Most of the literature about Africa’s growth, “Africa Rising”, “Lions on the Move”, etc., refer to the present or the future. An oft-quoted World Bank report said, “Africa could be on the brink of an economic takeoff, much like China was 30 years ago and India 20 years ago.”
Meanwhile, Alwyn Young has recently published a paper that claims that per capita consumption on the continent has been growing at 3.4-3.7% a year for the last two decades – about three to four times the growth rates documented in other studies. Instead of using national accounts data (which, as we know, suffer from several deficiencies), Alwyn adopts the Demographic and Health Surveys (DHS), which calculate the households’ ownership of assets and other indicators of well-being (ownership of a car or bicycle; material of the house floor; birth, death or illness of a child, etc.). Employing a variety of econometric techniques (including some robustness checks and sensitivity analyses), he then calculates an asset index across the various components of the DHS. Assuming a constant relationship among assets, educational attainment and consumption, he concludes that per capital consumption has been growing at 3.4-3.7% a year from 1990-2004.
Not so fast, say Kenneth Harttgen, Stephan Klasen and Sebastian Vollmer in a forthcoming publication in the ‘Review of Income and Wealth’. They find many reasons why growth in assets may not reflect growth in per capita consumption. For instance, households may not get rid of old assets, so that the average age of their assets is rising and the services from the assets (which contribute to consumption) growing much more slowly. Also, the price of some assets, such as cell phones, might have fallen sharply; increased ownership may not reflect increased consumption to the same extent.
They also question Alwyn’s use of educational attainment as a proxy for income, noting that education is also a result of public provision of schools. Finally, they look at the relationship between assets and consumption using a number of different data sources and find no correlation.
Perhaps one way of reconciling these two positions, as suggested by Stephan in personal correspondence, is to recognise that growth in per capita private consumption only captures part of the improvements in people’s well-being. The increased provision of education, health and infrastructure services also contribute to welfare, but are not captured in private consumption.
Young’s analysis tries to capture these factors in his asset index. However, this does not mean that per capita consumption was mis-measured. Rather, it was only telling part of the story. At the same time, when I look at the poor quality of the health, education and infrastructure services being provided, especially to poor people, I am not sure the asset indices are capturing real welfare gains.
We have a lot more work to do so Africans can realise their growth miracle.
Shantayanan Devarajan is the chief economist for Africa at the World Bank.