When there are 100 candidates for 50 seats or posts, the probability of success is 50 per cent. But when there are 10,000 candidates for 1,000 posts or seats the probability of success is 10 per cent.

This is what has happened. The number of seats or posts has increased tenfold but the number of candidates has increased hundredfold. This is the main problem. The supply of candidates is unlimited but the supply of seats or posts is not. It’s therefore not a solvable problem.

So each aspirant asks: how do I improve my chances? And the answer is that he or she can’t. But they can be made to pay for trying. That’s what coaching institutes do. They charge you to make you think that you stand a chance. They don’t charge you for guaranteeing success.

It’s exactly like investment advice or horse race tipping. There is thus a strong caveat emptor aspect to it that’s always ignored. Remember: no one ever forces someone to take a bet or join a coaching institute. It’s always voluntary. What coaching institutes do via fees is to offer the equivalent of odds. These depend on the probability of winning, which declines with each passing year. It’s like a lottery.

The trick lies in setting the odds — or fees — in such a way that there are enough people to take the bet. The institute is like a bookmaker except it never pays out. Coaching institutes are different from normal gambling in that respect.

And this, therefore, is the only regulatory problem. Regulation must ensure that the outcome and the fees are linked in some mutually acceptable manner. The risk must not be entirely borne by the candidates.

I think a money-back system would do the trick. Not the entire amount, but 60 per cent. The institute gives a post dated cheque for that sum to each candidate, which it can revoke for the successful ones.

Simple. Effective. Win-win.