One way to answer this question is to see if lending behaviour responds to the election cycle. In a recently published research paper (available for free on HBS’s website), I analyse bank lending throughout India, over an eight-year period beginning in 1992. I focus on state electoral contests, using a dataset compiled by the Reserve Bank of India. (India is blessed with some of the best banking data in the world.) Several interesting facts were thrown up.

First, public sector bank credit moves with electoral cycles. In particular, agricultural credit increases by 5-10 percentage points in a state in the year leading up to an election. Second, there seems to be significant geographical targeting: one observes the largest increases in credit in districts in which the elections are particularly close. Third, following elections, loan write-offs appear largest in districts in which the winning party enjoyed the greatest margin of victory.

This result is observed in agricultural lending, but not for lending to industry or other sectors. The effect is present in government-owned banks, but not private sector banks.

One might argue that these results indicate that politicians are merely making sure that lazy bankers do what they are supposed to, namely make profitable loans to farmers. If this were the case, one might expect to see an increase in agricultural output in states that experience credit booms. Using state agricultural output data, I test this hypothesis, but find no increase in lending.

Of course, this problem is not unique to India. Research by Serdar Dinc finds that, around the world, government banks lend more in election years than their private counterparts. And at least a few people have observed the recent controversial relationship between the government and the banks in the US with some Schadenfreude.

... contd.