Long-run causal effects of public investments : economic growth and the provision of schooling. Greece 1960-2015. Parametric and nonparametric approach

• Main Author: Κασάπη, Σοφία
• Other Authors: Συριόπουλος, Κωνσταντίνος
• Format: Thesis
• Language: English
• Published: 2016
• Subjects: Government expenditures; Economic growth; Time-series forecasting; Parametric; Non-parametric; Οικονομική ανάπτυξη; Παραμετρική μέθοδος; Μη παραμετρική μέθοδος; 379.495

The question addressed here has been whether public investments in the provision of statutory schooling promote economic growth, in terms of GDP per capita, for the case of Greece. This dissertation joins a vibrant discussion of whether public investments compose the necessary pedestal for the private investments to increase, whenever they are targeted towards new infrastructures, providing better market conditions for all business sectors. This study adopts the standard parametric econometric toolbox developed upon Granger causality methodology (1969) and goes one step forward in the direction of the nonparametric analysis. The non-parametric test used in this hypothesis testing is the Diks and Panchenko (2006) modified test statistic, initially proposed by Hiemstra and Jones (1994). The original HJ is not used due to its over-rejection problems. Consequently, the Diks and Wolski (2013) “data sharpening” extension to the DP nonparametric Granger causality test is being applied, in order to address the consistency issues of the basic DP test statistics in multivariate settings, regarding the kernel estimation bias. By applying nonlinear nonparametric Granger C codes on VAR and E-GARCH residuals, annual time series which were converted into quarterly data through cubic spline interpolation method. For the analysis of the data, MATLAB functions and C Codes were developed and used. The results taken are consistent with more general findings on the impact of public spending on economic growth (ECB, 2008), and do not justify these levels of public investment in schooling, in terms of growth. The promotion of free public schooling does not promote economic growth for the country in most levels, especially higher education and middle school, while other intermediate causal effects revealed show us an even more complex situation of interdependencies and synergies. Even according to the European Central Bank (ECB, 2005), public investment in human capital and R&D can be a main growth element, under certain conditions. It seems like this expansion in years of schooling alone, was not the only measure needed to provide for better economic conditions. The controversy here implies the importance of cognitive skills in promoting economic growth (Hanushek, E., 2013), and the need to shift our focus on quality for both basic skills and high achievers, either to alleviate poverty, or raise their level of income respectively. Data were collected from the Greek Statistics Bureau (ELSTAT) and AMECO for the 1960-2015 time. Unfortunately, the data collection process dealt with important data limitations. The lack of consistent and comparable data, and the unwillingness of the officials to cooperate led this study to shrink from a panel data analysis that was previously intended to a time-series one. Hopefully in the near future, another study will provide for more qualitative elements on the “uncertain role of human capital on growth”. Thankfully, the analysis gained robustness from the methods employed, and provided with a clear picture on public spending for schooling in the beleaguered Greek economy.