Search This Blog

Monday, October 12, 2020

Should we try to retain students going abroad for higher education

 Recently there were media reports pointing out that the top ranker in JEE Advanced has joined MIT (though still in India due to online classes). Stories have also come out that 4th ranker has joined UCLA. This caused the social media to react. Some of them criticizing them to leave India for US. I too poked my nose into it, and wrote on FaceBook that it was their personal choice and instead of worrying about 2 of them studying abroad, we should worry about more than 100,000 going abroad every year for under-graduate alone. Yes, every year, 100,000 students are going abroad for higher education, and the number is increasing every year.

I suggested that we can retain a lot of them in India if we could have high quality educational institutions in India. Since high quality institutions require a lot of expenditure, it is unrealistic to expect government to put in that kind of money. And, let us not forget, they have expanded IIT, NIT, AIIMS, IISER and other high quality institutions in the last decade, but government will have its limits. So, one will have to set these up in private sector, and that will happen only if the private sector can charge high tuition. And therefore, we must allow private sector to charge high tuition particularly when they have invested initial money and proven themselves to be of high quality. In particular, I suggested that a private institution with a similar NIRF rank as an average NIT should be able to charge about Rs. 5 lakhs per student per year, which is similar to the cost that is incurred at an NIT. (May be slightly less than that since it is assumed that private sector will be more efficient than government sector in managing expenses.) And a private sector institution having an NIRF ranking similar to an IIT, could charge Rs. 10 lakhs per student per year.

Why the number of students seeking foreign degrees increasing at a fast pace. I think as we integrate with global economy and more of us travel around, our aspirations are going up. As our economy becomes larger (ignoring covid related downturn in this trend), the ability to afford foreign education is going up. With ease of travel, ease of connectivity through video calls, and increasing numbers of Indian students on various campuses globally, the resistance to sending an 18 year old in foreign land is reducing. And as online education becomes mainstream, I suspect that many students in India will go for programs which allow him/her to spend may be 2 years in India at lower cost and 2 years on the campus abroad as that becomes easily affordable. So my prediction is that within 2 years of post-Covid times (say, 2023 Fall admission), there will be 200,000 Indians enrolled in foreign degrees (including those online from India).

Is this a good thing for India. An exodus because of poor quality of education cannot be a good thing for any country. Spending a massive 15 billion USD in foreign destinations instead of Indian campuses is depriving our economy of that much boost. Our economy desperately needs to reduce import and increase exports. And this is an avoidable import of service. Having high quality educational institutions in India would also attract foreign students (so we not only decrease import of service, but increase export of service). As these high quality institutions will not just be for these 1-2 lakh students, but for everyone else, we will also have a better trained manpower which is desperately needed by our industry. Many high tech companies are finding it difficult to recruit high quality personnel and the growth is slow because of that.

Seems like a win-win situation for everyone. And, of course, government has been talking about greater autonomy, including in setting up fees, attracting foreign students, becoming Vishwa Guru. The system of "Institutes of Excellence" and "Graded Autonomy" were started with these goals in mind.

But surprisingly, I found a lot of opposition to the idea. The arguments were primarily these:

1. These 100,000 are mostly going out because they want to emigrate and setting up high quality institutions in India would not stop them. A lot of them are any way low merit students (couldn't get high rank in JEE), and going to low quality institutions abroad. We shouldn't worry about them.

2. Even if somehow we can retain these 100,000 in India for the UG education, they will leave for jobs/higher education abroad. So our industry/society will not benefit from them. It is best they leave early, particularly the few meritorious ones in this group, because they would waste a good seat in India by leaving India after graduation.

3. High tuition will cause inequity. It is better to have everyone get poor quality education (except a few colleges like IITs which the government can afford to subsidize), than to have some colleges with poor quality education and some colleges with high quality education. The assumption here is that there can be no model of financing a high quality education in private sector, and hence private sector cannot be allowed to set up a high quality educational institution.

Let me answer these objections. The 2nd one is the easiest to handle. Would you want to have a car component factory in India if they are only exporting their products to a car manufacturer outside India. Of course, yes. Whatever economic activity we can do in India helps out economy. If a substantial portion of that 15 billion dollar can be spent in India, it is good for our economy.

As far as 1st objection is concerned, there is really no data. Everyone has different anecdotal experience. The argument that hardly anyone will stay back if there were high quality private institutions in India does not sound right because of my anecdotal experiences. When I talk to students at Ashoka University, for example, I do find many of them saying that they were considering universities abroad. Not only that, once we have something like Ashoka, we are able to attract a lot of foreign students. So even if only a few of these 100,000 will stay back in India, the high quality institutions will be good for economy by bringing in foreign students.

