Why is India so low in the Pisa rankings?

That is a request from J. and here is one recent story, with much more at the link:

A global study of learning standards in 74 countries has ranked India all but at the bottom, sounding a wake-up call for the country’s education system. China came out on top.

On this question, you can read a short Steve Sailer post, with comments attached.  Here are my (contrasting) observations:

1. A big chunk of India is still at the margin where malnutrition and malaria and other negatives matter for IQ.  Indian poverty is the most brutal I have seen, anywhere, including my two trips to sub-Saharan Africa or in my five trips to Haiti.  I don’t know if Pisa is testing those particular individuals, but it still doesn’t bode well for the broader distribution, if only through parental effects.

2. Significant swathes of Indian culture do not do a good job educating women or protecting their rights, even relative to some other very poor countries.  On educational tests the female students are at a marked disadvantage and that will drag down the average.

3. Countries taking the test for the first time may face a disadvantage in manipulating the results to their advantage; admittedly this cannot account for most of the poor performance.

4. Indian agricultural productivity is abysmal, in large part due to legal restrictions.  I discuss this in more detail in my next book An Economist Gets Lunch, due out in April.  That hurts the quality of life and opportunities for hundreds of millions of Indians, including of course children.

Overall, India has a lot of low-hanging fruit, but the country has further to go than many observers realize.  A quadrupling of per capita income would put them at what, the level of Thailand?


To wit:

Corruption's a big reason too. It's very common for unqualified people to pay a bribe to get teaching jobs, and others take a cut of their salary to fudge their attendance record. Varies quite a bit by state, and private schools are better, but it's still a significant factor.

A union-run dismal government school system and low agricultural productivity due to legal restrictions. India sounds like a perfect lesson for left-wing clowns who read your blog.

When I was growing up in India (1989-2007) in a middle class family, we employed about a dozen house maids over the years (and employing house maids was very common in India, even in lower middle class families). Virtually all of those maids sent their kids to *private schools*. They made like 500-1000 Rupees a month (a U.S. Dollar was like 35-40 Rupees during the 90s). Their only choice was to send their kids to private schools because the public school teachers won't even show up for work. The quality of education in government-run schools and healthcare in government-run hospitals was often the target of widespread cultural jokes.

Thailand's GDP per capita is 2.5-2.7 times higher than India's (http://en.wikipedia.org/wiki/List_of_countries_by_GDP_(PPP)_per_capita). Quadrupling India's GDP per capita would put it at about the level of Mexico or Turkey.

Thanks for the links. I would add this comment from my earlier post on the subject:

"India ought to be able to do better than score at sub-Saharan levels. Indians in other countries do better. For example, this same report has Mauritius, a mixed race country in the middle of the Indian Ocean where 52% of the population is Hindu, scoring like a Latin American country rather than a sub-Saharan African country.

"But, India itself has a long way to go. It's likely to take 1-2 generations to get India up to speed, and we don't really know what up to speed for India means yet."

http://www.outlookindia.com/article.aspx?278843 discusses the fact that India's development has been very uneven. Education is only one of the social indicators where they are not making the progress that might be expected from growth numbers.

By the way, since India v. China is obviously going to be one of the big stories of the 21st Century, how is the rest of China outside of Shanghai, which did outstandingly in 2009, doing on PISA? Tests were given in other provinces of China, but the results haven't been released yet. One hint came from a 2010 Financial Times article: "“Citing further, as-yet unpublished OECD research, Mr Schleicher said: “We have actually done Pisa in 12 of the provinces in China. Even in some of the very poor areas you get performance close to the OECD average.””

That statement is a little hard to parse reliably, but it sounds like other Chinese provinces did okay compared to European and North American countries, which would be impressive.

Allow me to add the caveat that an 20-80 rule applies to the accuracy of testing. It's not that hard to conduct tests that are roughly accurate, but it's quite hard to fine-tune the process to give accurate results to subtle questions, such as whether kids learned more between the 2006 PISA and the 2009 PISA or whether Chile is really doing better than Argentina, etc. A big question is how representative the samples of test-takers are compared to the entire population. (E.g., Argentina apparently made a stronger effort than most participants to round up 15-year-olds no longer in regular schools, so that may have depressed the results). Another big question is how motivated the test-takers are. PISA tests aren't supposed to have any consequences for the students, so lazier test-takers would be inclined to bubble in answers after they get bored or tired. Perhaps some countries do a more effective job than others of motivating test-takers? How much effect does this have on why, say, Finland leads Europe routinely in PISA scores (although not on TIMSS the one time Finland took it)?

