The release earlier this year of the UN High-Level Panel’s report calling for a “data revolution” has produced quite a stir among the international development community about what this actually means, why we need it and how we get on with it.
This was the topic of discussion today at an event organized by the Center for Global Development (CGD), with speakers from the Partnership in Statistics for Development in the 21st Century (PARIS21) (presentation slides are available here).
While we’re still a ways from a common definition, the idea behind a “data revolution” is that if the development community just had more data on how people are faring, where public – and private – money is flowing and the impact that programs are having, they could increase the pace of development, presumably through better, more informed policymaking.
I’m all about data and evidence-driven policymaking. But we all know that evidence is just one (some would argue quite small) driver of policymaking, particularly among poor, corrupt countries with limited capacity to deliver.
As Lant Pritchett aptly pointed out at a recent event on the current “fad” of randomized control trials (RCTs) as part of this broader movement towards evidence-based policy, one of the ironies of this movement is that its advocacy has been evidence-free – namely, there is little evidence (much less “rigorous” evidence) to show that evidence alone (in this case RCTs) affects policy in any significant way. Oxfam’s Duncan Greene has also blogged about this.
That’s not to say that evidence is not important – or that policymaking shouldn’t be more evidence-based. I happen to think quite the opposite is true. Without data, we cannot adapt programs, learn from our mistakes and improve results. The issue is that data – and technology – is just one component of the “data revolution.”
It’s also about politics. It’s about building capacity and strengthening institutions at the local level. It’s about changing cultures and mindsets. And it’s about getting data to be more demand-driven. At the end of the day, data is only as good as its ability to influence policymaking.
How can we incorporate these factors into a true “data revolution”? I think this is a question that too often gets pushed under the rug, including at this event.