Despite some differences between my definition and that used by ADR, they are highly correlated. In the dataset of parliamentary candidates, the correlation between the two is .79. In the dataset of state assembly candidates, the correlation is .78.
Because the affidavit database I employ does not fully extend to the present day (and is missing some elections), I often use ADR’s data (and classification of “serious” charges) when providing aggregate statistics. For instance, in discussing data from assembly elections after 2009, I must rely on ADR’s data since these elections fall outside of my database. However, for nearly all fine-grained, empirical analyses contained here I rely on my own coding. In either case, I always cite which data I am using.
Candidate Assets
According to ECI guidelines, candidates must disclose their movable and immovable assets, which one can aggregate into an overall indicator of wealth. Candidate wealth is a good proxy for a candidate’s financial capacity, which can be nebulous and difficult to estimate because it is likely to be correlated with social networks, fundraising ability, and overall spending power. Unlike pending criminal cases, which are a matter of public record, a candidate’s financial details are difficult to verify independently. How much money candidates have in family bank accounts or the quantity of gold jewelry in their possession is hard—if not impossible—to discern. To counteract the possibility of false reporting, the ECI stipulates that furnishing false information is grounds for criminal prosecution or disqualification. In practice, however, it is not clear whether the threat of such punishment is sufficient to deter false reporting.
The natural tendency of most skeptics is to assume that candidates regularly underreport the true value of their assets in order to hide any illegal income or ill-gotten gains. Indeed, there are several examples of candidates doing just that. Despite this possible incentive, the reported assets of winning candidates remains startlingly high, as do the increases in assets over time. Despite possible incentives to underreport, there have been many recent investigations of high-profile politicians, including several current and former chief ministers, on suspicion of possessing “disproportionate assets.”
Thus, to the extent that candidates provide inaccurate information, the incentives are to underreport.8 Given the substantial nature of the declarations candidates do make, there is likely a lower bound on this underreporting. If one assumes that candidates with links to illegal activity have more incentive to obscure the true nature of their resources, it is likely to bias the wealth advantage of criminal candidates downwards.
MODAL CRIMINAL CHARGES
The most frequently appearing criminal charges in the state elections dataset, disaggregated by “serious” and “minor” categories, illuminate interesting patterns (table A-2). Of the five most common “minor” charges, three are violations of “public tranquility,” commonly associated with civil disturbances. The five most common “serious” charges account for roughly half of all serious infractions. Four of the top five charges are human body offenses, including attempted murder. The average candidate charged with serious violations of the law faces 2.4 pending cases.
Data for the modal criminal charges in the national elections dataset, disaggregated by “serious” and “minor” categories, were drawn from the 2004, 2009, and 2014 parliamentary elections (table A-3). In terms of criminal charges, there is a significant amount of overlap with the MLA candidate dataset. Indeed, with just a few exceptions, the lists are nearly identical. The average candidate charged with serious violations of the law faces 3.7 pending cases.
TESTING FOR POLITICALLY MOTIVATED CASES
A principal concern about the data on criminal cases is that these cases are not a reflection of actual criminal activity but rather stem from political motivation on the part of rivals. To address this concern, beyond adopting the coding strategy described above, I conducted four, additional empirical tests.
Table A-2. Modal Criminal Charges in Dataset of State Assembly Candidates
Distribution of Candidates with Criminal Cases
If one is concerned that criminal cases against candidates are primarily driven by political motivation, we would expect to observe an “arms race” in the filing of cases. Imagine I want to sully a rival candidate by getting a false case filed against him or her and that I succeed in doing so. The obvious response from my rival would be to do the same against me. If one assumes many, if not all, candidates think this way, one would expect to find a substantial degree of clustering in the distribution of criminal cases across constituencies. Interestingly, there is no evidence of this in the data.
Table A-3. Modal Criminal Charges in Dataset of Parliamentary Candidates
For state elections, of the 5,001 state assembly constituencies for which I have data, nearly two-thirds (65 percent) do not have a single candidate in the fray who has a serious criminal case. Just over one-fifth (21 percent) have only one candidate with a serious case (figure A-3). This means that an “arms race” mentality, if it exists, is restricted to a small minority of constituencies (the remaining 14 percent). By and large, where criminal candidacy exists, the dominant form seems to be a single candidate with a serious criminal record.
Figure A-3. Frequency distribution of state assembly candidates with pending serious criminal cases, 2003–9. (Author’s calculations based on affidavits submitted to the Election Commission of India by candidates contesting state assembly elections between 2003 and 2009)
Data from parliamentary elections paint a very similar picture (figure A-4). Leaving aside the 46 percent of parliamentary constituencies that have no candidate facing serious cases, the next largest category—those with one serious criminal candidate—accounts for 28 percent of all constituencies in the dataset. This means constituencies with two or more seriously charged candidates account for just 26 percent of constituencies in the dataset, again a small minority.
These data, of course, cannot conclusively rule out the political motivation hypothesis, but they do suggest that the dominant model is not one of candidates cynically getting cases lodged against one another (or at least cases that result in judicial cognizance being taken).
