Assistant professor Morten Jerven rides the train in Kenya, where he was observing statistical collection methods.
African stats a 'numbers game'
International development and aid groups are making decisions and distributing funds to African nations based on statistics that are incomplete and untrustworthy, says economic historian Morten Jerven, a School for International Studies assistant professor.
“There is an unjustified numbers game with misleading air of accuracy,” says Jerven, who spent months observing statistical collection methods in countries including Ghana, Nigeria, Uganda, Kenya, Malawi, Tanzania and Zambia.
He found that little is really known about the countries’ economic status. “There’s a great variation across time and across countries on what kind of data is available for what kind of indicators,” he says.
That’s because statistics offices are chronically underfunded and tend to use windfall funding from strategic donors to produce statistics that only interest the donors. These are often social indicators related to health delivery, maternal health or school attainment, rather than broader, more useful economic statistics related to employment, industry, agriculture and economic growth.
As a result, countries’ statistical systems are incomplete and the type and reliability of data collected varies from country to country. Many data users are not aware of this variation in quality, says Jerven, adding most aggregate statistics on African economies involve a lot of guesswork.
“This raises questions about which of these indicators we should use to select countries for a particular project, and how we know whether a project is working or not,” he says Jerven has presented his findings to World Bank officials and several conferences and has a book on the subject coming out in late 2012 or early 2013. His aim is to initiate discussions with donors and aid organizations about what kind of statistics should be a priority and how to align them with the priorities of local African economies.
His project is funded by a three-year Social Sciences and Humanities Research Council grant.