Playing the Social Security Trust Fund Shell Game

There’s a simple way to calculate the size of the federal government’s debt at any point in the future:

D’ = D – R + S – T


D’ = Amount of debt at a future date

D = Present debt

R = Federal government’s revenues from all sources (including Social Security taxes), from the present to the future date

S = Federal government’s spending for all purposes (including SS benefits), from the present to the future date

T = Value of Treasury securities redeemed by the SS trust fund to defray the gap between SS taxes collected and SS benefits paid

(For an explanation of how the redemption of securities by the trust fund can reduce the debt, see below.)

The present level of debt (D) is approximately equal to the debt ceiling. Therefore, as long as the deficit (R – S) is greater than the value of securities redeemed by the trust fund (T) to pay current benefits, the debt ceiling must rise or spending must be cut. (Aside: The ability of SS trustees to redeem trust fund holdings and pay benefits is simply a mechanism for ensuring the payment of full benefits until the trust fund is exhausted, regardless of any budget crunch. The trust fund, itself, is nothing more than a set of numbers in a government ledger. It isn’t an asset, any more than swampland is an asset to a sucker who buys it sight unseen.)

In theory, the trust fund could be exploited to get around the ceiling, by redeeming more holdings than required for the payment of current benefits. It’s a ploy was used in the past but is now illegal, Michael McConnell explained it in 2011:

The Social Security Trust Fund holds over $2 trillion [now over $2.7 trillion] in special Treasury securities, which it is legally entitled to redeem when necessary for the payment of benefits. When the Treasury redeems those bonds, the public debt will correspondingly be reduced, which will enable it to auction new bonds to investors, without violating the debt ceiling. This is precisely what happened during the debt ceiling crisis in 1985. Then, it was a Democratic House of Representatives that refused to raise the ceiling at the behest of a Republican President (an episode conveniently forgotten by those who wish to paint the Republican House today as uniquely evil for insisting that a debt ceiling increase be accompanied by spending reductions). The Social Security trustees cashed in some $9 billion in special Treasury securities for the payment of benefits, and the Treasury auctioned off the same amount in new U.S. bonds, without violating the debt ceiling. Here is how the Comptroller General described the event:

The Treasury Department estimated that it would have insufficient cash on November 1 to pay social security benefits and other government obligations. In order for these payments to be made, the Treasury needed to borrow money from the public, and in order to borrow the money, Treasury had to reduce its outstanding debt below the statutory limit. Therefore, on November 1 the Secretary redeemed $9.613 billion of the Trust Funds’ long-term securities, and $1.9 billion of securities held by certain other government-managed trust funds, to permit public borrowing of about $13 billion.

In this way, the Reagan Treasury was able to continue to pay Social Security benefits without interruption, despite the failure of Congress to raise the debt ceiling at the time.

The Comptroller General ruled that these redemptions were lawful, except that the trust fund redeemed more securities than were actually necessary for the payment of benefits. Some years later, Congress passed a statute codifying the Comptroller General’s decision. Public Law 104-121, section 107(a), prohibits redemption of special securities held by Social Security prior to maturity for any purpose other than the payment of benefits or administrative expenses. This statute is significant for two reasons. First, it confirms that the trustees have authority to redeem the special securities prior to maturity for the payment of benefits, and second, it prevents the executive from using the trust fund as a massive kitty to avoid the effect of the debt ceiling.

What could happen if the law were repealed? This:

1. Given the estimated size of the trust fund at the end of 2014 (as reported here), trust fund holdings could be liquidated as follows: FY 2015 — $469 billion; FY 2016 — $536 billion; FY 2017 — $576 billion; FY 2018 — $627 billion; FY 2019 — $722 billion; FY 2020 — $144 billion. The payout in FY 2020 would exhaust the trust fund. The total ($3.1 trillion) exceeds the estimated value of the trust fund at the end of 2014 ($2.8 trillion) because the trust fund would be credited with interest on its remaining holdings while those holdings were being drawn down.

