Arnold Kling

Social Security Myths

PRINT
Health Care... Economists' Voice on Fiscal Po...

My latest essay on Social Security says,


The transition cost myth is one of four Social Security misconceptions that I want to address here. The others are the misconception that Social Security is a pension, the misconception that the Baby Boom is the main problem with Social Security, and the misconception that Medicare is in worse shape than Social Security.

...One way to eliminate the "transition cost" to partial privatization would be to first undertake a transition to better accounting. If the government were to put future Social Security obligations on its balance sheet as debt, then the accounting would be accurate. To borrow a locution from Warren Buffet, if promises to make Social Security payments are not a financial obligation of the government, then what are they? And if a financial obligation of the government is not debt, than what is it?

If unfunded liabilities to make future Social Security payments were counted as debt, then partial privatization would be nothing but a debt swap. The government would increase ordinary debt and reduce unfunded-liability debt by an equal amount. The transition cost would be zero.


Another recent post on Social Security is by Lancelot Finn.

People in the Soviet Union, Argentina and dozens of other countries worldwide have watched their governments have to break financially unsustainable promises. When the ground was washed out from under their feet, they had a long way to fall. Current trends suggest that will happen to us. If we want to stay afloat, we need to build the ship of Social Security private accounts and set out on the sea of the market.

UPDATE: See also this article by Phillip Longman.

For Discussion. Politically, raising the retirement age seems to be a sort of "third rail." Actually, you can retire at any age you want. Should we instead be talking about the "dependency age," meaning the age at which you become a government dependent?


Comments and Sharing


CATEGORIES: Social Security



TRACKBACKS (9 to date)
TrackBack URL: http://econlog.econlib.org/mt/mt-tb.cgi/150
The author at The New Libertarian in a related article titled TCS: Tech Central Station - Four Myths About Social Security writes:
    Arnold Kling's latest Four Myths About Social Security essay on Tech Central Station is a great starting point for anyone starting to grapple with the Social Security privatization issue. Even if you have read quite a bit on this subject [Tracked on October 27, 2004 10:02 AM]
The author at Vox Baby in a related article titled How to Reform Social Security, Part II writes:
    Arnold Kling recently wrote about the need to reform Social Security on his blog and at TechCentralStation and concluded that, "The retirement age is the most logical and effective policy lever." I think he is exactly right ... [Tracked on October 31, 2004 9:27 PM]
The author at WillPate.org in a related article titled Carnival of the Capitalists writes:
    Thanks and welcome to my first hosting of the Carnival of the Capitalists - relatively politics free. I've made a deliberate effort to avoid posts that were mostly based on politics. It may be November 1st on an election year,... [Tracked on October 31, 2004 10:53 PM]
COMMENTS (22 to date)
Jason Ligon writes:

"Should we instead be talking about the "dependency age," meaning the age at which you become a government dependent?"

YES!

rvman writes:

That brings us to my proposal - eliminate the retirement age entirely, and just have 65 and ups qualify the same way the rest of us do, by disability. An 80 y/o who can work shouldn't be on the government dole - if he wants retirement, he can save for his own. A 40 y/o who has been permanently blinded - why not pay him instead?

cb writes:

A couple of comments on your full article -

You mentioned creating debt accounts on the balance sheet of the gov't. It should be noted that the governments books are operated on a cash basis. From an economic standpoint I agree with you, but on a cash basis, the gov't deficit will increase by exactly the amount that is channeled to private accounts (assuming nothing else changes, ie the tax rates are the same). There will be an increase in demand for securities that private citizens want in their accounts, and an increase in supply of treasury securities. From the gov't's point of view, there is a transition gap, cash that was going into the Treasuries accounts will now be directed to the financial markets. It's not an economic cost perhaps, but either way you cut it the gov't is not getting the money, therefore the deficit goes up.

Regarding the Social Security and Medicare comparison you made, it's interesting to note that payments made by companies to the Treasury are made as one cash transfer. I'm sure the government keeps track of the breakdown, but I agree, the two programs go hand in hand. Together, they are 15 point something percent (up to a certain amount). Personally, whatever is considered SS and whatever is 'allocated' to Medicare, in my opinion, is arbitrary.

Lawrance George Lux writes:

Arnold,
Your Accounting argument lacks validity, because of the basic use of SS taxes over time. The Federal government simply spent the revenues. Such expenditure should be considered true liability, just like the National Debt. Social Security revenues should have been placed in a National Bank forum from the start, and assets used as lending capital. This would have led to more valid Government expenditures, and left the SS system secure. You, like the Government, pretend there is an excusable liability.

