Samuel Cohn: Global Social Trends

There is a giant debate in development studies about whether or not Europe is responsible for the general underdevelopment of Africa. That is a long and very complicated debate with multiple arguments on both sides. It is also a debate which plays out differently in different regions of Africa. Ethiopia was not colonized by Europeans for most of its history. (The Italian colonization was short and less impactful than Ethiopia’s relationships with Egypt and with the trading communities of India.) The Western African nations that were the sending communities for the global slave trade suffered greatly from that role. The heavily forested Western Congo would have been a slow developer regardless of the Belgians – since forested regions (and mountainous regions too for that matter) tend not to develop the intensive agriculture that is the starting point for economic development. (The one exception to the Mountain = Low Growth rule, Switzerland, is one of those exceptions that proves the rule. Unlike most other mountainous regions, Switzerland developed superlative agriculture in its valleys. Forested places that grew, such as Germany, did so by chopping down the forests first.)

For now, never mind the question of whether Europe underdeveloped Africa as a whole. There is new research out showing that Europe definitely underdeveloped Islamic Africa. Christian-friendly and traditional-animist regions of Sub-Saharan Africa did much better economically than did Islamic Africa. They did so because the European colonial powers put intentionally put fewer development resources into Islamic areas, than they did for regions with other religions.

Vincent Bauer, Melina Platas and Jeremy Weinstein have new research in the 2022 World Development that shows the under-provision of development resources to those parts of Africa that were historically Moslem.

When the colonial governments improved development in Africa, they customarily did so by providing public services. The most important public service was education. The role of education in stimulating development should be obvious. Colonial administrations also built roads, put in health clinics, took public health measures to reduce the amount of infectious disease, and reduced crime and banditry by imposing regular systems of law and order. The importance of these other measures should be clear as well.

In Africa, the colonial governments provided fewer public services than they did in their colonies on other continents. Particularly noticeable was their neglect of education. Building and running schools was generally left to missionaries. To be sure, the curriculum often had Christian components. But students who studied in mission schools became literate. Students were taught basic mathematics and were often given enough of a Western education that the most talented of them could consider studying abroad.

Leaving education to Christian missionaries was bad news for Moslem Africa. Colonial administrators rightfully felt that Christian schools would not be welcome in Islamic areas. So, they simply provided no education at all in these regions. The effect of depriving the Moslem areas of schooling was devastating. It lowered the ability of people in Moslem regions to integrate themselves into a modern economy. These areas became less attractive areas for foreign investment. There was less entrepreneurship and small firm creation in these areas. The capacity for both exogenous and endogenous development was stunted.

Secondly, the colonials avoided any kind of direct administration of Moslem areas. In principle, this ought to have been good. Traditional Islamic authorities were left in control of their regions to run them as they pleased. But with local autonomy also came neglect of government services. The colonialists did not provide the roads, health clinics, infrastructure or legal structures that would have been important preconditions for development. The Moslem regions were left to be free; they were also left to be poor.

What is remarkable about the Bauer, Platas and Weinstein article is not only the argument – but their methodological test. They come up with powerful statistical tests of their thesis – on the continent with the worst data on the planet. 

There are two issues. Both in the colonial era and in the present day, data collected in Africa are often dubious. Consider GDP, a measure which is important in a study of economic development. Measuring GDP requires the government counting all of the economic transactions that take place in a country. How accurate do you think Ethiopia’s counting of economic activity in the Tigray region is – a region which is currently in the middle of a civil war? Many other parts of Africa have internal political unrest or local distrust of national governments.

The situation was even worse in the colonial period when government authority extended only tenuously to most of the regions supposedly being governed. When the colonial governments “left the Moslem regions alone”, this meant that they were not sending government bureaucrats in to count pretty much anything. Yes, censuses exist but the quality of those censuses was sketchy at best. With no eyes on the ground, measuring what was going on in terms of social and economic development was nearly impossible.

PLUS, the tribal groups and regions that existed in the colonial period had little relation to the national boundaries of contemporary African countries. If a particular ethnic group was left alone to govern itself, it could be in a region that spans any number ofcontemporary nations. Present-day provinces, counties or administrative districts do not fit what the historical borders actually were.

A historical study of the legacy of missionary schools and indirect rule on the ancient regions of the colonial era ought to be impossible.

Yet this is just what Bauer, Platas and Weinstein actually do.

First, they reconstructed more or less exact boundaries for each of the major kingdoms that existed in Africa before 1900. This involved extensive research from the historical literature on African history. Overall, they found 71 large kingdoms, 20 Islamic and 51 either Christian or animist.

