Profile picture
Paul Novosad @paulnovosad
, 19 tweets, 5 min read Read on Twitter
🚨 New working paper 🚨 on Intergenerational Mobility in India. Biggest takeaway in 1 graph: Upward mobility for Muslims way behind forward castes, SCs and STs, and going down. Thread, with @thesamasher and @charlesrafkin. 1/11
Into the details. We want to measure equality of opportunity. If you start at the bottom, how likely are you to stay at the bottom? Is it easier to rise up than in the past? 2/11…
We get at this with a measure of intergenerational mobility: if you are born to a parent in the bottom half of the education distribution, what is your expected education rank? Let’s call this upward mobility. 3/11
(Calculating this is really hard in developing countries when you don't have linked income data and education is super coarse — we have a methods paper dedicated just to this topic)… 4/11
Here's upward mobility from 1950-1989 birth cohorts. Your chances of moving up in ranks relative to your parents are basically unchanged since before liberalization. 5/11
In other words, while India’s recent growth has made almost everyone a lot better off, it hasn't changed the rate of churn at all. If you started at the bottom, you're just as likely to finish at the bottom as you would have been in 1950. 6/11
(This said, in absolute terms, being at the bottom is a lot less bad than it used to be) 6.5/11
Let's see how this looks across some of the groups of people in India who have been marginalized historically: Scheduled Castes, Scheduled Tribes, and Muslims. 7/11
Here is that graph again.
1. Substantial upward mobility gains for Scheduled Castes and Tribes.
2. Substantial upward mobility *losses* for Muslims.
3. Basically no change for the combined Forward Castes and OBCs.
Muslims are now the least upwardly mobile group in India. Conditional on being born in the bottom half of the education distribution, Muslim children can expect to obtain the 28th percentile. This is really low! 9/11
For comparison: U.S. African Americans born in the bottom half of the education distribution on average get to the 34th percentile.
SCs and STs have similar mobility to U.S. African Americans, and Muslims are now quite a bit below. 10/11
Many more details in paper, incl. high-res geographic patterns of mobility. Here's a map -- the bright specks are towns/cities.
That's it; the rest of this thread (the thread appendix?) covers a few clarifications. 11/11
Thread Appendix 1: This is all about education ranks. Ed is a great proxy for income, and given available data, it's basically impossible to effectively measure changes in intergenerational income mobility. More details in the paper. 12/11
In consumption terms, our read on the NSS consumption data suggests that the Muslim, SC and ST gaps are constant from 1980 to the present. But we think ed is a better measure of opportunity going forward. 13/11 (I'm using latex numbering conventions)
Thread Appendix 2: X-group results are similar if you look at absolute outcomes, like ed levels, rather than zero sum ed ranks. Here's primary and high school completion, given parents in the bottom 50%: 14/11
As above, Muslims are falling behind substantially; in fact, in the bottom half of the distribution, Muslim high school completion rates are flat from 1975 to the most recent cohort. 15/11
Thread Appendix 3: What about OBCs? We’re working on it -- for now, we put them with Forwards / Others. OBC outcomes are harder to track over time b/c of classification changes over time; also, few/none identified in data as OBC in the 1950s/60s. 16/11
You made it to the end! Comments/questions very welcome. 17/11
Appendix 4: I should have mentioned all the data are from 2011/12, the most recent that such data are available. So this is not directly informative about the performance of the current party in office.
Missing some Tweet in this thread?
You can try to force a refresh.

Like this thread? Get email updates or save it to PDF!

Subscribe to Paul Novosad
Profile picture

Get real-time email alerts when new unrolls are available from this author!

This content may be removed anytime!

Twitter may remove this content at anytime, convert it as a PDF, save and print for later use!

Try unrolling a thread yourself!

how to unroll video

1) Follow Thread Reader App on Twitter so you can easily mention us!

2) Go to a Twitter thread (series of Tweets by the same owner) and mention us with a keyword "unroll" @threadreaderapp unroll

You can practice here first or read more on our help page!

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just three indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member and get exclusive features!

Premium member ($30.00/year)

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!