The data on the makeup of the two houses of Parliament, as of 2010, support this notion of a more “genteel” Rajya Sabha—although not one that is completely devoid of suspected criminality (perhaps a sign of just how entrenched crime and politics are). Rajya Sabha members are slightly older and more educated, as would be expected given the cultural and institutional characteristics of a “House of Elders.” Crucially, Lok Sabha members are significantly more likely than their Rajya Sabha counterparts to face criminal cases (30 percent versus 18 percent, respectively) as well as cases of a serious nature (19 percent to 7 percent) (figure 6.8). On average, Lok Sabha MPs have a greater number of pending cases, IPC violations, and charges warranting stiffer jail sentences (if convicted) than their upper house counterparts.
This difference is statistically significant and is reinforced by more systematic statistical analyses. Controlling for a range of individual factors, if one were to pick an average member of the Lok Sabha and move him or her from the lower to upper house, that move alone decreases the likelihood of possessing a serious criminal case by roughly 9 percent.69
Figure 6.8. Share of Lok Sabha and Rajya Sabha MPs with pending criminal cases, 2010. (Author’s calculations based on affidavits submitted to the Election Commission of India by Lok Sabha MPs elected in 2009 and affidavits of Rajya Sabha MPs compiled by the Association for Democratic Reforms in 2010)
GETTING DOWN TO THE VOTER
The data employed thus far provide further validation of the connection between the appeal of candidates with criminal reputations and the salience of social cleavages. Politicians with criminal records play less of a role in reserved constituencies as well as in indirectly elected bodies because in both instances the salience of identity politics stands diminished. These tests bolster confidence in the broader validity of the logic of demand for criminality. Be that as it may, these validation methods are still somewhat removed from the observable actions of voters themselves. While they tell us about where criminal politicians might hold the most appeal, they leave one question unanswered: do voters knowingly support politicians with criminal records because they can act as effective representatives?
Survey Evidence
The most direct way of resolving this lingering question is to ask Indian voters whether or not they support candidates with criminal records who can “get things done.” A group of researchers, including myself, had the chance to do exactly this in late 2013, in advance of the 2014 Indian general election.70 The survey, sponsored by the Lok Foundation, sought to measure the social and political attitudes of more than 68,000 randomly selected respondents across the country (see Appendix C).
To survey respondents, we posed the question, “Would you vote for a candidate who delivers benefits to you even if s/he faces serious criminal cases?” Twenty-six percent of respondents openly admitted that they would vote for a candidate who gets things done but also faces pending criminal cases of a serious nature. In other words, one out of four Indians in the Lok sample was willing to openly admit he/she would sacrifice probity for credibility. This is a sizeable percentage given the potential for social response bias (i.e., respondents might be inclined to say that they would not vote for candidates with criminal records—even if they truly believed otherwise—out of fear that the survey enumerator might morally judge their response).
To place this number in context, in the 2004 and 2009 parliamentary elections, the average parliamentary candidate with at least one criminal case of a serious nature earned roughly 17 percent of the vote in his/her constituency (compared to 7.5 percent for candidates not facing any cases of a serious nature).
Social Biases
The survey data support the proposition that a sizeable proportion of Indian voters reward politicians with reputations for engaging in criminal activity and for getting things done. The next piece of the puzzle is tying this finding back to the issue of identity politics. The credibility of criminality, as I have argued throughout, is inexorably tied to issues of ethnic politics and their salience: candidates with criminal reputations tend to have the greatest resonance in contexts where the salience of identity politics is greatest. That is, in areas where social divisions are most pronounced, the benefits criminals can provide are most desired.
As part of the same Lok survey, we asked two questions to gauge the degree of social biases among respondents. These questions were meant to pick up how “salient” social cleavages along caste lines actually are on the ground. The first question was designed to detect the degree of “negative” social bias present among respondents. Specifically, we asked respondents, “Would you be troubled if an [upper caste/OBC/SC/other community] candidate wins the election in your constituency?”
Because the survey recorded information on the backgrounds of each of the survey respondents, survey enumerators knew to which caste grouping each person belonged. With this information in hand, the enumerators asked respondents’ views on a candidate from another caste community. For instance, if the respondent belonged to an upper-caste family, he or she would be randomly shown either a prompt relating to an SC or OBC candidate. All respondents were given an option of a candidate winning the election who was not from their own community but hailed from another caste. We found that 36 percent of respondents indicated that they would, indeed, be bothered if a candidate from another community won the election. This provides a reasonable baseline estimate of negative social bias among survey respondents.
To detect “positive” ethnic bias, or the favoritism for one’s own group, we asked a second (slightly different) question modeled along the same lines as the previous query: “Is it important to you that [upper caste/OBC/SC/Your community] candidate wins the election in your constituency?” In this variant of the question, the prompt respondents saw corresponded to one’s own community rather than to an “out group’s.” Forty-six percent of respondents answered that it was important for them that someone from their own community should win the election in their constituency.
To sum up, almost half of all respondents declared that having a caste ally represent them in Parliament was a priority, and nearly two in five respondents demonstrated unease at the prospect of an MP from a different caste community winning the election.
