notch signaling As empirical research revealed that budget i

As empirical research revealed that budget institutions do affect the budget result of the government and its fiscal policy, it is essential to assess the impact of these institutions on the Brazilian budget performance, a topic, to our knowledge, underexplored in the Brazilian literature. Thereby, this article seeks to answer the following question: what impact do budget institutions exert on the Brazilian fiscal result? The analysis focuses on the impact of federal budget institutions on the primary deficit of the Federal Government between 1985 and 2009, a period for which there is data available for the primary deficit as a proportion of the GDP, within the concept ‘below the line’, and the main notch signaling is that the chances in the Brazilian budgetary institutions contributed to a better fiscal performance. One approach to assess the changes in the various rules that constitute the budget institutions, and thus to evaluate their impact on the fiscal result, is through the budget indices, which quantify and aggregate the various aspects that compose each institutional arrangement. Thus, based on the methodologies by Alesina et al. (1999) and Dabla-Norris et al. (2010), three sets of indices are constructed for this evaluation. The index proposed by Alesina et al. (1999) is the pioneer in this literature for developing countries and the basis for many subsequent studies. The methodology by Dabla-Norris et al. (2010) incorporates a wider range of institutional features and is designed to assess the budgetary institutions in low- and middle-income countries, allowing features typical of developing countries, such as Brazil, to also be considered. Both indices therefore capture different characteristics of the Brazilian budget process. Is it noteworthy that the majority of the empirical literature that evaluate the impact of budget institutions on the fiscal result through the budget indices compare different budget institutions across different countries in an given institutional framework. In contrast, this article also contributes to this literature by analysing the issue from a different perspective by assessing the effect of budgetary institutions on fiscal results in the context of institutional change, controlling for its possible endogeneity. The main results indicate that chances in the Brazilian budgetary institutions have contributed to a better primary result of the federal government.
Budget institutions The literature on political economy linked to the mainstream, as Gleich (2003, p. 7) summarises, starts from the idea that institutional structures have a systematic impact on the behaviour and strategic choices of politicians and may, therefore, influence the result of policies generated by a collective decision process. Regarding the budgetary process, two theoretical and interrelated issues explain its nature and result: the problem of the common pool and the principal-agent model. Concerning the first issue, the decisions on public spending are determined in a decentralised manner by the agents involved in the process, and the cost is financed by a pool of resources. As demonstrated by Persson and Tabellini (2002) and Velasco (1997), each agent sets the amount of expenditure in the utility maximisation process without considering the volume of expenditure set by others, i.e., without internalising the cost of that decision. The result is a level of spending above the socially optimal, generating deficits and debt accumulation in the long term. In the principal-agent model, in turn, the spending decisions are delegated by the principal to an agent, through explicit or implicit contracts. This relationship is present both among voters (principal) and the government (agent) as well as within government, between the Minister of Finance (principal) and the other ministers (agent). In a scenario of information asymmetry, distinct interest and different incentives, agents can select a volume and type of spending that is different from the preferences of the principal. As Dabla-Norris et al. (2010, p. 5) exemplify, resting potential problem arises when politicians can extract economic rents and appropriate public resources at the expense of voter preferences.