Oriana Bandiera est professeure d’économie à la London School of Economics. Elle est spécialisée dans la conception d’expériences de terrain qui permettent d’évaluer comment le comportement individuel est déterminé par des motivations économiques et des relations sociales. Ses travaux récents portent sur : - des expériences de terrain mettant à disposition des primes pour l’accomplissement de tâches d’intérêts public au sein de groupes de travailleurs en Zambie, - l’évaluation aléatoire de réduction de la pauvreté à grande échelle - l’autonomisation des femmes au Bangladesh, en Ouganda et en Tanzanie.
Featured Affiliate Interview
The success of TUP [Targeting the Ultra-Poor] indicates that sufficiently large one-off transfers set beneficiaries on a stable trajectory out of poverty and reduce the need for social protection in the future.
What got you interested in development economics, and particularly on female empowerment and targeting the ultra-poor?
What I find most interesting in economics is that it allows us to understand and to measure what motivates individuals and the constraints they face. This is especially interesting in low-income countries where the constraints are more binding and the consequences for human welfare starker. The two programmes you mention (ELA [Empowerment and Livelihood for Adolescents] and TUP [Targeting the Ultra-Poor]) provide a good source of variation to understand whether vulnerable girls and poor women find themselves in that situation because they intrinsically have a different “type” (less forward-looking, less able to exert self-control) or because of the constraints they face. Both ELA and TUP are large one-off injections of skills (ELA) or skills and capital (TUP) that lift the constraints. The fact that women, especially those receiving TUP, dramatically change their economic lives and continue on an upward trajectory years after the end of the programme casts serious doubts on the hypothesis that the poor have immutable traits that keep them in poverty. This has important policy implications: if the poor had immutable traits that keep them in poverty the only policies that can work to reduce poverty are regular transfers and similar social protection programmes. In contrast the success of TUP indicates that sufficiently large one-off transfers set beneficiaries on a stable trajectory out of poverty and reduce the need for social protection in the future.
What is one current research project that you're particularly excited about?
TUP programmes are very effective but also very expensive. A key driver of costs is frequent training and household visits; a programme with only the capital component would cost about half of the full package. Understanding whether capital or cash transfers alone can have the same effect is thus key to ensure that TUP can reach the largest number of potential beneficiaries. There are of course many reasons why cash alone might not work: spending it in small chunks might be more tempting, protecting it from relatives more difficult, training might be necessary to enable the ultra-poor to achieve a good return. We are currently evaluating a TUP programme in Pakistan where we introduced a treatment arm that gives beneficiaries the choice of cash instead of the standard TUP package. By revealed preference cash is better: almost everyone in this treatment arm chooses cash. We are holding our breath to see whether they are able to invest and transform their lives like the beneficiaries who receive the full package.
What is your "dream evaluation"? (It doesn't have to be feasible!)
Development economists have made enormous progress in designing and implementing increasingly sophisticated evaluations. Where we lag behind relative to other applied fields is on data. Most development papers have to rely on survey data to measure the effect of an intervention, yet administrative data would allow us to obtain a more detailed picture at a fraction of the cost. The problem is that even when these data exist, unfriendly software makes extraction very time consuming and merges of different data sets are often fraught with difficulties. My dream is that in the near future there will be investments in data capacity.
What is your most memorable story from the field?
When we ran our recent household survey in rural Zambia, the survey team would write regularly for feedback. When they started close to the district capital their questions were predictably of the kind "the village chief made us wait a day...can we extend our stay here.” As they ventured inside the districts the problems became more and more unimaginable including entire villages surrounded by water, boating in crocodile infested rivers, and lions on the loose. Luckily the team had been trained to put safety first (a message that we repeated over and over) and nobody got hurt. That’s when I really started thinking seriously about admin data.