Measuring Africa’s data revolution

Governments are producing more data than ever, but are the numbers any good?

It is the continent’s least sexy revolution. There are no stirring speeches, no victorious mobs, no humbled dictators. Instead, there are quiet academics and industrious researchers working amid piles of government gazettes and labyrinthine spreadsheets. And there are numbers. Lots and lots of numbers.

This is the African data revolution, but the lack of fanfare should not be confused with insignificance. Ultimately, all those numbers may effect more change than any number of political transitions.

For decades — for ever, really — African statistics have been a shambles (South Africa, with a large and highly skilled statistics office, is an exception to this rule). Data sets, if they existed, were incomplete and out of date. Sample sizes were tiny. Population censuses were haphazard and poorly conducted.

In his 2013 book Poor Numbers, Morten Jerven describes in great detail how official statistics for most African countries are so bad as to be almost meaningless. “This is particularly likely to yield nonsense or misleading findings, resulting in inaccurate economic histories of postcolonial Africa,” he says.

It doesn’t take a maths professor to work out why this is a problem. Inevitably, bad numbers lead to bad policy. “If you don’t have baseline figures, you don’t know if you’re making progress. Unless you have figures to back up what you’re doing, there’s no basis for criticism, there’s no basis for policy,” said Jean-Paul Van Belle, an expert in this subject at the University of Cape Town’s infor- mation systems department.

The contrary is true, too. Good numbers are likely to produce much more effective solutions to all kinds of problems, from poverty and education to healthcare and economic development. Which is where the data revolution comes in.

Over the past decade or so, a concerted effort by national statistics offices, along with various international organisations and nongovernmental organisations, has completely changed the African statistical landscape.

“Africa is undergoing a data revolution,” said the inaugural Africa Data Revolution Report, released this week by various United Nations agencies and independent research organisations. It documents the statistical improvements that have already been made and outlines the work that must still be done.

“While there are significant variations among them, many African countries are producing more, regular and better quality data ... Collectively [at the continental level] and individually [at the country level], there is high-level political will and commitment to harness data for accelerated sustainable development. There is also significant goodwill and commitment from a number of private-sector companies, NGOs, media entities, citizen groups and academic institutions,” the report said.

The statistics about the statistics are impressive, according to the report: there’s the fact that almost nine out of 10 people in Africa now live in a country that has conducted a population census in the past 10 years, and that almost all Africans (99%) live in a country that has con“Challenges ducted a household survey in the past decade.

But the data revolution is far from complete. Although there has been a massive increase in the quantity of statistics, the quality of the numbers is still questionable.

In its report on the strength of African statistics, the Mo Ibrahim Foundation, which publishes the Ibrahim Index of African Governance, a comprehensive source of continental data, said: remain in the frequency and quality of the data produced. Working towards more timely and reliable data through clear measurable initiatives represents the next hurdle for Africa. Moreover, data deficits still exist in crucial areas such as civil registration. Focusing on getting the basics right should be a priority for the continent.”

The foundation says there is plenty of room for improvement, particularly in national statistics offices, which are generally underfunded and understaffed. But making those improvements will be expensive. Jerven estimates that obtaining accurate statistics for the 169 developing countries covered by the United Nations’ Sustainable Development Goals will cost $254-billion, nearly double the total current expenditure on foreign aid.

Van Belle agrees. “African governments can’t afford to spend $20 per inhabitant to collect data. Collecting data is not a high priority if there are so many priorities perceived as more urgent. Do you hire one statistician or give 20 people a job? But if you don’t hire the statistician, then you don’t know if what you are doing is right or wrong,” he said.

Another major issue is that not all the data produced in Africa is freely available. “Data are still not significantly open by default in most African countries, nor are data largely accessible in usable form to most stakeholder groups,” said the Africa Data Revolution Report.

That these issues are being discussed at all is still progress, however. Although far from finished, there’s no doubt that Africa’s data revolution has arrived. And although this revolution might not be televised, it will definitely be measured.

Although there has been a massive increase in the quantity of statistics, the quality is still questionable