English or German or Both: Recipes for Developing Countries – Econometric Evidence from Aggregated and Disaggregated Data

Since its inception, Wagner’s law has gained the attention of researchers and is well-documented in literature. It deals with the relationship between the increase of government expenditure and improved macroeconomic performance over time. On the contrary, Keynesian theory purports an opposite causality between the two. This paper contributes to literature by investigating this long-term relationship between government expenditure (aggregated and disaggregated) and income per capita in developing countries. The Pooled Mean Group (PMG) approach to cointegration is employed on data of 76 developing countries for the years 1990-2012. As a heterogeneous panel estimation technique, PMG allows the slope and short run parameters to vary across countries. Results show the presence of a long-term relationship between government expenditures and its components and income per capita. Causality analysis is also conducted in this paper that provides insights into the relationship between income per capita and government expenditure (aggregated and disaggregated). Possibilities of Wagnerian and/or Keynesian causality(s) are explored. Recommendations are made on the basis of empirical analysis.