The election of a single Congress Member of Legislative Assembly in a district means it is 32% less likely that a riot will break out before the next election. This is the conclusion of a paper by political scientists at Yale who were examining the effects of incumbent politicians on ethnic violence in their area.

The paper, entitled “Do parties matter for ethnic violence? Evidence from India”, looked at elections in 315 administrative districts across the country between 1962 and 2000. Had Congress candidates lost all the close elections in their data set, India is likely to have witnessed 10% more Hindu-Muslim riots and had 46% more riot casualties over the 40 years that they had examined.

“The pacifying effect of Congress incumbency appears to be driven by local electoral considerations, in particular the party’s exceptionally strong linkages to Muslim voters during the period we investigate,” write Gareth Nellis, Michael Weaver and Steven Rosenzweig, all political scientists at Yale University. “Taken together, our findings point to a more important role for parties in developing democracies than existing scholarship tends to assume.”

Close elections analysis

In order to arrive at their conclusions, the authors of the paper used a close-elections regression discontinuity analysis to consider the likelihood of rioting and ethnic violence and its relationship with political partisanship. The numbers offer likelihood and probability, establishing a correlation between incumbent Congress MLAs and the reduction of ethnic violence, rather than suggesting that there is a causal connection.

The conventional wisdom in political science presumes that elections, particularly in the developing world, are less about the ideologies of parties than the personalities who head them. This should imply that, on average, it shouldn’t matter whether an incumbent Congress MLA is re-elected.

But the findings end up showing the opposite. Congress MLA candidates who manage to win elections lead to a 32% reduction in chances that the area would be hit by ethnic violence. Simply enough, the paper says, by electing Congress MLAs in close races between 1962 and 2000, the country has seen much less ethnic violence than it could have.

Because of the provocative statement that this is making, though, the authors lay out the potential caveats. For one, their analysis only looks at Hindu-Muslim conflict, rather than questions of religion, caste and economy. They also looked at the average impact of incumbent Congress MLAs and their effects on violence in the district, meaning anomalies are likely to have smoothed out by the averages. Finally, the study only looked at MLAs, although this is mostly appropriate because of their outsize influence.

In addition, because of the amount of time the Congress has spent in power, it is harder to examine what effects cause such riots, although statistics do point out that everytime India's grand old party loses an election violence increases. Despite these caveats, the authors still felt confident enough to say that electing a Congress MLA did reduce the chances of religious rioting, partly because minorities often voted very specifically with physical violence in mind.

“What emerges from our account, therefore, is that different parties facing the same party system and the same proportion of minority voters nevertheless confront divergent incentives for suppressing or fomenting riots,” the study said.

Congress effect on riots

Examining the effect further, the researchers looked into whether there is a change in the likelihood of riots depending on who the Congress candidate is up against or who might be in charge at the centre.

For the first case – whether the Congress effect changes if the candidate is up against an ethno-religious party like the Bharatiya Janata Party or the Shiv Sena or a “secular” party like the Samajwadi Party – the models constructed ended up with a similar result: the Congress MLA proving likely to reduce chance of ethnic conflict.

The authors acknowledge that this probably comes as as surprise to some analysts of Indian politics, who would hold the Congress responsible for much ethnic conflict in the country. It is important to note that the most prominent of the riots instigated by the Congress – the 1984 anti-Sikh pogrom in the aftermath of former Prime Minister Indira Gandhi’s assassination – aren’t counted in this study because they weren't a Hindu-Muslim faceoff.

One of the key findings is the flip view: whether Hindu-Muslim riots actually make it harder for the Congress to return to power, giving the party a reason to prevent conflict. “The outbreak of one additional riot in the year preceding a state assembly election depressed Congress's district vote share by 1.3 percentage points on average.”

“Taken in conjunction with the main result, our secondary finding that riots reduce subsequent Congress vote shares raises the possibility of a feedback loop or multiplier effect, whereby the outbreak of Hindu-Muslim violence causes Congress to lose votes and seats, which in turns leads to more riots, and so on in a vicious cycle,” the report concludes.

“Hence, this paper sheds new light on the puzzle of how democratic institutions have endured in India|the world's largest democracy|against challenging odds," the scholars write. "Democratic stability in divided societies depends not just on institutions or the nature of social cleavages, but on which parties citizens choose to vote into power.”