I personally like ruman's suggestion of relying solely on disability for benefits receipt. This alone could establish the financial viability of the SS system. Medicare will never be secure, until a limitation of benefits is derived. lgl

Bernard Yomtov writes:

I have many reservations about the privatization of Social Security. One has to do with the transition.

Let's accept, for the moment, Arnold's argument that there is no "transition cost," that we are merely replacing one liability with another of equal size.

We still need to meet the liability. As things stand we have a tax - the payroll tax - earmarked for this purpose. (Let's not divert into a silly argument about the trust fund, please). If we do away with this asset - possibly dedicating payroll taxes to private accounts in some fashion - we will still have this new liability to pay. How do we propose to pay it? Do the privatizers advocate a tax increase, or possibly a new earmarked tax, for this purpose? Or do we simply run up the deficit some more?

For all the claims of "no transition cost" I have not heard this issue addressed.

More deeply, if promised Social Security benefits are an "off-balance sheet" liability that will simply be replaced with an on-balance sheet liability, isn't the stream of payroll taxes an "off-balance sheet" asset? What will replace it?

rvman writes:

Bernard: We will have to pay this cost either way. If we privatize, we start paying it out sooner, but we start getting the returns on the excess savings this will encourage as well. If we don't privatize, we will just have to pay it out later, anyway, by raising the payroll tax or income tax or whatever.

Visualize the following (hypothetical numbers): current cost of Medicare/SS is 100. Assume no gains from interest on accounts, but also neglect time cost of money (they partly offset one another) The time series looks roughly like:

now: 100, 10 years: 120, 20 years: 140, 30 years, 160

Privatization looks like:
now: 120 , 10 years: 125, 20 years: 130, 30 years 145.

The extra money early is going to accounts to pay off the later years, kind of like the "Social Security Trust Fund" is allegedly doing, only at the end of the day, the assets in these private accounts will actually exist - putting assets in private hands will actually decrease future federal liabilities. (I am of the "The trust fund is just a way of hiding debt and is actually empty" school - the trust fund is full of federal bonds, when the SS comes to collect, what will happen has to be that the Feds will have to incur debt or raise taxes to pay SS just like they would if the TF didn't exist, but that is irrelevant here.)

So at worst you are equally well off as before - at the end of 30 years, the total cost is at worst the same. Preferably, the accounts are getting a return better than the time cost of money and are generating increased savings, and so the 30 year number should look more like 135 or 140 and privatization is long-run cheaper. It is just that you are BOOKING the cost earlier, to avoid a harder hit later. It LOOKS expensive, but in actuality the price is the same, or less.

What we are doing now is the equivalent of going to Best Buy and taking the "Buy now, pay nothing until 2010" option. As we all know, those options come with a price tag.

Lancelot Finn writes:

Lancelot Finn here. Thanks for the link, Arnold. You're one of TCS's solidest writers. I'll post links to this on my blog tomorrow; tonight blogger is slow.

Bernard Yomtov writes:

rvman,

Thanks. I do understand Arnold's argument that the promised benefits have to be paid one way or the other. What I am asking is this: If you put what are now payroll taxes into private accounts for workers, where do you get the money to meet the benefit obligations? If you're going to borrow it, where do you get the money to repay the loan?

As far as I can tell the privatizers have not seriously addressed this issue.

rvman writes:

And what I am saying is that that same question exists now, and is just as unanswered. My answer is "general revenues and debt". If we privatize correctly, the total cost in combined forced savings and taxes by 2040 should be lower than the cost of our current program. We WILL have to pay the taxes to cover SS/Medicare under the current program. We can just be in denial longer under the current program, and than we'll take a harder tax hit when "the day" comes. Privatization forces us to think about financing at a much earlier date, which should allow us to smooth the burden. (I prefer to eliminate payroll SS/Medicare taxes except for personal savings accounts, and fund the rump SS/Medicare out of general revenues - the Payroll tax is among the worst we have, with no deduction and a cap it is regressive, it deters work, it deters hiring, and creates extra paperwork for no benefit.)

Boonton writes:
Social Security revenues should have been placed in a National Bank forum from the start, and assets used as lending capital. This would have led to more valid Government expenditures, and left the SS system secure. You, like the Government, pretend there is an excusable liability. Thanks. I do understand Arnold's argument that the promised benefits have to be paid one way or the other. What I am asking is this: If you put what are now payroll taxes into private accounts for workers, where do you get the money to meet the benefit obligations? If you're going to borrow it, where do you get the money to repay the loan?

