The infrastructure Africa really needs is better data reporting

This table is smaller than our margins of error.
This table is smaller than our margins of error.
Image: AP Photo/Pablo Martinez Monsivais)
We may earn a commission from links on this page.

This week African leaders met with officials in Washington and agreed to billions of dollars of US investments and infrastructure deals. But the terrible state of statistical reporting in most of Africa means that it will be nearly impossible to gauge how effective these deals are at making Africans, or the American investors, better off.

Data reporting on the continent is sketchy. Just look at the recent GDP revisions of large countries. How is it that Nigeria’s April GDP recalculation catapulted it ahead of South Africa, making it the largest economy in Africa overnight? Or that Kenya’s economy is actually 20% larger (paywall) than previously thought?

Indeed, countries in Africa get noticeably bad scores on the World Bank’s Bulletin Board on Statistical Capacity, an index of data reporting integrity.


Bad data is not simply the result of inconsistencies or miscalculations: African governments have an incentive to produce statistics that overstate their economic development.

A recent working paper from the Center for Global Development (CGD) shows how politics influence the statistics released by many African countries.

One indicator the report examines is school enrollment rates. By comparing household surveys to administrative records, it finds that in 15 of the 21 countries examined, official figures overstate enrollment. In Sub-Saharan Africa, official numbers put enrollment 3% higher than the surveys do.


The report also looks at statistics the government has beautified for the sake of international donors. For instance government vaccination rates are higher than those reported by household surveys.

Optimistic statistics also attract investors. Following Nigeria’s overnight GDP explosion, for example, the Financial Times reported analysts saying that “the revision could serve as a catalyst for further investment”.

But in the long run, dodgy statistics aren’t good for anyone. They “distort the way we understand the opportunities that are available,” says Amanda Glassman, one of the CGD report’s authors. US firms have pledged $14 billion in trade deals at the summit in Washington. No doubt they would like to know whether high school enrollment promises to create a more educated workforce in a given country, or whether its people have been immunized for viruses.

Overly optimistic indicators also distort how a government decides where to focus its efforts. If school enrollment appears to be high, why implement programs intended to increase it?

The CGD report suggests increased funding to national statistical agencies, and making sure that they are wholly independent from their governments. President Obama is talking up $7 billion into African agriculture. But unless cash and attention are given to improving statistical integrity, he may never know whether that investment has borne fruit.