Why Africa’s development model puzzles economists

IT IS easy to buy a rolex in Uganda—albeit not one that will tell the time. Sold at ubiquitous roadside stalls, the Ugandan rolex is a greasy snack, made from an omelette wrapped in a chapati (“roll eggs”). Sellers compete side-by-side for the same custom. So do the motorcycle-taxi drivers, hustling for rides; or the countless small shopkeepers, stocking near-identical goods. In Uganda, as in much of Africa, the informal service economy is a crowded place to be. But it is hard to find work anywhere else.

Last year GDP in sub-Saharan Africa grew by just 1.4%. Income per person fell. But growth in itself is not the issue that troubles policymakers and intrigues academics: for most of this century, after all, African economies have been among the fastest-growing in the world. What has flummoxed observers is where that growth comes from. In 1954 Arthur Lewis, a Nobel prize-winning economist, argued that development occurs as labour shifts from an unproductive “traditional” sector—activities such as subsistence farming, or petty trade—into modern, capitalist activities.

Research by Margaret McMillan, of Tufts University, and Dani Rodrik, of Harvard, investigates how far Africa has followed this pattern. They distinguish two traditions of thinking about growth. One focuses on raising labour productivity within sectors of the economy, by adding capital or improving skills and technology. The other stresses structural change, as workers move between sectors. The output of the average African manufacturing worker is five times that of his agricultural counterpart. Move people from farms to jobs in factories or high-value services and growth will follow. As a thought experiment, consider changing the sectoral distribution of African workers to match that in the advanced economies, holding everything else constant. Productivity in Ethiopia would increase sixfold; in Senegal by a factor of eleven.

Things are rarely so simple, however. In the 1990s structural change in sub-Saharan Africa actually went into reverse; it was a drag on growth. In Zambia, for example, workers returned to their fields, as industries and mines shut down. But in the new millennium, momentum picked up again. Between 2000 and 2010 structural change accounted for almost half of productivity growth in a 19-country sample. The effect was especially strong in places with a lot of farmers, such as Ethiopia, Malawi and Tanzania. Overall, the proportion of Africans employed in agriculture fell by 11 percentage points.

This was no industrial revolution, however. For every ten workers to lay down their hoes, only two found their way into industry. The service sector absorbed the rest. Cities like Nairobi offered new jobs for skilled professionals in technology and finance. But most workers were more likely to be hawking phone credit than designing the next app; selling second-hand clothes, not stitching new ones. In the oil-soaked cities of Luanda and Lagos, they manned construction sites or waited on tables for the rich. “There’s been structural change,” says Yaw Ansu of the African Centre for Economic Transformation, a Ghana-based think-tank, “but not the type that really improves the lives of people.”

In East Asia both kinds of growth have occurred at once: workers have moved into more productive sectors, and productivity in those sectors has increased. So another puzzle is that in African countries that have seen large-scale structural change, productivity outside agriculture has often fallen.

From 19th-century Britain to 21st-century Vietnam, sustained growth has been built on manufacturing. Factories create lots of low-skilled jobs. And, as Mr Rodrik has shown, manufacturing productivity in poor countries tends to catch up with the most advanced economies, even in places with shoddy institutions or bad geography. But African manufacturing has stagnated. Its contribution to GDP has changed little since the late 1970s.

Orthodox remedies, focused on trimming regulation and improving governance, have lost their appeal. So there has been a revival of interest in active industrial policies. Ethiopia, where manufacturing employment has quintupled this century (from a low base), is experimenting with this approach. A new paper by Cornelia Startiz and Lindsay Whitfield for the Centre of African Economies, at Roskilde University in Denmark, describes how the government has encouraged Asian apparel exporters to open factories in industrial parks, while protecting local firms in the domestic market. Foreign investment helps, if multinationals connect with local suppliers and share know-how.

Making it harder

Yet the record of interventionist industrial policy elsewhere is mostly a sorry one. And old-style industrialisation is in any event becoming more difficult. Automation is transforming manufacturing, as it becomes a viable substitute for labour in countries at ever-lower levels of income per head. The result is that Africans are competing not just with low-wage workers in Bangladesh and elsewhere, but with even lower-wage robots. The development path followed by Japan, the East Asian tigers and, most spectacularly, China—moving from agriculture to low-margin labour-intensive manufacturing such as clothing and toys—may be fast closing. Trade patterns have changed, too. Instead of producing finished products in one country, African industries must slot into global supply chains.

Structural change is about more than factories. John Page of the Brookings Institution, an American think-tank, argues for the importance of “industries without smokestacks”: tradable, productive sectors, like cut flowers, call-centres and tourism. Uganda, with its mountain gorillas and the Nile, is promoting even its edible rolex to tourists. Africa can learn from the successes of other regions, such as East Asia. But it will take a different path.

^{Visit our Free exchange economics blog}