Economists, or more accurately, positional voting theories inspired by economic theory, offer little optimism about the ability of voters to sanction political parties that engage in pre-election violence. According to an influential theoretical study, pre-election violence credibly signals who the ‘strong’ party in the fray is—the one not averse to using violence in challenging election results against it. And hence, the median voter, in order to avoid the costs of post-election violence, will vote for this party, even if it has immoderate stances on important policy issues.
An empirical evaluation of this theoretical framework, based on a survey experiment in Kenya, offers a more nuanced view. Noting that voters in these settings may be normatively against violence, the study contends that they will nonetheless vote for an incumbent associated with pre-election violence if it has a strong track record of pursuing poverty-alleviation programmes.
In a recently published study, I evaluated these claims in the empirical context of the 2019 Lok Sabha election in the Junglemahal region of West Bengal. Given how deeply partisan violence is embedded in India’s electoral politics, it is surprising how little the Indian experience has figured in the global comparative research on the electoral legacies of political party violence.
Also Read: Junglemahal book on peace process with Maoists frightens the Indian state
The case of Junglemahal
On the face of it, the 2019 election in Junglemahal would seem to be an ‘easy case’ for positional voting models to explain. On the one hand, the 2018 Gram Panchayat election, which preceded the Lok Sabha election, was widely reported to have been among the most violent in West Bengal’s history. The incumbent Trinamool Congress party’s cadres were accused of unprecedented use of violence and terror tactics against potential Bharatiya Janata Party voters and even non-partisans. Incidentally, two of the most publicised acts of violence—the killing of two BJP workers, 18-year-old Trilochan Mahato and 32-year-old Dulal Kumar, both of whom were hanged—occurred in Purulia district in Junglemahal.
On the other hand, the TMC had also channelled substantial public resources into Junglemahal as part of its long-term strategy of post-conflict reconstruction in the region, a hotbed of Maoist violence from 2005 to 2011.
So successful was this strategy deemed to be in counter-insurgency circles that Jairam Ramesh, then Union rural development minister, lauded Chief Minister Mamata Banerjee for her ‘model’ handling of the Maoist problem. It would seem to follow axiomatically that Junglemahal would vote overwhelmingly for the TMC in the 2019 election. Instead, exactly the opposite happened. The BJP won in all but one of the parliamentary constituencies in the region. Why?
Instead of the positional voting model, I turn to ‘valence’ theories of voting to answer this question. Originally developed by Donald Stokes, this theory argues that under conditions of incomplete and imperfect information, which characterise most electoral environments, election outcomes are determined by what a majority of voters regard as the key issues for that particular election. Once an issue assumes this frame, parties are judged on the basis of their ‘competence’ in the domain. In the case of Junglemahal, security had become a valence issue in the 2019 election, and the TMC had ceded ‘ownership’ of that issue to the BJP.
Focusing on security as a valence issue is, however, insufficient to explain the decision to punish the incumbent, especially in weakly institutionalised polities where voting against an incumbent who fails to provide security, even in the presence of a viable opponent, is fraught with risks. The scales will ultimately tip against the incumbent in this setting only if the opposition party’s election campaign is able to invoke high-engagement negative emotions like fear and anger among voters that orient them towards information-seeking and a critical attitude rather than dispiriting emotions like sadness, which will simply reduce turnout.
Also Read: Violence isn’t new in BJP states. But Lakhimpur Kheri and Darrang mark a new turn
Fear aids concerns about post-poll violence
These observations lead to two testable hypotheses. First, voters who perceive a high level of threat from politically motivated violence will be more likely to vote against the incumbent party than voters who perceive a low level of threat. Second, the effect of perceptions of politically motivated violence on voting will be mediated by feelings of fear and anger.
The hypotheses were tested using data from an original survey that was conducted in February-March 2019 in Junglemahal’s 38 Maoist-affected villages and covered 1,140 respondents. Analysis of the data yields three main findings.
First, Junglemahal voters were able to discriminate between security and non-security issues. Their responses to a question about the importance of these issues for the upcoming election clearly separated into two clusters. They were more inclined to vote for the TMC if the provision of public goods received more importance. However, they swayed more towards the BJP when the issue was about politically motivated violence.
Second, the effect of the perceived threat from politically motivated violence was mediated by feelings of fear, anger, and worry about the general direction in which the state was going.
Finally, there was evidence to show a positive relationship between the level of campaign contact by the BJP and the level of negative emotions experienced by voters.
These statistical findings are also supported by field reports and case studies of the 2019 election. For example, one study notes how Union Home Minister Amit Shah launched the BJP’s 2019 Lok Sabha campaign in West Bengal with a trip to Purulia on 29 June 2018, which included, among other events, meetings with the families of the two BJP workers killed in the panchayat election violence.
Also Read: New study finds how Kerala-Bengal decided who to vote for between Centre and state in 2019
Lessons for parties that engage in violence
There are several important conclusions that flow from these findings.
First, voters did not spare an incumbent presiding over deteriorating security, that too, in the context of the least developed region within one of India’s less economically dynamic states. This contradicts the view that poor voters don’t punish candidates associated with pre-poll violence.
Second, election campaigns play a crucial role in inducing the specific emotional dispositions that enable voters to overcome the competing and powerful motivation to vote ‘insincerely,’ a factor that has not been studied adequately in extant research.
The immediate political implications of these findings are straightforward. There is little doubt that the BJP’s West Bengal unit is currently in a state of disarray. Even though the TMC was able to keep the focus on its social protection policies in the 2021 assembly election, which it won handsomely, violence by its cadres has been a recurring issue in the state’s political discourse since then.
Surely, the TMC leadership will know that voter behaviour in contemporary India diverges sharply between the assembly and Lok Sabha elections. Hence, it can be safely predicted that come 2024, the BJP will offer a more united front, and if the TMC fails to reign in partisan violence, it will almost certainly pay a heavy electoral cost.
Subhasish Ray is a Professor and Associate Dean at the Jindal School of Government & Public Policy, O.P. Jindal Global University, Haryana; and an editor for the Journal of Genocide Research. He tweets @subhasish_ray75.
(Edited by Srinjoy Dey)