Sunday, March 31, 2013

A sea change in the knowledge of the young in India

In 1887, roughly 14 million children were born in India, and we got one Ramanujan. It seems reasonable to think that there were 9 others who went undiscovered. We may guess that the Ramanujan rate is roughly one in a million. Applying this to the 28 million newborns per year that we have today, we may guess that there are 28 Ramanujans being born every year. Our challenge is to find them and nurture them.

So far, we are doing a terrible job of this. Pritchett & Viarengo suggest that while 18% of the 15-year olds in South Korea have a strong knowledge of mathematics, in India, this fraction is only 0.8%. Also see. We are doing roughly 20 times worse than we could, in achieving high quality mathematics knowledge among 15 year olds. Education in India is deeply broken and as long as the status quo remains in charge, improvements are unlikely.

India is experiencing a higher education revolution on quantity. Here is the age-specific probability of having graduated from college, drawn from the CMIE Consumer Pyramids database:

On the x axis are age groups such as age 20-25, 25-30 and so on. On the y axis is the fraction of persons who have graduated from college. The oldest available survey (December 2009) and the latest available survey (September 2012) are shown. We see a sharp surge in today's young, when compared with conditions one or two decades ago, when going to college was much more exotic: i.e. the 40 year olds of today are much less likely to have attended college (reflecting conditions 20 years ago) when compared with the 20 year olds of today. The age profile of educational attainment is a cross-sectional picture that expresses the time-series of historical experience, much like the rings in a tree.

What is remarkable is the increase seen from just 2.75 years ago (the oldest available survey) when compared with the latest data: in the overall population, the share of college graduates went up from 6.14% to 6.97% over this period. There is truly a revolution taking place in terms of the young going to college. The bottleneck is that almost all college education in India today is quite faulty.

Consider the software industry, which is believed to employ 3 million persons today. A wonderful story by Harichandan Arakali and Tony Munroe, from Reuters last week, tells us that the labour market is walking up the value chain; low-grade coding skills don't cut it any more. Most of the young people with freshly minted college degrees are not good enough for the firms. Entry level wages in computer technology have dropped; skill premia have gone up. On this subject, see my article Will BPO hit a staffing crisis? from 2005. I have long felt that it is dangerous for a young person to build job-oriented vocational skill: Broad intellectualisation is essential. What a person needs is the ability to think; the mere ability to code is no substitute.

Most young people in India today are not on a journey that gets them to think. As an example, IGIDR is arguably the best economics Ph.D. program in India today. Yet, I have been disappointed at how many students at IGIDR don't read books more broadly, don't read blogs, and don't read the Economist. While these problems afflict economics more broadly, we seem to face a greater danger of  narrowness of knowledge.

How is a young person to break free of mediocrity? For the first time, we now have a concrete set of answers: Plug into Internet-based education offerings, and read the great books of the world, across a diverse array of subjects. Plug into the blogs, and get connected to the great conversation about the world today.

The game changer is access to world class materials over the Internet. The Internet in general, and Internet-based education in particular, makes a significant difference to how we harness our human raw material. As a working approximation, an undergraduate education in Economics in India teaches little, but a student who will persevere and understand this blog will get somewhere. I recently wrote about the enormous gains that are possible for young people in India thanks to the new world of Internet-delivered education.

In this context, I found a fascinating fact about the nationality of the students at Coursera, which is one of the modern Internet-centric education offerings. Coursera has an enrolment of 2.9 million. The biggest single nationality is the US, at 27.7%. The second ranked country is India, at 8.8%. In other words, 255,200 persons in India are users of Coursera today. I would hazard a guess that across all the modern Internet-based education offerings, the enrolment from India might be 4x bigger than this, or a million. These are large numbers compared with the size of the knowledge workforce in India. And, this is only the beginning.

There is another angle through which the Indian labour market is getting connected up into the world, and thereby global skills, which is through Internet-based systems where the end-customer directly connects to the worker. These can be deployed for all problems where the task definition is unambiguous and payments can be done on a piece rate. While a lot of this is low end work, this is not necessarily the case. E.g. it is possible to pose pretty high skill challenges while retaining the idea of machine-driven verification that the work is done. I wrote a blog post about the fact that roughly a third of the persons working for Amazon's `mechanical turk' are in India. Ghani, Kerr and Stanton have a similar fact from a competitor to Amazon, oDesk. While the bulk of this work is low end, there is also high skill work here.

The overall picture that I see is one of enormous change in education and work in India, with four big themes:
  1. There is a surge of young people studying for the IIT JEE (and thus getting up to world class knowledge at age 17), and attending college.
  2. They are facing opportunities such as the mechanical turk and oDesk for low end work, and also some high skill work.
  3. Indian firms that engage with globalisation are pushing in favour of higher skills; there is a substantial mismatch between what college/university is doing and what the firms want.
  4. Internet-based offerings such as blogs and courses are giving access to the world of knowledge that go beyond what colleges and universities in India offer. 
In all four areas, the numbers are big. These are not fringe phenomena; they are of first order importance. World class knowledge is being demanded of the young, and for the first time, they have opportunities to build this knowledge on their own, without waiting for education policy makers to reboot the Indian education system. I feel that this phenomenon -- millions of young people bulking up with knowledge that is qualitatively superior to the knowledge of the old -- is the ultimate growth fundamental for India.


