The effect of legislature size on public spending: evidence from a regression discontinuity design

Abstract

What is the effect of legislature size on public spending? An answer to this question is provided by Weingast et al. (J Polit Econ 89(4):642–664, 1981), whose “law of 1/n” posits that an increase in the number of elected representatives always leads to an increase in public spending. Because elected politicians regard the tax base as a common pool from which they can finance specific projects for their constituencies, and these specific constituencies internalize the full benefits of the projects, but only bear a fraction of the costs (projects are financed from the common tax base), fiscal inefficiency will increase with the number of representatives. In this paper, I test the validity of the “law of 1/n” using a dataset of 9325 German municipalities between 2008 and 2010. Through the application of a regression discontinuity design, many of the methodological pitfalls of previous studies can be avoided and a valid estimation of the causal effect of legislature size on public spending for German municipalities can be determined. The results do not corroborate the positive findings of previous studies, which generally supported the implications of the “law of 1/n”. For the years 2008–2010, I find a negative effect of legislature size on public spending in German municipal councils.