As far as I can tell the privatizers have not seriously addressed this issue.

From what I understand the SSI accounts will be 'frozen'. In other words, if privitization comes tomorrow I will get an amount equal not to the present value of my SSI benefits but only what the present value of my SSI benefits would be if I stopped working today (I'm 31). Net effect would be to freeze the SSI liability as it is right now and not let it grow any more.

Social Security revenues should have been placed in a National Bank forum from the start, and assets used as lending capital. This would have led to more valid Government expenditures, and left the SS system secure. You, like the Government, pretend there is an excusable liability.

For all intents and purposes they were. By decreasing gov't borrowing the SSI surplus was in fact 'invested'. Interests costs to the taxpayers are lower today because SSI surpluses decreased debt in the past. This is a return, believe it or not.

Lawrance George Lux writes:

Boonton,
I may misunderstand you, but the fact the Federal government assumed a general liability, then spent the SS tax revenues, has presented the threat to the SS system and to the Retirees. The Government expending these revenues led to excessive Government spending, as the cost could be hidden, and forestalled confrontatiion of the Issue earlier. It will continue to be the case until the SS system Fund is separated from the general Federal budget. I do not favor private accounts, though, as they are a general avoidance of the total liability by those able to pay. The Social Security Fund should be set up on lines like Fannie Mae, or Freddie Mac. The Government could borrow the revenue, but only by short-term bank notes repaid from actual tax revenue. lgl

Bernard Yomtov writes:

And what I am saying is that that same question exists now, and is just as unanswered. My answer is "general revenues and debt". If we privatize correctly, the total cost in combined forced savings and taxes by 2040 should be lower than the cost of our current program. We WILL have to pay the taxes to cover SS/Medicare under the current program.

It is not unanswered. Current benefits are paid out of payroll taxes. If there is no payroll tax a substitute must be found. I don't understand your use of the word "we." Are you referring to the government, or to individuals?

It's true that at some point the trust fund (yes, there is one) will begin to cash in its bonds, and taxes will have to be raised to do that, but that problem would exist if there were no Social Security at all. My point is that for now at least we have the payroll tax available and it is more than adequate to pay benefits. What happens when the tax is abruptly cancelled, or redirected?

Boonton writes:
I may misunderstand you, but the fact the Federal government assumed a general liability, then spent the SS tax revenues, has presented the threat to the SS system and to the Retirees. The Government expending these revenues led to excessive Government spending, as the cost could be hidden, and forestalled confrontatiion of the Issue earlier.

It doesn't follow that the SSI surplus cause gov't spending to be higher than it otherwise would have been. Except for Nixon's increase in SSI benefits I don't think that is the case. What exactly is the mechanism that causes gov't spending to go up with the SSI trust fund surplus? How 'hidden' can the costs really be when the bond market is played by very well educated bond traders who are paid quite handsomely to worry day and night about what the US gov't's true credit risk is?

rvman writes:

"Current benefits are paid out of payroll taxes"

Well, sort of, yes. Money is fungible - the only numbers that truly matter to our long run debt level are "How much present-value revenue does the goverment collect from all sources?" and "How much (present-value) liability does the government incur in all programs?" There is no serious incentive link connecting payroll collections to current benefit payouts, so the dollar that went out to grandma has a heritage which could be from payroll, but it could be from income taxes, tariffs, whatever - it just doesn't matter. Privatization reduces the first number, but decreases the second number by more. It does also have a timing issue - a lot of the reduction of revenue is front loaded, and the reduction of liability is back-loaded to later liabilities. So yes, the government will have to beg, borrow or steal more in the short run to cover that. But on net, the TRUE government net liability - the present value of all future liabilities minus the present value of all future revenues - will be smaller with private accounts, because we will be taking advantage of compounding interest and all that.

What we are really discussing is a "writ large" version of the decisions we all face - do we start saving when we are young with relatively low income, or do we wait until our income increases, and it is "easier" to save? The responsible decision is to start saving as early as possible, but because it involves sacrificing the big screen TV and the 10,000 cable system, people don't do it. The responsible decision by government would be to stimulate savings by giving people control over their own SS, but it is costly in political terms, today. So I don't expect them to do it. Government rarely, if ever, does the right thing, after all.

