Dan Ariely had an interesting column in the latest issue of HBR, talking about how Chile forces its citizens to save money and annuitize their pensions:
When employees reach retirement, their savings are converted into annuities. The government auctions off the rights to annuitize retirees in groups of 250,000…
Institutionally, Chile has cracked an age-old problem with annuities. It’s risky business to predict how long people will live, so insurance companies charge a high premium to cover that risk, which makes for an inefficient market. Annuities also suffer from an adverse selection problem… By pooling the risk, the Chilean government makes annuities an attractive business with more competition and better prices. And since everyone is forced to annuitize, the adverse selection problem simply disappears.
This is rather clever, if it’s true. But a Chilean technocrat, Axel Christensen, responded on the HBR website, saying that Ariely misunderstood what he’d been told. The groups of 250,000 are allocated to fund managers, he said, not annuity providers.
At retirement, Chileans may choose between a fixed inflation-adjusted annuity offered by an insurance company or a variable annuity from by the same company that managed their retirement account. It is an individual decision, with no pooling as you stated. The insurance companies have to bid for the contributor´s savings that increases competition, but the system does has its flaws, like the adverse selection you identified.
This doesn’t clear things up a lot: it seems to me that if you have to pick an annuity, then adverse-selection problems are minimized even if there’s no pooling at all. After all, the problem with adverse selection is that people who buy annuities will live longer than people who don’t buy annuities. If everybody buys an annuity, there isn’t a problem.
And when Ariely republished the column on his blog after Christensen had made his comment, the column was unchanged. I don’t know what to make of that: maybe Ariely didn’t see the comment, or he thinks that for some reason it’s unimportant.
Ariely says that schemes like Chile’s wouldn’t go down well in the U.S., where Americans would consider it “heavy-handed and limiting”. I daresay he’s right. But it would be great if there were some way of allowing people to voluntarily commit to annuitizing their pension fund upon retirement. One of the problems with pension funds is that nobody actually needs some big multi-million-dollar nest egg at age 65. What they need, instead, is a healthy income in retirement. But converting a nest egg into an income is non-trivial. You want to maximize your income by spending principal as well as interest, but you also want to make sure you don’t run out of money if you live a long time.
Annuities solve that problem, but they do suffer from adverse selection: people who buy them live longer than people who don’t, and so insurance companies have to make allowances for that. If everybody in a big pool was committed to annuitizing, then the insurance company could ensure that people who died at a younger age would help to subsidize those who live a very long time — as should happen in any good pension system.
This, indeed, is one of the central problems with defined-contribution pensions rather than defined-benefit pensions. When we “save up for retirement”, we’re conflating two things: the savings, on the one hand, and our retirement income, on the other. If we die before we retire, then our retirement income is zero, but the savings are still there, and the only retiree they’re likely to benefit is our spouse, if we have one.
Are there any numbers on the amount of money which is paid in to Social Security against which no benefits are ever drawn? I’d include people working in the U.S. on temporary work visas, here, as well as people who die before retirement while unmarried. On top of that, of course, people who die early in retirement end up taking out of the system much less than they put in. And the benefits of that cross-subsidy accrue to the long-lived, who need money to live on in their 90s and beyond. It’s a humane and sensible system: the living need money more than the dead do.
Off the top of my head, I can’t think of a way of replicating anything like this on a voluntary basis. I could invest my retirement savings in a fund which automatically annuitizes with everybody else in the fund when I turn 65, for instance. But if I get cancer when I’m 64, I’ll surely move those savings into cash instead of meekly accepting a short-lived income.
So the Chilean system of mandating annuitization for certain types of retirement assets makes sense to me, if indeed there is a mandate there. Maybe people could have a choice: they can invest pre-tax dollars in retirement funds which are forced to annuitize, but if they want to retain control over whether or not they annuitize, they have to invest only post-tax dollars.
And then, of course, we’d have to look to see whether the insurance companies actually improved their annuity rates significantly in response to the new mandate. Is there any data from Chile on that? All of this government interference might make sense in theory, but the real world is nearly always much messier.