Literate sex ratio, 2011 census of India

Poking around the 2011 provisional data from the Indian census I noticed that literacy rates had gone up across the board. But, in many states there was the long standing disjunction between male and female literacy rates. So I decided to create some plots, what else?

First, below you see a simple plot of female vs. male literacy by Indian administrative unit. There’s a strong linear relationship, where variation in male literacy explains ~75% of the variation in female literacy. But again I’m a little wary of unweighted plots like this, since urban areas are not representative of the Indian society as a whole. So I also generated a bubble plot in R, where the size of the bubbles is proportional to the total population surveyed.lit1.png

lit3.png

It seems that the lower the average literacy the bigger the gap between males and females. Some of this is due to the fact that above you can’t get above 100%, so there’s going to be natural convergence at the tails of a distribution. So:

1) I converted the literacy proportions to log values. This rescales it somewhat so that extreme values don’t converge as much.

2) Then took the ratios of the log values. Concretely, I did the following for Kerala, for example:

(log male literacy)/(log female literacy) = ratio

Then I plotted the ratios against the log value of female literacy. You can see the results below:

lit2.png

Meghalaya is the outlier. It has an OK literacy rate, but relative balance between male and female values. Why? Those of you ethnographically aware will not be surprised:

One of the unique features of the State is that a majority of the tribal population in Meghalaya follows a matrilineal system where lineage and inheritance are traced through women. The Khasi and Jaintia tribesmen follow the traditional matrilineal norm, wherein the “Kha Khadduh” (or the youngest daughter) inherits all the property and acts as the caretaker of aged parents and any unmarried siblings. However, the male line, particularly the mother’s brother, may indirectly control the ancestral property since he may be involved in important decisions relating to property including its sale and disposal. In the Garo lineage system, the youngest daughter inherits the family property by default; unless another daughter is so named by the parents. She then becomes designated as ‘nokna’ meaning ‘for the house or home’.In case there are no daughters, then a chosen daughter-in -law (bohari) or an adopted child (deragata)comes to stay in the house as well as inherits the property. The tribal people of Meghalaya are therefore a part of what may be the world’s largest surviving matrilineal culture.

The last part is probably wrong. I think that the Minangkabau are the most numerous people who still engage in matrilineal practices, though readers can illuminate as to how pervasive the old customs are in Kerala and neighboring regions. In any case, the point is that culture matters, and that economics alone does not explain everything. Sometimes there is gold to be found in the residuals….

14 thoughts on “Literate sex ratio, 2011 census of India

  1. Very interesting data presentation. I especially liked the “bubble plot.” Keral wins hands down on literacy, but not on economy, right? What is the correlation of literacy to per capita gnp? Of course, one could argue that literacy is more important than per capita gnp.

  2. Indeed, Kerala may provide evidence that literacy is more the effect of economic development than its cause. Very interesting. See Randal Holme, Literacy: An Introduction, p. 27 for this theory that economic growth increases demand for literacy, but literacy itself doesn’t produce growth.

  3. Hello folks,

    SM actually has a hyperlink on the right side saying that the literacy rate mentioned is the “functional literacy rate”, which pretty much means the ability to sign one’s name. You can read about it here: http://www.dnaindia.com/opinion/main-article_the-unemployable-literates-of-india_1532691

    I would be very proud and impressed if the literacy rate truly increased to ~74%.

    So one of the best things about the gender imbalance is that the higher male population contributes to more literacy, since men are ~25% more likely to be literate than women. Just kidding, folks.

    • “SM actually has a hyperlink on the right side saying that the literacy rate mentioned is the “functional literacy rate”, which pretty much means the ability to sign one’s name.”

      True. they use the term literacy rather loosely. But we can take heart that something is improving. Hopefully the “true” literacy rate grew at approximately the same rate from a lower base.

  4. Thanks for the pointer, Razib. Fairly strong correlation–the interesting question is which way causation may run!

  5. Great article.

    I am a long time reader of sepia mutiny. I wanted to share some info about my new website zeole.com , with your and your readers.

    zeole.com combines blogging and location. In zeole.com writers always chose a location before they write. Because of this, content is high relevant. Writers get to reach the right audience directly. Check it out and let me know if you have any feedback.

  6. It’s a very interesting post. As another observation (perhaps validating the point about matrilineal system explanation of the graph as you offered) is the other outlier, the Rajasthan state. See how that is an outlier as well. Would be very interested to know of your explanation. Looks like this state validates your point about the states in the lower right quadrant of this graph.

    Brilliant data crunching, gotta say.

    Arin Basu New Zealand

  7. It’s a very interesting post. As another observation (perhaps validating the point about matrilineal system explanation of the graph as you offered) is the other outlier, the Rajasthan state. See how that is an outlier as well. Would be very interested to know of your explanation. Looks like this state validates your point about the states in the lower right quadrant of this graph.

    yeah, i highlighted it. i have no explanation since i don’t know the ethnography well. was hoping readers would illuminate.

  8. Razib, can you write something on hoe versus plough? I read the link you posted awhile back, but a lot of it went over my head. A Sepia summary would be nice.

  9. Is it possible analyze this data in terms of age demographics to drill down to the current generation?

  10. This is a brilliant post, could I reference your page in a paper I’m writing for an international development journal? Please contact me (I included my address).