As for the trust fund - it is just a tax increase with a funny name. They couldn't just increase the payroll tax as the extra costs come because that was irresponsible. They didn't want to say "we are going to increase the payroll tax and take the extra money into general revenues so the debt is smaller when the crunch comes" because that was impolitic, and would raise the question of "why not cut spending in other areas?", and they didn't want to say "we will pay for the SS deficit when it comes out of general revenue" because that destroys the fiction that SS is a pension, not a welfare program. So they created this fiction - the Social Security Trust Fund. The below are two scenarios:

Revenues from Payroll Tax = X1
Revenues from all other sources = X2

SS costs = Y1
Other costs = Y2

Contents of Trust fund =Z


Without trust fund: Deficit is total liabilities minus total revenues, that is Y1+Y2 - X1 - X2.

With trust fund: Trust fund balance Z = X1-Y1

"Other" deficit is Y2-X2. Government takes Z and applies against Y2-X2, replacing it with "bonds".

Deficit remaining is Y2-X2 - (X1-Y1) which is

Y1 + Y2 - X1 - X2, the same as without a trust fund.

the trust fund contains Z, government assets, in the form Z in bonds, which are government liabilities. Net contents, from the perspective of government = 0.

In other words, the trust fund was an accounting fiction to "cover up" a tax increase. They could have just as easily raised a less regressive tax, but they didn't - they knew people didn't want a government which ate that much of their incomes, so the creators of the Trust pretend they are keeping "our" money on deposit, while really just raising taxes.

And we wonder where Enron got their ideas...

Lawrance George Lux writes:

Boonton,
Your argument lacks some validity, as you assume the bond market and American taxpayers would not have responded to a total increase of national debt, if the SS tax revenues had been separated. Does Anyone know the exact liability owed from those expended revenues? I have heard some wild estimates, and some which are too low to be even marginally practical. It seems these liabilities should be in excess of $1 trillion, but less than $3 trillion. lgl

Boonton writes:

Aren't you contradicting yourself? If people generally save less than what they should then the problem doesn't seem to be lack of control but the need for control. Paternalism isn't always bad...unless you take a strong libertarian position.

But on net, the TRUE government net liability - the present value of all future liabilities minus the present value of all future revenues - will be smaller with private accounts, because we will be taking advantage of compounding interest and all that.

Aye but if the SSI Trust fund reduced the gov'ts debt then we are already benefiting from compound interest.

In other words, the trust fund was an accounting fiction to "cover up" a tax increase. They could have just as easily raised a less regressive tax, but they didn't - they knew people didn't want a government which ate that much of their incomes, so the creators of the Trust pretend they are keeping "our" money on deposit, while really just raising taxes.

In other words, gov't saved the money since running a surplus (or decreasing a deficit) is by definition increasing savings.

Your argument lacks some validity, as you assume the bond market and American taxpayers would not have responded to a total increase of national debt, if the SS tax revenues had been separated. Does Anyone know the exact liability owed from those expended revenues? I have heard some wild estimates, and some which are too low to be even marginally practical. It seems these liabilities should be in excess of $1 trillion, but less than $3 trillion. lgl

Lawrence, you've presented an argument that the SSI trust fund caused gov't spending to increase because it made gov't's deficits 'seem' smaller. This is an 'ATM' view of how gov't behaves. Imagine you go to the ATM and take out $20...the receipt shows you have $100 more than you thought so you say what the hell and take out another $60 not realizing the ATM is a day behind. That's not how gov't spending decisions are made IMO. IN fact, I suspect that surplus's actually end up constraining spending while deficits encourage it. After all, if you're already maxing out the credit card why not order desert with dinner?!

But to your other point, I'm not assuming perfect information or perfectly rational actors in the bond market but it is rather silly to think a simple account 'trick' could fool the entire bond market. If this is the case then you should be able to make a fortune shorting T-bonds because the market has been 'tricked' into thinking the gov't is at less credit risk than it really is.

Lawrance George Lux writes:

Boonton,
Turning the SS Fund into a financial institution in its own right, is not tricking the Bond market. It is obtaining a risk Banker who will not accept another IOU in liew of cash payment.

There is no way We can have gained from Compound interest in spending the SS Fund revenues. That is simply saying We did not accumulate more Debt, because We spent our Capital.

Do not think the Bond market would not have restricted Spending. The domestic market has long since given up funding the Federal Government freely and graciously, and foreign central banks are getting Dollar-poor trying to prop up Our currency in the face of unrestrained American importation. We could actually spend $1 trillion less next year, Y2006, simply because no one will lend the money. lgl

Boonton writes:

Lawrence,

Assume for the sake of the argument that gov't spending would have remained exactly the same regardless of how the Trust Fund was set up. By definition, if gov't borrowed less in...say...1978...because the surplus was used to partially cover the deficit then every year since 1978 the taxpayers have enjoyed less interests expenses than they would have otherwise had to incur. That's your compound interest.