In summary, PISA and the sometimes contrasting results on TIMSS show the same general broad patterns as seen in countless tests in the U.S.: Northeast Asians out-front on average, then Europeans, then Latin Americans. So the international testing results have some obvious prima facie validity, but it's hard to tell how much faith to put in their more detailed findings.

India still has major problems with nutrition, both in total calories consumed by the poor and in consumption of the micronutrients iodine and iron, which are needed to avoid IQ-sapping medical conditions such as cretinism. Fortifying salt with iodine and flour with iron became standard in the U.S. between WWI and WWII, which helped us a lot. Kiwanis International is the leading philanthropy in pushing iodine fortification in the third world, so if you are looking for a year-end philanthropy to give to, keep them in mind.

Further evidence to incriminate India's political system as a failure.

"Why is India so low in the Pisa rankings?"

Because of a plot by Pakistan, obviously.

Are you drunk or what? Few Indians attribute their poverty etc. to Pakistan.

It is just a mild joke. India has a tendency to attribute every terrorist attack on it to Pakistan, and refuses to take responsibility on their hands.

1. The main tests seem to have been in two states of India, Tamil Nadu and Himachal Pradesh. I don't know much about Himachal Pradesh, but Tamil Nadu is not one of those states that are that backward in womens' education. And indeed, the pisa guys' says that : "In Tamil Nadu-India, there was a statistically significant gender difference in scientific literacy, favouring girls.", so your argument doesn't apply. This difference doesn't seem to have been there in "mathematical literacy".

In fact, Tylers' point about girls not being educated - might well *raise* the averages, see point 2 below.

2. Is it possible that India has schools even in areas with very poor infrastructure; that, as a result, more students get education, but of a worse quality. May be, many other under-developed countries only start having many schools after developing better infrastructure. So ironically, this could be a result of free and compulsory education.

For instance, here is an article alleging that in the supposedly well developed state of Gujarat, school students are doing worse. However, I think it is because of Gujarat's success in enrolling girls and poor children in schools - such things lower the averages.

3. There are these things they about Indians having what is called a "chalta hai" attitude, i.e., high level of tolerance for mistakes. I have heard from some college teachers about how they have been forced to be too lenient while grading exams because the Government wants more students to pass or so.

India has a great education system...

...if you happen to be part of the top 0.1% (by income)

Well, I don't want to quibble, but - just focusing on India versus China - they're comparing Shanghai and Hong Kong - these are *Cities* - to Tamil Nadu and Himachal Pradesh which are *States*.

Do the same, if you want a fair comparison, for Delhi or Madras. I bet you all my (limited number of) dollars the story will be very different.

Well, these cities have their share of slums and bad schools. But I agree, most probably the results will be different. Also note that in Shanghai they probably forcefully evicted slum dwellers.

I may be digressing a bit here, but even for the richest, I don't think the schooling system in India is all that great. I grew up in India. I have had American high school kids either sitting in or crediting difficult undergrad honors classes I taught; many of them were excellent. The really good high schools of US are way better than the best schools of India. In India, everyone goes through the same syllabus, so there aren't as many options as in China or US to give the best students a syllabus that does justice to their potential. Compare with Iran - those who are among the nine students from whom six would be selected for their IMO team, are totally exempt from the last year of their high school, and can join any undergrad college of their choice.

True. But i feel that this point would not count for India where the illiteracy rate is quite high. The government's prime concern in India obviously is to improve the literacy rate rather than the quality of education. The Right for Education is a strong proof to validate my point. Added to this, easier syllabus encourages more number of students even from poor economic backgrounds to study. This would result in better employment and an improved standard of living. This is a continual progress. I am sure that in a generation or two, when the countries' literary rate reaches the desired level, the quality of education would definitely improve.

Shanghai is a Chinese province. Hong Kong have a status that is higher then a province.

