Public sector banks increase their rural lending just before elections, especially in marginal constituencies. Could this happen to the employment guarantee scheme as well?On December 15, writing in the Hindustan Times on the employment guarantee scheme, Abhijit V Banerjee of MIT and Pranab Bardhan of Berkeley cited a paper by a graduate student at MIT called Shawn Cole. Cole, who has been doing research on India, seems to have made a startling discovery. This is that public sector banks increase their rural lending just before elections "" that too in districts where the margin of victory for the ruling party was small. Moreover, no such increase is to be seen in the case of private banks. Even more striking is his finding, which goes against received wisdom, that "state governments did not spend more in elections years." I googled Shawn Cole and came across his paper*. It is a fascinating study, for more than one reason. But first, I wish to make a query to the RBI.Cole says "the Reserve Bank of India (RBI) provided access to confidential credit data from the Basic Statistical Returns of the Reserve Bank of India, from 1992-1999. These data include every loan made by every bank in India." My query is: would the RBI have made the same data available to an Indian researcher? Probably not, but that is the trivial question because as long as the information comes out, it doesn't matter how it does so. (One might also enquire: will the Right to Information Act apply to the RBI?) The non-trivial question: does Cole's thesis hold? The data and the analysis certainly seem to suggest so. The credit dataset he has used contains information for 412 districts in 19 states, with a total of 3,296 observations. His methodology appears fine and the tables and charts he has presented irrefutable.He compares the amount of credit extended in election years to the amount issued in non-election years. It turns out that there is a sharp increase in the election years."For agricultural credit, there is clear evidence of electoral manipulation (by politicians)... lending by public sector banks is about 6 percentage points higher in election years than non-election years." Not just this. It also turns out that "credit is targeted towards districts in which the majority party just won or just lost the election. This targeting is observed only in election years."And not just that, either because as he puts it, "agricultural loans in villages with government banks are much less likely to be repaid on time." Clearly, someone is telling the borrowers that if they delay repayment long enough, until the next election, the loan will be waived.Cole also clearly establishes that the costs of misuse of public sector banks by politicians are so great that the "additional government credit may have no effect on output". He shows that a substantial increase in government credit in villages whose banks were nationalised did not have any effect on observable agricultural outcomes. This has significant implications for the employment guarantee scheme.In other words, the loans are not loans but just handouts to favoured supporters. But lest they feel very pleased about this, they should wait for his final punch line, which should make small-time politicians and fixers sit up. This is that senior "politicians appear to care more about winning re-election than rewarding their supporters, and they do so by targeting 'swing' districts."Could it be that Sonia Gandhi is trying to push through something that will be used by the government to target the marginal constituencies? Are the nice guys like Jean Dreze being led up the garden path?Cole's paper would suggest so. It is up to Gandhi to prove otherwise and for Dreze to be wary of being used by politicians. It has happened once before when well-meaning people got similarly excited in the early 1970s over Indira Gandhi's garibi hatao stunts, only to be sadly disillusioned by her myriad cynicisms.*Fixing Market Failures or Fixing Elections? Agricultural Credit in India, November 15, 2004 (http://econ-www.mit.edu/graduate/candidates/download_res.php?id=79)
