Was colonialism good for growth?

Bill Easterly and Ross Levine add to the literature with a new paper.

A large literature suggests that European settlement outside of Europe shaped institutional, educational, technological, cultural, and economic outcomes. This literature has had a serious gap: no direct measure of colonial European settlement.

In this paper, we (1) construct a new database on the European share of the population during the early stages of colonization and (2) examine its impact on the level of economic development today. We find a remarkably strong impact of colonial European settlement on development.

According to one illustrative exercise, 47 percent of average global development levels today are attributable to Europeans. One of our most surprising findings is the positive effect of even a small minority European population during the colonial period on per capita income today, contradicting traditional and recent views.

There is some evidence for an institutional channel, but our findings are most consistent with human capital playing a central role in the way that colonial European settlement affects development today.

I read hastily, but see important new data and patterns. I don’t really buy the instrumental variables (sorry, Bill) but then again I don’t really buy any of the historical instruments people use to get around thorny causality issues. That doesn’t make me a total party pooper–I just think we have to take all the causal claims and mechanisms pretty cautiously.

Some people rankle over any rosy glow put on colonialism. Most of the authors of the long run growth papers know this acutely, but it bears repeating that “good for growth” necessarily applies to peoples not exterminated.

If you are still angry about the rosy glow, it’s also helpful to put in colonialism perspective: Development in most places in most of history has basically been a process of violence and coercion, either by your own elites or invading ones. When historical events are “good for growth” they are often very bad for the generation that experienced them, in Africa or elsewhere. So “good for growth” does not necessarily mean “good”.

This leads me to think: What is interesting about modern growth policy is that it is one of the first to try to respect human rights. I wonder to what extent growth take-offs require a trade-off between welfare of people alive today versus welfare of future generations. There are reasons to think there are some win-win solutions (e.g. education investments) but I am not so sure it is true generally.

I think a lot of development policy requires trickier trade-offs between those alive today and  unborn future generations than is commonly appreciated. More on that on that elusive day when I have more time to write.

55 Responses

  1. There are different sorts of colonialism. The colonialism of King Leopold’s Congo had precisely *no* benefits for the people living there then, and no benefits for the people living there now.

    The colonialism of the British in India arguably had some benefits, but then they didn’t disrupt and destroy the existing society with a meataxe like King Leopold did in the Congo; socially speaking, they simply imported some British tricks of bureacracy and management which have proven useful, and planted them onto existing princely states, mostly.

    This gets back to needing a genuine measurement for progress, and also gets back to the rule that there is no substitute for competence.

  2. With the caveats above, I’d say this is about right. Joan Robinson said it best, “The only thing worse then being exploited is not being exploited at all.”
    Right now I’m reading Horne’s A Savage War of Peace, the great book on the Algerian War. The Algerians freely admitted that the French had done things for their country they never could have; trebled the amount of arable land, provided medicine and public health that actually worked, imposed a somewhat competent administration that provided peace and order. They also pointed out that they got these benefits at the cost of political and cultural oppression and that almost all the benefits went to the pieds noire. And there’s the rub: greater growth that is appropriated by a foreign power simply isn’t acceptable, growth or no growth.

  4. I took a class on international finance in grad school. It was obvious based on the divide in the class as to who felt that colonization worked and who felt exploited because of it. All of the international students and many of the African American students were very critical of the European conquest of nations in their quest to “civilize” the rest of the world. Dr. Walter Rodney spoke of this in his book “How Europe Underdeveloped Africa”. But what many people fail to remember is that Adam Smith also wrote about it in his book “The Wealth of Nations”.

    Colonialism had one great influence and that is language. With the current digital integration of the world several nations have a commonality in language that can be utilized for international trade and the building of their own economies. So many people from developing nations have the ability to speak multiple languages but they are often shown to be ignorant or helpless even with an equal playing field. But ask yourself this: how many Americans are bilingual that were born in the U.S?
    Its amazing how foreign development (oops, i meant colonialism) can easily become misinterpreted to become something that it wasn’t designed to be at all. However, it is a pattern of history though. Wasn’t Columbus sailing west from Europe in hopes to find India?

  7. Apologies to Bill — not going to click on this paper. I’ll say what everyone knows: this entire genre of papers is shite. Easterly is no worse than the others, but it’s like he developed his views on development in 1956 and his econometric toolkit in 1985 and neither have improved much in the interim. He’s just not the sort that ever thought to cluster his standard errors geographically. And I’m no longer the sort that is going to read this garbage…

  8. Sorry, I read the paper, don’t buy it. Growth for who? “Distributed to future generation”? “Egalitarian institutions”(!)? We can talk about who colonialism raised the economic growth rate in a specific geographic area we now call a nation-state, but even in the “successful” neo-Europes, the people “who were not exterminated” AND not white hardly benefited. Look at contemporary population health of the indigenous (and imported) populations of Native Americans, Aboriginals, Maori, African Americans…even in areas where “non-extractive” colonialism occurred, the spoils went to, and continue to go to the victors. So the paper confirms that colonialism was good for Europeans. But we knew that already.

    So when we talk about “generational trade-offs” and development, history shows us that by in large this is really a trade off between the exploiters and the exploited, not different generations. If that is development, I’m out.

  9. “When historical events are “good for growth” they are often very bad for the generation that experienced them, in Africa or elsewhere.”
    I’m having trouble thinking of good examples. Could you provide some?

    Native Americans

  10. “When historical events are “good for growth” they are often very bad for the generation that experienced them, in Africa or elsewhere.”

    I’m having trouble thinking of good examples. Could you provide some?

  11. “When historical events are “good for growth” they are often very bad for the generation that experienced them, in Africa or elsewhere.”

    I’m having trouble thinking of some good examples. Could you provide some?

  13. . . . but at what cost? At the cost of a genocide? Yes, there is probably a railway line(built to exploit) and functioning trade somewhere in Africa after half-a-million inhabitants were wiped out? Is that included in the model? What about the cost of all the wars we see now across Africa, primarily with roots to the scramble for Africa, colonial borderlines ( e.g. Eritrea/ Ethiopia) or administration ( e.g. Sudan/ South Sudan) or divide and rule( e.g. Rwanda/Kenya)? Is that included as well?

  14. Just a remark on the introduction of the paper that says Ethiopia was a former colony. Well, if 5 years of occupation in Africa is considered a colony, then France in WWII was a German colony too. So let’s set the record right – Ethiopia was never a colony, and it earned it!

  15. “What is interesting about modern growth policy is that it is one of the first to try to respect human rights.”

    I guess the truth of this statement really depends on when you consider ‘modern growth policy’ to have appeared, and what you consider to be human rights. There is considerable evidence that economic growth-focused development strategies pushed forcefully during the Washington consensus years — which I guess falls within what is considered ‘modern’ — were pretty bad for certain human rights, such as social and cultural rights. Then again, Easterly — as unfortunately many development economists — don’t really think socioeconomic rights are rights at all…

  16. What about Acemoglu’s settler mortality instrument. It seemed to give a reasonable way to parse ‘good’ colonialism (the kind that left institutions and some infrastructure conducive to economic development) and ‘bad’ colonialism. It seemed to make a decent case both theoretically and statistically. Didn’t buy that one either?