Why are you so scared of annuities? – By Tim Harford – Slate Magazine:

Here’s what I like about insurance: You pay the insurers money when you do not desperately need it, and then the insurers pay you money just when you need it most.

Curiously, this is not what other people seem to like about insurance. Most people do not try to arrange for insurance payments to arrive when they will need them most. Instead, they arrange for insurance payments to arrive after bad luck.

If your house has just burned down, “when you need money most” amounts to the same thing as “after bad luck.” But what if your son has just been accepted by Harvard? That is when the money would be useful, but we are temperamentally more inclined to insure against the tragic death of a child. It goes against the grain to insure against “good news.”

Generally speaking, I enjoy Tim Hartford’s Slate column. This Saturday’s column features an otherwise interesting discussion of why annuities are so unpopular (a bit more on that below). But first, I want to look at Hartford’s prefatory comments on insuring against ‘good news’.

I think Hartford is being flippant and a bit naive when he suggest insuring against ‘good news’, except possibly good news which is completely outside of an individual’s control. As a colleague of mine pointed out, if we had the above insurance scheme for entrance to an ivy league (say), those who did not gain entrance would be subsidizing those who did. Talk about increased inequality!

Moreover, the typical insurance problem of moral hazard would be reversed and, I would wager, much more severe. An example of moral hazard might a driver who drives a bit riskier because his car insurance will reimburse him for damage to his vehicle. In this case, however, getting into an accident still has all sorts of negative consequences apart from the financial – from the risk of death to the inconvenience of needing to take one’s car to the shop. Getting into Harvard is quite the opposite, I would wager. A prospective student might well work even harder to get accepted knowing that her family will not face any sort of economic hardship if she does.

Another example that I think demonstrates the absurdity of the insurance model of healthcare, amongst other things, would be insurance for planned pregnancies. In this case, Hartford’s suggestion of insuring for when money is needed rather than against bad luck obtains in the real world – a planned pregnancy is a happy occasion, but one with that brings with it financial needs. Insurance seems like an incredibly poor model. Again, families who fail to get have children but were planning to would subsidize those who succeed.

Lastly, a comment on Hartford’s analysis of annuities. Hartford argues that our reluctance to choose annuities (over traditional pensions) is somewhat irrational – it prevents us from having money when we most need it. Hartford offers no real explanation of why people reject annuities, but does present this finding:

Using an Internet-based survey, [four economists] presented respondents with a series of comparisons between pairs of fictitious retirees who had made different decisions about funding their retirement. The survey asked who had made the better choice. Brown and his colleagues found that whether their respondents favored those with the annuities depended entirely on how the question was presented. Annuity purchases look attractive when described as sources of spending. For instance, when told that “Mr. Red can spend $650 each month for as long as he lives in addition to social security. When he dies, there will be no more payments,” respondents preferred Mr. Red’s choice (implicitly, an annuity) to Mr. Gray’s savings account, which was flexible but would run out of money at age 85 if he spent $650 a month.

It seems like a classic example of individuals refusing to give up the freedom to choose, even when it might be to their benefit. For more on this topic, see The Paradox of Choice, Stumbling on Happiness, or my favorite paper by Thomas Schelling. I would wager that these three sources will see more than a few citations from this blog. I might even insure against the possibility.