"In other words, only a little over one in six students in Tamil Nadu and nearly one in 10 students in Himachal Pradesh are performing at the OECD average."

"“If you look at the entire people entering the workforce, you may find lack of quality. But if you take the top 10% then they are perhaps the best in the world. This 10% is quite a large number which is giving India a competitive upper hand.”

Techically, I guess it is the top 15% on TN and the top 10% in HP. But that is....

That is 11M people in TN, and 600K in HP.

So, in other news, the creamy middle class in India is slowing gearing up to western standards, and the poor remain poor. Globalization works!

Maybe if they would let it work. Hard to argue with the results of the last decade world-wide.

A couple of poorer provinces took the TIMSS previously and also did poorly, although the top scorers didn't do too badly. In contrast, on the PISA, virtually nobody in the two tested Indian provinces scored at global elite levels. I don't know how to explain this contradiction between the tests. I would guess off the top of my head that the TIMSS results sounds more plausible (India's masses are poorly educated but elites are pretty well educated) than the PISA results (practically nobody in two moderate income provinces is at the level of the better Western college students), but I just don't know enough to say.

By the way, there's a similar conundrum involving Mexico's performance on the PISA tests. Overall, Mexico does not too bad by Latin American standards, trailing Chile but ahead of quite a few other countries. (Mexicans score worse on the PISA in Mexico than do Mexicans in the U.S., but, generally, American students do pretty well on the PISA once you adjust for ethnicity).

But practically nobody in PISA's sample of Mexican 15-year-olds in Mexico scores at international elite levels. Turkey, for example, has significantly more kids scoring a 6 on a 1-6 scale than does Mexico. I don't know if that says something about Mexico or whether it just says that PISA didn't get into elite schools in Mexico.

Here's the abstract of a paper by two academics who take seriously the idea that Mexico is badly missing in top scorers:

Producing superstars for the economic Mundial: The Mexican Predicament with quality of education
Lant Pritchett and Martina Viarengo
Harvard University, John F. Kennedy School of Government
November 19, 2008

Abstract. The question of how to build the capabilities to both initiate a resurgence of growth and facilitate Mexico’s transition into a broader set of growth enhancing industries and activities is pressing. In this regard it seems important to understand the quality of the skills of the labor force. Moreover, in increasingly knowledge based economies it is not just the skills of the typical worker than matter, but also the skills of the most highly skilled. While everyone is aware of the lagging performance of Mexico on internationally comparable examinations like the PISA, what has been less explored is the consequence of that for the absolute number of very highly skilled. We examine how many students Mexico produces per year above the “high international benchmark” of the PISA in mathematics. While the calculations are somewhat crude and only indicative, our estimates are that Mexico produces only between 3,500 and 6,000 students per year above the high international benchmark (of a cohort of roughly 2 million [which is about half America's cohort of around 4 million]). In spite of educational performance that is widely lamented within the USA, it produces a quarter of a million, Korea 125,000 and even India, who in general has much worse performance on average, produces over 100,000 high performance in math students per year. The issue is not about math per se, this is just an illustration and we feel similar findings would hold in other domains. The consequences of the dearth of globally competitive human capital are explored, with an emphasis on the rise of super star phenomena in labor markets (best documented in the USA). Finally, we explore the educational policies that one might consider to focus on the upper tail of performance, which are at odds with much of the “quality” focus of typical educational policies which are often remedial and focused on the lower, not upper tail of performance.

>>>American students do pretty well on the PISA once you adjust for ethnicity<<<

What exactly does that mean? How does one adjust for ethnicity?

White (non-Hispanic) Americans do okay in comparison to European whites, American blacks do okay in comparison to European blacks (or whatever), and American Hispanics do great in comparison to Mexico. Probably ditto for American Asians vs. Asian countries. But American blacks and Hispanics do so much worse than American whites and make up such a high percentage of the population relative to Europe/Asia that they drag the average down.

See the original post by Tino Sanandaji

"The amazing truth about PISA scores: USA beats Western Europe, ties with Asia."

Anyway, sorry to post so many comments, but I find international test comparisons interesting and important. An investment strategy of investing in countries where their relative test scores are higher than their relative per capita GDPs sounds plausible. But it also suggests that we ought to try to understand these tests more fully.

