The existing literature on the relationship between natural resources (NRs) and growth is inconclusive. To enrich this debate, some studies have investigated the role of institutions in the NRs-growth nexus. Unlike most previous work, which mostly considers the interactive effect of institutions, notably corruption, on the relation between NRs and growth, this paper determines the optimal threshold of corruption below and above which NRs affect economic growth differently. The aim of this paper is to investigate the effect of NRs on economic growth conditioned by the level of corruption in SSA. Using a panel data on 26 Sub-Saharan Africa (SSA) countries over the period of 1985 to 2014, this paper uses the Panel Smooth Transition Regression (PSTR) model developed by Gonzalez et al. (2005). Firstly, we found evidence of the existence of corruption thresholds that change the effect of NRs on economic growth. These thresholds are 0.94, 0.40, 2.33, 1.16 and 0.48 for public, executive, legislative, judicial and political corruption, respectively. Secondly, the relation between NRs and economic growth below and above each type of corruption gives mixed results. The sensitivity analysis, which led to the decomposition of NRs into forest and oil resources, confirms the divergence of the results found by the baseline specification. These results have significant implications for policy sequencing in SSA. To benefit from NRs-led growth, improvement of the institutional framework, including different political corruption reducing, should precede NRs management policies. Also, a certain diversification of the economies of SSA countries leads to a better efficiency of the NRs on economic activity.
