Abstract
One of the ideas claimed to justify the Osaka Metropolis Plan (Osaka-Tokoso) is that fiscal efficiency of Osaka city (or Osaka and Sakai cities) can be improved by dividing the city into several wards because the population of each ward will become closer to the optimal size of municipal population than that of Osaka city. Although previous studies have shown that local government spending is minimized in medium-sized cities, the methodologies of detecting optimal size and interpretations of the results still have problems to be solved. In addition, most of the studies focus on “expansion” through municipal amalgamation, so it is unclear whether the theories can be applied to cases of “down-scaling” by municipal division including the Osaka Metropolis Plan. In this study, we reviewed the methodologies and the results of previous studies that estimate optimal size(s) of municipal population in Japan, summarized the implications from them and discussed remaining problems. This review and discussion provides understandings as follows. Many studies have found ‘U-shape’ relationship between municipal population and public spending per capita, which means that a population growth leads to lower average spending in relatively small cities but the opposite it true in the largest cities. However, there have also been studies that do not support the ‘U-shaped’ trend (that implies no optimal population exists). Furthermore, causal mechanisms of increase/decrease in government expenditure per capita is not clear enough, so that we should be careful assuming that population changes by municipal amalgamation/division can improve fiscal efficiency. Particularly, if the increase in public spending in large cities is a result of urbanization as previous studies suggest, it is difficult to apply the theory of optimal size to the cases of municipal division.