They then reconstructed the history of some 773 ethnolinguistic groups in Africa, judging if any of them had ever been under the control of a Moslem state. Keep in mind that many ethnic groups were not under the control of any state because they lived in the interstices between existing African kingdoms.

Obviously, using all these groups leads to borders that are very complicated and do not relate well to any map of Africa that exists today. How does one measure GDP for these groups?

Bauer, Platas and Weinstein fall back on a clever trick that other development economists have used to study development in low-quality data settings. They use satellite data to measure the amount of light in a region at night. Luminosity measures have been previously used and found to be superlative indices of the overall level of economic activity of an area. They take satellite photos of Africa at night and superimpose those photos on maps reflecting the ancient African kingdoms and the borders of all of the ethnic groups. Using this, they get more or less precise measures of how economically developed these ancient kingdoms and tribal areas are in the present day.

Using these borders, they can also get measures of infant mortality and education. These use vital statistics and census data. The micro-location coding on birth and death rates and on education statistics is not too bad. So, mortality and education data can be broken down and reconstructed for ancient boundaries.

They then do very elaborate statistical analysis on the data looking at

a. statistical analyses for Sub-Saharan Africa as a whole,

b. matched comparisons of ethnic groups in Islamic kingdoms with neighboring ethnic groups outside of those Islamic kingdoms and

c. comparisons of the same ethnic group inside and outside of Islamic kingdoms.

What do they find?

1. Economic development as measured by luminosity (the amount of light at night) is lower in the areas of former Moslem kingdoms – regardless of which of the three tests is used.

2. Infant mortality is higher in the areas of former Moslem kingdoms – regardless of which of the three tests is used.

3. Education is lower in the areas of former Moslem kingdoms – regardless of which of the three tests is used.

4. Christian missions were less likely to be located in the areas of former Moslem kingdoms.

5. Quality of governmental institutions was lower in the areas of former Moslem kingdoms. Here, the measure of quality of government is frankly a little ragged. They use the accuracy of age reporting in the Census as a mark of how good a job the colonial government did of taking the Census. This works nicely in their study, but I would have preferred to see something more direct, like number of roads built, or presence of public health programs. Still – what they have supports their argument.

*  *  *

Now some American readers might be likely to think that the problem in the former Moslem kingdoms was Islam itself. Americans like to think that Western and Christian values are “rational” and “pro-capitalist” while Moslem values are “traditional” and “anti-capitalist”. Such thinking makes the very high economic rationality seen in Saudi Arabia and the Gulf States hard to explain, never mind the historical economic wealth of Egypt in the 1200-1700’s, and the successful economic development programs that one can observe in Morocco and Malaysia.

But for those readers who absolutely insist that only Christians can be economically rational, Bauer, Platas and Weinstein have a good counter-argument. They repeated all of their tests – running them separately areas which currently have populations that currently practice Islam versus populations that practice a different religion. Remember that these colonial legacies are based on where the Moslem kingdoms were over a century ago. Those kingdoms had Christians and Animists within their borders as well as Moslems. Likewise, Moslems can be found in areas outside of Moslem kingdoms. They looked at currently Moslem areas inside and outside the historic Moslem states, and they looked at non-Moslem areas inside and outside the historical Moslem states.

Regardless of whether the local population is Moslem or not Moslem, the residents of the former Moslem kingdoms have less luminosity at night, higher infant mortality and lower levels of education.

*  *  *

Bauer, Platas and Weinstein blame the nineteen and early twentieth century colonial governments for these results. I wholly concur with that analysis.

This analysis also suggests that missionary schools were good things. Whatever education there was, occurred because of Christian missionaries. There are other debates about whether Christian missionaries were good or bad. (Some authors accuse them of fomenting ethnic hostility.) But there can be little question that missionaries’ contribution to education was positive.

It is also the case that the colonies shouldhave provided widespread education to the populations as a whole. One set of colonialists who did provide education to the population as a whole were the Japanese. The Japanese also put in first rate administrative systems, and worked hard to create a solid economic base for their colonies. Those colonies - South Korea, Taiwan, and Northern China - are now wealthy today – far wealthier than the areas governed by Britain or France in Africa or Asia and wealthier than the areas governed by Spain in the New World. Colonialists varied in how much responsibility they took for the prosperity of their colonies.

We don’t think much about those issues when we think about the history of Europe or America and the Global South. But those legacies made a difference. The good that rich countries do for poor countries creates century-long effects. As for the legacies of colonial damage and neglect …

(You can finish this sentence yourself.)