Articoli

Council Size, Government Spending and Efficiency. Evidence from a Quasi-experimental Design for Italian Municipalities

Marco Alberto De Benedetto, 2018. Council Size, Government Spending and Eciency. Evidence from a Quasi-experimental Design for Italian Municipalities (doi: 10.1429/92122) Politica economica – Journal of Economic Policy (ISSN 1120-9496) Fascicolo 3, dicembre 2018

In modern democracies do politicians allocate resources efficiently?

A conspicuous body of literature has tried to understand how politicians make decisions, especially in terms of budget choices and how executive fragmentation (i.e. the number of spending legislators) affects governments’ budgets.

According to the Public Choice theory, especially relying on Tullock’s (1959) and Buchanan and Tullock’s approaches (1962), politicians, in order to obtain a greater electoral support, try to satisfy the needs of particular groups of constituencies to the detriment of the general community. In fact, government spending typically benefits a narrow segment of the population, whereas taxes are distributed broadly.

In a new piece of research, “Council Size, Government Spending and Efficiency. Evidence from a Quasi-experimental Design for Italian Municipalities” in Politica Economica – Journal of Economic Policy, Marco Alberto De Benedetto presents new evidence about the relationship between legislature size, government spending and the level of efficiency at the local level in Italy. Indeed, investigating the effect of legislature size on government spending, especially at local level, is a relevant issue in the current debate since municipalities provide public goods and services, such as the management of public utilities, transportation and nursery schools, assistance of elderly people and waste collection that have a big and direct effect on citizens’ daily lives.

The empirical analysis is based on a rich panel dataset on municipal budgets over the period 2001-2007, provided by the Italian Ministry of the Internal Affairs, of approximately 6,576 local governments, containing information on both inflows and outflows.

The available data on the number of councilors within each municipality allow to implement a Sharp Regression Discontinuity Design (RDD). Exploiting the Legislative Decree n. 267/2000, in fact, the Italian law establishes that the municipal council is composed by the mayor and by a variable number of local legislators, i.e. councillors, which depends on municipal population size. The idea behind the Sharp RDD methodology is therefore to compare municipalities just above the population size thresholds to municipalities just below it in order to distinguish the causal effect of legislature size on local expenditures; indeed, unobservable characteristics should not vary discontinuously around the cut-off and then the cut-off rule provides exogenous variations in the treatment «as good as a randomized experiment».

Overall, results highlight a negative and statistically significant effect of council size on total expenditures. In particular, if the council size increases by 1 unit, the total municipal expenditures per capita decrease by 0.4%. Furthermore, in order to reinforce the hypothesis that an agency problem in terms of spending could rise between legislators and bureaucrats, the author focused on expenditures that are more directly under control of bureaucrats, i.e. current expenditures per capita, after controlling for the share of employees within local governments, for a proxy of the quality of public goods and services provided by local governments (the ratio between waste recycling over total urban waste), and for a measure of population civicness, i.e. electoral participation. Also in this case, a negative effect of legislature size on government size at the municipal level in Italy emerges. Further results also highlight how council size positively affects some measures of efficiency/productivity at municipal level, namely the difference between revenues and expenditures, the speed of revenues collection and the speed of payment.

One possible interpretation of these findings is that an increase in the number of councilors leads to a better monitoring of bureaucrats in terms of spending, mitigating the agency problems between legislators and bureaucrats leading to a lower level of spending and to a higher efficiency/productivity of municipalities.

Future research on this topic suggests to investigate whether the relationship between legislature size and government spending is non-monotonic (U-shape relationship). In fact, it may be the case that when legislature size is small, then an increase in the number of legislators leads to a better monitoring of bureaucrats in terms of spending, heightening the efficiency of the entire public administration. Conversely, the efficiency of the local administration might tend to reduce if legislature size is big enough to guarantee a good management of the «public pursue».