توزیع مجدد شخصی و منطقه ای از طریق دارایی های عمومی در یک محیط فدرال

Following a partial equilibrium approach, this paper studies the effect of fiscal policy on income distribution in Argentina, based on budget information corresponding to the year 2004. Specifically, it aims to provide an empirical answer to a set of questions related to the responsibility of different levels of government (national or provincial) in ensuring the equitable distribution of income; how this responsibility is accomplished in practice, when this function is shared among two or more levels of government; whether an incompatibility arises among the policies from different levels of government; and the effect of fiscal policy on personal and regional income distribution.The main results of the paper are summarized as follows. At the aggregate level, both national and provincial budgets have a redistributive impact on personal income distribution, through a combination of progressive expenditures and (slightly) regressive taxes. Regional redistribution depends on two fiscal tools: the national budget and the revenue sharing regime. The progressive effect of expenditures and taxes interacts with the geographical effect of the revenue-sharing and the national budget, reinforcing progressivity in net-receiving groups and creating a trade-off between progressivity and (negative) regional transfer in net-financing ones. Provincial budgets have more impact than the national budget, both being compatible in pursuing the redistributive goal.

Redistribution of income through public finance has been an intensive area of research in the context of the welfare state. Economic analysis has concentrated on several dimensions, such as public policy variables (taxes/expenditures), the level of government (national/sub-national), the dimension of income (personal/regional), and so on. In the framework of a federal government, several questions have been investigated both in theory and in practice. Which level of government (national or provincial) should be responsible for correcting the distribution of income resulting from the market? (among others, see Musgrave, 1959, Oates, 1972 and Pauly, 1973). If two or more levels of government share this function, how do they shoulder such responsibility? (Boadway & Shah, 2009). Is there an incompatibility between the policies from different levels of government? (Tresch, 2002). On the other hand, which is the relevant dimension of the redistribution; personal, regional, or both? Do changes in personal income distribution coincide with or trade off against changes in regional income distribution? (Oates, 1972 and Tresch, 2002). This paper aims to provide an empirical answer to most of these questions, using budget information corresponding to Argentina in the year 2004, following a partial equilibrium approach (rather than a general equilibrium analysis). It performs a comprehensive analysis of the consolidated effect of budget on income distribution at both national and provincial levels of government, covering almost all expenditures (excluding social security), revenues and transfer regimes, and distinguishing between geographical source and destination of funds. The main results of the paper are summarized as follows. At the aggregate level, both national and provincial budgets have redistributive positive impact on personal income distribution, through a combination of progressive expenditures and (slightly) regressive taxes. Regional redistribution depends on two fiscal tools: the national budget and the revenue sharing regime. At the provincial level, the progressive effect of expenditures and taxes interacts with the geographical effect of revenue sharing and national budget, reinforcing progressivity in net-receiving groups, and creating a trade-off between progressivity and (negative) regional transfer in net-financing ones. Finally, provincial budgets have more impact than the national budget but they are both compatible in pursuing the redistributive goal. These results contribute to the debate on the design and assessment of public policy in Argentina. The 1994 National Constitution gave constitutional hierarchy to the revenue sharing – coparticipation – regime. However, a Coparticipation Law has not been passed yet because of political and economic reasons, including the lack of research contemplating the overall effect of revenue sharing and its interaction with the rest of fiscal policy, on income distribution. The paper contributes to international debate in several respects. First, the provision of public goods at different levels of government follows an evolutionary process, in which some countries are moving towards decentralization while others are returning to re-centralization.1 One of the dimensions to understand the expected effects of decentralization or centralization of specific taxes or expenditures is the identification of the impact on redistribution and the interactions between different levels of public budget and dimensions of income distribution. Second, provincial expenditure is a strong income-redistribution tool, but in federal countries provincial budgets may strongly rely on national transfers, which create soft budget constraint and incentives for fiscal irresponsibility. The international literature that studies the case of Argentina concentrates on the second effect, but omits the first.2 The paper is organized as follows: Section 2 situates the paper in context; Section 3 provides a brief description of the institutional framework in Argentina; Section 4 provides the basic definitions and the methodological framework; Section 5 presents the main results; and, finally, Section 6 concludes.

This paper studies the impact of national and provincial fiscal policies on income distribution using information of Argentina in the year 2004, by selectively pooling provinces into four groups according to their economic and social development (Advanced, Intermediate, Low Density and Lagged jurisdictions). This constitutes an advance in the field because the paper explicitly takes into account the interconnected relationships between national and provincial revenues and expenditures in a revenue sharing context, and their redistributive effect on personal and regional income distribution. At the aggregate level, the fiscal policy of both levels of government redistributes income (either Gini or Atkinson is used). For example, the Gini coefficient indicates a reduction of 0.077 points out of an inequality value of 0.483. The positive impact results from a combination of progressive expenditures and slightly regressive taxes. These impacts are different at the group level. In some groups, the size of the provincial expenditure is an important driver (Lagged and Intermediate groups), in others it is the progressiveness of national expenditure (Lagged Group), while in others it is the progressiveness of national taxes (Low Density Group). Provincial budgets are more important than national budgets, accounting for approximately 70% of the change in the Gini coefficient. At the level of regions, the positive distributive impact of expenditures and taxes interact with the regional redistribution from the revenue-sharing regime and national expenditure, reinforcing the positive impact in net-receiving groups but creating a trade-off with the (negative) regional transfer in net-financing ones (Advanced and Low Density groups). In the latter groups, however, the net effect is also progressive for intermediate to high inequity aversion coefficients. We do not find incompatibility between the redistributive policies at the national and provincial levels of government. In a vertical comparison, the most important impact comes from provincial budgets. In a tax-expenditure comparison, the provincial expenditure is identified as the most important redistribution tool. Compared with the papers on Argentina detailed in Section 2 our results are qualitatively similar in the direction of the effects (positive impact) but we provide a more comprehensive measure that considers all expenditures and taxes plus the revenue sharing regime and the joint effect on regional and personal distribution. The international literature mentioned in Section 2 refers to the impact on regional redistribution and risk sharing (using panel data) while this paper concentrates on the interrelationships of personal and regional redistribution (using data from a specific year). Aside from differences in the items of the public budget considered, the period of time, the countries, and the methods used, the impact of federal government policy variables on regional redistribution goes in the direction that we found for Argentina. Some final comments must be made. Firstly, the basic data is scarce and in some cases unreliable. In particular, we use gross domestic product that in several provinces could have large differences with the disposable income. Secondly, the estimates of incidence weights of taxes and expenditures (the γikn, γijn, τicn, τirn and τisn) imply very strong assumptions (some of them taken from other papers, as mentioned in the Appendix A) and therefore the effects of fiscal budget on the regional and personal distribution of income should be taken carefully. Thirdly, it is very probable that the effects change every year because of a different composition and distribution of some kind of expenditures (e.g., capital expenditures) or taxes (e.g., export taxes on oil, gas, agricultural products and their manufactures). In the fourth place, by concentrating on a specific year we are not analyzing the interactions between redistribution and risk sharing in public policy. Finally, we assumed that regional prices do not affect budget incidence.22 Lastly, there are tax costs through excess burden and efficiency costs in the public provision of goods and services (e.g., overflows or leakages), which are not addressed here.23