Working Paper: CEPR ID: DP3617
Authors: Riccardo Fiorito; Tryphon Kollintzas
Abstract: In this Paper, we investigate the relation between public and private consumption, by constructing a general government spending dataset, by function, for twelve European Union countries. In particular, we split government consumption into two categories. The first category includes defence, public order, and justice ('public goods'). The second category includes health, education, and other services that could have been provided privately ('merit goods'). Equations from a relatively general theoretical model of household behaviour are estimated by GMM. The estimates are fairly robust in showing that public goods substitute, while merit goods complement, private consumption, and that the relation between merit goods and private goods is stronger than that between public goods and private goods. So that, in the aggregate, government and private consumption are complements. It also suggests that the potential calibration/estimation bias by ignoring the composition of government consumption might be substantial.
Keywords: Dynamic Panel Estimation; Government Consumption; Merit Goods; Public Goods
JEL Codes: C33; E60; H30
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Public goods (H41) | Private consumption (D19) |
| Merit goods (H49) | Private consumption (D19) |