I always wondered what keeps the movements like the Narmada Bachao Andolan going, especially in the face of claims made that they might have hindered the possibility of the displaced being well compensated. However, this article was an eye opener. It also showed something more than that- the innovative use of an instrumental variable strategy in order to asses the impact of dams on rural poor.

It seems intuitive enough that dams will have a differential impact on down stream districts than the upstream districts or the district where the dam was built. This accords well with the engineering fact that dam construction requires some gradient for the river, a low but a non zero one more suited for irrigation dams. The authors exploit variation in dam construction induced by differences in river gradient across districts within Indian states to obtain IV estimates.

The main findings of the paper are as follows:

  1. Dam construction leads to a significant increase in irrigated area and agricultural production in districts located down stream and they also provide insurance against the rainfall shock in these districts. However, in the own district a dam induces something like a mean preserving spread in agricultural production.
  2. Each dam is associated with a significant poverty increase of 0.77 percent in its own district. However, poverty reduction in the down stream district is significantly lower than that. As per the authors there are 1.75 districts down steam and thus the poverty reduction in down stream districts is insufficient to compensate for the increase in poverty in upstream districts.
  3. The poverty impact of dam construction is accentuated in districts with a history of relatively more extractive institutions (as captured by their historical land tenure system). This is in accordance with Banerjee and Iyer (2000) who show that the ability of of the population to organize themselves and obtain public goods exhibited marked differences across regions with different historical land tenure legacies (as mentioned on pp. 637).

Ever wonder why the displaced choose to stay in their village than move else where? The answer lies in the relation between income and risk aversion. Risk aversion decreases with income. So if dams increase poverty in the own district, it means in absence of credit and insurance markets, the affected will not choose to migrate. Does that seem too much out of line?

This article is definitely a good read and all those who are interested in the economics of public policy should definitely browse through.


Duflo E & Pande R, (2007), Dams, The Quarterly Journal of Economics, May, pp: 601-646.

Banarjee A & Iyer L, (2005), History, Institutions and Economic Performance: The Legacy of Colonial Land Tenure Systems in India, American Economic Review, XCV (4), 1190-1213.