Roots of Political Centralization in Africa
Monday, June 25, 2012
Daron Acemoglu and James Robinson

Seen from a wider perspective, the development of the strong central state we described in our previous post in Rwanda is anomalous in African history. Though states did form in pre-colonial Africa, for example around the Niger bend in the late middle ages, and in many parts of West, Central and East Central Africa after the 17th century, it is clear that these processes lagged those which took place in Eurasia. One can get some quantitative picture of this via the data coded by Louis Putterman and his collaborators (see this paper).

The lagged development of political centralization in Africa is an important part of the puzzle about why Africa developed less slowly than the rest of the world (see also this paper and this paper).

What can explain this retarded political centralization? Scholars of European political development, notably Charles Tilly in Coercion, Capital and the European State, have advanced several hypotheses which, they claim, can explain the development of European states. Tilly’s boldest claim was “states made war and war made states”. He argued that it was the intense inter-state warfare of Europe which led to political centralization. Other scholars have instead emphasized high population density and trade and commerce.

A natural approach to explaining why political centralization lagged in Africa is therefore to argue that the factors that led to such centralization in Europe were absent in the African continent. Jeffrey Herbst in his book States and Power in Africa did exactly this, arguing that African had not developed powerful centralized states because it had low population density and inter-state warfare was absent.

Is this the right answer to the puzzle? One way of examining these ideas is to look at data from the Standard Cross-Cultural Sample, a dataset which has been built since the late 1960s by anthropologists and contains rich data on 186 cultures. Though this dataset is less than ideal and it only includes 40 cultures in Africa, it does contain several major Africa societies with differing levels of political centralization. These include famous hunter gatherers such as the Hadza of Tanzania or the Kung of Botswana, and stateless agricultural societies such as the Tallensi of Northern Ghana, the Kikuyu of Kenya, and the Tiv of Nigeria. It also includes some important politically centralized societies such as the Asante of Ghana, the Ganda of Uganda, the Hausa of Nigeria, and the Bemba of Zambia. Finally there are intermediate societies such as the Mende in Sierra Leone.

The next figure shows the positive relationship between population density and one measure of political centralization from this dataset, the extent of judicial hierarchy beyond local community, for the sample of cultures outside Africa.

But let us next look at the same relationship for the African subsample:

Here the relationship is much weaker, essentially nonexistent. Within Africa, it seems that it is not the societies occupying lands with greater population density that have achieved greater political centralization.

The data also show no correlation between warfare and political centralization, either in Africa or in the rest of the world, thus not providing much support for Tilly’s main hypothesis. This can be seen in the next four figures which show scatterplots of the relationship between the frequency of internal and external wars, and political centralization for the non-Africa and Africa subsamples (for the warfare measures, 1 means infrequent wars, 2 stands for frequent wars, and 3 for continual wars).

Maybe Africa needs a different theory of political centralization.

Article originally appeared on Why Nations Fail by Daron Acemoglu and James Robinson (
See website for complete article licensing information.