When the winners of the Nobel Prize in Economic Sciences were announced in October, pockets of applause erupted across Kenya. Nothing like what Eliud Kipchoge’s race against time had elicited weeks earlier. Yet within hours, small tightly knit groups gathered to toast the greatest acknowledgment of their work. Two laureates joined in by video link.
An American, Michael Kremer, French woman Esther Duflo and Abhijit Banerjee, an Indian-American, were officially awarded the Nobel earlier this month for “their experimental approach to global poverty alleviation”. This resonates with many Kenyans, and Africans, many of whom worked on the research and resultant programmes initiated by the laureates. The approach, and the movement it has spawned, were forged through research evaluations conducted in Busia County.
Organisations traditionally focus on measuring outputs, not impact. Outputs show whether a programme is implemented as intended – for example, the number of wells built, scholarships provided or people trained. Important, but the real question is what the initiatives accomplished. Did the wells reduce disease, the scholarships improve performance or long-term welfare or the training increase knowledge or skills?
It is much more difficult to isolate impact and attribute it to an intervention. The work of the laureates raised the bar for development work and inspired a global movement – led by organisations like Innovations for Poverty Action – focused on measuring impact and getting that evidence used.
The laureates applied randomised evaluations – a form of testing typically associated with medical research – to development work. By offering a programme to a randomly selected group and having a similar, random comparison group go on with business as usual, they established a way to confidently know if a programme made any real change to people’s lives.
Some critique this approach: Is it right to give a potentially life-improving programme to one group and not another? But NGOs, even governments, are almost always limited in how many people they can reach at once; randomisation simply leaves it up to chance. And even if they can deliver a programme on a large scale, testing it first offers a reliable way to learn whether it accomplishes its aim or needs adjusting before further scale-up. The question is whether not knowing, and potentially wasting resources, is better.
Others have questioned the usefulness of randomised controlled trials: Are their findings generalisable (applicable in other contexts)? Do they help? The answer is ‘yes’ on both counts. In Kenya in the 1990s, Michael Kremer and colleagues found that delivering textbooks to students, which seems intuitively helpful, did not substantively improve learning: It helped a few top performers but the majority saw no educational benefit.
Conversely, the same set of evaluations found that delivering cheap deworming treatment strongly impacted on education outcomes – reducing school absenteeism by up to 25 per cent. So, textbooks weren’t scaled up but the government, with technical support from some organisations, adopted a policy of mass, school-based deworming that benefits six million children annually and has significantly reduced parasitic worm infections, according to Kenya Medical Research Institute (Kemri).
A second example comes from early studies on improving water quality, which found that improving access to safe water by protecting wells didn’t necessarily result in the consumption of safe water. Drinking water often became contaminated when carried home and stored or handled in less-than-sterile environments. Today, a global non-profit, Evidence Action, leverages this evidence to deploy a safe water delivery model that improves water quality not just at its source but at its point of use. They reach roughly two million people in Kenya alone.
The non-profit Precision Agriculture for Development (PAD), co-founded by Michael Kremer, harnesses the power of mobile technology to empower smallholder farmers with actionable information to improve productivity and resilience. PAD continually tests messaging to deliver impact for Kenyan farmers.
The Nobel laureates’ approach is rooted in a healthy dose of self-scepticism while remaining optimistic that viable and cost-effective solutions to challenges confronting people in poverty can be found or forged. It’s a rare combination that is much needed in the development landscape, and one that we are proud to join hundreds of Kenyans in celebrating.
Mr Byatta is Regional Director at Evidence Action; [email protected] @PaulByatta. Mr Suleiman is regional director - East Africa for Innovations for Poverty Action; [email protected] @suleimanasman. Ms Nekesa is Chief People Officer at Precision Agriculture for Development; [email protected] @CarolNekesa5