Lant Pritchett wakes up each morning and worries about the state of India's government schools. Formerly an economist at Harvard and now with the World Bank, Pritchett is happy that 93 % of India's children are now in school as the SRI survey shows. However, digging deeper into the SRI data, Pritchett finds that 53 % of all children in urban India are in private schools. In some states the ratio is much higher, but urban India overall has amongst the highest levels of private primary education in the world. Chile privatised education in 1981, and after 25 years its private sector has achieved only 46.5 % share of enrolment. Even Holland, which has always believed in giving choice between private and public schools to its children as a matter of state policy, has only got a private school share of 68 %. This Dutch level has already been exceeded in six states of India. Whereas in Chile and Holland the government pays parents to send their children to private schools, it has happened accidentally in India because government schools have failed, and even the poor are exiting from them.

The de facto privatisation of schooling in urban India is confirmed by the government's own District Information System for Education website, which shows that 66.9% of children in urban Maharashtra are in private schools, 66.3% in Tamilnadu, and 65.1% in U.P. to name only three of India's largest states. This is supported by Samuel Paul's studies on people's satisfaction with public services. The states with the highest level of privatisation give the lowest rating to government schools. For example, only 1% of the parents in Punjab are satisfied with teachers' behaviour in state schools.

There is nothing wrong with giving parents a choice as Holland and Chile have done. If our government were to give the money that it spends on running schools to children in the form of scholarships, competition for the scholarship money would improve many government schools. Today, the government—the centre and states together--spends on the average Rs 4000 per child per year on primary education. Headmasters confirm that a child can get a decent education for Rs 4000. Thus, money is not the problem, and we ought to test this people friendly scholarship scheme in a few cities. However, we cannot give up on a million government schools. State schools do work in other countries.

The Kremer-Murlidharan study shows that one out of four teachers is absent from our state primary schools and of those present one out of two is not teaching. Thus, the heart of the problem is teacher accountability. And this failure is even more heartbreaking given the exalted status of the teacher in our civilization for whom inspiring young minds was his dharma. Many NGOs are making heroic efforts to improve the existing system, but given powerful teachers' unions, it is an impossible task. Hence, Lance Pritchett suggests that new teachers ought to be hired into a new professional cadre which is district based and offers incentives and promotions for good performance. It would be a Panchayati Raj institution and teachers would be accountable locally to parents instead of to bureaucrats at the state capital. The answer, thus, is not private schools alone, but to liberate government schools from the 'anti-teacher' grip of unions and babus. These are the things that matter in the end; thus will the poor benefit from our rising economy.