David R. Henderson  

Kevin Williamson on Social Security

PRINT
Divine coincidence?... John Gray on Hayek...

Note: I returned late Thursday from my 2.5-week vacation at my cottage in Canada. Thus the relative sparseness of my posts and especially of my answers to commenters.

In an otherwise excellent post, which I highly recommend, Kevin D. Williamson of National Review writes the following:

Social Security made sense in the context of 1935 demographics. It's as obsolete as those suitcase-sized portables that Sandberg-Diment was scoffing at 50 years later. But we're stuck with it, because there's not much evolution in political programs. Government programs don't die.

I disagree. It never made sense.

First, look at the context, specifically when it was introduced and when people started receiving payments. Many older people had lost their jobs and had little to fall back on. So, even if you favor having the government forcibly take money from some to give to others, wouldn't it have made sense to help them immediately, in the middle of the Great Depression? But that's not what FDR did. Instead, the first recipient of Social Security, Ida May Fuller, received her first Social Security check on January 31, 1940, over six years after the worst of the Great Depression was over. (By the way, this is the least powerful argument of my three. The unemployment rate was still in double digits in 1940. Suffice it to say that 5 years in the 1930s is a long time to wait for benefits from a Depression-motivated Social Security system.)

Second, in 1935, people who retired at age 65 did not live nearly as long as people today. So it made sense for them not to put away as much for their old age as people think reasonable today. So in 1935, it would have been a bad argument to say, "Look at the fact that people don't save much for their old age and so we need a Social Security system to provide for them." The problem was not the economic fortunes of the elderly per se; the problem was the lousy economic fortunes of a huge swath of the population, elderly and otherwise. So that's another reason that Social Security did not make particular sense even in 1935.

Third and most important, look at the structure of the Social Security system. It's a chain letter or a Ponzi scheme. It was set up that way. Franklin Roosevelt was someone who understood this. He stated:

We put those payroll contributions [sic] there so as to give the contributors a legal, moral, and political right to collect their pensions and their unemployment benefits. With those taxes in there, no damn politician can ever scrap my social security program.

The essence of a tax-financed Ponzi scheme is that those who get current benefits are paid from the taxes of those who are currently working. For the scheme to work, it has to exist forever. Otherwise there will be a generation or two who pay taxes and get nothing in return. So for Williamson to say that Social Security made sense in 1935, he would have to claim that it makes sense now. He doesn't, and rightly so. So he, and others persuaded by him, would do well to reconsider whether it made sense in 1935.

HT to Arnold Kling.


Comments and Sharing


CATEGORIES: Social Security




COMMENTS (29 to date)
JLV writes:

Life expectancy at age 65 has increased, but not really that much compared to overall life expectancy, less than 5 years since 1950 (earliest year the CDC data goes). And this is unevenly distributed - black men have seen less than 3 years of improvement.

SSA data suggests that there is a similar inequality with respect to income. Life expectancy at 65 for people in the top half of the earnings distribution increased by about 5 years (comparing birth cohorts from 1912-1941), while for people in the bottom half, the increase was a bit over 1 year.

BC writes:

I have an admittedly heterodox description of Social Security (SS) that actually would allow us to end it at any time without stiffing anyone. Most people describe SS as intergenerational wealth transfer. However, it actually does no more intergenerational transfer than the alternative: working children supporting their retired parents to the extent that those parents didn't save enough. Instead, SS socializes the retirement benefits of having children: instead of supporting their own parents, children support everyone else's parents and vice versa. This is literally what happens with SS taxes and benefit checks. The transfer is inter-family (just like regular welfare): wealth is transferred from families with many high-income working children to families (senior couples) with few or no high-income working children.

This is how we understand public education. It's not "intergenerational" transfer from parents to children. Compared to the alternative of self-financed private education, public education is transfer from wealthy couples with few or no children to poor parents with many children.

From this perspective, ending SS immediately "stiffs" no one. Even if you have been paying into the system your whole life to support everyone else's parents, everyone else has also been paying into the system to support your own parents.

My description has an advantage over the conventional one: it is consistent with both the observed decline in birth rates and the low savings rates. Why bother having lots of children to support your retirement when SS lets you rely on other people's children instead? (Some might object that neither they nor anyone they know has children nowadays for purposes of supporting their retirement. The idea is gauche. But, that supports my point: SS has indeed removed such motivations.) Re: savings, many people might be motivated to save responsibly to relieve their own children of the burden to support their retirements. Thanks to SS though, one can *never save enough* to relieve one's children of the burden to support everyone else's parents' retirements.

