Why a public register of insurance sellers won’t fix the problem

Some carmakers sell vehicles without a spare wheel, offering just a puncture repair kit instead. Nobody proposes a register of the salesmen who sold these cars because we know the choice was made above their level. The agent sitting across from an insurance buyer is in the same position, as he’s not the author of the incentive he answers to but merely its executor.

At the silver jubilee gathering of the insurance brokers’ association in Delhi last week, Insurance Regulatory and Development Authority of India (Irdai) Chairman Ajay Seth made a proposal that is hard to argue against: every policy should be traceable to the person who sold it, not merely to the bank or broking firm. A public register would hold this, so that when someone is sold an expensive endowment plan when a term cover would have done, the seller can be identified, and his record can follow him from one employer to the next. It sounds like the start of real accountability, and in a narrow sense it is.

To his credit, Seth did not stop at the register. In the same remarks, and in interviews since, he has spoken of tying the rewards of selling to the quality of advice rather than the volume of business, about making the unsuitable sale something an insurer is penalised for rather than paid for, and about a wider reform of the way these products are distributed. This is not a regulator who has forgotten that incentives exist; he has named both halves of the cure. The trouble is that the two halves are not equally likely to be implemented.

Incentive determines outcome

This reminds me of the late Charlie Munger’s most useful sentence, that if you show him the incentive, he will show you the outcome.

One of the two cures is, in the end, a database that records who did what. It costs the distribution business nothing as it threatens no existing stream of income. The other asks the industry to cut the very commissions that pay for the bank counters, the broker networks and the apparatus through which most policies are pushed. It is not hard to guess which of the two will be implemented, and which one will become a discussion paper, then a consultation, and finally a circular so soft that the sales machine will carry on regardless.

I’ve noticed recently that some carmakers are selling vehicles without a spare wheel, offering just a puncture repair kit instead. This is sort of acceptable in a city but a disaster on a highway after dark. Nobody proposes a register of the salesmen who sold these cars because we understand that the choice was made above their level to protect a price or a margin, and that the blame lies with whoever set the target, not whoever met it.

The agent sitting across from an insurance buyer is in the same position, as he’s not the author of the incentive he answers to but merely its executor. A register that captures him, however faithfully, leaves the person who designed the reward out of the picture.

This is why the promise that a seller’s history will follow him sounds more strict than it actually is. You can trace executors for as long as you like, but the next person hired into the same setup will behave the exact same way, because it is the setup and not the individual that is doing the work. When more than one in five complaints against life insurers now concerns the way a policy was sold, we’re seeing a system that is working as it was designed. No register will solve this problem. I made much the same argument here a year ago, suggesting that buyers might organise for themselves, since nothing else seems to work.

None of this is to say the register should not be built, only that we should be clear about what it can and cannot do. A regulator who meant the harder half would begin at the other end, making the unsuitable sale unprofitable before troubling to trace it, so that the register is used to confirm that a problem is already shrinking. I would be happy to be proved wrong on what I have said above, but the direction in which the money flows has always been a perfect predictor.

The Author is Ceo, Value research