Here's a thread for @Noahpinion on how economists have studied African development. We can divide this into i) long-run causes: (pre) colonial history, slavery, political institutions, ethnicity, and ii) proximate causes: mkt and govt failures. I'll focus on the first.
First, does pre-colonial history matter? Big time. The slave trade had long-lasting effects on development (Nunn 2008) dash.harvard.edu/bitstream/hand… . The likely mechanism was reduced trust among tribes. Nunn and Wantchekon (2011) aeaweb.org/articles?id=10….
But it goes back further than that. Michalopoulos and Papaioannou (2013) show that pre-colonial political centralization importantly predicts present-day development onlinelibrary.wiley.com/doi/10.3982/EC….
In understanding the origin of African precolonial states, Herbst (2000) is informative. en.wikipedia.org/wiki/States_an…. He argues that precolonial state formation is a function of geography and density—Africa’s was unfavorable.
If true, it means that geography can shape institutions, and these pre-colonial systems in turn affect present-day development. Fenske (2013) finds empirical support for this hypothesis onlinelibrary.wiley.com/doi/10.1111/ec…. The importance of interaction between geog. and inst. is a theme here
Precolonial history also looms large with respect to land. There’s not a ton of work on this, but customary tenure is extremely important in much of Africa. Insecure tenure is in turn associated with lower productivity and big gender gaps (Besley 1995, Goldstein and Udry 2008)
What about the colonial period. Certainly, Africa suffered AJR-style extractive institutions probably worse than any other region. The Scramble for Africa split numerous ethnic groups and caused present-day conflict (Michalapolous and Pappaiounou 2016) aeaweb.org/articles?id=10….
But legacy is complex. Evidence suggests colonial missions reduced polygamy Fenkse (2015) sciencedirect.com/science/articl…, and polygamy in turn is associated with greater child mortality (Fenkse and Arthi forthcoming, link.springer.com/article/10.100…?).
Nunn (2012) finds long-run positive impacts of missions, particularly protestant ones, on female education. scholar.harvard.edu/files/nunn/fil…
So while colonialism matters a lot, its legacies generally bad but also mixed. And it’s unclear whether it matters more than pre-colonial history. It all matters, and we're only now beginning to empirically tease out all the mechanisms.
What about ethnic diversity, an oft-cited scourge that is uniquely African. The scramble for Africa paper suggests that it matters for conflict. More importantly, ethnicity forms an overriding political concern that structures much of governance in Africa.
Francois et al. (2015) demonstrate this in the ubiquity and strategic logic behind the power-sharing coalitions that dominate governments across the continent. onlinelibrary.wiley.com/doi/10.3982/EC…
Ethnic favoritism is also common, which can result in misallocation of public resources: Burgess et al. (2015) find it in roadbuilding in Kenya aeaweb.org/articles?id=10…. Interestingly, democracy seems to counteract this effect.
But even in democracies, ethnic identity undergirds most political competition, generating a bad equilibrium where elections resemble ethnic headcounts. Eric Kramon has a ton of great work on why this happens, how it persists, and what its effects are. ekramon.com/articles/
We still don’t know why ethnic diversity matters so much. However, evidence points to political expediency rather than underlying preferences (Posner 2004 on Zambia and Malawi is a classic, citeseerx.ist.psu.edu/viewdoc/downlo….
Berge et al. 2015 measure ethnic preferences in the lab in Kenya and generally reject intergroup discrimination emiguel.econ.berkeley.edu/research/how-s….
However, a great paper by Hjort (2014) finds evidence of taste-based discrimination among teams in a Kenyan firm. It also lowers productivity, demonstrating another direct channel by which ethnic heterogeneity affects productivity academic.oup.com/qje/article-ab…
Still, worth noting that ethnic favoritism in particular is not just an African phenomenon, but holds more broadly (Hodler and Raschky 2014 academic.oup.com/qje/article-ab…). But African countries are uniquely disadvantaged given their artificial borders
Finally, geography. This is a favorite explanation of Collier and Sachs types. The first generation of this argument was all about malaria, disease burdens, and being landlocked. They didn’t nail the empirics convincingly; so many political economists dismissed it.
But the key is the interaction between geography and long-run institutional development. The resource curse literature suggests that natural resources seem to be only a curse in the context of weak institutions.
Furthermore, lot’s of the pre-colonial “causes” are themselves functions of underlying geography. Africa's rugged geography was an unexpected blessing: it prevented slavery from spreading much beyond coasts. This results in better development today, as Nunn and Puga (2012) show.
But geography cuts both ways. Fenske (2013) tests the classic Herbst hypothesis and finds that, land tenure and slavery, two key institutions with negative long-run economic effects, are predicted by Africa’s land-abundant geography.
That's it for now. We've learned a ton in the past 15 years about these long-run causes of development in Africa, but still lots we don't know. NB: tons of country-specific heterogeneity that I basically glossed over to give you the big picture.
