Sub-Saharan Africa

GDP Statistics: Are African GDPs Accurate?

by Elaine Schwartz    •    Jan 22, 2013    •    460 Views

With most people debating the validity of GDP statistics as a yardstick of well-being, I was surprised to hear a completely different conclusion in an econtalk podcast.

Before even considering validity, for sub-Saharan Africa, one scholar suggests we should check reliability. Compare for example, three respected sources of data: Angus Maddison, the World Development Institute, the Penn World Tables. For the year 2,000, ranking GDP per capita, they agree that the Democratic Republic of Congo is the poorest sub-Saharan nation. Otherwise, the disparities are considerable. For example, Penn World Tables said Liberia was the 2nd poorest country while the Maddison group placed it among the 10 richest.

The reason? For certain countries some of the data has to be estimated. Production can be a part of the informal economy. It can involve food production, barter, unrecorded transactions. And, to compound the problem, the statistics aggregators create their estimates from distant cities rather than visiting the entire countryside. You can see where this is going. With GDP data varying over time and within countries, the numbers are not necessarily dependable. And according to the econtalk interview, they might be 50% to 100% inaccurate.

For Africa, statistical accuracy can determine the size and kind of development aid. Beyond, countries use GDP statistics to shape economic policy. Especially remembering that seemingly indisputable numbers are comforting, shouldn’t we be questioning their validity more frequently?

Sources and Resources: Controversial and thought provoking, the econtalk podcast in which Russ Roberts interviews Canadian scholar (Simon Fraser University) Morten Jerven is worth listening to. To read more, I suggest going here to Morten Jerven’s links to his publications and the story of his research. For an entirely different and lighter look at statistics generally, you might enjoy reading Charles Wheelan’s book, Naked Statistics: Stripping the Dread From the Data.

Note: The title for this entry was slightly edited after it was posted.

« »