Thesis title: Experimental setups in poverty alleviation policies
Supervisor: Vololona Rabeharisoa
This dissertation focuses on the contemporary forms of intervention aimed at providing services, material help and other forms of assistance to people referred to as “the poorest of the poor”, inhabiting remote and underequipped areas in developing countries. These interventions present two main characteristics of interest. First, they take the form of experiments geared at evaluating the impact of various anti-poverty interventions while implementing them “in the field”, meaning in places where the poor live. In such experiments, the purpose of achieving improvements locally and the ambition of producing evidence about “what works” to reduce global poverty are completely entangled. In practice, it means that the implementation of whichever poverty-reduction program (e.g. distribution of solar lanterns, water chlorination, financial literacy trainings) is accompanied by a formal impact evaluation process based on data collection and quantitative analysis. More specifically, this evaluation process takes the form of a randomized controlled field experiment, or randomized controlled trial (RCT), a method that puts a strong emphasis on the correct attribution of causality. Second, these field experiments are, by design, particularly fit to test light-weight, small-scale, micro-level interventions, bringing minimalist responses to the lack of basic goods or infrastructures (e.g. solar lights). The smaller scale of these experimental projects is a full-fledge part of this new development paradigm: the size and shape of such devices define a particular political space and scope of action. The poor themselves, their behavior and their immediate environment are made into the main locus of poverty action.
The experimental approach and the proliferation of little devices are two distinct but entwined trends that together define an emerging paradigm for development aid and poverty-reduction policy. While the use of experiments for development purposes is by no means new, its contemporary forms are regarded as being part of a recent disruption in international development policy-making practices. This dissertation inquires into the effects of such development interventions. What does it mean and what does it entail to experiment on people to learn about and intervene on poverty? Using experiments is one particular way of producing knowledge, and it contributes to particular policy-making practices. I contend that the experimental approach has the effect of undermining relationalities between spaces inhabited by the poor and their outside. Restraining the spaces of causes in a limited perimeter and a-historic temporality obscures the possible role of exterior entities (e.g. other countries, multilateral organizations, multinational corporations) in producing poverty. Furthermore, this disconnection effect is also at play when it comes to imagine and design desirable futures: the perspectives of change proposed to the poor are completely disconnected (sans commune mesure) with the idea of the good life in richer places.