This paper considers the question of how one can translate evidence of nonlinearities and threshold effects in growth into policy recommendations. We argue that the current evidence of these effects, while important in terms of scholarly debates, does not readily lend itself to policy...