Integrating Social Biases and Criminality
Thus far the survey evidence reveals that criminal candidates who demonstrate competence elicit significant levels of approval and that voters harbor strong ethnic biases—in favor of their own as well as against members of other communities. Our next step was to integrate these two insights by taking individual responses to three questions—one on criminality and two on social biases—to see how much overlap there is. Although not all respondents who harbor social biases are willing to support candidates with serious criminal cases, those that do are significantly more likely to exhibit such biases (figure 6.9). The effects are larger for individuals espousing a positive co-ethnic bias.
Another way of visualizing this relationship is to aggregate these results up to the state level and then plot them against one another. When this is done, one sees a striking correlation, for instance, between the positive bias and criminality responses (figure 6.10). States in which a larger percentage of respondents are willing to electorally support candidates with serious criminal cases also tend to boast a greater share of respondents who harbor positive ethnic biases, a finding that is exactly in line with an identity-based explanation for criminality in politics.
Disaggregating the same relationship by a respondent’s urban-rural status indicates that while the relationship is strong in both settings, it is stronger for rural respondents—as demonstrated by the steeper slope (figure 6.11). These differences are perhaps an artifact of the diminished presence of civil administration in rural India, which allows for strongmen who fill in the gap in government capacity to win much greater popular support. Obviously, this is speculative; further research is needed to parse the exact driver of the variation. Yet in broa
d terms, the survey results provide broad support for the demand-side logic of the market for politicians tied to crime.
Figure 6.9. Social biases and popular support for criminality, 2013. (Neelanjan Sircar and Milan Vaishnav, “Ignorant Voters of Credible Representatives? Why Voters Support Criminal Politicians in India,” unpublished paper, Carnegie Endowment for International Peace, 2015)
THINKING ABOUT EQUILIBRIUM
The empirical support for the “demand” hypothesis suggests there are genuine reasons why candidates with serious criminal backgrounds gain traction in the electoral marketplace. Nevertheless, it is important not to push the argument too far. After all, while there is considerable electoral backing for such candidates—recall that 34 percent of MPs and roughly 31 percent of MLAs face ongoing criminal cases (21 and 15 percent with serious cases, respectively)—they have not entirely saturated the political system.
The market, in other words, seems to be situated in a kind of equilibrium where candidates with criminal reputations are a serious factor in politics but have not entirely swallowed up the system whole. This raises an obvious question: why do more candidates tied to criminality not win elections?
Figure 6.10. Patterns in social biases and support for criminality by state, 2013. The label markers represent individual states. (Neelanjan Sircar and Milan Vaishnav, “Ignorant Voters of Credible Representatives? Why Voters Support Criminal Politicians in India,” unpublished paper, Carnegie Endowment for International Peace, 2015)
The first and most obvious response is to point out the conditional nature of the marketplace. It is best to think of the marketplace for criminality not as a giant wholesale market for the country but as a series of hundreds—even thousands—of local markets. As such, the marketplace is far from monolithic, working in the same manner in all areas of the country. Rather, in any given marketplace support for candidates linked to criminal activity is contextual, varying according to local incentives and circumstances. For instance, my analyses in this chapter have taken advantage of quirks in India’s electoral design that—inadvertently—create variation in the importance of social divisions to demonstrate that criminality in politics is more prevalent in places where there are incentives for politicians (and parties) to mobilize along ethnic lines.
Figure 6.11. Urban/rural patterns in social biases and support for criminality by state, 2013. The label markers represent individual states. (Neelanjan Sircar and Milan Vaishnav, “Ignorant Voters of Credible Representatives? Why Voters Support Criminal Politicians in India,” unpublished paper, Carnegie Endowment for International Peace, 2015)
One data point that gives credence to this notion of “equilibrium” is the relative stickiness, or path dependency, in criminality within states over time. Data from ADR taken from the two most recent state assembly elections (as of August 2015) reveal that the share of MLAs with serious cases that won the previous election is a fantastic predictor of what that share will look like in the future (figure 6.12).
Second, it is also possible that the desire for self-financing politicians, which is one reason parties give tickets to candidates linked to crime, may vary across space. While I have assumed that this desire is relatively constant, one could relax that assumption, thus opening up the possibility than in certain areas, due to socioeconomic, geographic, cultural, or demographic reasons, elections might simply be less costly than in other locales. Where the cost of elections is relatively lower, parties might not weigh a candidate’s “money power” as heavily when determining ticket selection. If such were the case, the attractiveness of candidates with criminal cases would stand diminished.
Third, some voters might refrain from supporting candidates with criminal reputations on moral grounds even if doing so hurts their personal or their community’s perceived self-interest. For this class of voters, there might be a stigma associated with voting for candidates accused of breaking the law.71 Recall that the Lok survey data discussed above found that 26 percent of respondents would be willing to vote for a candidate who faced serious cases if he or she were capable of “getting things done.” Presumably this also means that the vast majority of respondents were not willing to make that trade-off. It is impossible to know how these voters would act in an actual electoral setting, but even if a sizable percentage voted their conscience, they could possibly prevent a large number of candidates tied to criminality from triumphing at the polls.
