We uncover evidence of substantial heterogeneity in the growth experience of countries using a structural threshold regression methodology. Our findings suggest that studies that seek to promote mono-causal explanations in the institutions versus geography debate in growth are potentially misleading