  1. On the one hand you mention that PhD students in IGIDR do not read. On the other hand, you expect that the various market forces will bring about a demand-driven change in the education level of the Indian student. Isn't there a contradiction here? Why does the Indian PhD student reach the rarefied heights of the PhD programme at an institute like the IGIDR and still not learn to read? It is because they believe they don't need to.

    Indian students will learn to "beat the system". They will study for exams, copy their assignments through college, get their degrees and then stumble badly when they get their first job and the first 12 months rip away the rose-tinted glasses. Those who don't get jobs after graduating get this ripping-off treatment immediately after graduation, and those who do, take those 12 extra months. I am told IIT BTechs (who aim to join IT departments of foreign financial institutions at Rs.20 lakhs/annum starting salary) too go through this unpleasant experience in their first year or so of work.

    In short, market pulls have been present for many years, but it doesn't change anything in the job aspirant -- not in the good ones who get those twenty lakhs, not in the bad ones who struggle to get a job. I believe our culture's faith in beating the system, in finding a short-cut to health, wealth and happiness, is very robust.

    The only end-result of all this will be the Indian employer. They will suffer because of mediocre, brain-dead employees like they suffer due to bad roads. This impacts, and will continue to impact, our economy.

    The only thing which will fix this will be a change in culture -- a change in the faith in "jugaad" which eats into everything we do as a people. I don't expect to see this change in my lifetime, irrespective of how much the Indian GDP grows.

  2. 'Yet, I have been disappointed at how many students at IGIDR don't read books more broadly, don't read blogs, and don't read the Economist.' - true but will IGIDR consider taking students who do all of these things? As long as the system rewards those who have 'narrow knowledge', that's what students will do.

  3. Can you please let us have links to websites (particularly Finance and Economics) where learning through internet facilities are available.

  4. Thanks, Ajay for the insightful post. A couple of considerations that might be worth thinking about -

    1. In order for students to partake of the benefits these courses might offer, they first need to have a "cognitive map" of what skills and knowledge they need to have when they enter the workforce. For example, in the case of economics, the cognitive map might indicate that a student needs to understand higher mathematics, history, behavioural economics, etc. I can envisage a situation where industry bodies like Nasscom publish these cognitive maps, so that students can equip themselves with the relevant skills even if their college isn't good enough. This could also help in addressing the problem you had highlighted with regard to IGIDR students not reading enough books outside the prescribed material.

    2. I believe the Internet is a double-edged sword insofar as education is concerned. I think it is hard for students to focus on their online courses, when the Internet is awash with potential sources of distraction - FB, Twitter, you name it. The only solution I can think of is to get students so motivated to learn, that they manage to pull themselves away from these distractions. The question of **how** these high levels of motivation may be inculcated among students seems to be worth researching.

    1. You have a raised a valid issue in point 2 above. This had been overlooked by Mr. Shah in his post.

      There are plenty of resources on the internet today. Mr. Shah has given a great example in 'Coursera'. However, the fact remains that even to complete online courses in specific areas, a person needs to be highly motivated. The course demands investment of substantial time and effort. In an age of tiny attention-spans, this kind of commitment is hard to agree to.

      I am myself a 23 year old software engineer having varied interests. But to satisfy my thirst for broad-based knowledge, I read articles from reputed publications, or participate in discussions on Quora.

      How far this effort can be extrapolated to long-duration, demanding coursework available on the internet, remains to be seen.

      Also, we need to consider the fact that most of the young Indians looking to broaden their knowledge base are working professionals, so demands of University courseware , even though it is available on the internet, may be hard to meet, in conjunction with regular job responsibilities.

      All the same, great article by Mr. Shah ! Insightful, as always.

  5. As Shuvam and JP both point out, I dont see how one can expect PhD students at any institute in India to have a liking for reading based on interest, when all their life they have been assessed on 1-3 hour exams. In this system of assessment, any time spent doing anything other than memorizing textbooks and solving 'sums' penalizes the student. Its a rat race and this has been pointed out many times now, including in mainstream Indian cinema.

    Another reason that leads to the intellectual tunneling of our young students is the nature of our higher education system. We dont like broad based universities like the great American universities of California/Michigan/Illinois/Texas etc. We like small, selective 'institutes of this and that', thinking that they will all become MITs. But there is a reason why,despite being a tech school, MIT has built top ranked programs in Economics and Linguistics. And after having been here for almost a year, I would say that, were it not for the brand name, the average MIT student might actually be worse off than the average University of Cali/Mich/Tex student, atleast in terms of the intellectual avenues he/she can explore.


Please note: Comments are moderated. Only civilised conversation is permitted on this blog. Criticism is perfectly okay; uncivilised language is not. We delete any comment which is spam, has personal attacks against anyone, or uses foul language. We delete any comment which does not contribute to the intellectual discussion about the blog article in question.

LaTeX mathematics works. This means that if you want to say $10 you have to say \$10.