Japanese Saving: An Economics Puzzle

Here's an observation on Japan that perhaps represents my best attempt ever to win a Nobel Prize in economics. It's about risk, and savings, and "economic man."

Start with a question: why do people buy casualty insurance? It can't save you any money. Suppose a loss will cost $1 and that the chance of loss is one in four. Then the premium must be 25 cents plus the cost of administration. Granted, people do self-insure some risks, particularly small ones (lots of people turn down the "extended warranty" coverage, for example). But consider bigger risks--fire, theft and casualty on the family jalopy, for example; or trip insurance on your vacation: lots of people could cover these losses on their own, but they'll pay out good money to pass them on.

Why? The conventional wisdom among economists is that people are "risk averse"--that they value certainty more than uncertainty, and will pay a positive sum just to smooth out their flow of wealth. This isn't an "explanation," of course; just an "observation"--at best an empirical observation, either true or not true as the evidence may disclose, just like the assertion that "all bachelors wear bow ties."

But in standard economic discourse, people tend to ignore the empirical part. They tend to treat it as an analytical, definitional ("all bachelors are single men"). I've done it myself, and I forgive myself: as a matter evidence, I think the statement is pretty wellsupported, and it is easy just to treat it as a universal.

Okay, jump cut to Japan--famous for having one of the highest savings rates in the world--down from its peak of a few years back, but still perhaps 10-11 percent of income, which is astronomic, at least by comparison with the US. Traditionally they've also received lousy returns. Why do they accept so much for so little? My friend Kagawachi-san explains:

Japanese face many risks. You may lose your home to an earthquake, for example, or a typhoon. You need to have money on hand for an emergency.

So, why don't they buy insurance? Kagawachi-san continues:

So, why don't they buy insurance? Well, they could buy insurance. But then what if they paid premiums all their life than never had a loss? They would have nothing to show for it. Japanese would rather save and hope to avoid the loss, and and enjoy the opportunity of having something to show for their efforts at the end.

In other words, the Japanese attitude to risk seems to be just the opposite of the American: they'd rather not trade uncertainty for certainty: they'd rather bear the risk of a downside loss to keep the upside opportunity. Fine so far; but recall, risk aversion is built into standard economic thinking as a quasi-analytic proposition. If Kagawachi-san is right, then it is just another empirical hypothesis, leaving the underlying issue--why do people (ever) buy casualty insurance?--unanswered.

Afterthought: As I write, I can think of a couple of complications. One, I can think of one area where Americans tend to think like Japanese. That is pension planning: most folks figure out sooner or later that they need to provide for their old age. They could buy annuities, and some few people do. Burt most do not; why not? One possible reason: returns on annuities always strike me as unrealistically low, which might be reason enough to look for another choice. But another reason is that people seem to find it unbearable that they might have to give up all "their" money: no matter how much they need, they want to be able to leave some kind of estate for their survivors.

Another complication: traditional life insurance. Traditionally, life insurance has usually come packaged with some kind of "investment/savings" paackage. For many potential customers, this never did make any sense: insurance is one thing, savings another, and no need to mix them. People who do want to mix insurance and savings are probably saying something just like the Japanese.