Political Vendettas
Second, if cases are politically motivated, one observable outcome might be that successful politicians are more likely to be susceptible to framing of false charges made by envious rivals. To analyze whether popular politicians disproportionately face criminal cases, I take advantage of the fact that seven states in the dataset (plus the national parliament) have experienced two elections under the affidavit regime (in 2003–4 and 2008–9). Thus, I can examine candidates at two time periods and test whether the presence of a serious case in time t is related to the political success the candidate experienced in the prior election in time t-1. Unfortunately, constructing a dataset of re-contesting candidates presents its own challenges for a host of reasons.9 After using a name-matching algorithm to identify the potential pool of re-contesting candidates over two election cycles, I used two unique identifying fields—candidates’ fathers’ names and their home addresses—to identify exact matches.10
Figure A-4. Frequency distribution of Lok Sabha candidates with pending serious criminal cases, 2004–14. (Author’s calculations based on affidavits submitted to the Election Commission of India by candidates contesting the 2004, 2009, and 2014 parliamentary elections)
To test the proposition that politically successful politicians are more likely to face serious cases, I investigated the correlates of having a serious criminal case at the time of the most recent election in time, t. The correlates are prior electoral performance (captured by vote share in the previous election), a binary measure of a candidate’s criminal status in the previous election, and a binary measure of candidate incumbency. Whether I use state, national, or combined state and national data, I find no evidence of a relationship between prior electoral performance and a candidate possessing a serious criminal case. In fact, the stronge
st predictor of a candidate’s criminal status is the presence (or absence) of a prior serious case declared before the previous election.
Incumbency Advantage?
A third way of explicitly testing for politically motivated charges is to study differences between incumbent and opposition politicians. If charges are easily manipulated, one would expect that the party in power would manufacture pending cases against its political opposition while simultaneously squeezing the judiciary to drop cases against ruling party politicians. To investigate this claim, I examine data from the north Indian state of Bihar, which is a poor state with a weak bureaucracy that could be vulnerable to political interference. According to data from Bihar’s November 2005 and 2010 elections, there does not appear to be any systematic pattern of political targeting: candidates from the incumbent party (marked in bold) are just as likely as opposition candidates to contest elections while facing serious cases (table A-4).
Timing of Criminal Charges
A final method of testing for politically motivated charges is to explore the timing of charges filed against politicians. For instance, if most charges are filed against politicians around election time, it would suggest an underlying political motivation. In their affidavits, candidates are required to disclose the date on which a judicial body has taken cognizance of each pending case. What we would like to know, however, is the date the initial charges were filed (as there is typically a lengthy lag between when charges are filed and when a court takes cognizance of a case). In 2006, the Allahabad High Court (which has jurisdiction over Uttar Pradesh) asked the government to provide information on the criminal records of all sitting politicians in the state. The report, which I obtained from the court, discloses the year in which charges were filed against Uttar Pradesh politicians with pending cases (and was current as of 2006).11
Table A-4. Share of Bihar State Assembly Candidates with Pending Serious Criminal Cases, 2005 and 2010
From these data, it is clear that the majority of charges against incumbent MLAs were not filed in election years: charges filed in an election year account for roughly one-quarter of all charges. While there is a sharp increase in charges filed in 2002 (the most recent election year in the data), there are also a substantial number of cases filed in the years before and after this election.12 A second interesting finding from these data concerns the pendency of cases: as of 2006, nearly 50 percent of cases against sitting MLAs were at least ten years old (with one case dating back as far as 1968). This reinforces the point made earlier that convictions are few and far between due to inefficiencies in India’s judicial system.
Appendix B: Details of Bihar Voter Survey
THE BIHAR POST-ELECTION survey took place immediately following the conclusion of voting in state assembly elections in October–November 2010. The survey was conducted in collaboration with the Lokniti Programme of the Centre for the Study of Developing Societies (CSDS). Lokniti is the only social science research organization in India that conducts post-poll surveys of voters following each state and national election. The data were collected as part of a special module Lokniti carried out in parallel with their standard post-election survey of voters.
To construct the sample, Lokniti first randomly selected 40 assembly constituencies (out of a total of 243 in the entire state). Assembly constituencies were selected using the probability proportionate to size (PPS) method. Within each of these 40 constituencies, four polling stations were then chosen at random as sites to carry out the survey. Polling stations were again sampled by employing the PPS method. After sampling polling stations, 20 respondents were randomly selected from the electoral rolls provided by the chief electoral officer of Bihar. Respondents were sampled using the systematic random sampling (SRS) method, which is based on a fixed interval ratio between two respondents. A sampled respondent’s list for each polling station was prepared, which is a comprehensive list of selected respondents with their complete name, address, age, and gender. Lokniti does not allow substitution in its surveys. Thus, out of a target sample size of 3,200 respondents (40 constituencies x 4 polling stations x 20 respondents), Lokniti successfully interviewed 2,333 respondents (response rate of 73 percent). Survey enumerators, trained by Lokniti and residents of the state, conducted face-to-face interviews at the respondent’s place of residence in the respondent’s native tongue.