How about equity. I strongly believe that it is possible to come up with a model which allows people from financially weak families to study in expensive universities. For example, why can't and why shouldn't government say that anyone from a weak background will get a voucher to study and they can take that voucher to any of the good quality institutions, whether government or private. Already many of the schemes for SC/ST students allow studying in private institutions, we could extend that for EWS as well. So that pays for a significant part of the cost. The universities can get philanthropic funds to provide some scholarships. Some part of the cost can be taken care of through bank loans. There could be newer models like income sharing agreements. When we compare 100% poor quality versus 90% poor quality and 10% good quality, the latter is bad only if these 10% are all from privileged backgrounds and would cause the gap to only expand. But if we can find ways to ensure that there is representation of under-privileges students in this 10%, then it is definitely better for the country to have more well educated citizens.

Thankfully, despite the objections listed above, the government is going ahead with its policy to attract good quality private institutions even by allowing higher tuition. Some states have started allowing high quality institutions to charge high fees. And states which are rigid on this issue will not attract quality institutions and their residents will suffer. Remember quality institutions not only provide high quality education, but also do research, their alums tend to setup companies in the neighborhood of the college.


Monday, October 5, 2020

JEE Counseling in Covid era

 I have been receiving emails and phone calls ever since the results of JEE Mains were announced about three weeks ago. Some of them weren't going to take JEE Advanced and wanted to know where to apply. And some others were sure about their ranks in JEE Advanced and wanted to know where to apply. Each one of these parents (and never a student) would tell me that everyone gets a better rank in JEE Advanced compared with JEE Mains and there is a rule of thumb. If you had 2X as your rank in JEE Mains, your rank in JEE Advanced is likely to be around X. Go figure.

I would start off with my favorite statements. Don't think about placement. Think of your interest. And if you are normal and therefore, haven't found your passion yet, just get admission to the best college you can get into. Also, travel to a few institutes which you are considering seriously and talk to some faculty and students there.

But as I finished discussion with 20th parent (just two students so far), I have come to realize that Covid has changed the way we should consider choosing a college. And the change is not only about the inability to travel.

Till last year, if someone said that they are interested in Computer Science (and 90% people said that), one would quickly get into a discussion of Civil at an old IIT versus CS at a 2nd generation IIT, or Meta at an IIT versus CS at an NIT, and so on. I believe the discussion ought to be different today.

The concern in the recent past has really been this: I don't know my interest and you are telling me not to consider placement numbers. What if I just take admission in an unpopular discipline in an IIT, and then I don't like the discipline and I will be forced to change career through an MBA route. It was a serious enough concern. A 12th class student hardly knows about various disciplines and has no easy way to find out what s/he would like. So over the last decade or so, most progressive engineering institutes have been having a large number of electives in their curriculum, which could be used to do a "minor" program within your degree program. So if you were interested in CS, you could at least do 4-5 courses in CS. But that was not enough to gain sufficient confidence and sometimes even capability to compete with those who are doing may be 15 courses in Computer Science. And hence the minor programs ensured some relief, it wasn't enough. You needed another 3-4 courses to really get into that discipline.

This is where Covid has come as a savior. Today, the online education is more widely available, the quality has been improving, and most importantly, the market has started to appreciate online courses. So now, if you join an unpopular program, and you realize that this is not what you want, just do the minor program offered by the college in the discipline of your choice, and do an additional 3-4 courses online in the summer or even in parallel with your semester courses. With 7-9 courses, you have pretty much all the skills of that discipline that you need in the beginning of your career and you can keep learning on the job or online.

So earlier I would suggest that if you are interested in CS, study CS even if that meant taking admission in a somewhat "lesser" college. Now, I would suggest that you better be passionate about CS to take admission in a somewhat "lesser" college. The better strategy is to take admission in another discipline in the top college, do a minor in CS, do a few online courses in addition, and you are as good as anyone else.

This means that the most important question that you must ask is "how easy it is to do a minor in a popular discipline?" All IITs, I am sure, would proudly display on their website that they offer minors and lots of electives, flexibility in curriculum, etc. But is that flexibility for real. I know some CS departments would argue internally that if everyone were to be allowed a minor in CS, their teaching load would go up substantially. Why should they do this for students of other departments.  You shouldn't be joining these IITs if the goal was to keep the option of CS (or any other discipline) open.

