A Study on Regime Type and Globalization in Simultaneous Equation Framework

In this study we build a simultaneous equation model in which the measures of different aspects of globalization (attributable to KOF) and different aspects of democracy (attributable to EIU) are related in seven structural equations. A bi-directional relationship between democracy and globalization is visualized. The model is estimated by the conventional 2-SLS as well as a modified 2-SLS in which Shapley value regression is used at the second stage of 2-SLS. On the basis of our analysis, we document several findings. First, we find that democracy and globalization promote each other and hence there is a bi-directional causality with positive relationships running both ways between democracy and globalization. At a national level, there may be various intermediary conditions that modify the relationship as well as set in motion a complex of positive and/or negative feedbacks to accelerate or retard the pace of globalization and democratization in a country-specific manner. However, when a large number of countries are studied, a clear relationship emerges out. Second, there is a need to estimate the structural coefficients of the model cautiously since the regression equations may be suffering from collinearity among the predictor variables. The Shapley value regression based 2-SLS has performed better than the conventional regression in estimating the structural parameters of the model. Third, the system methods of estimation of the model gives better results than what are obtained by the single equation methods of estimation of structural parameters of the model