Cold stare of hard data

{Please click on the images to view a larger version.}

This is the data available from the NSSO Report 1999 - 2000 for education levels across social groups. This is the latest data available for educational attainment, although more recent figures are available as far as the OBC count is concerned. The first two segments (urban and rural) are quoted verbatim from the report. I calculated the total figures, by using a population weighted average.

Let's recap the NCBC education criteria for qualifying as OBC:

1. Castes and communities, whose literacy rate is at least 8% less than the State or district average.
2. Castes and communities of which the proportion of matriculates is at least 20% less than the State or district average.
3. Castes and communities, of which the proportion of graduates is at least 20% less than the State or district average.

For the moment we will ignore the district and the community part. Too vague, besides no district level data is available anyway. We also take a conservative estimate and assume an OR condition for the above although NCBC does not elaborate. For Mandal, you had to qualify in all the above.

If you look at the image above, you will see that the OBCs actually lead (or just marginally lag) the state average in almost all counts, except in the numbers for graduates per thousand. Overall in TN there are 38 graduates per thousand OBCs as opposed to the state average of 42 per thousand which is about a 10% lag. This clearly indicates that some OBC castes will not qualify the NCBC criteria. Only in the urban areas, OBCs lag the state average by about 25% in terms of graduates per thousand. But that fact alone does not qualify all OBC castes for quotas, even by the conditions laid down by NCBC.

A summary of the above is as follows.

I did some further calculations, just to arrive at a ballpark estimate of how many OBC castes might actually qualify. Absent the exact numbers for each caste, this seems to be hard to do. However, you can easily assume a normal distribution for the graduates per thousand figure for different castes. This is valid under central limit theorem. We have the overall average, which is 38 per thousand. We also know that to qualify, a caste needs to have less than 33.6 graduates per thousand (42 state average, less 20%). We also know TN has about 181 distinct OBC groups. Thus, all you need to do is assume different standard deviations and determine how many castes might qualify.

Obviously, you can not really go beyond a standard deviation of about 13, but I computed it anyway. As you can see, at best 60 - 70 caste groups can qualify.

I did the same thing, this time for the urban numbers only. The relevant numbers are: 68 per thousand (OBC), 91 per thousand (state average) and 72.8 (91 less 20%) to qualify.

Here also there is no need to go beyond 23 for the StDev. If we just take urban figures, about 100 out of 181 caste groups just might qualify.

A caveat of this analysis is that much of the numbers include those who are currently availing quotas. In which case there is no way to find out unless we repeal the quotas for a few years, which I am sure is not going to happen.

Of course everyone knows that in TN the quotas are being cornered by a dozen or so rich and powerful OBCs. But the socialists will never admit to it nor will they let anyone conduct a proper statistical survey of all relevant information. The quotas will remain a slap on the face of the truly backward.