India has a world-class elite sitting on top of hundreds of millions with low skills and literacy. This accounts for the title of a new study on India’s educational tragedy — ‘India Shining and Bharat Drowning’ — by Jishnu Das of the World Bank and Tristan Zajonc of Harvard University.

What does this imply for India’s future? Das and Zajonc say that if Indian firms adopt a Ford Model-T approach — where a few skilled managers use many unskilled workers — then India Shining (the upper crust) can act as a rising tide that lifts all boats. Bharat Drowning will become Bharat Rising. But, if Indian firms progressively become skill-intensive, India may end up with islands of privilege in an ocean of deprivation.

The researchers have not characterised our educational tragedy fully. The skilled upper-crust has been educated mainly in private schools, and the dregs in moribund government schools. The Indian Institutes of Technology and Management represent successes in public education, but the students who get in are mainly from private schools. So, a better title for the study might have been “Private Schools Shining and Government Schools Drowning”.

Das and Zajonc use a new methodology to compare maths tests across countries. They compare maths test results of Grade 9 students in 2003 in two states, Orissa and Rajasthan, with students in 49 countries. These two states perform worse than 43 of these countries. Few readers will be surprised.

But while most students in Orissa and Rajasthan fare poorly, the upper crust fares very well. The top 5% of children scored 577 in Orissa and 544 in Rajasthan, well above the US average of 504. Assuming that these two states are representative of all India — an underestimate, surely — then for every 10 top performers in the US, India has 4. However, for every 10 low performers in the US, India has 200.

The top performance in 2003 came from Singapore (605 marks), with Korea second at 589 marks. The US ranked only 18th (504 marks). The performance of middle-income countries — including oil-rich ones — was surprisingly bad. Indeed, Orissa (404) and Rajasthan (382) fared quite well compared with much richer places like Egypt (406), Bahrain (401), Chile (387), Morocco (387), and the Philippines (378). Far below came Saudi Arabia (332) and South Africa (264), both of which have substantial per capita incomes.

India is the second poorest country on the list, and Orissa and Rajasthan are among the poorer Indian states. So, the data may give the impression that India is doing well for its income level. This is probably a mistaken impression. The study considers only kids at school. And in India only 53% of kids are enrolled in secondary school, against 90% in South Africa. If we consider all children rather than those at school, Orissa and Rajasthan may come even lower down in the list.

Let us return to the big issue Das and Zajonc pose. Will India follow the Model-T path of development that uses lots of low-skilled labour, and hence uplifts the poor? Or will it follow the high-tech path, based overwhelmingly on skills, leaving out the unskilled poor?

Pessimists have a case. Technological change means that fewer and fewer people are required for production. Tata Motors and Bajaj Auto have doubled their production while halving their workforces. Infotech companies export $40 billion but employ only two million people, a tiny fraction of India’s workforce of 450 million. This accounts for fears of Bharat Drowning.

However, i am an optimist. Despite our educational catastrophe, we have averaged almost 9% growth for five years. Bad education is a constraint, but not a binding one. A slowdown is coming, yet India should average 7% over the next 25 years. With good education we would do better, but 7% is still miracle growth.

Because of its huge population, India’s top 5% equals 55 million people, a large pool of skills. Despite this skill shortages have arisen, and so every part of the private sector is investing in training. This looks capable of improving most skills, sustaining rapid growth for two decades and giving time for education to catch up.

Consider the infotech industry. For every high-skilled software engineer or BPO worker, four other workers provide support services in transport, catering construction and maintenance. Most of these are decent jobs. So, a Model-T approach is not essential. An infotech workforce of 50 million can create ancillary jobs for 200 million.

Infotech is only a small part of the India Shining story. Our 9% growth has been driven much more by trade, transport, restaurants, construction, social services, finance and communications. High skills will help in these sectors, but the majority of jobs will need only modest skills.

Vested interests make government schools difficult to reform. But as incomes rise and people progressively send their kids to private schools, skills will improve. Four states have proposed school vouchers, and perhaps a new trend has begun. Governments may soon fund millions of poor people to send their children to better private schools. Possibly Das and Zajonc will one day write another paper titled “Government Schools Drowning and Bharat Rising”.