Optimal threshold taxation : an empirical investigation for developing economies

In this empirical study we assess both linear and nonlinear relationship between total taxation and several tax items with real per capita GDP growth rates for 43 developing countries between 1990 and 2019. We use panel data techniques to evaluate the effects of taxation on economic growth for both short and long run perspectives, and to find optimal tax threshold values. We obtain evidence of nonlinear relationships between all tax items, except for corporate income taxation, as well as an optimal value for total tax burden around 23,5% of GDP for the whole sample. When the sample is subdivided by countries' income levels, we find threshold values for all tax items and an optimal tax burden around 23,6% of GDP for high income countries and 21,3% of GDP for low income. Our results provide support regarding the existence of nonlinearities and about policies focused on raising certain tax revenues, as a percentage of GDP, without hampering economic growth.