Thread by @arpitrage

Thread on the India chapter in *The Narrow Corridor* by Acemoglu and Robinson, which focuses on caste 1/n

amazon.com/dp/B07MCRLV2K/…

First, let me say I really liked the book. The interplay of Roman (state capacity) and German (participatory democracy) was interesting and I think shapes local outcomes today

I also emphasize that I am not defending caste or minimizing its human costs. I enjoyed these two Dalit autobiographies and there is a lot of other excellent work out there:

amazon.com/dp/B007QXVUSU/…

amazon.com/Untouchables-F…

The key argument in the book is that caste is entrenched in India; and its prevalence explains poor political participation, low state capacity, and ultimately bad economic outcomes.

I think this is a little simplistic.

Let me start by highlighting some evidence, not in the book, which supports the point. @suannaoh's job market paper highlights task stickiness by caste. Hseih et al. document how less discrimination in the US -> better outcomes.

nber.org/papers/w18693.…

However, see this paper by Cassan, Keniston, Kleineberg. Argument: though caste has large micro effects; it doesn’t affect aggregate productivity that much (one reason: caste restrictions between different kinds of low-skilled labor don’t have big macro consequences)

Another natural comparison is with Pakistan and Bangladesh; where caste like systems are much weaker that in India. Yet outcomes aren't much better, as Tyler Cowen mentions, and Abhijit Banerjee suggests caste may be in fact be a countervailing force:
conversationswithtyler.com/episodes/abhij…

You can also compare within India. In 1890, Swami Vivekananda found Kerala to be a “madhouse of castes” and it likely was the most caste stratified region in India.

Yet now it is hailed as a development miracle, and is quite progressive:

outlookindia.com/magazine/story…

We can also look at Bihar, Acemoglu and Robinson's poster child for lots of caste and bad governance with Lalu Prasad Yadav as the key villain:

Yet Bihar has since racked up some of the highest growth rates in India! Under a chief Minister, Nitish Kumar, who represents a coalition with a heavy contribution of so-called “lower castes”

finance.bih.nic.in/Reports/Econom…

So if you thought Bihar was going to be perpetually poor, you would have really gotten this wrong.

Part of what’s going on here is “all politics is identity politics.” Given that caste is an important part of identity in India, it is an important part of political mobilization

This doesn’t necessarily imply that Indian politics is, as a consequence, inherently tribal and unable to incorporate ideological parties. See Pradeep Chhibber and @rahul_tverma, who emphasize the ideological nature of Indian democracy

amazon.com/dp/B07GSDQQZX/…

So I think there are good reasons to doubt that caste is an inherent characteristic of Indian society solely responsible for India’s political and economic outcomes.

But I think there are deeper questions about *why* India has caste

Rather than seeing caste as just a weird thing that Indians do; it’s important to recognize kinship groups and marriage networks as important elements of all societies.

In Ancient Greece here, for instance, you see people identifying along their tribal ("demic") identity. Rome, too, organized political participation along tribal lines.

https://twitter.com/arpitrage/status/1175779135463055361

In many societies, growth in state capacity -> withering away of kin groups.

But where the state was weak, they flourished and acquired an occupation linkage. Think guilds; caste in W Africa; caste among the Yazidi; and in Madagascar. Shatter zones!

https://twitter.com/arpitrage/status/1167461111161077761

Caste solves the problem of how to organize complex economic production and division of labor absent state authority. Why is caste so important in India? Because unlike other regions, India has frequently been in conditions of political anarchy in its history.

Acemoglu/Robinson provide evidence themselves that points to this; in highlighting the jajmani system linking castes within an interdependent division of labor.

Rather than seeing this as pure deadweight loss: put this in context of thousands of years of Indian history featuring constantly shifting political overlords, and a complex and sophisticated village economy which manages to enforce contracts, etc. without state authority

*across* castes the hierarchy enable trade

*within* caste kinship encourages greater insurance and self-supporting networks

https://twitter.com/anup_malani/status/1167470507496787969

I think looking at it this way helps shed light on the persistence of caste in India.

With markets and the state still shallow; people rely on their caste networks for housing, jobs, insurance, etc.

So more public goods -> less caste (prediction)

PS the book equates Jats with Shudras with OBCs (bureaucratic distinction for affirmative action).

This is a decision up to each state; in Haryana (highlighted in book) Jats are not OBCs

PPS the book emphasizes continuity in the village panchyat representation, and implies this is just dominated by upper castes (typically Brahmins)

But the prevalence of these local panchayats is in question. Luis Dumont emphasizes we have better evidence for *caste* panchayats regulating internal affairs. The "dominant" caste panchayat had more influence, but not always Brahmins - see attitudes of Rajputs in Rajasthan

cc @shrikanth_krish @anup_malani @nebuer42 @omarali50 @razibkhan @pseudoerasmus @teasri