Historical Social Research: An International Journal for the Application of Formal Methods to History, 35(1) pp 363-372

This paper studies persistence properties of total factor productivity (TFP) from Bayesian perspective. Emphasizing that classical unit root test for TFP cannot determine the probability whether a stochastic shock to the series is permanent, we design a Bayesian unit root test for TFP. Examination for a set of African economies’ TFP data show that the probability of having a unit root is very high for majority of countries. The evidence of high-persistence has implications for perpetual growth and business cycles.