Three essays in panel data econometrics

Abstract

This paper contributes to the literature on regional growth-cycles by developing in a Bayesian framework different notion of multivariate cycle synchro- nization and proposing an encompassing model combining three dimensions: panel, Markov-switching and multivariate synchronization. The approach is used to study the chronology of provincial growth-cycles in China’s economy from March 1989 to July 2009. The estimations reveal the importance of a third regime of rapid-growth in growth-cycle dynamics at both the national and provincial levels. The first regime of recession is of two types: classical recession and growth recession. The third regime of rapid-growth is also of two types: catching or bounce back. We obtain evidence of geographical clusters and qualitative differences among growth rates between coastal and interior provinces. Moreover, this retrospective view sheds some lights on China’s ‘new-normal’ of slower growth since early 2010.