India has about 4,300 ‘large dams’, and ranks third behind USA (6,600) and China (22,000) in dam construction. [Source: The Economist]

In a 2007 paper, Esther Duflo and Rohini Pande analyzed the effects of large irrigation dams in India. They compared the impact of dams in districts where they were built to districts that were located downstream to the dam. Their main findings were:

  1. Dams improve irrigation and agricultural production in districts located downstream to the dam, but have no significant impact in the district where the dam is built.
  2. In districts where dams are built, rural poverty increases. This increase is sharper than the fall in poverty in districts located downstream to the dam.
  3. The impact on poverty is a function of state institutions. (Districts with extractive institutions fare worse than other).

The paper has a lot of data compiled from novel data sources. For instance, consider the trends in dam construction collected from the World Registry of Large Dams, maintained by the International Commission on Large Dam (ICOLD):

“Between 1971 and 1999 the number of large dams quadrupled from 882 to 3,364 [...] (46 percent of the districts had no dams in 1999). There was significant regional variation in dam construction [...] Dam construction was rapid between the mid-1970s and late-1980s but slowed down considerably in the 1990s.”


The following maps illustrate the trend even better:

Source: http://econ-www.mit.edu/files/796

The paper also points out that dam construction in 1970 and 1999 was primarily concentrated in western India, with relatively little construction in north and north east. I was wondering if someone had data on small dam construction in the north east? (My guess is that large dam construction in the north east would be difficult due to geographical constraints, but there may be many small dams).

Source: Esther Duflo & Rohini Pande, 2007. “Dams,” The Quarterly Journal of Economics, MIT Press, vol. 122(2), pp. 601-646, 05. (Click here for ungated access to a 2005 draft version of the paper)