Given that SS is really just wealth transfer between families, why not subsume it into the regular welfare system for the poor?

Andrew_FL writes:

Perhaps Williamson means by "makes sense" not "was a good deal at the time" but "looked like it could work in principle at the time."

By which I mean, not that it would achieve good results, but that it would at least fail to completely collapse under it's own weight given a high enough sustained rate of population growth.

Williamson would still be wrong, because actually the population growth rate in 1935 was lower than it has been in recent years, but he can perhaps be forgiven for making this mistake, as that's not a fact which is common knowledge and it's somewhat counter-intuitive.

Jeff writes:

Yes, it's forced, but let's put that aside for a moment.

I'm not sure I follow #2: If "it made sense for them not to put away as much for their old age as people think reasonable today" because they weren't going to live that much longer, then the concept of Social Security kind of makes sense when viewed as (socialized) insurance against outliving your savings for retirement. Of course, good intentions don't necessarily make policy. We don't need to argue here about whether the market is preferred for such insurance or government (we know what the response is in this forum!), but the idea of insuring against outliving your savings surely seems relatively uncontroversial, no?

As for #1, Social Security was not intended to be "welfare", but a socialized insurance system. As an "earned" benefit, you had to work first in order to receive it. So to essentially argue that it is a poorly designed welfare system is kind of forcing a questionable, even if popular, interpretation of the program.

Bob Murphy writes:

Hi David,

Thanks for linking to that article, if only because that 1985 NYT essay on laptops was hilarious...

Regarding Williamson's point, I agree he didn't justify it, but I assume what he meant was something like, "Back then, the ratio of workers to retirees was X, and that made the model workable. But today, the ratio is Y, which makes no sense at all."

Of course, your objections are still relevant and arguably trump his point (assuming I am right about what his point *was*), but I imagine he would think you weren't really addressing his point.

JK Brown writes:

From Life magazine: Social Security article, 1939

The joker in this is that the Government has been spending Social Security tax money for ordinary expenses and putting its own I.O.U.'s (i.e., bonds) in the reserve fund. Thus when the time came to pay old-age annuities partly out of the interest on the bonds the money could be raised only by taxing the people a second time.

When President Roosevelt signed the Social Security Act just four years ago this month, he hailed it truly as a milestone in American history. Nobody, however, regarded the Act as much more than a long first step toward its objectives, a tentative plan to be revised and expanded with experience.

The scam started before the first benefit payment, as well as the "theft" of the reserve fund, "trust fund" or whatever they call it in the future.

Greg G writes:

If you hate government enough, its successes will bother you even more than its failures. Social Security was designed to sharply reduce poverty among the elderly (over time, not right away) and to maintain its political support. It has been a huge success at both and is one of the most popular programs in the history of government (present company excepted).

For 80 years, the opponents of Social Security have been predicting it will lead to our economic downfall but we continue to grow and enjoy a level of economic success unimagined in 1935 by either its supporters or opponents.

>---"The essence of a tax-financed Ponzi scheme is that those who get current benefits are paid from the taxes of those who are currently working. For the scheme to work, it has to exist forever. Otherwise there will be a generation or two who pay taxes and get nothing in return."

The essence of a Ponzi scheme is that its financing is deceptive and non-transparent. Social Security is financed in a way that is transparent to anyone who wants to learn about it.

Many people today "get current benefits (which are) paid from the taxes of those currently working" when it comes to any federal government spending you want to pick. Why don't we ever hear this Ponzi scheme criticism made of military spending or the federal courts? Because it has its own dedicated tax, Social Security comes closer to being paid for by its recipients than many other Federal programs.

David R. Henderson writes:

@Greg G,
For 80 years, the opponents of Social Security have been predicting it will lead to our economic downfall
Your “the” is way too general. I’m an opponent of Social Security and I’ve never claimed that it would lead to our economic downfall.
The essence of a Ponzi scheme is that its financing is deceptive and non-transparent. Social Security is financed in a way that is transparent to anyone who wants to learn about it.
Notice that you felt the need to put “to anyone who wants to learn about it.” That undercuts your point. People are still often told that there’s a trust fund there with real assets in it and not just special bonds that put a claim on the Treasury.

