In a few days the municipal elections will be held throughout Spain. More than eight thousand city councils will renew their local political representatives after knowing the verdict of the polls. At Piedras de papel we have always wanted to bring some of the debates, findings or hypotheses of the academic world closer to current issues, and this is an excellent opportunity to talk again about the phenomenon of political economic cycles (political business cycles in English).
In short, these cycles reflect a pattern of behavior by governments in terms of taxes, spending, indebtedness or many other policies that accommodate their application to the times of the electoral cycle. For example, a tax reduction, an increase in subsidies, an increase in positions for public officials, the start-up or inauguration of works (parks, roads, schools), all in a period close to the electoral date. As is logical, this strategic behavior aims to maximize the probability of winning elections at any level of government (whether local, regional/autonomous or state) trying to manipulate the feeling of social and economic well-being in the context that occurs.
With regard to local governments, intuitively we can think of actions of great importance, such as the inauguration of large public works, but also actions with less shine, such as the inauguration of a roundabout. The truth is that there is an extensive literature that offers evidence on how local governments in Spain and in the rest of the world make a clear strategic use of time in their policies to maximize re-election. I will only mention two references for the most curious. The first is the work of Bastida, Beyaert and Benito (2013) where, using data from 238 Spanish municipalities between 1992 and 2005, they show that the level of indebtedness increases as local elections approach, particularly in the richest municipalities and in local entities without an absolute majority. The second, in a post on this same blog, signed by Pablo Fernández Vázquez, in which he shows, with data for the group of more than eight thousand municipalities in our territory between 1999 and 2014, how municipal governments act on politics. prosecutor to try to ensure his survival in office.
In this post, we will contribute to adding some evidence by extending the data to the electoral appointment that awaits us just around the corner, concentrating on the municipalities of the Community of Madrid. On this occasion, using data from the Ministry of Finance and Public Function for the period 2010-2022, we will contrast the Total Budgeted Expenditure of Local Entities in pre-election years vs. the years No pre-election We will analyze the budgeted expense instead of the liquidated one to, precisely, be able to extend the analysis to the last pre-election year, that is, 2022, for which there are still no liquidation data. Likewise, we will explore the differences between political parties when making use of the budgetary electoral cycle. To do this, we will divide the local entities between those governed by the PP (57%), the PSOE (23%), IU/Podemos (or similar candidacies such as Ahora Madrid, for example) (3%), Ciudadanos (2%) and candidacies independent (15%). When a municipality has had more than one mayor/mayoress from different parties in the same legislature, we have stayed with the party that has governed during the pre-election year. And to control the possible effects of the size of the municipalities, we will divide the total expenditure by their number of inhabitants, resulting in the variable under study being the total expenditure in euros. per capita, although we will refer to this measure as spending in euros.
To get an idea of the magnitude of what we are talking about, it is useful to know that the average expenditure in our database is approximately 1,400 euros per inhabitant, with a wide standard deviation (1,250 euros), and that a municipality in the median of distribution spends 1,000 euros. The first sign that there is a higher level of spending when the time to vote for our local representatives approaches is seen by comparing the average spending in the pre-election years of the Madrid municipalities (2010, 2014, 2018 and 2022) with the years No pre-electoral (2012, 2013, 2016, 2017, 2020 and 2021; election years are not included here). For the former, the average expenditure per capita it is 1,455 euros, while for the others it is 1,355 euros, that is, 100 euros less per inhabitant. This represents 10% of average spending.
As we can see in graph 1, some parties already exhibit, in terms of average, some differences in spending in relation to the electoral cycle. Almost all parties spend more in pre-election years than in election years. No pre-election The PP and the PSOE do not show great differences between these two periods (the first only spends an average of 23 euros more, while the second 81 euros more), but the municipalities under the command of independent candidacies do (435 euros more the previous year to the elections) and those governed by the parties to the left of the PSOE (203 euros more). The exception are the municipalities governed by Citizens who, against it, spend 200 euros per capita less the year before the elections. However, this contrast of means is still a broad statistical test and only reflects statistically significant differences in the case of independent candidacies after performing a test of means. To more carefully explore the effect of electoral cycles and their differences by party, we will carry out a regression analysis where we can control for the potential influence of the particularities of each municipality (fixed effects) and temporal trends.
Said analysis shows, firstly, that the pre-election years, in contrast to the years No pre-election, have an effect on total spending of 114 euros per inhabitant (somewhat higher than our mere mean test). Graph 2 shows the effect of the pre-election year when we interact with the different parties in local government. Exploring the variation within each municipality and controlling for the time factor associated with the different contexts of the Spanish economy, we can now see that, again, all the parties -with the exception of Ciudadanos- increase the spending of local entities the year before the elections. The PP spends approximately 100 euros more, the PSOE 70 euros more, the independent candidacies 260 euros more and the governments of Izquierda Unida/Podemos 190 euros more. The councils governed by Ciudadanos, going against the current, spend 180 euros per inhabitant less in the years prior to the municipal elections. These differences, however, are particularly significant from the statistical point of view only for the PP and for the independent candidacies. In the case of the IU/Podemos and Ciudadanos governments, we may not find statistical significance due to the low number of observations (together they only represent 5% of municipal governments).
Based on these results, we can make three notes about the electoral cycle of spending in the municipalities of the Community of Madrid. The first, and most obvious, is that increasing spending in pre-election years is a widespread practice with virtually no partisan distinction. Governing and manipulating the times of political action seems natural to the exercise of power itself. This issue -as the literature on credit economic cycles has been doing- opens the debate on the convenience of some type of regulation when this exercise of budgetary opportunism undermines the solvency of our public accounts. It is something complex to elucidate, but, not for nothing, there is a prohibition on inauguration acts, as well as institutional campaigns once the elections are called. The second point is that contrary to the discourses around public spending policies that we usually assume from conservative and progressive parties -and specifically from the two main formations of our party system- we find that the party of the right with With the greatest representation in the Community of Madrid, the PP does not spend less than the PSOE, and which also opens the faucet of the municipal coffers especially in the years prior to the elections. Third, the independent candidacies -generally centered on local platforms, with an anti-traditional party discourse- do not escape the opportunistic dynamics of the exercise of power. This result is quite interesting since it is surprising how, with a relatively low number of observations, we can capture a positive effect, of relevant magnitude and statistically significant, when we contrast their spending pattern in pre-election years with other years. Without being familiar with this -surely- heterogeneous world of independent candidacies, I can only provide a series of questions by way of hypothesis. Are those formations without a strong brand -recognizable parties- the ones that most need to use the instrument of spending to be competitive in electoral contests? Are they more prone to the electoral spending cycle because they are more clientelist candidacies or with more personalist leadership? Ideas to ponder.