Figure 6.12. Path dependency in the share of state assembly candidates with pending serious criminal cases. The x-axis represents the share of state assembly candidates in the most recent election (as of August 2015) with at least one serious criminal case. The y-axis represents the share of state assembly candidates in the previous election with at least one serious criminal case. Each dot represents a state. (Author’s calculations based on data from the Association for Democratic Reforms)
On occasion, there is also a point when a politician’s criminal linkages become a liability. When voters in India perceive that candidates are no longer delivering on their promises, they have a track record of throwing such nonperformers out of office. While candidates with criminal cases have a statistically significant electoral advantage, they do not possess an indefinite hold on power. The competitive nature of Indian elections denies them such a luxury.
Furthermore, there are times when parties are forced to part ways with candidates if their alleged criminality attracts excessive attention beyond their local constituencies. Criminality in politics is largely salient at the local level when it is clearly rooted in the social context of a given constituency. Once a politician’s criminality turns into a larger state or national issue, parties often get nervous because they fear such attention might draw investigative scrutiny and/or because it could create a backlash among voters in constituencies (or among demographics) where criminality holds little (or less) appeal.
There are many instances of parties dropping politicians once their alleged criminal exploits get too “hot.”72 For instance, a senior BSP leader and onetime health minister in the Mayawati cabinet, Babu Singh Kushwaha, was forced to resign in 2011 when authorities uncovered a significant scam in his ministry. Auditors claimed that a network of politicians, led by Kushwaha, engineered a massive fraud in which they embezzled billions of rupees from public health funds and later conspired to cover it up. The alleged conspiracy involved the murders of several top health officials with knowledge of the corrupt activities.73
At a time when Uttar Pradesh was gearing up for elections, Mayawati did not want to risk the possibility that Kushwaha might damage her party’s fortunes, and he was soon forced out of the BSP. His departure was hailed as a big loss for the party given his status as a leading backward-caste leader. Interestingly, the very moment the BSP ousted the tainted minister, he became the subject of an intense bidding war among the BSP’s main competitors in the forthcoming 2012 elections. The BJP quickly moved to induct him into the party, citing his caste appeal: “Kushwaha represents a very backward class community. He has come into the party and we have welcomed him. . . . SP, BSP and Congress are not serving the Most Backward Castes. . . . OBCs feel their rights are being taken away.”74 The party was forced to suspend his membership once it too came under national media fire for embracing him.75
PAINTING WITH A NARROW BRUSH
The dynamics of support for criminally suspect candidates are not limited to Bihar or any one state in India. The data presented in this chapter demonstrate that the marriage of crime and politics is a much broader phenomenon. The identity-based logic suggests that criminality will be lower in reserved constituencies and among indirectly elected legislators, and, indeed, the evidence is consistent with these hypotheses. These findings point to two paradoxes. It appears that while voters regularly elect legislators allegedly engaged in illegal activity, when legislators themselves are given the opportunity to select their peers, they are less likely to embrace such candidates. This is not because politicians are more v
irtuous than voters, but rather because the two actors face different incentives.
Second, although the elite discourse in India often points to the mobilization of lower castes as contributing to a coarsening of Indian politics, politicians affiliated with the lowest groups in the social hierarchy are significantly less likely to be under criminal scrutiny than their peers higher up the ladder.76 This is not due to inherent differences in criminal propensities, or socioeconomic or cultural factors per se, but rather the way electoral politics is structured in reserved versus open constituencies.
Writing for the New York Times, one commentator noted in 2012, “Criminals and strongmen have long been a feature of Indian politics. Their ill-gotten wealth provides easy campaign cash, and they often control constituencies with strong caste or religious loyalties.”77 It is as concise an explanation as can be given for the supply and demand of suspected criminals in contemporary Indian politics.
III
7 Crime without Punishment
From Deep Roots to Proximate Causes
AT FIRST GLANCE, Arvind Kejriwal cuts an unlikely figure as a rabble-rouser. As an interviewer once noted, the middle-aged, bespectacled Kejriwal “comes across as a simple, but stubborn man, with no material taste. He wears trousers that seem a size too big, his shirts are what government clerks in small towns wear.”1 The only thing remarkable about his appearance is the cap he is fond of sporting: a white hat inscribed with the words Main Aam Aadmi Hoon (I am a common man).
Before 2011, few in India had any clue who Arvind Kejriwal was. Like many other bright young men and women who enjoyed solid middle-class upbringings, Kejriwal studied engineering at one of the prestigious Indian Institutes of Technology. After graduating, he worked for a leading business house before gaining entry into the Indian Revenue Service, one of the many branches of the Indian civil service. Frustrated by the confines of the Indian bureaucracy and the sclerotic pace of government decision making, Kejriwal eventually quit the service and devoted his full attention to social activism, immediately making waves by stridently advocating for the enactment and full implementation of the Right to Information (RTI) Act, a landmark government transparency initiative passed by Parliament in 2005.
When Crime Pays Page 28
