The quiet revolution in the Indian workforce

“You should remove ‘agricultural worker’ from the list of options of parents’ occupations in Question 11,” said the senior government bureaucrat. He explained, “It is impossible for the child of a farm laborer to enter an engineering college.” That statement was made on May 8 in Delhi this year, while he – the chief advisor on higher education to the national government – reviewed a questionnaire for final year engineering students. The questionnaire is to be filled by the graduating cohort of engineering students at various Indian universities this coming year. Its purpose is to discover job mobility across generations and relate that to the cost of education, location, public versus private provision, and various other factors. It is part of a broader study supported by FSI that colleagues at Stanford University and I, along with research groups in India, China and Russia, have initiated to compare the quality of the engineering workforce in three countries – China, India and Russia – with each other and with the United States.

A few days later, on May 14, I was with the head of a medium-sized private college in Bangalore, which had administered the pilot version of the questionnaire to graduating students. As he handed me 450 completed forms, I glanced at the first few. There, right on top, I read the first student’s response to Question 11. A female, she had chosen “agricultural worker” as the father’s occupation. Combined with information on her family’s income (which was in the lowest tier), this was clearly someone who contradicted the bureaucrat’s assumption.

As heartwarming as it was to see that response on the questionnaire, it reminded me, not for the first time, about how little government officials can sometimes know about their constituents. In 2004, I had studied, jointly with a division of the Ministry of Information Technology, how rural users might best use information technology. Our expectation (prior to the study) was that e-mail for personal and business purposes and Internet searches and transactions for farm work would be the main uses.

Instead, what people wanted was government services – health care and other welfare services, postal services, accessing titles and other official records, and government jobs. When I presented our findings to the country’s Minister for Information Technology, he insisted that we were wrong and that our initial hypotheses were correct. It was only when his own division head, who had conducted the study jointly with me, stated (firmly) that he stood by the results that the Minister started to change his views.

Perhaps one should not be too harsh on a bureaucrat when a political master, the minister, could be so ignorant! But, there is another reason for leniency: the higher education revolution in India has still not been understood, even within India, perhaps because of the speed of its happening. A revolution it undoubtedly is. For example, in engineering studies, the number of students enrolled in full-time 4-year undergraduate degree programs has risen from 250,000 in 1997 to 1.5 million in 2007, and is currently growing at 25% annually. Most surprisingly, the higher education sector has moved from a primarily state-provided service to private provision within a decade. 95 per cent of the above increase comes from enrollment in privately-run colleges, which now account for 80% of total enrolment. The storied state-owned Indian Institutes of Technology, which made up 10% of national engineering enrolment in 1990, now account for less than 2%, and graduate 5,000 students a year.

How this happened is too long a story to go into here. Briefly, the national government has increasingly yielded control over higher education to the individual states over the past ten years. The states have, in turn, allowed the private sector in, something that the national government resisted when it was in charge.

One of the desirable outcomes is, as demonstrated by the response to Question 11 above, increased access. Ten years ago, the child of an agricultural worker was, if educated through secondary school, likely to have studied only in the vernacular – and would thus have been excluded from the higher education engineering degree, which is taught only in English. Even if there was money in the family till to pay for tuition, the nearest college was probably too far to allow the student to stay at home; even if she had the money for staying away from home, competition for the limited number of available seats would likely exclude her from even the least meritorious college.

Today, even though the private colleges charge, on average, fifty thousand rupees ($1250) a year for tuition, which is three times the tuition fees at the comparable state college, affordability has increased. This is for two reasons. First is the proliferation of colleges. Thanks to the blanket coverage being provided by the private sector, there is a college, most likely two or three, in most small towns. Bangalore, with 290 engineering colleges – almost all private – tells the story of the rest of the country.

So, even small-town students no longer need to live away from home, thus saving on living costs. This can be a significant savings: in Bangalore, rent for a single room more than makes up the difference in private and state tuition fees. Second, the private colleges have built linkages with banks, so bank loans will usually cover half the tuition costs.

The democratization of higher education in India has removed the impending shortage of talent for the IT exporting sector. It has also brought into question the importance of the IITs to the eco-system, which – according to the recruiters I have interviewed over the years – was always overstated. Let’s examine both of these in the current context.

For the top IT exporting firms in India, such as TCS, Infosys and Wipro, the private providers are a boon. Together, the top three firms will, even in today’s difficult global economic environment, add 70,000 persons to their payrolls (net of attrition) in 2008. 70 per cent of these recruits will be fresh graduates. Private college graduates will account for the overwhelming majority of their recruits, followed by state colleges (not IITs).

Of course, these firms would like to recruit the top IIT graduates. However, the best IIT graduates either go abroad to study or work (a third do so, though that ratio is declining), another third join an MBA program in India, and the rest are recruited by the Indian operations of western firms like Google or Yahoo!, or join Indian startups like Tejas Networks or Telsima. Such firms pay starting salaries that are double the $7,500 starting wage offered by the Indian IT majors.

