Reform reversals and output growth in transition economies*

This paper tests whether there is a macroeconomic cost of a reform reversal during transition. A reform reversal is defined as a downgrading in the level of an average reform indicator. This is important both from an empirical and a theoretical point of view. In the standard empirical framework the current level of reform affects growth negatively, while the lagged level affects growth positively. This nonlinear effect is shown to imply a counterintuitive, short-lived, or at best an insignificant, positive effect of a reversal. From a theoretical point of view however, most models assume a reversal to be costly. The existence of reversal costs is even crucial for gradualist strategies to dominate big bang strategies in the presence of aggregate uncertainty. In a simultaneous equation system with growth and the level of reform as dependent variables we explicitly introduce a reversal parameter. Empirical results suggest that a reversal generates an immediate negative contribution to real output growth. Taking into account the level of reform a country achieved, a reversal is found to be more costly at higher levels of the reform indicator.