This study aims to estimate the contribution to the macroeconomic stabilization made by the autonomous communities’ budgets. The object of study is a novelty, since there aren’t any works about the stabilizing capacity of the autonomous’ budgets in the Spanish literature.An antecedent is Hierro and Atienza (2014), but focuses on the stabilizing effect of regional financing. In this paper we delve into this research but focusing now on regional budgets.
Indeed, there is an extensive literature on measuring the effects of central government budgets on the regional stabilization. In this work we apply the methodology widely applied to central budgets to the case of regional budgets. Perhaps this scarcity is attributed to theoretical prescription in favor of centralizing the function of macroeconomic stabilization.
Therefore, we estimated the stabilizing effect exerted by the budgets of the common regime autonomous communities (without Canary Islands) jointly in the period 2002-2008. The reasons for this timeline and the regions analyzed are based on the realization of estimates from stable structure budgets and homogeneous data. Since there are two regional financing regimes, the common and the foral, the latter producing per capita resources well above the first, we focus only on the regions of the common system. We have also chosen to exclude from the analysis the Canary Islands since they enjoy specificities in their tax financing (the Economic and Financial System of the Canary Islands) and even more since 2002 with the decentralization of indirect taxation that occurs with the new model of regional financing. Moreover, although the budgetary data source we use offers information from 2000 to 2011, we chose 2002-2008 as the analysis period for two reasons. First, since obviously the financing model affects both the overall amount of the revenue budget (and consequently the expenditure budgets) and the distribution of resources among the various taxes and transfers, we focus the analysis in the period 2002-2008, in which the same regional financiang model prevailed. Another reason, directly related to the change of the financing model, to choose 2002 as the first year of analysis is that in this year there was also a major change in the budget structure of the autonomous communities, with the decentralization of a competence with such a quantitative importance as health to ten autonomous communities of the article 143 of the Constitution.
The methodology used is based on the work done by Bayoumi and Masson (1995) and Mélitz and Zumer (2002) on measuring the effects of central budgets on regional stabilization, in which the estimates are made using a regression between regional income before and after the action of the public administrations. Specifically, we have used the two econometric specifications presented in Mélitz and Zumer (2002). We believe it is a relevant methodology for our purpose because, with it, we try to estimate how tax flows smooth cyclical oscillations in regional primary income, regardless of the fact if these flows come from central or regional budget.
As for the data, as the initial regional income we have used Gross Value Added data at basic prices from the "Regional Accounts of Spain. Base 2000. Homogeneous series 1995-2010" and published by the Spanish National Institute of Statistics. Moreover, to obtain the final income in the different scenarios analyzed, to the values of the autonomic GVA we have added or subtracted the budget item under consideration in each case, following the accrual criteria, according to the data relating to the autonomous communities’ budget settlement.
Both the stabilizing effect of the regional budget as a whole and total expenditure and revenue as the main items of revenue and expense have been estimated. Consequently nine spending scenarios, 13 revenue scenarios plus another based on the budget as a whole (total balance) have been defined. To choose these scenarios, based on the expenditure’s functional classification and revenue’s economic classification and excluding the revenue and expenses which have a financial nature, we have followed the following criteria. First we have taken each of the function of expenditure and revenue chapters, except the function "general actions" because it includes items of dispar economic nature and it is nonsense to interpret the result of its stabilizing power. Second, we have taken within each expenditure function or revenue chapter, those items with more quantitative entity, setting as the eligibility threshold those items which represent more than 5% of the total expense or revenue budget for all the autonomous communities, as appropriate. Thus, these items are, for the expenditure: health, education and economic regulation of the primary sector and for the revenue: personal income tax, VAT, excise duties, the tax of capital transactions and the Sufficiency Fund.
To estimate the two econometric specifications mentioned we have used a panel data of 14 autonomous communities and 7 years, 2002 to 2008. The existence of heteroscedasticity, autocorrelation and contemporaneous correlation issues have been detected from the application of Ward’s modified test for heteroscedasticity, Wooldridge’s test (2002) for panel data autocorrelation and Pesaran’s test (2004) for the contemporaneous correlation. Such problems can be solved jointly with Feasible Generalized Least Squares estimators (FGLS), so this has been the method used to calculate the estimates of the stabilizing effect of the autonomous communities’ budgets.
Moreover, to avoid spurious regression problems due to the existence of non-stationarity in the panel, we applied Levin, Lin and Chu (2002), Im, Pesaran and Shin’s (2003) test of stationarity, and tests similar to Fisher’s using Augmented Dickey-Fuller (ADF) and Phillips and Perron (PP) proposed by Maddala and Wu (1999) and Choi (2001). In every case, the results tell us that we can reject the null hypothesis of unit root, from which we can deduce that the panels are stationary and therefore it’s not necessary to make any cointegration analysis.
The results confirm that regional budgets have not exerted a stabilizing effect against asymmetric shocks (the total balance has a value of 4% in a specification but 0.6% in the other). Having said this, at a disaggregated level, it is observed that total expenditure has indeed maintained a significant counter-cyclical power of 11% -12.6%, mainly due to spending on merit goods. In other items of expenditure a significant amount is not reached. Regarding revenue, the total has maintained a certain destabilizing power of 5% -6.5% due to both the total taxes and the transfers. At a more disaggregated level it is shown that the various items of revenue have remained virtually no impact.
Finally we perform a comparison exercise of our results with other studies that estimate the stabilizing effect of the Spanish central budget. We observe, regarding expenditure, that the total expenditure has a similar effect in our work and in Bosch, Espasa and Sorribas’ work (2002), as well as pure public goods. But expenditure on social protection is strongly stabilizing in Capó (2008) and also, although less in Bosch and Espasa Sorribas (2002), while in our estimates it’s virtually zero. In this sense, it should be noted that in the works that serve as comparison unemployment benefits have been included, which are known for its status as an automatic stabilizer. As for the expenditure on merit goods the stabilizing effect is similar to that obtained from the central budget in Capó (2008).
Regarding revenue, the total is very close in absolute values, but is of opposite sign, procyclical at regional level and countercyclical in the central budget case. As for direct taxation, and in particular personal income tax, in regional budgets it has no stabilizing effect and yet it does in the central budget, although in Capó (2008) it is clearly destabilizing. The same applies to VAT. In short, we can conclude that our results are in line prescribed by the fiscal federalism theory in that it must be the central government which holds the competence of macroeconomic stabilization. However, if we focus only on the expenditure budget, we conclude, in this comparison with other papers focused on the central budget, the Spanish autonomous expenditure is countercyclical to a similar level to the central expenditure, particularly as regards expenditure on merit goods. Classification-JEL: R1 Keywords: Presupuestos autonómicos, Regional budget, Efecto estabilizador, Ciclos económicos asimétricos, Stabilizing effect, asymmetric economic cycles Pages: 131-160 Volume: 2 Year: 2015 File-URL: http://www.revistaestudiosregionales.com/documentos/articulos/pdf-articulo-2472.pdf File-Format: Application/pdf Handle: RePEc:rer:articu:v:2:y:2015:p:131-160