Non-Linear Dynamics of Macroeconomic Variables: The Case of Greece
This paper is part of an ongoing doctoral dissertation. It examines the long run relationship between government spending on education, GDP growth and educational enrollments in all levels, for Greece in the years 1960-2012, leveling up to capture more complex dependent dynamics in time series. OECD Indicator A11 shows that, life satisfaction, civic engagement and health. are at the same level for adults with tertiary education and adults with upper secondary education for Greece. Having conducted Linear Granger tests, this study goes beyond, following the methodology of both non-parametric testing for asymptotic Granger causality in C, and non-parametric bootstrapped bidirectional Granger causality, an MS Dos application in C. We apply a new nonparametric test for Granger non-causality by Diks and Panchenko (2005, 2006) as well as the conventional linear Granger test on the time series. The secondary data used in this present study, have been processed through MATLAB interpolation function. There is a strong symmetric nonlinear causal relationship detected between government spending on education and enrollments in higher education, and vice versa, .An even stronger, both symmetric and asymmetric nonlinear relationship between higher education enrollments and Gdp growth is validated confirming a spectacular case of “adverse causality”. Government spending in education is causally related with Gdp growth in a symmetric as much as an asymmetric nonlinear relationship. Also, lagged values of secondary enrollments are a major predictor of Gdp growth.
Keywords: Economic growth, Greek economy, Granger Causality, Non-linearity.