The limits of plug-and-play development

THE hard-fought contest for the World Bank's presidency reignited the debate over differing models for development. On one side is the "big development" camp, which advocates institution building and large-scale investments in health, education and infrastructure. On the other stand the "small development" proponents who focus on micro-policy questions that impact a particular community or issue. Divergent as the two models are, development projects, both big and small, increasingly have one common feature—the use of technology as an agent of change.From hi-tech gadgets (laptops and cell phones) to low-tech widgets (solar powered lamps and inexpensive syringes), engineering innovations are being deployed to tackle social problems in the developing world. The anecdotal evidence offers grounds for optimism: telecenters that brought the internet to rural India, mobile phones that link farmers in remote African villages to buying centers, cheap chlorine dispensers which are widely used in Kenya. But what is the impact of technology in alleviating poverty and spurring development? Two recent studies, both using randomised field trials, evaluated the outcome of technology-based initiatives. The results are striking.First consider the One Laptop Per Child (OLPC) initiative. The American charity has an ambitious mission—transform the quality of education in the developing world by giving every poor student a laptop. Targeting a $100 laptop, OLPC succeeded in creating a usable computer at a very low price point (the actual number was closer to $200). Unfortunately most of the attention in the project was focused on the technology and not enough on its efficacy. In the first rigorous evaluation of the programme, the Inter-American Development Bank (IDB) found little evidence that the laptops influenced educational outcomes. The study, conducted in Peru four years after the programme was launched, found no improvement in math or language. While the computers did lead to some gains in cognitive skills, the authors concluded that access to a laptop didn't improve attendance. Neither did it motivate students to spend more time on their homework.There is similarly disappointing news on cooking stoves. The World Health Organization estimates that indoor pollution from primitive cooking fires contributes to 2m deaths annually. One solution is to use clean cooking stoves. At a cost of $12.50, these stoves are an inexpensive way to reduce respiratory ailments and improve air quality. The Global Alliance for Clean Cookstoves (GACC), a public-private initiative, is making a big push for 100m homes in the developing world to switch to clean stoves by 2020. But a new NBER paper by Rema Hanna from Harvard University and Esther Duflo and Michael Greenstone from MIT, questions the long-term health or environmental benefits from this programme. The authors evaluated a clean-stove programme in eastern India, covering 15,000 households over five years. Their study found that after the initial year, enthusiasm for the stoves waned and households didn't make the necessary investments to maintain them. As a result, the programme had very little effect on respiratory health or air pollution.Both these projects highlight some common misconceptions in using technology for development. For one, solving intractable social problems requires fundamental changes in the target population. It also needs a supportive institutional framework to reinforce the right behaviour. Technology can complement this process, but it is no substitute for the human element. In Peru, simply adding laptops to the classroom, without investing in teachers who were proficient in computer-aided education, meant that the academic impact was limited. The IDB paper rightly points out that in poor countries where wages are low, development money may be better spent on labor-intensive education interventions than on expensive tools.There is also a big difference between well-controlled field trials and real world usage. As engineers and scientists, it is easy to fixate on the technology. It is a lot harder, however, to predict human behaviour and how that interacts with technology. With the clean stoves, Ms Hanna and her colleagues found that previous evaluations of the programme relied on trained fieldworkers to inspect and repair the stoves regularly. Not surprisingly, households readily switched to the newer stoves and the results were positive. But in cases where owners were responsible for the upkeep and proper use of the stoves, take-up of the programme was slow. What's more, in households that did make the switch, use of new stoves declined over time, broken stoves weren't repaired and households ultimately reverted to the polluting cooking fires.This isn't to say that technology isn't useful in development. One only has to look at the success of M-PESA, Kenya's mobile money platform, to appreciate the impact of technology that is well integrated into people's lives. But the belief that countries can leapfrog on economic and social issues solely on the basis of technology seems optimistic. The poor can be frustratingly stubborn to an economist for failing to conform to a rational-agent model. Instead of expecting the poor to "do the right thing", a better approach may be to design devices that fit into their lives with minimal effort. Else, despite good intentions, these programmes won’t affect meaningful change other than the addition of a shiny new toy.