This is a really cool paper by Kristof De Witte, Benny Geys, and Nanna Lauritz Schönhaged in Journal of Urban Economics. Mayoral salaries in Belgium are based on a population schedule, so you might expect politicians to favor policies that can bump up their pay grade.

Political economists have long maintained that politicians respond to both (re-)election and financial incentives. This article contributes to the latter literature by analysing whether, when and how local office-holders respond to the economic incentives embedded in exogenously imposed population thresholds leading to an increased number and remuneration of local politicians. Building on insights from the urban economics and public finance literatures, we argue that local politicians may strategically adjust fiscal and housing policies to stimulate in-migration when approaching a population threshold where their remuneration increases. Using data from all 589 Belgian municipalities over the period 1977–2016, our results confirm that approaching important population thresholds causes lower local tax rates and the granting of additional building permits (particularly for apartments). These policy changes occur early in the election cycle and, at least for housing policy, are restricted to incumbent mayors themselves expecting to benefit from crossing the population threshold.

The paper is written in an interesting way that I generally describe as the “detective novel.” The paper walks through a series of findings, asking interesting questions and answering them along the way. To some extent, the next question explored is conditional on the finding from the previous question. This is a different from a paper that asks a couple of different questions defined at the start because they can be independent from one another. I find these papers more compelling, but they are risky to write as an author because revisions have downstream effects on the paper, so the authors are to be commended for it.

Here is an outline of those questions and my summary of their findings:

1. “Is there evidence of population bunching around the threshold?” Not really, not to the eyeball or a few statistical tests, anyway. So perhaps they are not good at it because they don’t have a really sharp mechanism for manipulating the population figures.
2. “Is there evidence of attempted bunching from a plausible mechanism?” In building permits for apartments, local governments just below the threshold are more permissive than those above, albeit the confidence intervals are somewhat large. However, there aren’t comparable changes in placebo areas like renovation permits. On tax policy, local income and property tax rates suggest some bunching around thresholds, but they don’t seem to be particularly good at this approach in that the rates stick to either side of the threshold.
3. “Is the timing of the threshold determination date is important relative to the election cycle?” In other words, did the mechanisms demonstrate heterogeneity based on whether we are about to undergo another election? De Witte et al.’s results suggest the cycle does matter, as the results converge towards zero as the election date approaches. Perhaps mechanisms like building permits are too slow to be of any value within a few months, and are likely to disturb your current voters in the meantime, or you don’t want a big raise just as you are about to go up for reelection.
4. “Does it work to get you re-elected?” This is kind of important for guessing at the welfare inference. If getting across the population thresholds increases the probability of reelection, then this is less policy manipulation to get a raise and more doing what voters want. After all, a mayor who is doing well should attract new residents, and you could plausibly think voters would want to reward that. But it should not uniquely matter at the thresholds if that is the case. Indeed, mayors are no more frequently re-elected when they cross thresholds than not, so it seems unlikely that seeking rewards from the electorate is at least not a valid justification for mayors to attempt this (albeit, they may be doing so on the mistaken belief).

It is a neat paper. I would also recommend Ed Glaeser’s “The Curley Effect: The Economics of Shaping the Electorate” if you wanted to think about how to distinguish competing hypotheses over how political actors are motivated in altering the demographics of their populace.