Information Processing

Just another WordPress.com weblog

Archive for the ‘india’ Category

To have, and have not

with 4 comments

This NYTimes article starkly illustrates the gap between rich and poor in India. It depicts a gated high-rise community with its own water, power, security and health systems, surrounded by slums from which are drawn the 2.2 servants per affluent occupant family. The high-rise is visible in the background of the picture above (slideshow).

Many of the relatively few Indians who enjoy first-world living standards, including the family in the article, do so as a consequence of globalization.

…“Things have gotten better for the lucky class,” Mrs. Chand, 36, said one day, as she fixed lunch in full view of Chakkarpur, the shantytown where one of her two maids, Shefali Das, lives. “Otherwise, it is still a fight.”

When the power goes out, the lights of Hamilton Court bathe Chakkarpur in a dusky glow. Under the open sky, across the street from the tower, Mrs. Das’s sons take cold bucket baths each day. The slum is as much a product of the new India as Hamilton Court, the opportunities of this new city drawing hundreds of thousands from the hungry hinterlands.

…For those with the right skills, the good times have been very good. Mr. Chand, 34, a business school graduate who runs the regional operations for an American manufacturing firm, has seen his salary grow eightfold in the last five years, which is not unusual for upper class Indians like him.

The Chands are typical of Hamilton Court residents: Well-traveled young professionals, some returnees to India after years abroad, grateful for the conveniences. Some of them are also the first in their families to live so comfortably.

…Mrs. Chand, a doctor who decided to stay home to raise her children, trained in a government hospital. Her other maid told her recently that her own daughter had given birth at home, down there in the slum.

Sometimes, Mrs. Chand said, she thinks of opening a clinic there. But she also said she understood that there was little that she, or anyone, could do. “Two worlds,” she observed, “just across the street.”

The following Bloomberg article describes a recent World Bank study. The results are amazingly consistent with an estimate I mentioned in a previous post, that the “effective” first-world population of India in human capital terms might be around 100 million (4 high performers for every 10 in the US).

The study links the results of a test given to 6,000 teenagers in two states — Rajasthan in western India and Orissa in the east — to students’ performance on the same exam in 51 other countries.

The researchers conclude that mathematical abilities of India’s 14-year-olds vary widely between the worst and the best students.

If other states are similar to the ones studied then it would mean that 17 million Indian students don’t meet the lowest international benchmark of “some basic mathematical knowledge.” That’s 22 times the corresponding figure for the U.S.

At the same time, the depth of India’s math talent — those whose test scores are considered to be of an advanced level — is also significant. “For every 10 top performers in the U.S., there are four in India,” the World Bank economists say. That’s 100,000 students, or more than any European country.

This latter group is supplying the bulk of India’s scientific, technical, managerial and entrepreneurial talent and is responsible for the country’s growing clout in the global knowledge economy.

Advertisements

Written by infoproc

June 10, 2008 at 2:49 am

IIT uber alles?

with 19 comments

I recently came across this interesting web site maintained by Kamal Sinha, an IIT (Indian Institute of Technology) Bombay alum who has worked at Mitsubishi in Japan and in Silicon Valley.

It has been widely claimed (e.g., CBS Sixty Minutes) that IITs are the most selective universities in the world — each year about 300k applicants compete for about 4000 spots. To enter the most competitive (e.g., EECS) departments, applicants must score amongst the top few hundred! I know several theoretical physicists in the US who were “toppers” on the IIT-JEE (Joint Entrance Exam), including one who placed first in all of India his year (“first ranker”)! Perhaps ironically, the first ranker didn’t attend IIT — he chose Caltech instead.

Despite the hype (see below) Sinha seems to think IIT is roughly comparable to other elite national universities like University of Tokyo, Seoul National University or Taiwan National University. Note he estimates the effective population base (the number of people who have access to first world educational resources in K-12) of India as only comparable to that of Japan (about 125 million; see here for a similar estimate by a well-known physicist). The estimates that lead to the conclusion that IIT is the most competitive in the world usually normalize to the entire Indian population of nearly 1 billion. I would say that China’s effective population (in this sense) is around 200-300 million people (and growing rapidly), so perhaps Beida (Beijing University) and Tsinghua are the most competitive universities in the world.

