Allocating Collective Expenditure: The Case of Education

Creating distributional national accounts (DINA; e.g. Piketty, Saez, and Zucman 2018) requires the allocation of all government expenditure to individuals in order to compute their post-tax, post-transfer income. A sizeable part of government expenditure is in-kind spending, either in the form of individualized transfers (e.g., Medicare and Medicaid) or of collective consumption expenditure (e.g., education, defense, and the general legal and administrative infrastructure). Because of data limitations, the existing DINA studies allocate the collective consumption expenditure either proportionally to post-tax cash income (in which case it is distributionally neutral) or as a lump-sum transfer. In this paper we provide evidence on the way some of the collective consumption expenditure is actually distributed. We focus on public spending on education, which makes up about 5% of national income in most OECD countries. We find that, in Germany at least, education spending tends to go disproportionately to the bottom half of the post-tax cash income distribution, so the proportionality assumption adopted in the DINA literature does not work very well in the cross-section. However, this regressivity is driven by strong age effects. Moving beyond the cross-section, we find that individuals with higher lifetime earnings or better educated parents have indeed received substantially more in terms of public education spending.