A National Bureau of Economic Research (NBER) paper issued this week by Xavier Sala-i-Martin and Maxim Pinkovskiy shows that: poverty in Africa is falling; that it is falling at a faster rate than expected; and that this holds true for a wide range of African economies. What’s more, if this trend continues, Africa will achieve the first Millennium Development Goal – halving the proportion of people living on less than one dollar a day – if not by the 2015 deadline, then shortly after. This is surprising news, and good news. But is it too good to be true? 

The findings are controversial for economists involved in the painstaking business of constructing consistent regional and global estimates of poverty using nationally representative household surveys. Sala-i-Martin and Pinkovskiy use a very different method – anchoring their poverty measures to national GDP data adjusted for purchasing power parity (PPP) to assess average income per person per household. Because the measurement draws on GDP, it includes not only household consumption but also private investment and government spending.

Estimates generated by Martin Ravallion and colleagues at the World Bank use the poverty lines actually found in low-income countries through household surveys. These consumption-based estimates do not include allowances for investment and government spending and are grounded in the most basic food and other consumption needs.  Differences in method mean that the estimates generated by Sala-i-Martin and Pinkovskiy on  the number of people living below the poverty line are consistently lower than those generated by World Bank researchers, and they are apparently a lot more bullish than the Bank about the downward trend.   

So, what are we to believe? 

• First, alternative methods for measuring poverty must have their place in the debate about where we are headed. Monotheism is unhealthy and I expect that this work will generate another round of intense debate on how best to monitor poverty trends. This, in the end, is a strength and not a weakness.
• Second, Sala-i-Martin and Pinkovskiy’s results are based on standard GDP series that are used widely by economists in the study of economic growth. However, their figures tell us what has happened to poverty based on changes in average household income, rather than on the ability of a household to achieve a basic basket of consumption goods. The latter is more exacting and more likely to reflect the genuine conditions on the ground for many people living in poverty.
• Third, as William Easterly has argued, Africa’s high initial poverty rate (and low poverty elasticity) at the MDG start point of 1990 makes the achievement of MDG1 that much harder. Africa needs higher economic growth than other regions to reach the goal. The result, of course, is that even with growth rates of 5% or more, progress on poverty in Africa is almost always off-track. Sala-i-Martin and Pinkovskiy’s approach gives us at least some cause for optimism about the direction of change in Africa, even if we are more cautious about the magnitude of change and what is driving it. 

The findings cannot, however, be a signal for complacency. Many Africans continue to live on the very margins of existence, in countries or regions afflicted by conflict, in situations of extreme scarcity and fragility. The study is clear that the overall trend decline in poverty does not apply, and is in fact reversed, in countries embroiled in conflict. We must remember that while rising per capita incomes can drive poverty reduction, poverty is about more than income. It is also about insecurity, vulnerability to shocks, discrimination and lack of political voice. And there is far more to be done to tackle the many faces of poverty in Africa.    

The MDG review summit is looming in September this year. While we shouldn’t lose sight of this good news story, we need to use the review process to reinforce what needs to be done to build more resilient economies and societies that can secure a better life for all of Africa’s citizens.  

"