Incidencia de factores políticos en los convenios de inversión del Estado con las Comunidades Autónomas

Resumen: Los convenios de inversión constituyen unos recursos de las Comunidades Autónomas (CC.AA.) cuya distribución se basa en la potestad discrecional del Estado y el poder de negociación bilateral de cada Comunidad Autónoma (C.A.). El objetivo de este trabajo es explicitar las variables que dan concreción a esa discrecionalidad en el reparto de estos recursos. En primer lugar, consideramos al beneficio electoral como variable que sustancia ese reparto, como postula la porkbarrelpolitics. Segundo, la coincidencia ideológica entre gobierno central y autonómico y la existencia de partidos nacionalistas/regionalistas que se constituyen en pivote para la estabilidad del gobierno central. Y tercero, también se incluyen variables no políticas de reparto, ahora desde una perspectiva de economía ortodoxa. Los resultados muestran que la coincidencia ideológica da contenido a la discrecionalidad así como el apoyo electoral recibido por el gobierno central en cada C.A., existiendo también, en efectos fijos, un efecto de los votantes indecisos (swing) a favor de un reparto favorable. Asimismo, el reparto es regresivo, beneficia más a las CC.AA. más ricas y las CC.AA. forales reciben muchísima menor cuantía. Abstract: The present study attempts to answer the question of what the unrevealed political motives are which guide the central government to divide among the governments of the Autonomous Communities (ACs - Autonomous Communities, or political-administrative regions) those transfers over which it has a broad discretionary power. For the majority of the transfers which form part of regional financing in Spain there exist explicit distribution criteria. These distribution criteria are usually variables of an economic nature and representative of efficiency and equity, as postulated by the normative theory of fiscal federalism. However, there exists a source of financing, the joint investment agreements, for which distribution criteria do not exist and which are negotiated bilaterally and without prior public information. Obviously, through this path of distribution we can tell intuitively that the power of negotiation of each AC and the discretionary power of the state are the fundamental criteria for distribution. But what distribution criteria are implicit in that discretionary nature and different negotiation powers of the ACs? In reality, if we are talking of discretionary arrangement funds, it is not absurd to consider that factors of a political/electoral nature determine the result of the distribution. The fundamental objective of the present study is to attempt to determine whether this is true, ascertaining what would have been the political-electoral motives which would have incised upon the distribution of those discretionary resources among the ACs. To perform our analysis we adopt the theoretical framework of vote-winning politics, or `pork barrel politics’, for which the variables of an electoral character incise upon the territorial distribution of resources and services. The proposal is that the objective of the central government being to remain in power, electoral profitability is determinant in the territorial distribution of central expenditure. This said, it is necessary to determine whether the governing party estimates that such profitability may be greater if the money is spent in the safest electoral districts, with greatest government support, or whether, by contrast, it is more profitable to favour those districts with a greater number of “swing” voters (those who change their vote for one political option or another on the basis of the profit the voter perceives). Subsequently, the theory of pork barrel politics is centred on the expenditure directly aimed at the citizens/voters in the form of public services, infrastructure or economic transfers, while in the present study we concentrate on intergovernmental transfers of a discretionary nature. Furthermore, the fact that in our case there exists an intermediate political institution between the government conceding the transfer and the citizens/voters, may imply theoretical predictions different to those indicated by the theory of pork barrel politics. In effect, according to Grossman (1994), the government uses transfers to buy influence from regional politicians and pressure groups, with the objective of positively influencing electoral support towards the central government. But if the two levels of government are of distinct ideology, what incentive would the conceding government have to make the transfer? Only the existence of a need for political support in the parliaments, as occurs in the Spanish case when the party winning the elections lacks an absolute majority and requires support from nationalist or regionalist parties, could lead the government to compensate regional governments of a distinct political hue with special treatment via transfers. On the other hand, according to Khemani (2007) and Arulampalamet al. (2009), when the case is one of intergovernmental transfers which bring with them an increase of expenditure by the recipient government, the electoral incentive for the central government diminishes because the citizen is not fully aware of the origin of the financing. All depends on the level of information of voters, yet even with a high level of information, part of the electoral benefit of the transfer is shifted towards the recipient government. Consequently, given that division of the electoral benefit between the concessionary and recipient governments, it is clear that the former would prefer to assign more transfers to a government of the same political hue. These arguments lead to our second objective, which is to test the hypothesis that the ideological coincidence or affinity between the two levels of government, and/or the existence of pivotal regional parties, are the important political variables in the distribution of transfers. The results show, on the one hand, that from the political point of view both the considerations of ideological affinity and those of electoral support are relevant. This can be demonstrated that when both the PP and the PSOE govern simultaneously in the central and regional governments, those regional governments obtain more resources through investment agreements. Similarly, although it is clear that the greater is the electoral support of the party in the central government in a specific AC, the greater resources it receives, which appears to confirm the hypothesis of Cox and McCubbins (1986), on the other hand, in the fixed effects model, the opposite hypothesis, propounded by Lindbeck, Weibull (1987) y Dixit, Londregan (1998) is also corroborated. Curiously, the pivot variable is not significant in the basic scenario, but it is in two scenarios of the fixed effects model with a negative sign. This introduces a certain contradiction into the results, since it would appear to indicate that the allies which supply political support to sustain the central government receive fewer resources. With regard to the explanatory factors of an economic nature, representative of normative hypotheses, the results show that these transfers are characterised by their regressive character. Finally, from the results we can also deduce that this type of transfer is used as a compensatory system for the distributive results of the remaining transfers. Thus, as the system of autonomous financing favours the ACs with lowest GDP, above all in the financing of investment, it appears that the central government utilizes the discretionality permitted by the investment agreements to compensate those who lose. This conclusion is reinforced when we observe the opposite bias affecting the ACs in the foral system, which barely receive transfers of this type. This may be the consequence of the fact that, as the foral system produces financing distinctly greater to that of the ACs in the common system, the central governments are not willing to transfer additional resources by the path of investment agreements. With respect to the possible partisan and political use of these discretionary transfers, there is also in play the fact that homogenised data regarding them ceased to be published precisely when they reached their greatest level. The study is structured as follows: after the introduction section, the next section reviews the empirical literature, followed by a specification of the econometric model employed, with a description of the variables and data sources; subsequently we present the results and we finish with a brief discussion section.