There is an endless literature on the determinants of development trying to figure out what is more important: endowments (geography, climate, resources, diseases) or institutions (colonization, legal framework, corruption). Countless cross-country regressions have not been able to settle this question once for all. This may not look that surprising, given the quality of the data and the pitfalls of cross-country regressions (collinearity all over the place). This is why it is important to highlight some innovative approaches.
Raphael Auer still uses the cross-country regression approach, but puts the finger on one very important aspect: institutions are to some extend endogenous. In particular, endowments can have an impact on them. We highlighted earlier how slavery influenced institutions and then development even today, and that slavery was limited to accessible area. Thus geography matters in indirect ways. There are other examples.
Basically, Auer uses an empirical strategy to disentangle the impact of endowments and institutions. He notes that the impact of endowments on development is different in former colonies and non-colonies. This calls for including interactions in regressions. But you want to do that intelligently, for example by including an interaction for former colonies only. He does that by including a term that reflects the impact of endowments on colonial institutions.
The results indicate that both endowments and institutions are significant. Looking at the details, disease seems to be particularly important, not just because it impacts directly the economy, but also because it influenced colonization policies. Does this mean that disease eradication will solve all problems in Africa? Not necessarily, because disease has this impact also through colonial institutions, these institutions need to be addressed as well.