Towards the De-criminalization of elections and politics

Abstract

We study the role of crime and money in elections and politics in the context of several recent far reaching judgments of the Supreme Court and the Central Information Commission. Based on publicly available data of over 62,800 candidates who contested National and State Assembly elections from 2004 to 2013, we show that both crime and money play an important role in winning elections. Thus average assets go up from Rs.1.37 crore per candidate to Rs.2.03 crores for third place candidates, Rs.2.47 crores for runners up, Rs.3.8 crore for winners, Rs.4.27 crore for winners with some criminal record, and Rs.4.4 crore for winners with serious criminal records. About 30% of elected representatives in Lok Sabha and State Assemblies have criminal cases against them. Similarly, while only 12% of 'clean' candidates without any criminal record win, 23% of those with a criminal record, and again, 23% of those with serious criminal records win. Several candidates and winners have multiple cases, and on average there is 1 serious case for every 5 candidates and 1 for every 2 winners. Women candidates and representatives clearly show a much lower level of criminal records. With rising inequality, proliferation of political parties leading to winners with smaller percentages of the votes, and the role of big money in elections, votes can be easily 'bought'. We end with some tentative recommendations for improving the situation. These include legal, and administrative changes, and the role that civil society and voters can play. We look at options of moving beyond the current first past the post 'British' system to other systems including the American Presidential system, the European proportional representation system, the French style run-off elections and the Japanese system of multi member constituencies. Immediate electoral political considerations might promote ad hoc changes that would lead us astray. A long public debate and discussion is needed to try and evolve a reasonable consensus.