2. The redemptions in 2015-2019 would entirely offset projected budget deficits for those years; the redemption in 2020 would offset about one-fifth of that year’s projected deficit. (See Table 1 here.) Thus it wouldn’t be necessary to reduce federal spending from currently projected levels until some time in 2020.

3. Under present law, however, depletion of the trust fund means that SS benefits must then be cut to a level that can be sustained by SS taxes. The cuts would be relatively small at first, but would grow steadily through the years. (Go here and compare the columns “non-interest income” and “cost” for the years 2020 and beyond.) For example, benefits in 2021 would have to be reduced to 90 percent of the level currently planned for that year; by 2030, benefits would have to be reduced to 80 percent of the planned level.

The good news — if the ploy could be executed — is that the Social Security crisis would be brought forward to the near future, instead of being deferred until 2033, when the trust fund is now expected to vanish. The undeniable urgency of the situation might compel Congress and the president to act — and perhaps to do something about the federal government’s entire fiscal mess — instead of continuing to kick the can down the road.

The bad news is that the federal government would run huge deficits for the next several years (at least). Programs that are unaffordable in the long run would be kept alive to acquire larger constituencies. Accordingly, it would be harder to curtail or kill them.

The real solution, of course, isn’t fiscal trickery; it’s fiscal responsibility. Let’s hope that 2017 brings with it a Congress and White House controlled by the non-RINO wing of the GOP.

*     *     *

Related posts:
Economics: A Survey (also here)
Why Are Interest Rates So Low?
Estimating the Rahn Curve: Or, How Government Spending Inhibits Economic Growth
America’s Financial Crisis Is Now
“Social Insurance” Isn’t Insurance — Nor Is Obamacare
The Keynesian Multiplier: Phony Math
The True Multiplier

The Social Security Trust Fund Is Not a “Get Out of Jail Free Card”


David Friedman, drawing on an op-ed by Thomas Saving, suggests that

Obama may have a $2.7 trillion dollar get out of jail free card, a way of spending that much additional money without exceeding the debt limit.

How does that work? Friedman explains:

Suppose no agreement is reached on raising the debt limit. Obama instructs the relevant people to spend the income from Social Security on the war in Afghanistan, bailouts, whatever he thinks needs money. He then instructs the Social Security system to cash in as many bonds as are required to meet its obligations to Social Security recipients, say $700 billion. He then instructs the treasury, since the national debt is now $700 billion below the debt limit, to borrow $700 billion. The net effect is that he has increased total expenditure, Social Security included, by $700 billion without exceeding the debt limit. The trust fund is currently at about $2.7 trillion, so he can do it for four more years.

Friedman’s scheme would work only if total federal receipts (including Social Security taxes) remain greater than or equal to total federal outlays (including SS benefits), from the point at which federal indebtedness hits the statutory ceiling. But that is not the situation.

Let us say, for the sake of argument, that the ceiling will be reached at the end of FY 2011. The president’s budget for FY 2012 shows total outlays of $3.729 trillion (including SS benefits of $0.761 T) and total revenues of $2.626 T (including SS taxes of $0.660 T). (See tables S-1 and S-3, here.) In other words, if Congress passes the president’s budget exactly as it stands, the debt ceiling must rise by $1.103 T ($3.729 T – $2.626 T) in FY 2012. And the SS trust fund, no matter how large it is, cannot alter the arithmetic.

Here is why. Suppose the feds spend all $0.660 T in SS taxes on things other than SS benefits (as they will, in effect). From an accounting standpoint, that reduces the non-SS deficit for FY 2012 from $1.001 T (non-SS spending less non-SS receipts) to $0.341 T. But the folks at SS are faced with a bill for $0.761 T in SS benefits. To pay the bill without having received a dime in SS taxes, the SS folks must go to the SS trust fund and grab U.S. treasury bonds with a value of $0.761 T, which they must then present to Tim Geithner for payment. Geithner thinks, “Who are these fools? Do they imagine that I’ve got that much unencumbered cash lying around, when I’m over my head in debt and sinking fast?” But being a good Obamanite, Geithner gives the SS folks their $0.761 T, and they go away happy.

