Geography, slavery and development

There's a widespread view (pdf) that Africa's lack of economic development is due in part to unfavourable geography. However, a fascinating new paper (pdf) suggests this might be wrong, in at least one respect.
Authors Nathan Nunn and Diego Puga looked at the link between one geographical feature - the average hilliness of a country - and income.
In principle, hilliness should retard growth. It makes it harder to build roads, thus inhibiting trade. It makes irrigation of land difficult, and it limits the growth of cities and hence the gains of agglomeration.
And the data corroborates this, on average. But only for non-African countries.
In Africa, they found, the net effect of the ruggedness of its terrain is positive; countries with one standard deviation hillier landscapes, on average, have 27% higher per capita income.
The reason for this is straightfoward. The ordinary negative effect of ruggedness is offset by a positive effect. Hilly countries escaped the worst predations of the slave trade, because Africans could hide from slave traders in the hills. And other work (pdf) by Nathan Nunn has found that slave exports between 1400 and 1900 are strongly negatively correlated with income even today.
In protecting Africans from slavery, a rugged terrain therefore helped the economy.
But not entirely. Nunn points out that because attempts to escape the slave trade led Africans to populate rugged areas, subsequent growth has been lower because of the negative effects of bad terrain upon economic activity.

Comments

[Alex]

Or is that because South Africa and Kenya are both rich and mountainous, and are dragging the average about?

[dearieme]

Hard to say which is sillier, the proposition they are scrutinising or their argument agin it.

[pseudonymous]

Presumably the same hilly terrain also protects people from modern day predators, such as rebels or, in Africa, the legitimate goverment.

I trust Nunn allowed for this effect.

[Sam H]

Wouldn't hillier areas also get more rain due to relief rainfall? Thus reducing likelihood of drought/famine?

[Praguetory]

Was about to post the same as Sam. What a numbskulled piece of work.

[chris]

Sam H - your theory only works if hilly areas in Africa get (significantly) more rain than hilly ones outside Africa.
Why else would there be a negative link between hilliness and prosperity outside Africa, but a positive one in Africa?

[Marcin Tustin]

Chris, your comparison is flawed. The key difference is whether African hilly regions get significantly more rainfall than regions where more rainfall would increase wealth. Africa has many such regions. Most regions have little land in such a state.

[Sam H]

Well, I think the terrain, would ensure regularity of rainfall, so even if it isn't a ot more rainfall, at least they are sure of some rain every year. In other words, average rainfall isn't important but the potential for drought to occur, which is low in a hilly area, the rest of the world doesn't really have this problem.

[Sam H]

Something else, didn't the slave trade concentrate on West Africa because that was the shortest route both to Britain and America (not including the French northern territories) Couldn't it just be a conincidence that these areas aren't very hilly?