While it has been known for long that politicians exploit the state machinery for their narrow political gains, for the first time, we have systematic evidence informing us of such misuse. In his study of how the lending patterns of Indian banks coincide with national elections, Professor Shawn Cole of the Harvard Business School finds that the lending by public sector banks tracks the electoral cycle, with agricultural credit provided by the public sector banks increasing by about 10 percentage points in an election year. The author employs his district level data on bank lending from the Reserve Bank of India together with the data on elections from the Election Commission of India. The author investigates and reassures readers that this effect of elections on agricultural credit is not due to aggregate annual shocks to agricultural investment. Nor can this pattern be attributed to budgetary manipulation since state governments were not found to spend more in the election years. Also, there is no such pattern coinciding with elections in non-agricultural credit. And, such correlation with election years does not show up in the credit provided to agriculture by the private sector banks or foreign banks.

Is it possible that when the threat of a re-election looms, politicians are just ensuring that banks are fulfilling their legal obligation to provide credit to the poor sections of society? If this were the case, an increase in agricultural credit would be observed across all districts that face an election. However, the author finds that such increases in agricultural lending by public sector banks in the election year are particularly targeted in the swing districts?districts in which the election ended up a posteriori to be particularly close. Districts whose populations were strongly in favour or strongly opposed to the incumbent majority party did not receive this largesse in the election year. Of course, this outcome is expected since the marginal value of directed credit is more in those districts where the election is expected to be close. Since politicians would have a good assessment a priori of which districts would be the swing districts, directed agricultural lending to districts that eventually turned out to be so is consistent with political manoeuvring of public sector banks. Since it is the public sector banks that are controlled by the politicians and a significant part of the electorate is involved in agriculture, the author convincingly makes the case that this directed credit by public sector banks is an unhealthy outcome of the race for the ?Supreme Parliament?.

The author finds a different pattern with respect to forgiving of agricultural loans coinciding with election years. In the years following an election, in a district where the margin of victory was about 15 percentage points, write-offs of agricultural loans is approximately 27 percentage points higher. In contrast, in districts where the ruling party lost, there is no increase in write-offs of agricultural loans. Note that the author is not using across-the-board loan waivers like the UPA-1?s loan waiver scheme in 2008-09. Instead, these are specific loan write-offs in some districts but not in others. Thus, politicians reward their supporters immediately following elections by causing public sector banks to write off loans to borrowers in constituencies in which governing politicians enjoyed the greatest support. These patterns stand in contrast to those for lending, where only the marginal districts are rewarded. From an economic point of view, such targeted agricultural credit and forgiveness of agricultural loans are essentially money down the drain. The author finds that neither does such credit increase agricultural investment in the targeted district nor does it increase its agricultural output.

Note that such targeted lending does not constitute public spending of an anti-cyclical nature. If this directed lending were indeed anti-cyclical, then such spending is useful. However, election times do not necessarily coincide with the troughs in economic cycles. In fact, if at all possible, the incumbent government would choose to have an early election precisely when the economy is doing well (the 2004 early election recommended by the NDA government is a case in point). In other words, election times are more likely to coincide with good economic times. Public spending during election times cannot necessarily be labelled ?anti-cyclical public spending?. It is more likely to be pro-cyclical public spending that is purely a transfer from the taxpayers to some fortunate farmers.

What is the magnitude of the loss to the taxpayer from the political machinations indulged in by governing politicians? To estimate this amount, note that the total amount of direct credit provided to agriculture by public sector banks is more than R2,65,000 crore (figures for 2010 from IndiaStat database). Most agricultural loans are short-term credit with maturities less than one year since these are loans provided for the purchase of inputs such as fertiliser and seed. Thus, the total stock of credit is also equal to the flow of credit to agriculture in a year. Therefore, a 10% increase in the election year translates into more than R26,500 crore of public money down the drain. If we use the figures for all forms of agricultural credit (including indirect credit to agriculture as well), the loss amounts to a staggering R37,000 crore. If we go by the CBI?s estimate of the loss due to the 2G scam to the national exchequer of about R50,000 crore, we lose more than three quarters of the 2G scam amount every five years. The fact that such politically motivated redirection of credit did not occur in private banks clearly pinpoints the egregiously negative effects that government ownership has on public sector banks in India.

The author is a PhD in finance from the University of Chicago and is currently faculty in finance at the Indian School of Business