On the optimal size of town government

How big should a town be? Should it merge with its suburbs? Should neighboring villages agglomerate? What is the best outcome for inhabitants?

Large towns may count on some economies of scale, say in policing and in task that require a substantial fix cost (administration, sports facilities, schools). However, they are more likely to have bloated bureaucracies, as voters have less influence to keep them in check.

Small towns often have the tendency to free ride on larger neighbors for facilities. This is especially true for suburbs, leading to the problematic situations of the core city providing services but be sucked dry of people moving for lower taxes or better taxes. This situation thus calls for a forced merger of suburbs with the core city. How much to merge needs to take into account the administrative inefficiencies mentioned above.

Instead of mergers, a two layered systems can be put in place, where towns remain largely independent but a still forced to participate in some regional structure for some tasks, like public transportation, police, environment and zoning. This avoids the free riding, allows to exploit economies of scale where they are and still leaves local control of most of the budgets.

Why do I write about this? Observing the huge income and service disparities within metropolitan areas in the United States, one has to wonder how such ghettoization could happen. People vote with their feet, and in most cases they did this by fleeing to the suburbs, where lower taxes and better demographics reinforced each other and attracted the good risks. This is only possible if the remaining poor city core is small enough. By agglomeration, the city core becomes less poor, and the attraction towards the periphery lessens. Merging is the only chance for the American city to become attractive again.