How Nigeria became twice as rich overnight

This post appeared in Dutch on the Oikocredit website.

One Sunday in April, something remarkable happened: all of a sudden Nigeria became almost twice as rich. The Nigerian statistical agency had recalculated the official statistics and the GDP turned out to be 89% higher than previously thought. Overnight, Nigeria had become the largest African economy, surpassing South Africa.

www.photocontestinsider.com/2013/04/winners-of-the-malaria-photo-contest-announced

Had someone messed with the numbers? After all, companies and other parties have an interest in a larger economy. For years, South Africa had been the African country to invest in; this revision could divert investors’ attention from South Africa to Nigeria. However, economists seem to agree that not the new, but the old statistics were unreliable.

The old statistics could not be trusted because Nigeria was using severely outdated methods for its national accounting (read a more detailed explanation here). This is not just a Nigerian problem: Morten Jerven writes in his book  Poor Numbers that African statistical agencies generally lack people and knowledge, causing them to publish inaccurate numbers. It is not only economic statistics that are often of low quality in Africa, but also other types of data. For example, education and health statistics are often incorrect.

Why should we care about good data? Isn’t that a problem for number fetishists in rich countries? Aren’t there more urgent problems in Africa? It’s probably not the sexiest topic, but data quality is essential for poverty reduction and development. Without good data it is impossible to know where money should be invested, how health care is doing, and what the impact is of a new education policy. Without good data, new policy measures are merely shots in the dark.

Claire Melamed mentions the example of malaria. Malaria is one of the foremost causes of death in poor countries. Nevertheless, good data is scarce. To fight the disease effectively it is important to know which areas are confronted with malaria, how many patients are there, and what is the quality of local healthcare facilities. Moreover, the results of eradication programs can show what works and what doesn’t. For example, should mosquito nets be free or is it better if people pay for them? (Answer: hand them out for free.) Such knowledge doesn’t fall from the sky: local organisations report figures or special teams collect the information needed. This costs money, but such investments have high returns: less people fall ill, less people die.

So how can data quality be improved? In order to figure that out, it is key to understand why statistics are unreliable (and sometimes completely absent) in African countries. First of all, there is a lack of resources. Jerven tells about a visit to the Zambian statistical bureau, where one man is responsible for the entire national accounting. “What happens if I disappear?” he wonders. (To contrast, ninety-nine employees work at the national accounting department of the Dutch statistical agency.) Secondly, perverse incentives sometimes encourage organisations to report wrong numbers. A recent article by Justin Sandefur and Amanda Glassman shows that governments exaggerate the number of vaccinated children in order to receive more donor money. Or schools report higher enrollment figures, because the administration pays them per student. 

Data quality can improve by tackling these two causes. There are initiatives, such as Paris21, that focus on enhancing ‘statistical capacity’ of developing countries by giving money and training. However, that doesn’t solve the problem of wrong incentives, which requires fundamental changes in the way development progress is rewarded.

On that Sunday in April Nigeria showed that it is indeed possible to produce better data. Of course, better data is a not an end in itself. The poor man in a Lagos slum didn’t see his income double overnight. Nor did the ill woman in the countryside all of a sudden have access to a better hospital. Only if good statistics are actually used for better policies, can they make a difference.