Determinantes del gasto social de las Comunidades Autónomas en el período 1996-2011

Résumé

ABSTRACT This study aims to investigate the determinants of social spending throughout the period between 1996 and 2011, for the seventeen Autonomous Communities (CCAA) that exist in Spain. Social spending will be divided into three main sections: Education, Health and Social Protection. For this purpose, we take budgets for each of the CCAA, bearing in mind their functional classification, that is, the final destination of the financial resources budgeted. To achieve this, we firstly carried out a descriptive analysis of the autonomous budgets and then we focused on the study of the political, social and economic factors that determine or influence these policies. From the statistic analysis carried out on the relative percentages of the budget areas and the per capita expenditures, we highlight the increase in real budgets until 2009, which was higher than the GDP growth, despite the increase in population in that period. We have also to note the high relative weight of the area "Key Public Goods Production", exceeding 55% of the whole budget. It should also be noticed that in the order of relative importance of the various items of autonomous expenditure, Health expenditure stands first (33%), followed by Education expenditure (22%), being the Social Protection policy in the fourth place until 2008, only to come up a position during the crisis years, 2009 and 2010. As for the distribution of the different per capita budgets for Health, Education and Social Protection, we should mention the lack of homogeneity among CCAA. The econometric analysis is based on a revision of the literature about the expenditure determinants for each of the analyzed policies and, according to this, we have chosen the variables to be included in our regressions. Once we have defined our sample, we note that these data consist of observations on an individual unit sample repeated over time, which is a well-balanced panel. In order to work with these data, an efficient and unbiased option is to use the instrumental variables and the generalized method of moments (GMM). We conducted multiple regressions for each of the sections of expenditure. It is worth noting that among the results of Health expenditure, the lagged dependent variable turns out to be significant; that the population variables measuring certain age groups have limited influence; and that the proportion of foreigners does condition positively a higher Health expenditure. Finally, the income level and a higher expenditure on technology have a less relevant influence than it was initially expected. From the results of the analysis of expenditure in Social Protection it is worth noting the positive influence of both progressist governments and the emergence of the economic crisis, as well as the positive impact of a higher rate of unemployment and the higher rate of over-65-year-old population on this type of policies. Finally, with regard to spending in Education, we should mention that former educational spending does have an influence on the current one, but in a negative sense; that left-wing governments spend more and that a plural society means more Education spending and that a higher economic standard also involves a higher expense.