The distinction between domestic versus foreign lenders is a distraction in this discussion. In both cases lenders are putting their money at stake, there is no reason to think they would be fooled by a simple accounting trick into thinking that the gov't's debt was any different than what it really is.

Ken writes:

"Aren't you contradicting yourself? If people generally save less than what they should then the problem doesn't seem to be lack of control but the need for control. Paternalism isn't always bad...unless you take a strong libertarian position. "

That's assuming that people "generally save less than they should" because of a lack of control, as opposed to other reasons, such as a 15% reduction in take-home pay, the promise of future money reducing the need to save, and so on.

Let them keep more of their income, and promise them less future money, and they'll be a lot more inclined to save.

Boonton writes:
Let them keep more of their income, and promise them less future money, and they'll be a lot more inclined to save.

1. Money put into 401K's or IRA's reduce your tax bite.

2. Money put into Roth IRA's eliminate taxes on capital gains and other earnings on that money.

3. Capital gains are not taxed until realized as opposed to being taxed for paper gains (or credited for losses) thereby making it cheap to 'buy and hold' (i.e. save).

4. There's a slew of tax free savings plans for medical care, sending your kids to college and so on.

With all these incentives, why hasn't the savings rate gone up?!

Ken writes:

"With all these incentives, why hasn't the savings rate gone up?!"

Gone up from where? And how much free money has been added in the meantime?

Get a clue writes:

The SS issue is a huge problem that accounting will not fix. I've already decided not to count on SS being available when I retire, even though I've contributed tens of thousands to the program over the years.

This is the issue - When SS started, 42 people contributed to the SS check of 1 senior. Today it's less than 4. By the time I reach retirement, it will be 1 or less than 1. - How can any senior citizen expect that a single person making an average salary contribute enough $$ to support one senior until death - much less 1.1 seniors? It's not economically feasible under any circumstances. The senior will either suffer because SS benefits will be cut drastically, or the worker will suffer under the burden of taxes. I guess we should not be borrowing from Peter to pay Paul, huh?

The second problem that hits us is the fact that as a society we are living longer but not dying very quickly. Our life-saving strategies (curse the Hippocratic oath) keep people living years past when an illness might previously have killed the senior. Instead of having a life expectancy of 67 (read this to be a 2 year SS income) as we did when SS started, we now have a life expectancy of 78 - a 16 year SS income, and this is only getting longer. So the people who are contributing to SS are expected to pay more $$ for longer periods of time because the numbers of senior citizens are increasing while the numbers of workers are not.

Fixing the SS was partly on politicians' minds when they gave illegal aliens a free ticket to our country. They figured that would increase the tax base. They didn't figure on two things - that the new citizens would continue finding jobs that pay "under the table" and so wouldn't contribute to SS, that the new citizens would shop for products mainly at garage sales - thus eschewing state sales taxes, and that the new citizens would continue to send their cash out of the country to their families. (Keep in mind - this is a general picture - not everyone has held to this, but a lot of them do.) So, the ploy failed at upping the tax base and helping out SS.

I'd say we levy a national sales tax of 1% and levy a special tax on anything tourists do (rental cars, hotels, state parks, etc.) and use the income to pump up SS. (Why not levy against tourists? Just go to Italy and see the enormous taxes you have to pay on renting a car. They use it for their citizens, why shouldn't we?)

Medicare is a health care insurance program that has not kept pace with health care costs. It's too bad that we have had no caps on lawsuits against doctors, which could have reduced E&O insurance premiums and in turn not required those costs to be passed through to patients. Do you know that Medicare pays about $38 to a doctor or hospital if a nurse cuts that patient's toenails? You can get an entire pedicure for about half that. It's really too bad that we've set up a system where there are multiple middlemen between the pharmaceutical companies and the patients. We've got to peel back those layers in order to reduce the costs of medicine. We could also have the government create a program for young people to apply to become doctors - pay for their education as long as they spend 15 or 20 years as a Medicare doctor with a controlled salary, a promise of a decent pension and experience to go into private practice after the fact. We could insist that Medicare patients use only Medicare doctors, if they do go out of the Medicare network, then they would be penalized and required to pay for a much larger portion of the medical costs. Why not? That's what private health insurance does. It's how you control your spending. It's really the only way to do so.

Comments for this entry have been closed
Return to top