While non-educational factors like nutrition, poor healthcare, discrimination against girls etc. may be playing a role in the poor showing of India, the single most important factor is the completely skewed education system and process.

The focus is only on enrollment, attendance and completion of schooling and not at all on learning outcomes. The incentives for the policy establishment, administrators, teachers and even parents are not aligned towards learning by understanding.

The quality of teachers and teaching is poor, even in the private unaided schools in the bigger cities catering to the middle and upper classes. I say this from the personal experience of my own child's schooling in a major metro.

While the curriculum itself may not be bad, the focus is on completing the curriculum irrespective of whether children are learning or not. The evaluation and assessment model incentivises knowing how to answer questions on a set pattern. Children are used to knowing the kinds of questions that will appear on a test and knowing how to answer those specific questions and nothing else, leading to learning by rote rather than understanding. Parents are not demanding better learning outcomes either as long as they feel their children are able to crack the current assessment model.

Hopefully, the PISA results will spark a debate towards urgent reform to create the right incentives and accountability. If not, the demographic dividend is likely to become a demographic disaster.

Somebody needs to send all these scratchings to Tom Friedman and ask him to write a column on them (snark snark)

To compound problems, our media doesn't even cover stories like these! We, Indians have the attitude of an ostrich - hiding our heads in the sand instead of dealing with our problems. Our elites do not like to be confronted with our own ills. Its weird - its like a self-imposed propaganda that our elites indulge in. All you will see in our media is self-congratulatory articles about how Obama said Indian kids are doing well in science, or how some Indian-origin scientist in a western research organization did well (as if that means anything for India and its institutions). And, not just this, if one criticizes, they are ostracized as anti-Indians even though its constructive criticism born out of a desire to do better! Its really frustrating. See this for example:

Yes, Obama Administration speeches on American education typically contain ritual invocations of the outstanding education received by students in China and India, which can be misleading, to say the least.

The PISA study compares test performance in a couple of states of India, Tamil Nadu and Himachal Pradesh with elite and better developed countries: almost all the rest of the countries in the sample ( including Krygzsthan) have literacy rates close to 100%.
In comparison Tamil Nadu in 2011, has a literacy rate of 80%, Himachal Pradesh of 83%. Also in 1981, Himachal had a literacy rate of 51%, Tamil Nadu of 54%.
A large portion of the kids in school today in India are first generation literates: and first generation literates won't find it simple to do well at sophisticated tests.
Yes, problems of the Indian education system which is huge on budgets ( Education is the highest expenditure item for the government) but poor on measurable outcomes are glaring and malnutrition may be a factor too ( though Tamil Nadu based its school enrolment programme on a successful Mid-Day meal scheme), but even if these could be overcome, it shall be atleast 20 years before a Tamil Nadu competes with a Mexico on PISA .

And, to the people arguing that the elite in India are well-educated. Not really - India has 1 university in the top 400 (http://www.timeshighereducation.co.uk/world-university-rankings/2011-2012/top-400.html) and that too ranked somewhere in the 300s. And, competition is intense for the few good institutions. Most of the elites send their kids outside the country for higher education.

Tamil Nadu, has 70% Affirmative action covering 90% of the population, meaning 90% of the population is low IQ
The remaining 10% consists of brahmins, some merchants and some landlords, and these are the one competing globally


Has detailed interactive tables

For Tamil Nadu- Math

60% is Public school and 40% is Private school

Public School = 334
Private School = 378

The following is not malnourished or from illiterate background
It shows even private schools for the rich are lousy

Long-term, Tamil Nadu ought to equal Mauritius, with scores of around 440

With Educated Mother ( top 44 percentile ) = 371
Owns Family Computer ( top 22 percentile ) = 396
Owns 2 Family cars ( top 10 percentile )= 370
Uses e-mail ( top 15 percentile ) = 369


The only missing link is the caste background , but it ought to reflect the 90% low-IQ affirmative action castes

Hey Steve,

If you just took the top 5% of Orissa students in the TIMSS study, they would be #5 compared to the median scores of other countries.

The top 5% averaged well but that doesn't mean they performed at Level 6 proficiency. Just 2% of American students score at Level 6! So estimate 2% of the top 5% of Orissa score at Level 6.

