Who Benefits and Who Suffers from Politically Based Lending?

When the cabinet of the new Indian state Andhra Pradesh (whose borders were made official last June) offered substantial loan forgiveness to farmers in 2014, some central bankers feared it had created a problem: a vicious cycle of bad-credit culture, with banks showing less willingness to lend to the whole industry in the future. The bankers’ concerns may be valid—research suggests that this kind of political meddling in bank lending can be costly.

Researchers are learning about such effects by studying politically motivated interference in India. A 2009 study by Harvard’s Shawn Cole found evidence that politicians influence banks to increase lending to farmers before elections. And to determine the real economic cost of such distortion, Chicago Booth’s Nitish Kumar, a PhD student, examined the effects on bank credit being made to the agricultural and manufacturing industries.

Kumar studied state elections between 1999 and 2008 from all 30 states in India, each of which follows its own five-year election cycle. He compared bank lending and firm activity in states with upcoming elections (those both one and two years out) to those without upcoming elections. This helped to rule out other economic variables, such as changes in monetary policy, which could have driven the results and would have affected all firms.

According to the study, which relied on lending data from one of the largest banks in India, banks increased their agricultural lending during election years by 9.4 percent, relative to nonelection years. The biggest increase occurred, predictably, in locations where a higher percentage of voters are involved in agriculture. Additionally, the effects were more pronounced in areas where elections were tightly contested. “Wooing farmers with cheap loans during elections seems attractive [to politicians],” says Kumar.

The increase in agricultural lending, however, meant that the manufacturing sector received 2.7 percent less in loans during the same period, compared to nonelection years. Why did politicians favor agriculture over manufacturing? The answer lies in the census data. According to the 2001 census, 57 percent of the total workforce was engaged in agriculture. Manufacturing clocked in at just 13 percent.

Among manufacturing firms, Kumar calculated mean firm leverage in an election year to be 3.3 percent lower than two years postelection. As a consequence of this reduced access to credit, during election years manufacturing firms both drew on their cash reserves and decreased their investment by 3.6 percent, relative to nonelection years.

Moreover, due to the credit squeeze, manufacturing firms were forced to fire many temporary workers, translating to a loss of 52,000 jobs during election years, or a 1.2 percent drop in total employment.

Ultimately, political influence forced firms to cut down production during election years, such that their output was 3 percent higher two years before an election. It also stifled new firm creation, with the share of new firms during election years being 8.8 percent lower than during off-election years.

This distorted lending negatively affected India’s GDP, in large measure due to a fall in plant-level production efficiency. When plants produce less during elections, they have to leave some of their productive capacity idle. And since the manufacturing industry represents a considerable portion of the Indian economy—its average contribution to the GDP between 1999 and 2008 was 27 percent—even a small drop in firms’ production efficiency significantly impacts the economy. Kumar calculates that, had it not been for this politically motivated skewing of resource allocation that reduced production efficiency, India’s GDP would have been 0.2 percent higher.

Political distortion may affect other industries and economic areas beyond the ones explored in this paper. “Quite often, when politicians dole out freebies to politically important segments of voters during elections, the argument is, ‘What is the harm in providing additional support to some sections?’” says Kumar. “The answer is: more often than not, it comes at the cost of someone else.”