Ethnic Conflict and Civic Life : Hindus and Muslims in India

I like Varshney’s book. Unlike earlier books (with the exception of Golden and Gould), he writes very clearly and coherently about his larger theoretical framework and connects his empirical analysis quite carefully to the theoretical framework.
What’s more, the basic thesis makes a lot of sense. He even makes some effort to draw out the causal mechanisms through which cross-sectarian civil society associations manage to defuse growing tensions short of violence or manage to disrupt escalating cycles of violence. In his anecdote about media coverage of his research in Aligarh and Calicut and the differences in terms of the level of paranoia in the press, he also hints at how a virtuous civil society environment can actually prevent an atmosphere of tension and potential violence from arising – although a more detailed analysis of the role of the media while riots or potential riot situations were evolving would have been very nice to see.
I think he does well initially with his argument that between-city variations in basic socioeconomic indices like wealth or income and literacy doesn’t explain anything about variations in ethnic violence. It is disconcerting that there is not a single word of mention of the 1984 anti-Sikh pogrom carried out in Delhi by criminal elements of the Congress Party in revenge for the assassination of Indira Gandhi in which 3000 people were killed. Although his focus is on Hindu-Muslim violence, this is so much larger than all the other episodes of communal violence in Delhi put together that it seems it should be taken into account in some way.
Methodologically, his attempt to bracket out questions regarding the form and operational style of the state by looking at within-country comparisons makes a lot of sense – although, given the lack of centralization of the Indian state, two out of his three pairs (Aligarh and Calicut and Lucknow and Hyderabad) actually involve significantly different modes of governance. Only the pair of Surat and Ahmedabad controls for differences in provincial as well as national government and it’s no accident, I think, that his connection of the empirical with the theoretical is weakest in that case.
In particular, when he is faced with the challenge of Surat’s having actually evolved communal violence during the 1992-93 riots, he is rather unconvincing. After having invoked the decline in Congress Party organization as Indira Gandhi eviscerated intraparty democracy as a major causative factor in the rise of ethnic violence in Gujarat, he just tosses out a claim that this decline was much less in Surat than in Ahmedabad. It’s very hard to swallow this as an explanation simply because the collapse in party organization (and often replacement by thuggery as a mode of spreading political influence) was so total on a national scale that it’s hard to understand why Surat, one of the cities with the worst municipal government in the country (at least as far as the well-off provinces like Gujarat, Punjab, and Maharashtra are concerned) would somehow buck this national trend to a degree that it would be significant in preventing communal violence.
Varshney has done a lot of work to put flesh on Putnam’s rather casually developed, throwaway notion of bridging capital (while at the same time carefully emphasizing the ethnic/sectarian dark side of bonding capital), both through a useful study of correlations and through a specification of detailed causal mechanisms.
Still, the relentless focus on civil society and on “scientifically testing a hypothesis” lead to potentially very significant omissions.
First, although Varshney has in general successfully bracketed out the role of the state through his argument that it generally responds to the imperatives of civil society and only acts in accordance with the law if political interests align with human rights and law and order, the argument is still open to specific counterexamples. In particular, it is absolutely shocking that, in his lengthy discussion of Calicut, he says almost nothing about the terribly important fact that Kerala’s provincial government has more often been not been formed by the CPM (Communist Party-Marxist); West Bengal is the only other state where this has been the case, and the CPM quite clearly played a more transformative and “left” role in Kerala than in West Bengal.
There is a throwaway paragraph on p. 152 where Varshney casually acknowledges this, but it doesn’t color his analysis at all. He focuses on the evolution of “reading rooms” before independence and on something called “civil society” afterward without anything about the tremendous role of the CPM, using state resources, in building the very civil society organizations that he is talking about. He mentions at some point that Calicut, unlike Aligarh (or almost any Indian cities), actually has a union of “head-load” porters. The only reason something seemingly so improbable by standard economic analysis is possible is the Communist government in Kerala. The Science for the People movement, which is one of the largest civil society organs in Kerala, is wholly a creation of the CPM-run state government. One could go on.
Second, in his two best-argued cases regarding lack of sectarian violence, he brings up alternative explanations, admits their strength, and then completely ignores them. In Calicut, Hindu-Muslim violence is dampened because the true conflict is Hindu upper-caste vs. lower-caste; in Lucknow, the true conflict is Shia vs. Sunni.
It is entirely possible to correctly identify civil society as a causal factor in depressing ethnic violence and still to completely miss the picture. This is because this sort of analysis is very “microcausal;” it may well be that the proximate cause of defusing tensions and disrupting cycles of violence is the work of civil society organizations, but that may simply be a convenient mechanism for the playing out of some larger imperative. In Kerala, it may be that the provincial government is committed to anti-communalism; it’s well known among the political cognoscenti (and many others) in India that it is only the left parties that are resolutely anti-communal. That commitment may manifest in creation of “bridging” civil society groups, either directly at the behest of the state, or through the provision of space by the state, which then act to disrupt potential sectarian problems.
Or it may be that in Calicut caste tensions drive the creation of Hindu-Muslim civil society organizations and in Lucknow Muslim sectarian tensions do. And then those organizations act in the way Varshney has talked about.
In either case, though, if it did play out that way, a focus on civil society would be missing the forest for the trees and missing the deep causation for the proximate causation.
In the case of his remarkable ignoring of facts that everybody knows about Kerala’s Communist government, I have to suggest that the fact that Varshney is in political science, writes in Foreign Affairs (he has a fairly insipid piece about the tensions between democracy and economic growth in India in the latest issue), is a professor at Princeton (where they actually think Woodrow Wilson was a principled friend of democracy and self-determination) must have something to do with it. I think there is an obvious political bias as well as the intellectual bias created by the dominance of naïve Toquevillian notions of civil society within political science.