I’d be interested in the opinions of other IIT graduates! Here is some detailed discussion of the JEE exam by an IIT-Kanpur professor (link provided by a commenter). The professor suggests that the test is too hard: beyond the first few hundred or thousand rankers, noise dominates signal (i.e., even many admitted students have very low absolute scores, in which luck may have played a role).

Hype:

“This is IIT Bombay. Put Harvard, MIT and Princeton together, and you begin to get an idea of the status of this school in India.” (Lesley Stahl, co-anchor on CBS 60 Minutes)

“And it’s hard to think of anything like IIT anywhere in the world. It is a very unique institution.” (Bill Gates, Microsoft)

“Per capita, IIT has produced more millionaires than any other undergraduate institution.” (Salon Magazine)

Sinha on IIT acceptance rate:

Admission to IITs is extremely difficult. Only the top 2 percent of the applicants are admitted and to get into a decent department, about half a percent is a reasonable corresponding figure. Here I will explore whether IITs are the hardest school to get into and later I will check if high selectivity results in higher quality. “Hard” facts will be supplied when they become available.

Extremely low Acceptance Rate?

Having results of a single entrance examination determine whether one would be accepted or not is a common feature among the educational institutions in East Asian countries. I worked in Japan for six years and therefore being somewhat familiar with them will compare Japanese figures with that of IITs. All figures ae based on certain assumptions.

Selective Admissions in Japan

While it might not make the CBS news, Tokyo University, or Todai, an abbreviated form of Tokyo Daigaku is the place Japanese moms start thinking of to send their children for undergraduate studies before they are even born. There are 8 national universities like the Tokyo university, Todai being the most coveted one, and a few prestigious private schools like the Keio and Waseda, and these are the schools where almost every graduating school senior hopes to get into. Among technical schools Tokyo Institute of Technology (part of those 8 national universities) leads the pack. Each year news of a few students committing suicide on failing to secure admission into one of these schools is not uncommon.

Tokyo University admits fewer than 1500. My guess is that all the top private universities and the eight national universities combined admit fewer than 15000 applicants. How many students applied for these seats? About one and a half million which is about the total number of graduating seniors. Means about one percent!

Applicant Pool Size

‘Wait a minute’ I can hear you saying. Unlike in Japan where almost everyone takes the test to get into an an university, in India not everybody applies to IITs. Most of the applicants who take the JEE are quite good, there being a self-selection process. In response I will point out that if we compare the potential applicant pools, the following factors stand out:

IIT JEE is taken mostly by middle class applicants from urban areas with total population about 125 million. Japan has about same population with lower percentage of test-takers compared to that of India because of lower birth rates. There almost all eligible seniors take the test to get into these prestigious universities, meaning the potential applicant pools are almost equal.

Engineering schools like Tokyo Institute of Technlogy and various engineering departments in other universities are more selective than the average department. Assume those to be twice as selective.

IITs accept about 3500 of applicants. Given the above assumptions it is about (15000/2)/3500 = two times more selective than the average engineering department in these Japanese universities.

Take Tokyo University for comparision. An overwhelming majority of the grduating seniors choose Todai as their first choice. This means that Tokyo University is about twice (3500/1500) as selective as the IITs and more likely at least four times as selective than the IITs when engineering departments are compared.

Quality of Potential Applicant Pools

Japanese seniors in schools perform near the top in international tests in sciences and mathematics. (Seniors from Korea, Hong Kong and Singapore perform equally well.) Indians are not included in most comparison studies but there seems to be some evidence that the average Indian students would have performed near the average, probably somewhat below it. Moreover, after graduation many Japanese students take time off to study for the entrance exam and their dedication has to be seen to be believed. It leads me to believe that their potential applicant pool of of higher quality.

Other Asian Countries like Korea, Singapore, Hongkong, China?

