Growth Miracles and Failures in a Markov Switching Classification Model of Growth

This paper interprets economic growth as a sequence of transitions between four growth regimes: crisis, stagnation, stable growth, and miracle growth. Countries experience distinct patterns of growth because they switch between the growth regimes with different frequencies. The aim of this paper is to identify countries with similar growth dynamics in order to determine the driving forces behind different within-country growth experiences. To that end, a traditional Markov switching model is estimated and extended by a classification mechanism that endogenously assigns countries that feature similar dynamics into the same, and countries that feature distinct dynamics into different clusters. Three clusters of countries are obtained: successful countries that are characterized by persistent periods of stable or miracle growth, moderately successful countries where growth is frequently interrupted by periods of stagnation and failing countries that stand out for their highly volatile growth rates with frequent periods of stagnation and crisis. It turns out that both a well-educated workforce and a geographically advantageous location reduce the probability of countries to belong to the cluster of failing countries, even if institutions are not well developed. However, in order to belong to the cluster of successful countries well-functioning institutions are essential.