So the TIMSS suggests that about .1% of students are scoring at PISA Level 6. Not much disagreement in the results.

Also TIMSS studied two below average states in India. PISA looked at two above average states. Above average states are probably well organized enough that 90%+ of the 15 year olds are going to school. In below average states, maybe half start high school?

That difference in attendance could account for why India is near the bottom in TIMSS compared to the very bottom 3 in PISA.

This might be of interest:

India's Demographic Outlook: Implications and Trends
An Interview with Nicholas Eberstadt

Two excerpts:

1. Perhaps most importantly, China has a dramatic edge over India on mass educational attainment. As of today, almost everyone in China’s working-age population is at least literate. By contrast, roughly a third of India’s working-age manpower has never been to school. India is about half a century behind China in eliminating illiteracy. Even posting steady educational progress, India will still lag far behind China in attainment levels twenty years from now.

2. Measuring scientific-technological capabilities is a complex proposition. One useful aperture on “knowledge production” is the number of international patents a country earns in relation to its manpower with higher education and its income level. In general, every doubling of per capita income tracks with a quadrupling of patents per person with a higher education. At this juncture, India is punching way above its weight in patent generation. Over the past decade and a half, the U.S. Patent and Trade Office (PTO) awarded India over three times as many patents as would have been predicted on the basis of its income level and educational profile. China, on the other hand, does not seem to be punching above its weight, but rather performing more or less as a country with its income and education profile would be predicted to perform. Whether China can emerge as an indigenous center of knowledge production is a huge question for the future of Asia, and the world. India, on the other hand, looks to be already on course to accomplish this.

"India, on the other hand, looks to be already on course to accomplish this."

China fifth in UN patent filings, India way behind
February 09, 2010 09:03 IST

China tops US, Japan to become top patent filer
‎Reuters - Dec 21, 2011

1) Actually, education quality has been a major issue for the civil society in India for a while. In fact, the NGO Pratham has been bringing out far more extensive and detailed reports on education quality in India for past five years, http://www.pratham.org/M-20-3-ASER.aspx http://www.asercentre.org/ngo-education-india.php?p=Download+ASER+reports

2) Needless, to say the reports highlight substantial shortcomings in the education quality in India's schools, but they also provide us with a time series of the quality.

3) Teacher quality is a really critical issue. There has been a massive expansion in enrollment in the last decade or so, and schools have not been able to recruit quality teachers. There are some efforts being made to improve teacher quality through remediation programs, but the scale of these programs is still limited.

4) The English language news outlets in India wont give much space to such issues since their audience already has a good station in life, and their kids will probably go to a 'good' private (or semi private) school anyways. As for the vast Indian language media (which is how the majority get their news), I think most of their readers will already know that government schools are not that great. They deal with this inefficient, corrupt and insensitive government almost everyday.

5) Just a general note, I think the rest of the world really has to scale down their hopes from India. Compared to most other countries, India's social system is quite backward. This is the sort of thing that takes generations to resolve. The good thing is that its *formal* political institutions are reasonably developed and its overall security, and politico-economic outlook is fairly stable (normalized for its population of course). But that doesnt mean its going to be at the overall level of Europe, the US or Japan any any time soon. As for the comparison with China, that issue is a bit more complicated than simply comparing stats comparing economic production or patents etc. Of course, China's human development record is much better than India's but then again, given India's peculiar social systems, this has probably been true for most of history. In fact, seen from a historical perspective, the human development gap between India and China (and indeed the global average) has probably been the lowest in the post 1950 era.

Small correction to your comments: India's constituent units are states, not 'provinces'

Just went through the test protocol and the assessment booklet that a student is expected to use to answer the PISA test .....
The kids in Tamil Nadu and Himachal Pradesh must have freaked at just the presented booklet!!!!! LOL
The elementary challenge for global investigations remain in the realm of comparable samples......Shanghai does not equal Himachal and TN : The Tower of PISA has a tilt that cannot be fixed easily!!!

India Education is dealing with volume delivery issues, period. That is known and needs tackling.... but we dont need to start another tsunami of policies that will focus on OECD outcome tracking devices as the holy grail of performance either!! God knows we have the IIT Entrance Exam Obsession headache, already!!! :)

Comments for this post are closed