How much evidence is enough before we know that a global development intervention works for people? How much evidence is enough to know that a program is worth scaling to millions of people because it works and benefits lives across multiple settings and contexts?

These are great questions that Michael Hobbes raises in an article in the most recent issue of The New Republic. In fact, we at Evidence Action think a lot about this. Our mission is to scale programs that have been proven to work so they benefit millions of people.

Unfortunately, Mr. Hobbes used a poor example to raise these questions by focusing on deworming. In the case of mass deworming of children and our Deworm the World Initiative, the policy has followed the (rigorous) evidence. Deworming has a well-proven, clear causal chain from intervention to effect.

Large-scale deworming has been shown by numerous studies from different settings to improve health AND educational outcomes in children. Mr Hobbes cites a review of studies that questions the strength of this evidence, but in fact this analysis is an outlier withfindings influenced by studies with methodological weaknesses and the exclusion of other, methodologically strong, evaluations. Organizations such as the Copenhagen Consensus, GiveWell, and the Poverty Action Lab have conducted reviews that find deworming to be highly cost-effective and an important area for investment. Regular mass treatment for intestinal worms in countries where they are endemic increases children’s health during a critical period of their development which, in turn, increases their cognitive functions and their ability to attend school.

Large-scale deworming is good public policy recommended by the World Health Organization, and governments agree. Evidence Action’s Deworm the World Initiative’s partnership with governments in Kenya and India treated over 35 million children in 2013/14, and more governments are making school-based deworming programs a priority. And even though large-scale deworming has been more than ten years in the making we have a long way to go: Hundreds of millions of children still need regular treatment at low cost.

Would we love to see more rigorous research on the educational impacts of deworming? Other strategies for improving schooling? Of course. Indeed, we actively support researchers who focus on generating additional evidence when that makes sense in parallel to service delivery. However, we don’t think every country considering deworming needs a local trial first. The drugs are cheap; the WHO recommends the program; the evidence base is solid. There is room for more research, but to leave kids with worms untreated now would be a serious mistake.

We agree with Mr. Hobbes that we need more evidence about how to tackle tough problems in global development. We ourselves are optimistic about the potential for progress today by building on the evidence base that we have.