Greg G writes:

David,

I agree with both of those points. The first one is especially relevant because I am often the first to object when someone lumps everyone who feels the same way about some issue into too broad a generalization.

I don't feel that the point that people are "often told" misleading things about Social Security carries the same weight. People are "often told" all kinds of things. The facts are easily available here which is not the case in a real Ponzi Scheme.

And I don't see why it's so objectionable for the Social Security Trust Fund to buy Treasury Bonds. If I plan for my retirement by buying Treasuries in my IRA or SEP all economists will say I am being prudent. If I plan for my retirement by buying the exact same bonds through the Social Security Trust Fund suddenly people think that is reckless.

David R. Henderson writes:

@Greg G,
I don't feel that the point that people are "often told" misleading things about Social Security carries the same weight. People are "often told" all kinds of things. The facts are easily available here which is not the case in a real Ponzi Scheme.
You’re right to call me on my lack of precision. What I have in mind is that the people running the scheme made this part of their message for decades. I recall receiving notices from the Social Security administration telling me that the trust fund was in good shape and never mentioning that it depended on taxing future generations. As an economist, I knew better but many bright people around me didn’t. You could argue that they should know better and would know better if they inquired. All true. But the same could be said of Madoff’s scheme.
Re your Treasury Bonds point, it’s individually rational for you to buy Treasuries, although I would be careful if I were you because of the risk of default. But that’s a separate issue from basing the biggest pension scheme in the world on Treasuries. Also, you CHOSE to buy Treasuries.

Greg G writes:

David,

I just don't see why it's such a scandal that Social Security in particular relies on taxing future generations. Isn't that true of ALL deficit spending? Why not make this criticism of military spending? Much MORE so than other Federal spending, the bulk of Social Security is paid for by a dedicated tax paid by the recipients.

For all of human history, people have been paying for the raising of their children through current income while hoping their children would help fund their retirements. This is just a more successful and reliable (not perfectly reliable, just MORE reliable) version of that.

David R. Henderson writes:

@Greg G,
I just don't see why it's such a scandal that Social Security in particular relies on taxing future generations. Isn't that true of ALL deficit spending?
Yes, it is true of all deficit spending. I’m critical of deficit spending too. Check out my piece for Mercatus on Canada’s Budget Triumph.
Why not make this criticism of military spending?
I think it’s a good criticism. Those who favor deficit spending to finance wars could argue, and have argued, as Robert Barro did, that the wars should be paid for by future generations since they benefit from the protection that war creates. I don’t buy it because I think the vast majority of wars create future problems rather than solving them.
For all of human history, people have been paying for the raising of their children through current income while hoping their children would help fund their retirements. This is just a more successful and reliable (not perfectly reliable, just MORE reliable) version of that.
It’s profoundly unfair because it also supports the childless.

ThomasH writes:

I think Henderson has very good points about how SS could have been better designed.

Cohort benefits could have been tied to cohort tax receipts (plus interest).

Alternatively benefits could have been funded from general revenues or some other non-wage tax.

The age at which benefits start being collected should have been adjusted as the age at which people become less able to work increased.

Perhaps a system of individual retirement accounts, subsidized for low income earners could have been instituted with general revenues financing of benefits paid to those who have not had a lifetime of contributions.

Whether Social Security made sense in 1935 or makes sense today, one must look at alternatives.

ThomasH writes:

A technical quibble: A "Ponzi scheme" works only by continued expansion. A Social Security type system with no changes in the tax or benefit rules has an equilibrium state if demographics are right and if the demographics are not right the rules can be changed to achieve equilibrium.

Greg G writes:

ThomasH,

>---"Alternatively benefits could have been funded from general revenues or some other non-wage tax."

There are tradeoffs here. A wage tax, especially one an employer pays part of, may be a disincentive to employment. But basing the system on a wage tax does require people to pay a higher percentage of the benefits they actually consume than is the case with most other government programs.

>---"The age at which benefits start being collected should have been adjusted as the age at which people become less able to work increased."

This has already happened and could and should happen again.

>---"Perhaps a system of individual retirement accounts, subsidized for low income earners could have been instituted with general revenues financing of benefits paid to those who have not had a lifetime of contributions."

The private investment markets are, and should be, systems where resources are transferred from those who are bad at investing to those who are good at investing.

That is a good way to decide which ventures get funded but a poor way to provide for the retirements of the majority who are not good at investing.

ThomasH writes:
People are still often told that there’s a trust fund there with real assets in it and not just special bonds that put a claim on the Treasury.

