Why It Matters That Donors Evaluate Their Work

Redwood City, Calif.

Could a national effort to fight malnutrition, backed with significant money from the World Bank, have no impact?

Yes, according to Howard White, executive director for the International Initiative for Impact Evaluation, a nonprofit dedicated to determining what works in international antipoverty efforts. Mr. White was among more than two dozen speakers who participated in the opening day of the Global Philanthropy Forum, an annual meeting for donors.

His speech underscored the importance of testing whether programs work before putting more money into expanding them.

Mr. White described the Bangladesh Integrated Nutrition Project, an effort to fight malnutrition in Bangladesh supported by the World Bank. The program provided mothers with counseling on nutrition, supplemental feeding to severely malnourished children, and other kinds of support.

But the program didn’t work: A study by Mr. White and others, who were brought in after questions had been raised about the effort, found that there was no difference in malnutrition levels between people participating in the program and those who weren’t.

Why not? First, he said, the program focused on the wrong people. It sought to change the way mothers thought about their children’s nutrition—but in rural Bangladesh, it’s the men who go shopping for food and the mother-in-laws who make most household decisions. So trying to change the behavior of mothers had very little impact on family health, Mr. White said.

Second, changing people’s behavior takes time. “The idea that you can go into a country and in 18 months change its practices is rather optimistic,” Mr. White said.

Also, the neediest children were often being left out of the supplementary feeding programs, while children who didn’t qualify were often included. Mr. White said that nutrition experts running the programs weren’t able to assess which children ought to participate.

Finally, in a quarter of the cases, women were giving their children the supplementary food but then skipping their regular meals.

Lack of Comparisons

So why did the World Bank back the program in the first place? It did conduct an evaluation, Mr. White said, after it decided to expand the program across the country. But the study looked only at the health of people in areas where the program was being run; it didn’t compare those people’s health with the health of people in areas that weren’t part of the program.

And at the time, malnutrition rates were falling across Bangladesh, most likely due to broader changes in the economy, Mr. White noted. But because the bank didn’t compare malnutrition levels across the nation, he said, it erroneously credited the program with reducing those malnutrition rates.

Mr. White urged donors to remember the consequences of throwing money at bad programs: “Children died because the World Bank spent money in the wrong way.”