So the ordering has been simplified, thanks to Covid. Order all IITs based on whatever criteria you like. Whether you prefer big city or small city, whether closer to home or away from home, whether your friends joined there last year or not, whether they already have most of the academic infrastructure built or not, and so on. Then those IITs where you have a strong chance of getting at least the least popular program, find out how much flexibility they have in curriculum. Specifically, do they have a minor program in all disciplines you might be interested in. Second, how many students graduated in 2020, and how many of them were able to do a popular minor like CS. If this ratio is small, remove that IIT from your list. And now, you can specify all the programs of the remaining IITs.

Friday, October 2, 2020

Are Engineering College Placements Correlated with Programs

This is a question that I have been thinking for a very long time. Why should average salaries for graduates of mechanical engineering, be higher than average salaries for graduates of, say, metallurgy. Is there something about mechanical engineering that market values more than metallurgical engineering. (I am just taking two random examples, nothing specific about them.) This would happen if the jobs that require the knowledge of mechanical engineers pay more than jobs that require the knowledge of metallurgy. Or if we consider non-core jobs which are common today, say, finance, somehow there is something that we teach in Mechanical that is indirectly useful for those non-core jobs.

Now, if you compare the salaries offered for core jobs (that is, jobs which require training of a specific engineering discipline), one notices that CS jobs often have higher salaries, may be some jobs in Electronics and Communication have higher salaries too, but after that the core jobs don't have too much variation across discipline. And we can't really think of skills that we impart to mechanical engineers that we don't impart in Metallurgical engineers but are useful for non-core jobs like finance. In fact, all these non-core companies don't seem to discriminate across disciplines. They typically have the same eligibility (in terms of CGPA) and the salary offers for selected candidates is same irrespective of discipline.

So, it would appear that there shouldn't be any significant difference in average salaries of different disciplines, except where the core jobs pay well and that happens in Computer Science and to some extent Electronics and Communication. But there is no denying that average salaries of various programs are different. For example, we at PEC recently announced the following average salaries (in lakhs) for those who graduated in 2020 with a BTech degree.

Computer Science: 14.6
Electronics & Comm: 11.7
Electrical: 9.1
Mechanical: 7.6
Production: 7.5
Aerospace: 6.9
Civil Engg.: 6.7
Metallurgy: 6.1
So we dug deeper into this, and this is what we found. The average CGPA of the graduating students also declined from Computer Science to Metallurgy. We then looked at the average salaries offered vis-a-vis their academic performance (as reflected through CGPA). And this is what we noticed:
9.0 - 9.5: 17.0
8.5 - 9.0: 11.3
8.0 - 8.5: 10.6
7.0 - 8.0: 8.9
6.0 - 7.0: 7.1
(Too few students in 9.5+ and <6.0 to be statistically significant)
One notices a much stronger linkage between CGPA and salary than between department and salary. if one sees the salary of folks with similar CGPA, the numbers are perhaps no longer statistically significant, but the spread is less (and again mostly due to CS/ECE) and students get similar salaries irrespective of their departments.
 This is, of course, on expected lines. As we started by saying that the difference between salaries should be either based on some core jobs paying higher, or non-core jobs valuing some part of training in some disciplines higher. With very few students outside CS/ECE taking up core jobs, there is no reason for salary differential.

So basically what is happening is that Mechanical students have higher average CGPA compared with Metallurgy and that is the primary reason for their better salaries, not their intrinsic mechanical training.

Why do they have higher average CGPA? A few things cause that to happen. One, the higher JEE rankers prefer Mechanical (in the mistaken belief that there is something about that training that gives better placement). That would put a small advantage in terms of CGPA (because the difference in the input quality is frankly not that diverse given that a few marks in JEE means a large difference in ranks). Second, the branch change based on their performance in first semester (or first year, as the case may be) ensures that those who perform really well in the less popular programs shift to more popular programs, and that really affects the average CGPA of a department. The third impact comes from motivation. People who join less popular programs are constantly nagged about it. They are told that they would have been better off if they had joined the more popular department in some other college. That demotivates them and causes reduction in their academic performance.

So the bottom line is that there is no statistically significant difference between the salaries offered to graduates of different engineering disciplines (except where core jobs are well paid and in plenty, which is currently true for CSE/ECE). The salary will depend on your performance in their test/interview, which will be impacted by how much you have learnt, your academic preparation. So, in this admission season, don't worry about discipline and join a good college.

No, it does not mean CGPA is the only thing in life. If you have done well in non-academic areas, you will still get a good job. But most people with poor CGPA also don't do well in other areas. So, please focus on learning and your interests (if any) and not worry about which degree will give you better salary.