fbpx Big Data: Vital Statistics for Development | Asian Development Bank

Big Data: Vital Statistics for Development

Article | 25 November 2013

Writing exclusively for Development Asia magazine, Bill Gates calls for a concrete strategy to inspire a revolution in development data. Better data should be a cornerstone of the post-2015 development agenda, he says, helping to save lives and empower those left behind.

Hunger, as in the physical sensation, is easy to identify. But hunger the statistical category is more abstract.

When the UN's Food and Agriculture Organization (FAO) calculates national rates of undernourishment, they don't do what you might expect: take a large sample of people and determine how many aren't eating enough. Instead, they estimate how much food is produced and imported in a given country.

Then, they develop a "coefficient of variation" to estimate how those calories are actually distributed among the population (since not everybody has the same amount to eat). Finally, they calculate a "daily energy requirement." What comes out, after some arithmetic, is the undernourishment statistic.

This is no way to track an important piece of the first Millennium Development Goal (MDG). The countries of the world committed themselves to cutting hunger in half, but they don't know who is, in fact, hungry. As a result of data gaps like these, the leaders debating the agenda that will succeed the MDGs, which expire in 2015, have called for a "data revolution."

"When you're trying to reach a goal, data not only tells you if you're succeeding, but it also suggests which activities you should do more of in order to improve your results. "

- Bill Gates

That's an achievement in itself. Few people believe in the power of data in the way I do. When you're trying to reach a goal, data not only tells you if you're succeeding, but it also suggests which activities you should do more of in order to improve your results. Ultimately, the better the data available in the development field, the higher the quality of people's lives in poor countries.

Still, "data revolution" is a vague term. It means different things to different people. Now that there's excitement about the idea, the next step is a conversation about specifics that can lead to a concrete strategy for improving the way we collect and use data for global development.

At the very least, a "data revolution" needs to accomplish three goals: In each country, decide more rationally which data needs to be collected; invest in developing countries' ability to collect good data over the long term; finally, ensure that data is widely available and informs public policy.

On the first goal, right now there is little coordination and therefore lots of fragmentation among the governments and organizations collecting data about development. Understandably, funders pay for surveys that serve their particular needs. As a result, however, some statistics are gathered several times in different ways while other statistics aren't gathered at all. In other important areas of development, such as the provision of vaccines, the world has created coordinating systems like the GAVI Alliance to help donors and developing countries set joint priorities, vastly increasing efficiency (and saving millions of lives). I am not suggesting a GAVI for data, necessarily; it is for the data and development communities to figure out the best solution, but there must be a way to harmonize the chorus of voices asking for data.

Investing in developing countries' ability to collect good data over the long term - the second goal of a data revolution - recognizes that national statistics on development issues are a classic public good; the people who pay to collect them can't reap all the benefits, so they're poorly funded.

There are two parts to getting this right. First, we have to invest in data-gathering capacity in developing countries. These investments will continue paying off long after the big push to establish baselines for 2015. Second, we have to invest in technology for better data. Right now, too many statistics are tracked as if digital technology didn't exist. To measure crop area, for example, a person with a tape measure and compass walks a farmer's field, jotting down angles and measurements. The results can be accurate if the person is well-trained, but even then the tape and compass method takes up to 15 times longer than walking around a field with a handheld GPS. Some countries are using satellites to do this work even faster. There are many similar opportunities to take huge leaps forward in data collection, and it is past time to seize these opportunities.

Finally, we must ensure that data is widely available and informs public policy. As the world gets better at collecting data, we have to get correspondingly better at making sure it's open and easy to use. Some governments resist releasing data if it seems to be bad news that might generate criticism. Government data should be open to civil society and entrepreneurs. We also have to make sure those who make policy have the capacity and the incentive to use the data properly. As this article argues, the lack of data for development is a serious problem. But so is the fact that data users don't take full advantage of the data we do have. As 2015 approaches, world leaders will debate which priorities should be reflected in the next set of global goals. No matter the outcome of these debates, it is clear that investing in data will be essential. Because the need for accurate, actionable data is common across all development objectives. If the data revolution happens in the right way, the world--and, in particular, developing countries--will get better at everything: saving children's lives, increasing agricultural productivity, empowering women, and on and on and on. That kind of impact does merit a steady, thoughtful revolution.

This article was published in Development Asia magazine with the title Vital Statistics

Development Asia is a publication of the Asian Development Bank.