Case Study 2 — The RCT That Changed Development Policy: Deworming in Kenya
In 2004, Michael Kremer and Edward Miguel published one of the most influential papers in development economics: a randomized evaluation of a school-based deworming program in Kenya. The study found that treating children for intestinal worms cost about $0.50 per child per year and produced enormous benefits: a 25% increase in school attendance, significantly better health, and — in long-run follow-ups — higher adult earnings.
The intervention
Intestinal worms (hookworm, roundworm, schistosomiasis) infect about 1.5 billion people worldwide, mostly in tropical developing countries. The worms cause anemia, malnutrition, cognitive impairment, and chronic fatigue. Treatment is simple: a single dose of albendazole or praziquantel, costing pennies.
The RCT design
Kremer and Miguel randomly selected 50 of 75 schools in the Busia district of Kenya to receive free deworming treatment. The other 25 schools served as the control group (they would receive treatment later). The randomization ensured that the treatment and control schools were statistically identical — any difference in outcomes could be attributed to the deworming.
The results
- School attendance: 25% higher in treatment schools
- Health: significant improvement in weight, height-for-age, and hemoglobin levels
- Spillover effects: untreated children in treatment schools also benefited (because worm transmission fell in the community)
- Long-run follow-up (2016): treated children earned 13% more as adults, worked 12% more hours, and were more likely to hold skilled jobs
The cost-effectiveness
Deworming costs about $0.50 per child per year. The long-run earnings gains are estimated at $30–50 per dollar spent. This makes deworming one of the most cost-effective development interventions ever measured — and it provided the evidence that persuaded governments (Kenya, India, Ethiopia, and others) to adopt national school-based deworming programs reaching hundreds of millions of children.
What the study teaches about RCTs
1. Randomization is powerful. By randomly assigning treatment, the study eliminated the possibility that differences between treatment and control groups were due to pre-existing differences rather than the intervention.
2. Cheap interventions can have outsized effects. A $0.50 pill produced a 25% increase in school attendance and a 13% long-run earnings gain. The lesson: in development, the binding constraint is sometimes not money or infrastructure but a simple health intervention that unlocks children's ability to learn.
3. Follow-up matters. The initial study (2004) showed attendance and health gains. The 10-year follow-up (2016) showed earnings and employment gains. Short-run studies miss long-run benefits.
4. RCTs have limitations. The deworming study works beautifully for a specific, well-defined intervention. It cannot tell you whether Kenya should democratize, liberalize trade, or reform its judiciary. The big institutional questions are not amenable to RCT testing.
Discussion questions
- Deworming costs $0.50/child and produces $30–50 in long-run benefits per dollar spent. Why isn't every country doing it?
- The study found "spillover effects" — untreated children benefited too. What does this tell you about deworming as a public good?
- Can the RCT approach be applied to questions beyond health (e.g., education quality, governance, justice)?