Photograph by oneVillage Initiative on Flickr; Illustration by Bloomberg View
Photograph by oneVillage Initiative on Flickr; Illustration by Bloomberg View

Did the slick animations in the Girl Effect video that’s been viewed 3.6 million times compel you to buy soccer cleats for a South African girl? Maybe Matt Damon persuaded you to invest in Water.org so you can supply clean water and toilets to the world’s poor. Perhaps you’ve bought something (RED) to fight AIDS?

If you haven’t, you’ve missed out on a big trend. Overseas development assistance is going private. More and more, private charities, foundations and nongovernment organizations are joining or displacing government organizations such as the U.S. Agency for International Development and multilateral groups like those of the United Nations to help the poorest of the poor. Celebrities vouch for various projects, and the Internet makes giving simple. From 2005 to 2010, overseas aid from U.S. private groups jumped by 164 percent, while official development assistance grew by only 8 percent, according to figures tracked by the Organization for Economic Cooperation and Development.

Apart from making Americans feel good, does the money they donate to global causes ($22.8 billion in 2010) actually do good? The results of a study published last month highlight how surprisingly hard it is to answer that question. It involved measuring the effects of installing innovative cookstoves in 2,600 households in 44 villages in India’s Orissa state.

The clean-cookstove movement has considerable momentum. Championed by Hillary Clinton and Julia Roberts, the Global Alliance for Clean Cookstoves has raised $135 million to reduce smoke exposure from cooking indoors with primitive stoves or open fires: The World Health Organization estimates that such exposure causes 2 million premature deaths a year. The alliance hopes to distribute 100 million smoke-reducing stoves, which come in endless varieties, by 2020.

The researchers behind the study, Rema Hanna of Harvard, and Esther Duflo and Michael Greenstone of the Massachusetts Institute of Technology, expected to confirm the health benefits of the $12.50 mud-based, chimneyed cookstoves installed in Orissa. In laboratory experiments, clean cookstoves have been shown to release fewer pollutants and burn more efficiently than traditional cooking methods. However, such tests can’t predict what will happen in the real world. That’s where a well-designed randomized trial is irreplaceable.

In Orissa, households were randomly assigned to three waves of stove construction, and researchers measured a meaningful reduction in smoke inhalation in the first year after a stove was installed. Over a longer period, however, they saw no health benefits and no reduction in fuel use. That’s because once the stoves and chimneys developed cracks, the villagers generally chose not to fix or maintain their new devices but instead went back to their old, smoky ways of cooking.

This doesn’t suggest the clean-cookstove campaign should be abandoned so much as slowed down. It would be wise to test various designs in real-life settings, and, where necessary, take more time to human-proof models. Clean-cookstove advocates need to develop incentives for families to stick with the stoves, and they need to study why many villagers in the India trial embraced the devices yet continued using their conventional cooking fires as well. Otherwise, the innovative stoves of today could wind up in the same junk piles as models from efforts decades ago.

Those castoffs are a reminder that, however well-intentioned, many assistance programs for the developing world can prove fruitless. One obvious area for similar rigorous evaluation is microcredit, the fastest growing field in global poverty reduction. According to the Microcredit Summit Campaign, the number of families receiving microloans grew from 7.6 million in 1997 to 137.5 million in 2010. A handful of studies on whether these programs ultimately reduce poverty have come up with ambiguous findings, pointing to a need for more, and longer-term, trials. A related initiative that needs field-testing is insurance for the poor against sickness or economic calamity -- for example, insufficient rainfall for farmers.

The purpose is not to debunk the idea of helping poor countries. On the contrary, such work bolsters the case for aid by ensuring that dollars are well spent. Generally, such trials point out the need to alter, not ditch, assistance programs. Frequently, they shine a light on programs that have proved successful beyond expectation. A series of trials in Kenya, for instance, showed that young adults who had been dewormed as children were more productive and earned more than peers who didn’t get the treatment.

Groups such as Duflo’s Abdul Latif Jameel Poverty Lab and Innovations for Poverty Action are dedicated to such studies. Their work deserves support, and their conclusions demand attention. Long the standard for medical programs, randomized trials can help sort promising projects in foreign aid from truly effective ones, and speed our way toward a better world.

Read more opinion online from Bloomberg View.

Today’s highlights: Albert R. Hunt on the next Kennedy superstar; David Aaker on marketing brands; Edward Conard on what drives the U.S. economy; Aaron David Miller on safe zones in Syria; Simon Johnson and Peter Boone on the euro and banks; Rachelle Bergstein on wedges and World War I.

To contact the Bloomberg View editorial board: view@bloomberg.net.