Deep determinants of economic growth – empirical verification with panel data models

We verify the impact of the so-called deep determinants on the level of economic real GDP per capita for an unbalanced panel of 207 economies within the period 1996-2004 using the Hausman-Taylor method of estimation. Institutional variables are detected to be endogenous. The results confirmed the assumed impact of deep determinants on the observed disparities in economic development. In most cases the basic specification of the model suggested by the empirical literature (log of openness, rule of law, distance from equator) is statistically significant and the impact of the variables has the anticipated direction. Several other specifications are tested and they perform pretty well. As the distance from equator has been detected not to be statistically insignificant in several specifications (for Asia and Europe) a combination of exogenous geographical variables enters the model with positive results. The basic specification of the model fits well the context of Africa and South America. It however performs badly for Asia. The quality of institutions is of prime importance for southern hemisphere economies as well as for former (currently economies in transition) and current socialist economies. The permanent improvement in the quality of institutions is the key determinant of success of economic transformation – underperformance in this area leads to smaller gains in terms of GDP per capita levels attained.