Poor Economics - Esther Duflo at the LSE

The lecture began with Duflo setting out the pair’s intentions of looking beyond ideology. The ‘Is aid good or bad?’ dilemma was used to illustrate how ideological battles, have in the past, drawn focus away from effective policy making. Duflo argued that social experiments on a micro level have been able to teach us far more and ultimately are more useful in informing us on poverty reduction strategies.

During the jointly presented lecture, the audience were taken through six such experiments to examine the conclusions that could be drawn. Firstly, the issue of immunisation was addressed. With 9 million infant deaths globally each year, this certainly is a key area of policy making in developing nations.

Experiments were carried out in Rajasthan, India, to determine why, despite large amounts of aid being given, immunisation rates remained low. The researchers had hypothesized failure of bureaucracy and cultural factors may be to blame. Randomised controlled tests were set up to scrutinise these predictions using immunisation camps (held at regular and advertised times) and camps with additional incentives for mothers to immunise their children. The incentive used was a kilo of lentils per child immunised, enough to provide a ‘nudge’ to prevent procrastination but not of sufficient value to change behaviour if there was strong feeling against the jabs.

The results showed that regularly advertised camps increased immunisation rates three times compared to the control comparison. However, small incentives were highly successful at encouraging immunisation, rates increased six times the control. The research proved the importance of thinking wisely about these types of problems and how testing intuitive thinking on a micro scale could provide useful evidence about a policy’s likely impact. Counter-intuitively, the experiment also showed the cost-effectiveness of the policy; the cost of immunisation per child fell from $50 to $27, as lentils were given away. The fixed costs of the camp and the nurse were spread over many more patients – giving away money to save money!

Professor Banerjee went on to explain how orthodox thinking on nutrition and food aid was fundamentally flawed, citing the example of an experiment in India that had subsidised the price of rice by 10%.

Rather than consumption rising, as might have been expected, demand fell by 2.3% as individuals upgraded to more luxurious foodstuffs; the demand for shrimp within the locality rose! The unintended consequences of a policy may not be known until after funds have been spent.

Further examples examining education and enterprise were used to test accepted thinking. The importance of the political economy was also highlighted by assessing the effects of Brazil’s decision to replace butterfly voting slips with an electronic system. The new system could be better understood by those with low levels of education; 18% of poor voters were re-enfranchised, more representatives elected from poorer backgrounds and more public money was spent on healthcare. The mechanisms of democracy are often as important at the democratic institutions themselves.

Banerjee finished by identifying ‘The 3 I’s’. He believed that the experiments had suggested ideology, ignorance and inertia had to be overcome before serious gains in poverty reduction could be felt. Lazy and generic thinking should be eradicated; Banerjee and Duflo called for a more pragmatic, targeted approach where policy makers should think beyond the obvious and logical perspective.
The talk was engaging, and whilst many questioners sought to challenge the idea that lessons learnt at such a micro level could be scaled up to have macro effects, it was clear that for poverty reduction to be more successful than it has thus far been, enlightened thinking will need to drive policy making in the future.

Jon Andrews