Political risk, export credit insurance and trade: a gravity model analysis of South Africa.

Abstract

This study adds to the literature on political risk, export credit insurance, and trade. Specifically, the study aims to consider the effect political risk and export credit insurance may have on South African exports using an extended gravity model framework and regression analysis from 1996 - 2018. Gravity models and estimation techniques differ in the literature. Methods include the Fixed Effects (FE), Random Effects (RE), and more recently, using the Poisson-Pseudo Maximum Likelihood (PPML). The PPML is the preferred estimation technique of this study as it reduces the bias problems resulting from zero trade flows and potential heteroscedasticity (Santos Silva and Tenreyro, 2006). In addition, this is further justified as Santos Silva and Tenreyro (2011) and Martin and Pham (2020) argue that the PPML estimation technique is also more appropriate than the traditional FE and RE estimation techniques. Thus, considering these limitations, using the preferred PPML estimation technique, this study finds that political risk significantly undermines trade. However, the export credit insurance variable is insignificant and does not play an influential role in South Africa’s bilateral trade. Additionally, other attributes, including the role of market power as measured by Gross Domestic Product (GDP), distance, and the regional groupings of African states through SADC, have important bearings. An important implication is that African countries' inability to harness the gains from trade might be rooted in factors beyond trade finance constraints and more related to political risks and geographical constraints.