A Right Step Towards Risk Pooling

A couple of months ago, I suggested here (not wholly originally) that the major problem of the CPF system was not that returns were low, but the lack of risk-pooling to guard against "longevity risk".

First on the issue of returns. Based on the sketchy information that has been brought to the public in the two days since N-Day rally, CPF returns would indeed be boosted by giving members a higher "expected" rate that would fluctuate according to market conditions. Obviously, higher expected returns would be accompanied by some risk. The interesting question is - if an individual is more risk averse, can he opt for a fixed 4% interest instead?

The bolder proposal being discussed is the "longevity insurance" - citizens contributing towards a national pot, from which senior citizens aged 85 and above would be paid from. The idea of this is simple enough but it represents a key shift for it is the first time a form of national insurance for retirement (either cross sectionally or inter-generationally) is introduced. When my mother-in-law (a youngish 52) heard the proposal as it rolled out at the 10 o'clock news, her first reaction was, "Why should I be paying for others above 85". As Singaporeans are brought up with the idea that CPF account belongs to the individual, it may be difficult to accept that setting aside a small fraction into a common pot for insurance is in fact beneficial to every one.

From an economic point of view, the proposed policy is indeed a right step forward - and a very smart one in fact. First, the scheme is likely to be mandatory. Otherwise, since an individual is most aware of his own health status, only those who think they can outlive 85 will participate in the scheme (averse selection) making it less sustainable. Some critics however argue that mandatory participation would in fact disadvantage the poor since they have a lower life expectancy and the scheme would be actuarially unfair to them. But my sense is that this problem is small compared to old age destitution when retirement funds run out.

Second, notice that the pay out age of 85 is also set above the average life expectancy. Suppose a worker stops working at 65, he would have to live on his own retirement funds for another 20 years before he could dip his finger into the common pot. In other words, the worker would still have to save up for his own retirement (of 20 years at least) before the insurance element kicks in.

In economic parlance, the 20 years is the "excess" - like how we often have to pay for the first $500 of damage when we file a car insurance claim. The purpose of this "excess" is to reduce the moral hazard problem that always occurs with insurance. Because he still have to worry about 20 years of his retirement, the worker still has strong incentives to build up his personal retirement account even though he is insured past 85.

As expected, the loss of control of a portion of retirement fund has given rise to some unease ("coerced annuity"). Rather than seeing it as coercion, a better way would be to understand the true nature of this proposal, which is essentially a social safety net for the advanced old. But the nature of having a social safety net is as such - there must be contributions by all to support and sustain this safety net. In principle therefore, building in some element of risk-pooling into our CPF system is the right step forward. The proposal also cleverly reduces the moral hazard problem. The government should be given some credit for creative thinking.