It may be assumed that the student quality and the selection rates are similar to that in Japan, if not better. Means it appears that IITs, however difficult they are to get into, could be overshadowed by institutions in neighboring countries with more difficult admission standards.

There is a difference though. While graduates of universities like Tokyo are quietly working hard to bring their countries up to top and compete with the West, India with its population of a billion or so, through its IIT and other engineering college graduates, seems destined to become a country where the developed world can chooose its low-cost subcontractors to do the jobs they don’t want to do or have a shortage of workers.

Admissions in the USA

While it seems true that admission rate at IITs is less than even the most selective US school like the CalTech, it does not mean IIT recruits students of higher caliber. In a country like the USA, educational resources were well developed and the enrollment capacity for engineering majors is kept about the same as the number of seniors intending to enter those programs, if not more. It means less desparation. Moreover, there are lot of top-notch schools schools of about equal caliber which decreases their selectivity figures. My guess is that the top 50 engineering schools in the USA exceed IITs in almost all respect and another 100 or so other schools are not far behind.

Written by infoproc

June 3, 2008 at 2:00 am

Skills shortage in India

with one comment

Thanks to a correspondent who sent me the link. I get the New Yorker in the mail, but strangely the last issue or two seem not to have arrived!

I think the main advantage India has over China is linguistic: thanks to their colonial history and the similarity between Hindi and English (both Indo-European languages with lots of cognates), it is much easier for Indians to operate comfortably in the lingua franca of business and science. While the elite of Indian science and engineering are fantastic, the article below points out that depth is still somewhat lacking. I suspect this is less true for China — university-trained engineers there are relatively close to world standards.

Previous related posts.

New Yorker: James Surowiecki April 16, 2007

The economic transformation of India is one of the great business storiea of our time. As stifling government regulations have been lifted entrepreneurship has flourished, and the country has becom a high-powered center for information technology and pharmaceuticals. Indian companies like Infosys and Wipro are powerful global players, while Western firms like G.E and I.B.M. now have major research facilities in India employing thousands. India’s seemingly endless flow of young, motivated engineers, scientists, and managers offering developed-world skills at developing-world wages is held to be putting American jobs at risk, and the country is frequently heralded as “the next economic superpower.

But India has run into a surprising hitch on its way to superpower status: its inexhaustible supply of workers is becoming exhausted. Although India has one of the youngest workforces on the planet, the head of Infosys said recently that there was an “acute shortage of skilled manpower,” and a study by Hewitt Associates projects that this year salaries for skilled workers will rise fourteen and a half per cent, a sure sign that demand for skilled labor is outstripping supply.

How is this possible in a country that every year produces two and a half million college graduates and four hundred thousand engineers? Start with the fact that just ten per cent of Indians get any kind of post-secondary education, compared with some fifty per cent who do in the U.S. Moreover, of that ten per cent, the vast majority go to one of India’s seventeen thousand colleges, many of which are closer to community colleges than to four-year institutions. India does have more than three hundred universities, but a recent survey by the London Times Higher Education Supplement put only two of them among the top hundred in the world. Many Indian graduates therefore enter the workforce with a low level of skills. A current study led by Vivek Wadhwa, of Duke University, has found that if you define “engineer” by U.S. standards, India produces just a hundred and seventy thousand engineers a year, not four hundred thousand. Infosys says that, of 1.3 million applicants for jobs last year, it found only two per cent acceptable.

There was a time when many economists believed that post-secondary education didn’t have much impact on economic growth. The really important educational gains, they thought, came from giving rudimentary skills to large numbers of people (which India still needs to do—at least thirty per cent of the population is illiterate). They believed that, in economic terms, society got a very low rate of return on its investment in higher education. But lately that assumption has been overturned, and the social rate of return on investment in university education in India has been calculated at an impressive nine or ten per cent. In other words, every dollar India puts into higher education creates value for the economy as a whole. Yet India spends roughly three and a half per cent of its G.D.P. on education, significantly below the percentage spent by the U.S., even though India’s population is much younger, and spending on education should be proportionately higher. …

Written by infoproc

April 13, 2007 at 11:34 pm

Posted in globalization, india