My old friend Jamie Lumsden recently posted the following on my facebook wall:
This is a great question. On the surface it’s just an idle curiosity – why are the electronics I buy made it this poor, far-away place over here and not this other poor, far-away place over there? But really, this is about why some countries are growing richer and others are not, and about the associated differences in income across places. As I never tire of pointing out, average income is the single strongest determinant of all kinds of important stuff, from health to happiness to the probability of a civil war. And the cross-country differences in income are staggering: in Malawi, >the typical person’s income isn’t a whole lot different from what it was in 1800. In America, for all our hand-wringing about inequality within our own country, even people near the bottom of the income distribution are better-off than monarchs were a few hundred years ago. Fixing this problem – achieving growth in Africa to rival what has happened in America and is happening in China – is of enormous importance. As Nobel laureate economist Robert Lucas put it in a 1988 paper,
The consequences for human welfare involved in questions like these are simply staggering: Once one starts to think about them, it is hard to think about anything else.
What can be done about this? In fact, “making iPhones” (and other such high-value-added activity) is the only possible solution to Africa’s long-standing poverty. While we don’t know a whole lot about how to advance economic development, one thing that we do know for sure is that the answer is never going to come from continued small-holder farming. In Malawi around nine of every ten people is a farmer, and anecdotally even people with other jobs often still have farms. The comparable figure for the United States is less than 1%. This is no accident – for farming to become more efficient, we need to use labor-saving technologies. These necessarily push people into other fields. Think about it this way: if Americans were all farmers at current levels of productivity we would produce over 100 times as much food as we consume. Instead, all that spare labor has to work on other things, and it is those things that make us rich. The path of economic development is one of moving the labor force into manufacturing and services, where people can earn good incomes producing stuff that we actually need, instead of continuing to scrape an existence out of the soil. Worrying about the well-being of factory workers in developing countries is a noble ideal, but it’s important to bear in mind that if you shut down such a factory, the outside option is subsistence farming, where a bad drought can literally kill you.
Identifying that the basic problem is that people need better jobs is a long way off from solving it. The field of development economics has historically been focused on this very issue – why is that China makes iPhones and Africa does not – but that history is a long series of sadly over-hyped failures.* What follows is a list of explanations that I take seriously. I’m leaving out reasons that (in my view) have been discredited, like the fad in the early 2000s of blaming everything on government debts**:
- Poor institutions. “Institutions” here is political economy jargon for stuff like corruption and the rule of law. Anecdotally these seem bad in Africa, and Acemoglu, Johnson and Robinson have argued that you can account for the entire income gap between Africa and other places based on institutions. The legal and social environment seem like they could be important, but I don’t think I’m alone in doubting that everything can be explained by those things. That’s equivalent to saying that stuff like malaria and AIDS have no negative effect on income at all. Indeed, the statistical strategy of the linked paper assumes that malaria isn’t responsible for current poverty in Africa, even though other research has argued that it’s very important. One reason is that childhood malaria, if survived, can cause stunting and brain damage, severely limiting adult earnings. Moreover, there is obvious two-way causality between institutions and economic development; poorer people have a stronger reason to engage in corruption, for example. Their empirical strategy accounts for that reverse causality issue, but Michigan Econ’s own David Albouy has raised some doubts about its robustness. I would say this matters but it’s unclear how much.
- The legacy of colonialism and slavery. This is very closely related to the above. I don’t want to diminish the magnitude of the harm done by colonization and especially by slavery, but I don’t like this line of reasoning very much. First off, it’s very hard to prove – we’re basically relying on cross-country comparisons using possibly-untrustworthy colonial data. Second, there seem to be cases, like Ethiopia, that aren’t doing very well despite not having been colonies, and ones like Botswana that are fairly rich despite a colonial past, so at the least we can say that other factors must be critical as well.
- Geography. Africa is hot and dry with highly-variable rainfall. This makes agricultural success harder than in other places. I don’t buy this as a complete explanation because there are economically-successful places with very little agriculture. More important is that Africa has a fairly high average elevation but no continental mountain ranges and few low-lying areas as well. That means a lot of the continent can be thought of as a high plateau. One effect of this is on agriculture, because it means that it’s harder to find rivers that can be used for irrigation. Stephen Carr has made this point for Malawi in particular but it’s likely to be true in general – you need water flowing downhill to do gravity-fed irrigation, and hence more mountains and hills than you tend to find in Africa. But of course I argued earlier that the key is to get away from agriculture, and it is in moving into manufacturing in particular that Africa’s geography hurts. It has virtually no navigable rivers, so it’s very hard to produce stuff in the African heartland and ship it to the world. The US economy’s engine used to be the automotive industry of the Midwest, taking advantange of our rivers and digging a number of canals to allow cars to be shipped around the world. Malawi lies in Africa’s own Great Lakes region, but it’s not particularly plausible to ship vast amounts of goods to the ocean .* That means that for almost any area not right on the coast, manufacturing is pretty much out of the question until the roads get much, much better. The current dirt road system that covers most of the continent is hamstrung by the intense wet season that makes many of them impassable (a concern for my own project). That means much of Africa may need to follow the “India model” of a move directly into services: a viable plan, but it’s unfortunate to have the options limited in that way.
