumption-wealth channel of monetary transmission, as suggested in Pastore (1996). Only weak evidence in favor of this "traditional view" is found. Second, an alternative channel is proposed in which the demand levels for floating rate debt publicly announced by the Treasury could be serving as an indicator of "hidden" inflation expectations of economic agents, forcing the Central Bank to react appropriately. By estimating a backward-looking modified Taylor rule that includes lagged changes in the share of floating rate bonds as an explanatory variable it is found that, even after controlling for measures that capture default risk and exchange rate risk, changes in the share of floating rate debt can help to explain subsequent changes in the nominal short term interest rate.