Effects of the General System of Royalties on municipal fiscal performance in Colombia: a dose-response analysis

摘要

There is an extensive literature on the impact of nonrenewable natural resources (NNR) in the fiscal performance of countries. Overall, with few exceptions, the studies suggest an inverse relationship between these two variables. According to the IDB (2013), the presence of NNR can lead to the so called 'resource curse', where the abundance of income that comes from this source adversely affects institutional capacity, governance, and economic growth. There is not too much literature about the impact of the relative abundance of NNR on the fiscal performance of sub-national governments and conflicting results have been found. One of the most recent changes in regional policy in Colombia was the creation of the General Royalties System (SGR by its Spanish acronym) in 2011. Before the new scheme, royalties were distributed to those territories where NNR were extracted and to the maritime or river ports used to transport these resources. With the creation of SGR, these resources began to be distributed between all municipalities and departments through various funds and according to the socioeconomic conditions of each territory. As a result of the increased mining and energy production, royalties increased from 0.6% of GDP in 2002 to 1.66% in 2012. These resources are an important source of funding for projects in sub-national governments. For municipalities, these funds are twice the amount collected by two of the most important municipal taxes, property tax and industry and commerce tax. This is an excellent opportunity to assess the impact of the new system on the fiscal behavior of local authorities. This paper evaluates the effects of the implementation of the SGR on the fiscal performance at the municipality level in 2012, employing a dose-response analysis based on Hirano and Imbens (2004) for a sample of 1,025 municipalities. Unlike conventional impact evaluations comparing treatment and control groups, the dose-response analysis compares municipalities with higher and lower allocation of royalties. The non-randomness of assignments is controlled by estimating the generalized propensity score. The results indicate that a level of 20% allocation of royalties in the total revenue of the municipalities represents an important threshold performance of these localities. It was found that in the 93% of municipalities, where the proportion of royalties in their total revenues are less or equal to 20%, the fiscal performance measured by several indexes worsens, while the dependency on the royalties increases. On the other hand, if such proportion is higher than 20% the fiscal performance improves but the local investment falls. Given that the reform assigned resources but did not guarantee their appropriation by the municipalities, these results can be explained by the low execution of royalties during 2012.