Why you should care
When the size of Nigeria’s economy almost doubled overnight, it became clear: Developing countries have a big data problem.
In April, Nigeria revised its books. Overnight, GDP nearly doubled, and Nigeria shot past South Africa to become the continent’s largest economy.
This wasn’t gimmickry or dodgy accounting. According to Nigerian officials, it was a correction — one that’s likely to be repeated in developing countries around the world. Just last month, Kenya recalculated its GDP upward, to the tune of 25 percent.
The episodes highlight what those in the development world have known for decades and mostly ignored: Africa has a data problem. A big data problem. Official GDP figures for most African countries, even relative powerhouses like Angola, are dramatically understated — literally off base, for they’re measured against baselines so old they’re obsolete. Nigeria was calculating its income against a baseline set in 1990, which meant it undervalued fast-growing sectors like telecom and movies.
Developing countries aren’t alone in their accounting woes — even wealthy economies must revise their GDP calculations — but theirs are likely the worst. World Bank data indicate African countries have the weakest statistical capacity in world. And the stakes of “Africa’s statistical tragedy” are quite serious. We don’t really know continental vaccination rates, or how many kids go to school in, say, Rwanda. Most important, we don’t know whether the approximately $150 billion spent annually in development aid is actually helping.
Demand for Africa’s data is surging, and national statistics offices — some of them musty, one-man affairs — haven’t kept up.
These stakes will shift as aid-dependent countries transition to emerging markets, experts say. Against lowballed GDP figures, key investor metrics like national growth tend to be overstated. To compensate for higher risk, businesses in developing markets must offer higher premiums. Expect a good deal of confusion and inefficiency as Ghana, Nigeria and Kenya enter capital markets.
“Data in Africa is becoming more and more important, and it’s likely to become more contested,” says Morten Jerven, an economic historian and author of Poor Numbers: How We Are Misled by African Development Statistics and What to Do About It (2013). The book argues that Africa’s income and growth accounting is awful, few admit it’s awful, and that most economic decisions regarding Africa — aid or investment — are ill- or misinformed. “More than half the rankings of African economies up to 2009 may be pure guesswork,” Jerven wrote.
Mostly, this is a problem of capacity, not intention. Demand for Africa’s data is surging, and the continent’s national statistics offices — some of them musty one-man affairs with marginal electricity — haven’t kept up. A July report from the Data for African Development Working Group found that most national statistics offices in Africa lack reliable funding and, in fact, depend on aid donors for their budgets. The aid donors are generally happy to fund their statistical priorities, like household surveys, but not so keen on national statistic “building blocks” like births and deaths, growth and poverty, and taxes and trade. Governments aren’t stepping in.
Still you can’t insulate data from politics, the report argues. “To the extent that data is used to hold people accountable, it will be misreported,” says Amanda Glassman, an author of the report and a researcher at the Center for Global Development, a think tank in Washington, D.C. Another study she co-authored this summer found that across Africa, when school administrators are paid by the pupil, they report higher enrollment rates. Surprise, surprise.
There are even rational reasons to lowball GDP, Jerven argues. The poorer you seem to be, the easier it is to score concessional loans from the likes of the IMF. And a low GDP can make governance look a lot better than it is: Donors are in the habit of using tax collections to GDP as an indicator of the state’s health, and the lower the GDP looks, the better the ratio does.
“How on earth could it even pass that it took a quarter of a century for Nigeria to update its statistics? That was because nobody cared about the numbers. And if they did care, it made sense to look poor,” says Jerven.
With the United Nations calling for a worldwide “data revolution,” plenty of geeks are lighting up at the prospect of solving Africa’s data woes. But technology alone won’t fix it. There’s not enough valid data to mine, for starters, and for all the hullabaloo around gadgetry, “we’re not going to solve these problems by collecting satellite information,” says Glassman, referring to the trend of using satellite data on light intensity to predict GDP. “New technology can make the system more efficient, but it doesn’t overcome the perversions of misreporting.” What’s needed, she says, is the “kind of routine administration and data collection we take for granted.”
And then there’s the coming tide of investors, lured by Africa Rising. More and more, says Jerven, countries don’t want to look poor. They want investment, and they want to be able to offer investors lower risk and lower returns. These days, says Jerven, when Ethiopia’s rulers want a new dam, they don’t beg the World Bank for project finance. They float a bond and go to big commercial banks instead. That opens up an interesting possibility: Private investment could introduce rigor and resources into statistics infrastructure.
When Jerven’s book was published in February 2013, before Nigeria’s rebasing, Jerven hoped it might gain notice at think tanks and universities. But the first query he got about his book came from the private sector: Standard Chartered Bank. “They wanted to know just one thing: What is the GDP of Nigeria?”
This OZY encore was originally published Oct. 6, 2014.