This paper provides a sensitivity analysis on the prior choice for interaction terms for the results of Masanjala and Papageorgiou (Rough and lonely road to prosperity. Journal of Applied Econometrics 2008; 23: 671–682), which has been criticized for its implementation of interaction terms by Crespo Cuaresma (How different is Africa? Journal of Applied Econometrics 2011; 26: 1041–1047). We find that the discussed priors have similar predictive power and we also evaluate alternative prior solutions that account for the special nature of interaction terms. Copyright © 2013 John Wiley & Sons, Ltd.