Is this a big loss for the Indian IT industry? No, say the recruiters, pointing out that the IIT graduating cohort was always a small proportion of their recruits because of overseas migration. What is important, they point out, is that other providers are rapidly catching up with the IITs in quality. Given their reliance on fresh graduates and their scale of recruitment (for example, between June and August of this year, TCS will make one thousand job offers a week and recruit 85% of its offerees), the Indian IT firms make precise calibrations of schools and rank them. The top quartile of the graduates of the top local private colleges in Bangalore are now considered equal in quality to those at the 50th percentile in the IITs. The top quartile at national colleges, such as the National Institutes of Technology, are deemed equal to the 75th percentile of the IITs.

The rank is based on various factors: alumni recruited by them in earlier years, internal factors such as laboratory and library infrastructure, and course content, their interaction with faculty in research projects, and student performance in internships. A thousand colleges (of the four thousand that offer engineering degrees in India) are deemed to meet the standards of the top three IT firms and their graduates are thus eligible for recruitment. According to one of the IT firms I spoke to, a decade ago, there were only fifty colleges that met their standards.

In consequence, in states where they are concentrated, eg., Infosys and Wipro in the state of Karnataka (whose capital is Bangalore) and TCS in Tamil Nadu (whose capital is Chennai), the ranking by the top 3 IT firms is critical for the colleges. A corporate recruiter from a smaller firm seeking IT talent from a Chennai college will demand to know its “TCS ranking."

This, in turn, is invaluable information to incoming students, which, in its turn, influences how colleges invest in faculty and infrastructure. As a result, in a way that was unforeseen by government planners and even the World Bank (which, in 2000, argued that market failure was likely in case private provision in India became important), a thriving market for engineering education has been created and quality has improved.

As recently as 2001, a report on IT education (which included a study of the IITs) by the Ministry of Human Resource Development noted that “The barest minimum laboratory facilities are available in many of the institutions and very little research activity is undertaken…Engineering institutions have not succeeded in developing strong linkages with industry…The curriculum offered is outdated and does not meet the needs of the labor market.” Around that time, when I had interviewed the director of one of the IITs, he had supported this finding, noting that almost all the engineering students at that IIT did their final year thesis projects in laboratories within the IIT (rather than, as intended, in companies).

Today, an engineering graduate from any of the thousand colleges that the IT services industry deems eligible for recruitment will always have completed several internships with industry prior to graduation, including the final semester thesis project – in other words, this is a sea change from just a few years ago.

Of course, there are caveats to the story of higher education. One of the concerns stated by regulators is that, as control has shifted from New Delhi to the states, the weak states have not been able to keep up with the strong states, thus increasing the intellectual gap between them. This appears to be true, on first impression. My conversations with recruiters of IT firms in Bangalore in May indicated increasing regional selectivity. Karnataka, Tamil Nadu, Kerala, Andhra Pradesh, Delhi, Maharashtra and West Bengal were the regions of choice, while weak states such as Bihar and Uttar Pradesh were falling behind.

A second genuine concern of policymakers is that the private colleges have no research agenda. Of course, what policymakers do not state is that the IITs have historically had no research agenda either. The good part of the present situation is that, with the burden of providing mass education off its backs, the national government is using its limited resources to support centers of excellence for research.

A final caution is on replicability in other countries. The higher education system that has resulted in India was not foreseen and caught the nation’s education planners by surprise. No one expected that the private sector would respond as it did. Planners designed the system to allow only non-profit private providers. Planners expected that those private providers that would enter the system would be philanthropic. They would exist at the margins of the then larger state-system. Accordingly, planners encouraged them, through incentives, to set up their institutions in smaller towns.

Instead, the private providers stormed into the big cities first, preferring to ignore the incentives, and have only recently spread to smaller towns. They have made profits through the back-door (by charging an upfront fee, the capitation fee).

A key factor was rising federalism: strong states like Karnataka and Tamil Nadu were able to provide the regulatory support that made private sector entry possible. The second key factor was the IT industry’s willingness to be the market maker, as described above. In this, the role of the large Indian IT firms, as noted, was critical. It is unlikely that an industry characterized by a large numbers of small firms would have been able to play the role of market maker.

So, there are some unique factors in India. China offers an alternative, perhaps more replicable, model: an entirely state-run system in which tuition fees, which average $800 per annum, pay for 50 per cent of costs. It, too, has grown rapidly: for example, 5 million students are currently enrolled in undergraduate engineering programs. The share of the burden per student appears to be higher in India. In India, the state and “aided” private colleges (these are privately owned and managed, but accept state-aid to pay for costs such as infrastructure and faculty salaries – in return, they must charge the same tuition fees as state-run institutions) account for 40% of total enrollment and charge fees that cover 30 per cent of costs. The unaided schools, as noted earlier, recover full costs through tuitions (endowments insignificant). Hence, the share of total national costs of education borne by students in the system is over 70%. This may be important for achieving long-term sustainability, although, in the short-term, it may adversely affect enrollment.

For the moment, though, the Indian IT industry, earlier starved of talent, has been saved by one of its own – the for-profit private education sector.