As the strongman of Siwan, Mohammad Shahabuddin walked out of Bhagalpur jail with his swagger not just intact but now enhanced, one wondered what effect it would have on the district that once trembled at the mere mention of his name.
Fear will clearly be back. But Siwan’s local economy could be affected, too, going by the findings of a May 2016 research paper Do Criminally Accused Politicians Affect Economic Outcomes: Evidence from India. The authors— Nishith Prakash of the University of Connecticut, Marc Rockmore of Clark University and Yogesh Uppal of Youngstown State University— argue that the election of criminal-politicians negatively affects economic activity in the constituency and could even pull down national gross domestic product (GDP).
The paper uses the intensity of night-lights— data from satellite images that are widely used as a proxy for economic activity, especially at the local level— to make its point. The main point— the annual growth of night-time lights is approximately 22 percentage points lower in constituencies that elect a criminally accused candidate against those that do not. This, the paper further argues, could even translate into a 2.3 percentage point lower GDP growth a year.
The findings in the paper are based on a study of assembly elections in 20 states between 2004 and 2008. Going by affidavits filed with the Election Commission and analysed by the Association of Democratic Reforms, the study homed in on constituencies which have elected people facing criminal records and compared economic activity with constituencies that have not.
The study also uses an alternate measure of economic development and a proxy for provision of public goods— the number of incomplete roads projects in the constituency, using data on the Pradhan Mantri Gram Sadak Yojana. Accepting that roads projects get delayed across the world, the authors say that, in India, delays as well as poor quality of roads projects are often linked to weak governance and corruption.
Here’s what the study concluded, apart from a reduction in night-light activity.
One, the negative effect of criminal politicians getting elected is largely confined to certain states, which are development laggards since these are also likely to have weaker governance-related institutions and higher corruption levels. The study identifies these states as Bihar, Jharkhand, Odisha, Uttar Pradesh as well as Arunachal Pradesh and Assam, which span three categories of states— BIMAROU (with Odisha added to the traditional BIMARU acronym), least developed states and high corruption states.
In contrast, criminal politicians have less harmful effects on economic activity in the states that are non-BIMAROU, relatively developed and have low corruption levels. These states, the authors say, may have stronger institutions and so criminal politicians may be bit more restrained here.
Two, constituencies with criminal politicians facing serious charges, especially for financial crimes (involving loss to the public exchequer), have lower economic activity than those with criminal politicians facing non-financial and non-serious charges.
Three, the number of criminal charges against a politician also influences economic activity. The costs of electing representatives facing five and more charges are higher than of electing those facing two or more charges.
Four, the negative economic costs kick in after a lag. The authors guess that this may be because the criminal politicians take time to form a nexus with local bureaucrats in order to indulge in dodgy activities. The effect of neglected public infrastructure could also take time to slow down economic activity, the authors argue.
Five, the number of incomplete road projects increases in constituencies represented by criminally accused candidates; it is even higher in cases where politicians are accused of serious and financial charges. The number of incomplete projects is approximately three times higher in the high corruption and least developed states than in other states.
The study can be contested on a few points. Change in night-lights could have various other reasons, which a local don may not be in a position to influence, like the quantity and quality of power supply. The authors concede that changes in night lights can be influenced by supply-side and demand-side factors but say that there is still a correlation, though not one-on-one.
Also, an elected criminal may not need time to develop a nexus with the local bureaucracy. That nexus may already be there and be quite entrenched. In fact, that may be one factor influencing electoral success— not only will the official machinery help him in subtle ways, but the voters may be more willing to vote such a person because of his ability to ‘get things done’.
The extrapolation of local night-lights to GDP may also be a bit stretched. The study looks at elections between 2004 and 2008. It also makes the point that the share of members of Parliament facing criminal charges has increased from 24 percent in 2004 to 34 percent in 2014. But 2004 to 2008 was also the period of the boom years of the economy.
Nevertheless, the study does show that the costs of electing criminals could be more than lawlessness. This lawlessness has serious economic costs.
Siwan reportedly now has a thriving remittances-driven economy, as this report points out. Much of it seems to have happened during the time Shahabuddin’s writ was somewhat constrained (though it was not eliminated entirely). With his brazen comeback, it will be interesting to see how the night-lights over Siwan fare.