Arnold Kling  

Why is it Life Insurance?

Outsourcing our Carbon Emissio... Italy's Black Friday Sale...

I claimed that if an event is very likely to happen, then it is not insurable. Some commenters responded that we have life insurance, and yet everybody dies.

The reason that we can have life insurance is that the timing of death is unknown. If you were known to have exactly two years to live, then you could not obtain life insurance.

If I am 50 years old and my family depends on my salary for its well-being, then I might buy life insurance. That way, in the unlikely even that I die before age 65, my family will be provided for. The insurable event is not my death per se. It is my death at an early age.

What that suggests is that life insurance policies should expire about the time that you quit work (you can do this with term life insurance). Beyond that, it just becomes a form of (tax-advantaged) saving that only benefits the purchaser to the extent that the tax savings make up for what the life insurance company chews up in administrative costs and profits. Given that there are other tax-advantaged long-term savings options available, I think that the case for buying life insurance that goes beyond the point of retirement is pretty flimsy.

Comments and Sharing

COMMENTS (12 to date)
Sean Brennan writes:

I didn't see the other post, but on this one I think again the timing of the death is still the case for insurance after retirement....especially for people who owe taxes, or have to depend on savings for their retirement.

Two quick examples:

Suppose you will owe taxes based on your net worth. When you die your net worth says you will owe 100,000 in taxes, but when your family gets around to liquidating assets to pay for your taxes they only get 70 cents on the dollar.....the government doesn't care, they get their taxes. So life insurance is there to ensure the funds are available when needed.

The other example is that you won't owe taxes, but you and your wife are living on retirement savings. Unless you are a financial genius you probably underestimated how much you needed, overestimated how much you can earn in the markets, and again underestimate the rate of inflation. Statistically your wife will live about 10 years longer than you, so it might be nice to refill the funds when you go, or cash it in for the cash values if you need them yourself.

Lord writes:

Whoever argued there is a case for doing so? People do many things, while not optimal, are accessible and convenient that overcome some of their inherent behavioral deficiencies and satisfy some of their anxieties.

Marc A Cohen writes:

Without advocating for or against Life Insurance, the way a whole-life insurance policy is structured can give a lesson on how to design permanent Long term Care Insurance policies.

Whole life insurance policies have level premium for the life of the insured. Compared to Term, they are MUCH more expensive initially, but stay level over time once in force. They do that by setting a portion of the initial higher premium aside into a reserve fund. As you get older and your annual chance of death increases, the reserve fund also increases to reduce the $ amount at risk for the insurance company, so the premium remains the same. Eventually, the reserve grows larger than the face amount of the policy, and the insured is paid off whether they are dead or not.

A Long Term Care policy could be structured the same way - a higher premium in early years, with a growing reserve fund to offset increasing risk of debilitating illness or injury.

No, I don't sell insurance or work for an insurance company.

ed writes:

Another example: futures and options contracts function pretty much just like insurance, for events (in this case price changes) that are not uncommon.

I agree that long term care insurance doesn't really work, but not for the reasons you give. A large problem is the uncertainty about government policy in the future. Government already pays for a lot of care through Medicaid, and in future decades it may pay much more or less, which makes my needs very hard to predict.

In general, what makes an event un-insurable is either (1) risks in the insured population are highly correlated so pooling doesn't help (e.g. earthquake insurance), or (2) risks to the insured cannot be reliably observed or estimated.

Bob Murphy writes:

I have actually co-authored a book on the benefits of whole life insurance. When I get a chance I'll try to respond to Arnold's points directly, but for now anyone who is curious can go here. (The PDF is available for free.)

Peter H writes:


I just read over your chapter on insurance, and while I found it quite interesting, I think you are mistaken that a whole-life policy is a better thing to buy (from a personal finance perspective) than term life. This is because the savings needs of a person vary dramatically and predictably over time. E.g., when I am working and have children, my need for insurance against my death is quite high, since my income will be sorely missed by my husband and kids. However, when my children have established households and incomes of their own, and my husband and I are retired and drawing on savings to live, then the costs of me dying (financially, not emotionally) are near zero, basically just the cost of a funeral. So I do not need life insurance when I am retired.

Now, why does this mean whole life is a bad idea? As you stated, whole life policies last until you're 100 or 121 generally*, and either pay out their face values at that age, or pay out earlier than that when you die. The problem is, that you are paying premiums pursuant to a life insurance contract for periods of your life when there is next to zero chance of life insurance being very useful for you. Let's say I expect to retire around age 65, and at that time my children will be over 30. I can buy a term policy lasting until I am 75, covering 10 years of later retirement (just in case), and it will cost much less than a whole life policy, since the whole life policy would have to cover at least the period from 75-100, which is when I am statistically most likely to die.

*Only one person in recorded history has lived past 121, and a tiny minority (roughly 1/4400 in the US) live past 100.

Silas Barta writes:

@Peter_H: Please be sure to read Bob_Murphy's chapter on objections to whole life insurance. While I have my own reservations about the strategy[1], I think you are missing that the payments into a WL policy are effectively going straight into your savings, available for your withdrawal. So it is not quite right to say that there are periods of your life where the WL insurance is "doing nothing for you", any more than there are periods were simple saving is "doing nothing for you". Plus, this kind of savings is largely tax-free and-or tax-deferred and earns a more stable, market-uncorrelated return than investments that have similar expected returns.

[1] I don't like how WL basically means investing in the bubblicious ultra-long-term bonds of governments and government-entangled entities, nor the strange tendency of WL promoters to think it's so cool that you can borrow from your own savings.

Steve writes:

Comparing whole life to term is like comparing apples to oranges. Neither is better; they serve different purposes.

There is a limit to tax advantaged savings vehicles, and for people who are maxing out their IRA and 401(k) and whose risk profile is one that would seek bond-type investments, where else will they go for tax advantaged savings? Sure, you can defer gains in a stock fund but with bonds the coupon interest would normally be taxed at your ordinary rate (35%).

HOI CHI IAN writes:

I agree that life assurance is not insuring the insurant's life, but the beneficiary, most likely his or her family members. When my family depends on my salary and I die in a sudden, then the amount of money that the insurance company pays to my family will be an important sum to support my family at least for a while to mainland their basic spending. As time passes, they are more responsive to my death and monetary condition by seeking jobs. Besides, the insurance companies provide job opportunities. More people having a job increase a country's productivity and enhance economic growth. Personally, life insurance is a good form of saving as it is not only benefits to my family, but also has a good contribution to the society.

[URL removed--Econlib Ed.]

Peter H writes:


The problem is that the savings component vanishes upon your death if you die. At death, you get the face value of the policy, minus any loans you've taken against it. The savings building alongside are taken by the insurance company upon your death. Additionally, you're much less likely to get a good deal on WL as opposed to TL because you have to shop on 2 factors, rate of return and premium for the life insurance itself. With term, you're just shopping on premium for insurance, with RoR removed.


You can get a similar advantage by using a variable annuity and buying bonds within the annuity.

Dan Weber writes:

One big reason to purchase long-term case before you need it, even if it is sure you will need it, is that you are in a better negotiating position now.

Hank writes:

Good points. This is why term life insurance is generally the ideal choice for most families. It is easy to understand, affordable, and covers beneficiaries while they are dependents. Life insurance should never be considered an investment. By doing so, people over buy and over pay. Thanks for the post.

Comments for this entry have been closed
Return to top