Strategic behaviour by Italian mayors differs by age

In my ECONS102 class, we cover a little bit of the theory of public choice. One aspect of that is the way that politicians behave. Of course, voters want politicians to act in the voters' best interests. However, there is a principal-agent problem here. Voters are the principal - they have engaged the agents (politicians) to act on their behalf. The problem is that politicians have their own motivations, which don't necessarily align with the goals of the voters, and if they aren't monitored closely, politicians will act in their own best interests (and not necessarily those of the voters).

That disconnect, between the goals of voters and the motivations of politicians, explains a lot of the reason why politicians act in ways that voters don't necessarily like. And, voters may let them get away with it because of rational ignorance. Monitoring the behaviour of politicians is costly to voters. If the costs of any negative behaviour of politicians is less than the cost of monitoring them, then it is rational for voters to avoid the monitoring costs and be rationally ignorant of the politicians' behaviour.

One of the things that politicians may do, which may not align with the goals of the voters, is engage in strategic behaviour designed explicitly to get the politician re-elected. They may provide 'favours' or re-direct resources to some groups of voters, or engage in high profile but costly (in terms of public welfare) activities that give them good press coverage. This strategic behaviour is likely to intensify in the lead-up to an election.

That brings me to this new article by Alberto Alesina (Harvard), Travis Cassidy (University of Alabama), and Ugo Troiano (University of Michigan), published in the journal Economica (ungated earlier version here). Alesina et al. use data on Italian mayors over the period from 1998 to 2014, to test how younger and older mayors behave differently.

In theory, there are a number of reasons to believe that younger mayors will behave differently from older mayors. Alesina et al. note that:

Younger politicians may differ from older ones for at least five reasons. One is that they have a potentially longer political career ahead of them and therefore have stronger career concerns. The second is simply that, as younger citizens, they have a longer horizon and therefore they may have an incentive to adopt more long-term policies... The third and more mundane reason is that younger politicians may be more energetic and productive at work. The fourth reason is that there could be different self-selection patterns by age: because people of different ages have different opportunities in the labour market, this may affect the decision to become a politician. The fifth reason is that politicians of different ages may have different political connections—innate or accumulated doing the course of their previous work.

The career-concerns reason is interesting to consider. Younger politicians have a potentially longer political career ahead of them. So, the benefits of engaging in strategic behaviour to ensure their re-election are higher for younger politicians than for older politicians. Alesina et al. find some support for this. After showing that younger candidates are statistically significantly more likely to be seek re-election, and more likely to be re-elected, than older candidates, they consider whether these results are indicative of younger mayors enacting better policies. They investigate this by looking at house prices (because areas with better local government policies would be attractive places to live in, so house prices would be higher). However, they find that:

...there is little evidence that house prices respond to the age of the mayor in office.

They also show that the age of the mayor has no effect on the speed of public goods provision (another indicator of the quality of governance). So, having established some evidence that younger mayors are not better politically than older mayors are, they consider how younger and older mayors differ in terms of local government spending. They find no statistically significant differences in local government revenue per capita or current expenditure per capita. However, with capital expenditure they find evidence of a political cycle in spending. That is, younger mayors spend more in the year immediately before an election, such that:

...a typical younger mayor will increase capital expenditure before the election by 9.66 euros per capita, while a typical older mayor will actually decrease capital expenditure before the election by 2.1 euros per capita.

Capital expenditure is attractive for political candidates to be strategic about, because it is highly visible (voters will remember the politician who cut the ribbon on their new expressway), and because unlike current expenditure it is less subject to balanced-budget rules.

If mayors are engaging in strategic behaviour, it is reasonable to consider how voters can respond to try to reduce this behaviour. Closer monitoring is unlikely to be effective because of rational ignorance. So in my ECONS102 class, we discuss alternative solutions to the principal-agent problem, one of which is paying an efficiency wage - effectively paying a higher wage to the agent, to increase the incentives for them to engage in behaviour that is aligned with the principal's wishes (otherwise, they lose their job and have to go work elsewhere for a lower wage). Interestingly, in supplementary results, Alesina et al. find that:

...the effect of age on political budget cycles is smaller when the mayor’s wage is higher...

So, maybe there is some support for increasing mayoral wages to reduce this strategic behaviour.