- Education. The impact of educational attainment on income is arguably the single most-studied topic in economics, with hundreds or even thousands of papers on the financial returns to education. Africa has historically had lower quality education and lower attendance, but big gains have happened, especially recently. Ted Miguel has said that Africa’s recent increases in GDP growth may be attributable in large part to the spread of education to more of the population since decolonization in the 1960s. Proving a link between national education and GDP growth is much harder than showing the relationship between individual education and income, but it’s plausible. You need to be pretty smart and well-educated to work in an electronics factory or at a call center. China has more and better education, and hence its people have moved into higher-skill jobs.
- Health and the burden of disease. Africa is fairly infamous for the wide range of diseases here, many of them very nasty. It is the continent where people have lived the longest and where we are in the closest contact with our primate relatives, who are veritable fountains of horrible illnesses. The tropical environment is also a factor. (Although I did stay at a Holiday Inn Express last night read a bunch of books on this stuff, I’m not a real epidemiologist, so I may be missing some explanatory factors here). A large body of evidence suggests that disease, especially early in life, causes descreases in both physical and mental human capital. The most compelling research is on intestinal helminths (hookworm, etc.) by Hoyt Bleakley, looking at US history, and a team of researchers led by Ted Miguel and Michael Kremer, in present-day Kenya. They find that eliminating these parasites, and the associated drop in anemia, causes huge gains in educational attainment and wages. Bleakley argues that these account for much of the reduction in the former North-South income gap in the US, and followups on Kremer and Miguel’s work have found similar results in Kenya. There is also feedback between poverty and disease: richer people stay healthier, so it’s possible that societies may get stuck in a low-income, high-disease equilibrium.
- Export-led growth. Many of Asia’s growth successes appear to lean heavily on exports, often with heavy government encouragement. I’m not a trade economist, so I don’t know why this approach works, and even the trade economists I’ve asked about this don’t seem very sure. One bit of evidence on this (which I can’t think of a citation for now, but which I’ve seen raised in several talks) is that firms that begin exporting appear to become more efficient and profitable; international competition may spur more effort, and learning from one’s new, international peers probably brings big benefits. Inasmuch as this really works, Africa doesn’t have a whole lot of it going on. I suspect that truly broad and deep economic growth in Africa will come only when regions of the continent find their export niche (think of textiles in Bangladesh, for example) and start investing heavily in it.***
- Random chance and path-dependence. There’s suggestive evidence that once economic success in a particular industry gets rolling in a region, it has a lot of momentum and even picks up speed. The reason we make iPhones in China now is not that they have skilled labor at a cheap price (wages there are actually getting pretty high) but that they have a supply network there, with skilled workers, easy access to parts, institutional knowledge of good production techniques, relationships with other companies, and reputations for quality and efficiency. If fortune blesses an area with a head start in a given field, it tends to keep that advantage. This is probably the most important reason why you see some areas specializing heavily in certain things.
This list is not comprehensive, but it contains are the main factors that came to mind when I was writing this. My view on these (which is perhaps obvious from my research agenda) is that we should focus mainly on reasons 4 and 5. Items 1-3 are very good reasons for Africa’s relatively slow development but are terrible solutions. It’s unclear what we can do about the continent’s poor institutions, history of slavery or geographic misfortune other than to say “that’s really terrible,” to apologize for the past sins of our culture in the case of slavery. There’s no level we can pull to change that stuff. Number 6 is still a little ambiguous – it’s not entirely clear what works, and I still suspect that exports may be a result of growth and not a cause. Number 7 isn’t something we can take advantage of directly, but it means that if we can get some momentum going in any industry in Africa, it make develop very quickly. 4 and 5 tend to be very cheap and cost-effective, and moreover they are valuable in their own right. Indeed, things like health and education are some of the most important goods that we want to buy with the income gains from economic growth. One way that I get to sleep at night is by remembering that even if whatever I’m working on doesn’t make an appreciable dent in poverty in Malawi, at least some of what I do is worthwhile in and of itself.
So why aren’t iPhones made in Africa? I’ve tried to offer the best answers I can, but the truth is that we don’t know – and if we did know, and the reason was something tractable, we might be able to make huge improvements in the lives of lots of people.