The determinants of economic growth of transition economies: Economic reform versus initial conditions

Abstract

This study analyzes factors for economic recovery of transition economies in Europe and the Commonwealth of Independent States for the period of the 1990s. Covariance structure analysis is employed to estimate the structural equation system, and exploratory factor analysis is conducted to measure initial conditions and economic policy as latent variables. The result of analysis shows that the effect of initial conditions is negative and the impact of economic reform on growth is positive. However, the negative effect of initial conditions had overridden the positive impact of economic policy as of 2000. The reason that transition economies could not recover their pre-transition GDP level (even after ten years of transition history) seems to stem from the negative influence of initial conditions on growth rather than the slow speed of economic reform.

Keywords:

• Transition economy
• growth
• economic reform
• covariance analysis
• initial conditions

Acknowledgment

The authors are grateful for the financial support of Yonsei Center for Global Studies, the Graduate School of International Studies, Yonsei University.

Notes

¹Campos & Coricelli (2003) well describes the economic performances of the transition economies in the 1990s.

²Some studies on transition economies, which employ simple regression analysis, such as Berg et al. (1999) and Polanec (2004), can include this kind of qualitative variables as independent variables. However, this study is using structural analysis of covariance, which needs a few latent variables as causal variables extracted from so many independent variables using factor analysis. In factor analysis, it is very hard to extract meaningful principal components from the data set in which qualitative variables are intermingled with quantitative variables.

³The areas on which EBRD bases itself in assessing the transition indicator are ‘market and trade,’ ‘corporate sector,’ and ‘financial sector.’ The transition indicator uses a scale from 1 to 4 + according to the success of economic reform in each area, and it uses the average scale of each area as its indicator. If scale is 1, it represents no change under the rigid socialist economy, while 4 + represents a role model for the market economy. See EBRD (1999: 16).

⁴For interpretation of the statistics of the analysis results, see Noh (2002: 256–283). The AMOS 4.0 package automatically produces non-standardized regression results and standardized regression results.

⁵The coefficient estimates shown above are obtained from standardized regression in covariance analysis in which the sign and the size of the estimates are important. The reason to use standardized regression is to focus on direction and relative influence of each latent variable, instead of on quantitative prediction of the dependent variables. In the structural equation model of covariance analysis, the significance test of coefficient estimates for null hypothesis is conducted using critical ratios (CR) of non-standardized regression results, which have more or less the same statistical characteristics as t-values in general regression analysis (see Noh, 2002, chapter 10.). The values of CR of the coefficient estimates are shown in the appendix.

⁶If we compare the GDP changes by region, Mid European countries are achieving continuous growth after exceeding the pre-reform GDP level while the CIS and Baltic states are still struggling to reach pre-reform GDP level.

⁷IMF (2000).