Dweep you have done an interesting analysis of this issue on several accounts. However, it seems that one basic fact is missing in most of your analysis. “Investment in primary education in India has one problem that the present legal restrictions force companies providing education to invest back the profits and utilize the funds for their stated objective of providing education. Trusts, societies and charitable companies are the most-preferred legal structure for educational institutions and promoters are barred from taking profits out of such institutions.” (http://prayatna.typepad.com/education/2008/07/primary-educati.html)

Creative capitalism works at its best when the fields are leveled and the investor has control over his profits and methods of re-investing (This perception may change with the current financial crisis but that may take time as “the show is not yet over”)

This can be one of the reasons for lack of innovation in private schools and hence the quality. Who cares about quality when whatever I gain have to be re-invested here itself ?

I am not sure if that is the case with mushrooming coaching centers around the country. Even though these centers are not judged by the “formal system” yet if you look from different perspective they are the churners of students being admitted to elite Universities of India like your alma mater. The ratio of students admitted by the students coached by them is no small number. Assuming this admission to be a criteria of quality, I feel given a free hand these private houses (for competition, more profits or whatever reason) can give quality output too.

I think then only the public-vs-private comparison will be fair.

Oh yes. It is because this is not easy that I asked for better analysis. Hope some academic takes a shot at it, it’s badly needed.

Incidentally, the point about selection bias also applies with regard to our much touted IITs. I would think that a substantial part – how much I don’t know – of the reason why IIT graduates do so well is that the IIT input is very selective in the first place. So comparing IITs with other engineering institutes is not straightforward either.

However, even if private schools are superior in terms of measurable output – test scores and the like – there is also a large “unmeasurable” output – externality in the economist’s jargon – which is very important and would tend to support the case for public schools. This externality comes in the form of benefits to society from having children of all sections (all castes, religions etc.) learn in a common environment.

Some may argue that this is not important. However, if one looks at the bitterness underlying the current quota debate – where each side seems to have the worst possible image of the other side – I can’t help but think that may be, some of that bitterness could have been reduced if only there had been more interaction at the school stage. In our current, more or less private education system, most times an urban “upper caste/class” student gets to meet someone from a rural background is when he/she goes to college by which time attitudes have hardened substantially.

I am not suggesting that a public school system will miraculously yield solutions to the caste-religion problem any more than “busing” has solved the American race problems. But the benefits should not be understated, either.

Just some random thoughts; apologies for the long comment.

You could argue that since the comparison is with private schools, there should not much difference in the intake – in both cases, we would expect students to have better parental support etc. True, but there may be other things going on. In any case, the difference in pass percentages is not much – some 95.6% for KVs with 91.8% for private schools. Note that the government schools have a pass percentage of 70% or so.

The pass percentage itself may be hiding a lot. Between the pass mark of 40% and the maximum of 100% there is a big range. It might be interesting to look at other figures – median, mode, standard deviation and other measures of dispersion to get a better sense of the results.

Secondly, there is some dispersion with regards to the quality of KVs as pointed by many commentators, some from the KV system itself. They range from the bad to the very good. The general suggestion has been that there is a correlation between the percentage of “upper class” students and the quality of the school. A proper analysis of the functioning of KVs is yet to be done.

I am not arguing for privatization or anything – just for better analysis from all concerned. In any case, from what I can infer, the KVs, when they do function well, do so for particular reasons which are not really capable of replication.

Second, as you point out, schools with above average students will have good results. Have you factored in that private schools already have this selection bias? In fact, and ironically, a comparison between JNVs and private schools is better than one between private and government schools!

“Among the different categories of schools, the government-run Jawahar Navodaya Vidyalayas topped the examinations with a 96.41 pass percentage followed by Kendriya Vidyalayas (Central Schools) with 95.64 percent.

While 92.33 percent of students at Tibetan Schools tasted success, private independent schools scored a pass percentage of 91.81 and the government schools were at the bottom of the pool with a 70.36 pass percentage.

From this, your evidence, you assert that Government schools are better than private schools. However, as you must know, Jawahar Navodaya Vidyalaya’s are Indian schools for talented children. They form a part of the system of gifted education. The objectives of the scheme are to provide good quality modern education to the talented children predominantly from the rural areas, without regard to their family’s socio-economic condition. So we are hardly comparing like with like!

In a large sample of government schools we would expect some in the tail of the distribution to outperform the average of the private schools (anything can be proved with a single sample) and we would expect selected gifted children in Government schools to outperform private schools whose intake took a cross section of the local children.

The website you refer me to states the government schools were at the bottom of the pool with a 70.36 pass percentage. So the issue is not that a few government schools that cherry pick pupils do well but why do the vast majority of government schools fail. More to the point, if the government initially intervened in education to help the poor, then why are the majority of failing government schools located within low income areas?

Incidentally I would prefer my children to go to a school which was good academically rather than one which had good toilets.