I'm not sure in what sense people are "told" that and I agree that there is a popular misunderstanding about passing costs and benefits to "future generations," but even so, why imply that US government bonds are any less real that any other kind of financial asset? Would the program be fundamentally different if the Treasury took SS taxes and bought bonds or stocks, put them in a "fund" and payed benefits "from" the "fund?"

The essence of SS is a set of taxes on wages and rules about who can receive benefits when. Any effort to improve these taxes and rules will, of course have to take account of people's expectations.

Greg G writes:

David,

>---"It’s profoundly unfair because it also supports the childless."

I'm not sure I understand what you mean by this. Does it follow that Social Security would be fairer if it didn't support the childless?

Is it that the childless have some unfair advantage because they have not produced children to help with the future burden on the next generation?

David R. Henderson writes:

@Greg G,
Is it that the childless have some unfair advantage because they have not produced children to help with the future burden on the next generation?
Yes. Look back at your own reasoning.

Greg G writes:

David,

Thanks for the clarification.

By your logic then, it is even MORE of an outrage against those of us with children that the childless benefit from the many other government programs which are funded by borrowing but have no dedicated tax to fund them. Social Security should be seen as less objectionable than these because a higher percentage of it is paid by the actual beneficiaries.

David R. Henderson writes:

@Greg G,
Thanks for the clarification.
You’re welcome, Greg.
By your logic then, it is even MORE of an outrage against those of us with children that the childless benefit from the many other government programs which are funded by borrowing but have no dedicated tax to fund them. Social Security should be seen as less objectionable than these because a higher percentage of it is paid by the actual beneficiaries.
Absolutely.

Greg G writes:

David,

We have some disagreements but I really appreciate how fair minded and consistent you are when you respond to comments. I'm still a bit amazed that the internet has evolved in a way that a total layman like myself can enter into a conversation with a semi-famous economist like yourself.

And all at no cost to me. Those of us who enjoy your blog probably don't thank you enough for all the work you do to maintain it.

David R. Henderson writes:

@Greg G,
Thanks so much. Comments like these help make it worthwhile.

RPLong writes:

JLV - Be careful when using life expectancy as an explanatory variable. Most of the gains in life expectancy over the last 100 years have been due to decreased infant mortality, not great longevity of those who reach adulthood.

bobroberts170 writes:

@Greg G

Hear, hear!

David R. Henderson writes:

@RPLong,
Thanks, but I think your message to JLV is JLV’s message to me. I should have pointed out what he pointed out.
@bobroberts170,
Thanks.

RPLong writes:

Prof. Henderson - Hmm, perhaps you're right about JLV's comment, but now I'm confused. A substantial increase in overall life expectancy, such as we've seen since the 1930s, would certainly be expected to strain Social Security's feasibility, even (or especially) if it is comprised mostly of reduced infant mortality. But a 1-5 year discrepancy in seniors' life expectancy at different wealth brackets over the same period would potentially be much less significant to the matter. (The latter extends a commitment to a static pool of individuals while the former increases the size of the pool.)

So while your inclusion of life expectancy seems appropriate to me, now I don't think I'm sure what JLV's comment intended to point out. Am I in vehement agreement with someone again? :)

David R. Henderson writes:

@RPLong,
A substantial increase in overall life expectancy, such as we've seen since the 1930s, would certainly be expected to strain Social Security's feasibility, even (or especially) if it is comprised mostly of reduced infant mortality.
Yes, it would strain it but only a little. The reason is those babies that don’t die grow up to be workers who pay payroll taxes for approximately 40 to 45 years before they collect the benefits. The increase in life expectancy that can bankrupt the system more easily is one at the upper end: people who would have died at, say, age 80 but now die at age 83.
I understood JLV to be pointing that out.

RPLong writes:

Aha, I've got it now. That is a good point. Thanks!

AS writes:

There just needs to be an "opt-out" option. Let people who like SS pay into it and receive benefits from others who paid into it. Let people who don't like SS, for whatever reason, opt-out and plan their own finances. We don't need to get bogged down into debates about particular details of SS and whether it's fair or efficient or sustainable. If you focus on high level rights like the right to "exit" or say "no", then the issue is quite simple.

And if you like the idea of SS but you don't like the government's particular implementation of it, you could create a private SS company that structures its own plan. There's no reason we all have to be pooled into the same top-down plan.

Comments for this entry have been closed
Return to top