(In)stability in the dynamics of the cross-country distribution of income per capita
摘要
Studies of the convergence hypothesis employing the distribution dynamics approach usually assume that the cross-country distribution of per capita output evolves according to a time-homogeneous, first-order process. Using a panel of 102 countries from PWT 10.0 covering 1970–2019, we examine the veracity of that assumption by comparing transition kernels estimated for different time periods and using an implication of such a process’s Chapman–Kolmogorov equations. We find that the process was time homogeneous and first order in the 1970–1995 period during which the distribution dynamics are consistent with convergence clubs. Following an apparent break in the process in the late 1990 s, the 2000–2010 distribution dynamics are consistent with recent claims of short-term