Evaluating Development Interventions: Esther Duflo

Esther Duflo shares her insights as MIT economist and director of the Poverty Action Lab, where she and colleagues have developed randomized evaluations to answer critical questions relating to poverty alleviation. Duflo was mentioned in today’s post on teaching resources and the policy prescriptions implicit in Kony 2012.

Video | This entry was posted in Africa, Development, Evaluation. Bookmark the permalink.

3 Responses to Evaluating Development Interventions: Esther Duflo

  1. Joseph says:

    Good point. I still wonder, given my own experiences in Tanzania naively thinking that each village was “similar” only to find glaring differences in politics, history, class structure, etc., how assessments of similarity are made without considering power. Incentive-based development agendas create power dynamics, too, in the process of awarding certain behaviors. The example of her work where she lures teachers to the school to gather their daily pay that would otherwise be lost in their absence is highly political in nature (aka highly power-laden). My point isn’t that power should be consider in order to negate it – this is impossible; even for Amizade students who are in Karagwe to learn are influencing the politics of the community in unpredictable ways. I guess maybe now that I’m heading back to academia, I’m trying to think in my “constellational” ways. If I submitted a Duflo-esque Geography dissertation proposal to my PhD advisor, whereby I was proposing to enter two communities where I didn’t speak the language, offer incentives to people to do certain things then record what they do, apply empirical data collection, then write about why one community does one thing and another does something else, without knowing the language, my adviser would simply tell me to stop in my tracks and go apply for a job at the World Bank. I guess I can understand why Duflo is appealing to people like BIll Gates and Bono and why an MIT economist is an expert in solving the world’s problems, given that neoliberal solutions to development are in style.

  2. Joseph says:

    Duflo’s solution of “clinical” development sounds sexy, but she misses one vital ingredient: power. The analogy to medicine trials is far flung, considering that enzymes in cells ALWAYS react a certain way to chemicals in all human beings (with some rare exceptions, of course). Pakistani, Nigerian, American, it doesn’t really matter, because the cells have no experiences, memories, histories, or politics. People do. The enzymatic reactions are not conscious. People are. I distrust a policy panacea that claims to be medicine-like and therefore universal. Policy programs in Nigeria, Pakistan, or the U.S. play out differently because of things like power struggles, history, cultural baggage, etc. Identifying a control is problematic because inevitably an outsider will be assessing the community to see if it is “missing” the variable that exists in the experimental group. Outsiders can’t really “see” politics or power and how these affect decisions at the local individual’s level. This is where ethnography becomes very key, and why people spend the necessary time for good research to happen and learn the language.

    • Thanks for leaving a reply, Joseph. I agree completely that the absence of power is a flaw in Duflo’s work, as does this reviewer on the Guardian’s Poverty Matters blog: http://www.guardian.co.uk/global-development/poverty-matters/2011/apr/11/duflo-banerjee-rethinking-fight-against-poverty?INTCMP=SRCH – I think it’s a good review. Duflo does not, however, suggest that she brings universal insights. The work she and her compatriots at the Poverty Action Lab advance emphasizes comparisons between similar communities – because they’re aware that community constraints, culture, and behaviors are likely to vary across contexts. And in that sense, for what she’s evaluating (does this policy ‘work’ here, in the way it’s intended?), she doesn’t need to know about power (which would help her explain why it may or may not work). But in a broader sense – does everyone have the access to a fully human experience, empowerment, the right to live a life how they would chose to life, etc. – Duflo’s randomized evaluations can tell us little or nothing.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s