EU Economic Growth and the Age Structure of the Population

In this paper we test the robustness of correlations between age structure and economic growth in the old EU countries. The model is based on a modified Solow model in transition to steady state. We control for the catch-up effect not only by the conventional initial income but also introduce a technological convergence or absorption factor. In practice this is achieved by weighting the regression with the gap between GDP per capita in USA and each country included. Data are organised as a panel with 5 year periods and estimated controlling for fixed effects. A hump-shaped relation is found with the age structure where increases in the dependent age group shares are associated with decreasing GDP growth rates. This is in accordance with findings in most other work. In addition we confirm previous findings that the peak of the hump is in the upper part of the working age population implying that early retirement practices may be extremely harmful to growth. We find that models estimated on data up to 1990 agree quite well when data for 15 more years are added. Sensitivity tests, instrumented regressions and addition of control variables does not change this conclusion. The model parameters that we estimate are country- and time-specific (due to the weighting