The 2010 elections were held in six phases, with a unique set of constituencies going to the polls in each phase. Lokniti carried out its survey with the target sample within a few days of the end of polling in each phase. In order to obtain information on respondents’ vote choice, Lokniti begins its surveys by simulating a mock election. For respondents who affirmed that they were able to vote in the election, enumerators then produced a dummy ballot box and asked voters to cast their vote by marking the symbol of the party they voted for using a dummy ballot paper slip and to deposit the ballot in the ballot box. The dummy ballots are tagged with a unique identifier, which allows for easy merging with the remainder of the respondent’s information.
More information about the survey can be found on the author’s website, http://www.milanvaishnav.com.
Appendix C: Details of Lok Social Attitudes Survey
IN THE FOURTH quarter of 2013 (September–December), the Lok Foundation carried out a survey of the political attitudes of a broad cross section of Indians in advance of the May 2014 Indian general election. The survey piggybacked on top of a standing quarterly panel consumer survey of more than 150,000 households conducted by the Centre for Monitoring Indian Economy (CMIE).1 This survey was carried out in collaboration with a team consisting of Devesh Kapur, Megan Reed, and Neelanjan Sircar, based at the Center for the Advanced Study of India (CASI) at the University of Pennsylvania. Collectively, the team assisted Lok and CMIE in the design of the survey instrument on political attitudes.
The survey solicited responses from 68,516 respondents across 24 states and union territories. The sample was drawn as a cluster random sample in the following way. The 24 states and union territories under study were broken into 98 “homogeneous regions.” A homogeneous region is a set of contiguous districts (the largest administrative unit inside a state) that satisfy certain similarities in agro-climatic conditions. Within each homogeneous region at least one city (and often multiple cities) were selected. A set of villages was chosen randomly in each homogeneous region. Selecting a random position in the village and selecting every nth household in a randomly chosen direction was used to select households. Cities were broken into wards, which were further subdivided into census enumeration blocks (CEBs). The wards in the city were stratified by average asset wealth (as determined by the 2001 Indian census) and selected randomly. The CEBs (which are of roughly equal population) were selected randomly from each ward. Within each CEB, a randomly selected position and every nth household from a randomly selected direction was used to select households. Within each household an individual over the age of 18 was randomly selected.
Although within each locality respondents were randomly selected, the sampling frame itself is not representative of India. The sample includes data from each city that had at least 200,000 inhabitants as of the 2001 census of India as well as a smaller rural sample. As such, the data are skewed toward larger cities in India and the data are biased toward urban respondents. India is 32 percent urban and 68 percent rural according to the 2011 Indian census; using the same classifications, the Lok sample is 62 percent urban and 38 percent rural.2 For purposes of drawing representative inferences, estimates were reweighted by urban and rural population within each state or union territory.
More information about the survey can be found at: https://casi.sas.upenn.edu/lok-survey-social-attitudes-and-electoral-politics/lok-survey-social-attitudes-and-electoral.
Notes
CHAPTER 1. LAWMAKERS AND LAWBREAKERS
1. While five of the six jailed lawmakers were expected to support the government, the sixth (Umakant Yadav) was expect
ed to vote with the opposition.
2. “A Penchant for Guns, Horses and Cars,” Tehelka, October 2, 2004.
3. Ibid.
4. According to the affidavit submitted by Ateeq Ahmed to the Election Commission of India, he faced 36 pending criminal cases at the time of his election in 2004.
5. Edward Luce, In Spite of the Gods: The Rise of Modern India (New York: Knopf, 2010), 134.
6. Ibid.
7. “A Penchant for Guns, Horses and Cars,” Tehelka.
8. Venkitesh Ramakrishnan, “Canker of Criminalisation,” Frontline, February 12–25, 2005.
9. “Former SP MLA in Police Custody for Murder of Raju Pal,” Indian Express, May 12, 2011.
10. Pappu Yadav declared 27 pending criminal cases in the affidavit he submitted to the Election Commission of India in advance of the 2004 Lok Sabha election.
11. Sankarshan Thakur, Subaltern Saheb: Bihar and the Making of Laloo Yadav (New Delhi: Picador, 2006), 153.
12. Ajit Sahi, “The Gangster’s Last Gamble,” Tehelka, May 23, 2009.
13. Ibid.
14. Sankarshan Thakur describes Pappu Yadav as “part gangster, part contractor, part caste-lord, part political profiteer.” See Thakur, Subaltern Saheb, 153.
15. Nalin Verma, “Importance of Pappu, the Politician,” Telegraph (Calcutta), September 21, 2014.
16. Kumkum Chadha, “She Married a Don . . . a Criminal,” HT Just People (blog), March 24, 2009, http://blogs.hindustantimes.com/just-people/2009/03/24/just-people/ (accessed March 19, 2014).
When Crime Pays Page 38
