Wagner’s Law, Public Employment, and Spanish Regions

Abstract

Abstract Following Peacock and Musgrave’s rediscovery of Wagner’s Law, the latter became a standard tool used in research on the relationship between growth of public spending and the factors by which it is influenced. However, conventional empirical tests are based on a specification error related to Wagner’s definition of the public sector, which he considered in its totality, including public companies. The present article attempts to correct this error and obtain an approximation to the size of the public sector by considering public employment as a whole, both in public administrations and services and in public companies. To this end, panel data for the Spanish autonomous regions are used in addition to data for the overall public sector. The empirical test is performed utilizing cointegration techniques and unit roots in panel data. Similarly, the possibility of structural breaks in the data is taken into consideration and they are estimated using fictitious variables. JEL classifications: H11; H50; E62 Keywords: public employment; gross domestic product; unit root; cointegration; panel data