Rethinking NEET, and medical education
by Vijay Kelkar and Ajay Shah,
Business Standard, 8 July 2024
Failures in the operations of NEET in recent months have aroused much ire. It is possible to think about the NEET problem statement and improve its implementation. We should peel the layers of the onion and go deeper. Why should every medical college in India be subjected to central control of how it admits students? Why is there so little medical education?
Our first instinct is: we have a failure of operations, so let's go do the operations better. We can add more policemen and armoured cars to protect the papers from leaking. This is a bit of a mugs game, given the extreme supply demand imbalance in medical education and the high incentive in favour of examination fraud. There are too many civil servants involved at too many points in the process. It is not wise to ask flawless performance of government organisations in India. As with most policy problems that we see in India, we need better thinking and not mere execution of poorly thought out concepts. How could we do better than the worms eye view?
A better centralised examination
Why should there be a high stakes examination, organised in person all over the country, on one date? In fact, putting an identical question paper in front of multiple different candidates is well known to be an inefficient statistical estimator of the candidate quality.
Computerised testing based on `item response theory' obtains a better estimator of each candidate. This is how the GRE examinations work. If there was any one fixed exam, there would be a bunch of people with a perfect score (and the people slightly behind would not be materially different from the bunch at the top). Modern statistical methods tease out the difference between the 95th percentile and the 99th percentile by probing them with difficult questions. These are administered from a large question bank of questions which are each the subject of research and measurement based on the past performance of the question.
In this arrangement, testing would take place all through the year, with the test taker making an appointment to take the test. This avoids the bad luck of floods, heat waves or illness. It is operationally easier: instead of having 2.5 million people being tested on one day in a year, we break up the problem into 10,000 persons being tested every day. From a management point of view, it is more feasible to establish a process of data release, research, criticism and then iterative policy modifications, so as to learn how to do 10,000 tests per day correctly, 250 times a year.
This requires organisational capability at the level of the ETS, the private American firm which runs the GRE. It would be good to build the policy reforms to get testing activities out of the Indian state.
This would help, but can we do better?
A better college admission process
How should a university choose from many applicants? This is the legitimate choice of the managers and the board of the university. The purpose and strategy of each university would shape the mechanisms that it uses to choose students. There is no one-size-fits-all. Some universities might like to take in STEM propeller heads, some might emphasise multiple personality traits, some might like to use allocation based on lotteries. It might surprise us in India, but a significant portion of doctors in the US have backgrounds outside of science majors. Doctors require strong communication capabilities, critical thinking, and problem-solving skills, which are fostered by a liberal arts education. A wise leadership of a medical college is best placed to choose an admissions algorithm that is consistent with its strategy.
Medical college admissions is, then, one of a piece with the array of situations all across the Indian landscape, with excessive centralisation and excessive government control. There is no reason for union government control of medical admissions: each medical college should do as it pleases. We need the policy reforms to get the union government out of medical admissions.
This would help, but can we do better?
A better medical education system
There is a massive shortage of doctors, e.g. to the point where Indian students are going to China or Ukraine. Domestic requirements, medical tourism and telemedicine services delivered to overseas users: all these imply that a 10-fold or 20-fold increase in the number of doctors would be readily absorbed by the economy.
There are no shortages of telephone connections or scooters in modern India. What happened to those queues and corruption? The solutions lay in injecting superior ideas, not superior execution, into the policy process. The root cause of conflict around admission to medical college is the shortage of seats. This shortage is caused by mistakes in the rules made by government which hold back entry.
There is an analogy in this field. The emergence of the Indian software industry was threatened by a shortage of engineers. This was solved by policy reform with the entry of private educational organisations. The education market and the labour market know how to distinguish good from bad in the colleges; we don't need government involvement in helping students or employers combat asymmetric information. The Indian private sector knows how to build and run educational organisations, what is needed is a large dose of removal of restrictions. We draw on Agarwal et. al. 2023 on the solutions. Government restrictions hold back myriad universities in India from launching medical schools, which many could do in partnership with private hospitals.
It is efficient for the best private hospitals to add medical education alongside their core operations. A medical university must have faculty members who are not just practitioners: they need to be researchers. This is good for the private hospital as they would command more respect and be sought after. This requires a different way of life compared with conventional private hospitals. The role of the government is to establish a research funding mechanism through which grants go to such researchers.
Back up to Ajay Shah's 2024 media page
Back up to Ajay Shah's home page