If the analysis stops there, Friedman is correct. The treasury has just redeemed $0.761 T in bonds held by the SS trust fund, and the total debt of the federal government has magically dropped by $0.761 T. But the analysis cannot stop there, because the treasury does not have the $0.761 T in unencumbered cash. It must now borrow $0.761 T, to cover the redemption of the SS trust fund bonds, plus another $0.341 T, to cover the amount by which non-SS outlays exceed total receipts (including the SS taxes that it intercepts). With rounding, that comes to $1.103 T, which just happens to be the amount by which total federal outlays exceed total federal receipts.

Under what conditions would Friedman’s fix work? Here is a list (perhaps not an exhaustive one):

  • The debt ceiling will not be reached, given current projections of federal outlays and receipts (including SS benefits and taxes).
  • The debt ceiling has been reached but will not be exceeded, given current projections of federal outlays and receipts (including SS benefits and taxes).
  • The debt ceiling has been reached, but the surplus from non-SS programs will offset the deficit in SS accounts, or vice versa.

What about the SS trust fund? As long as the federal government is in debt by at least the face value of the SS trust fund, the trust fund has no real value. There is one (unlikely) saving condition, which is that the government’s net worth — represented by real assets — is equal to or greater than the face value of the trust fund. Such assets would have to be authorized for sale, by law, and would have to be valued at their quick-sale price on the open market. Given the reluctance with which Congress and federal agencies part with valuable assets (mainly land), it will be a cold day on the Equator before the SS trust fund is more than a valueless collection of accounting entries.

UPDATE (07/25/11)

The crux of my objection to Friedman’s scheme is found in the original post and my reply to his first comment; viz.:

If the analysis stops there, Friedman is correct. The treasury has just redeemed $0.761 T in bonds held by the SS trust fund, and the total debt of the federal government has magically dropped by $0.761 T. But the analysis cannot stop there, because the treasury does not have the $0.761 T in unencumbered cash….

*     *     *

3. This intra-governmental transaction does not affect the revenues that SS and non-SS receive from third parties.

4. Total spending by SS and non-SS must therefore equal their total receipts from third parties.

I ended my reply with this observation:

If you disagree with this analysis, then you and/or I must be making some assumptions (perhaps inadvertently) that remain hidden from view….

As it turns out, Friedman was making a hidden assumption that allows his scheme to work. That hidden assumption is revealed in a note appended to Friedman’s original post. The note was not there when I published this post, and I was unaware of it when I replied to Friedman’s first comment. In fact, I was unaware of it until late yesterday, when I revisited this post and Friedman’s after receiving his second comment. The note reads:

Some readers seem puzzled as to where the Treasury, in my story, is to find the $700 billion that it is to pay to the Social Security Administration, once the debt limit is reached. The answer is straightforward. With or without a debt limit, the federal government is continually collecting money and spending it. In my scenario, the government takes (say) $50 billion that it was supposed to pay as salary to federal employees, pays it to SSA instead. SSA cancels $50 billion in trust fund bonds. The national debt, which includes the debt owed by the federal government to the SSA, is now $50 billion below the limit, so the Treasury borrows $50 billion and pays out salaries to federal employees. Rinse and repeat as many times as necessary.

This is too clever by half. It requires exquisite timing on the part of the Treasury; otherwise, payrolls are not met, vendors are not paid, and existing debt is not serviced. In other words, the federal government would be in constructive default and violation of the debt limit. Moreover, it most certainly would not allow the federal government’s outlays to exceed its revenues over an extended period, which is why Obama seeks a higher debt limit in the first place. I could stop there, but there’s more to say about the scheme.

It resembles check-kiting, and may be just as illegal. But even if it is not illegal, it amounts to a patent evasion of the debt limit, and the evasion soon would be obvious to knowledgeable observers. Among other things, financial markets probably would react as if the federal government were in default — because the scheme could sooner or later result in a default of some kind (especially if outlays are rising as revenues stay flat). It would not take an act of Congress (over Obama’s veto) to put an end to the scheme; financial markets would do the job, as Treasury would be unable to refinance existing debt, except (possibly) at exorbitant interest rates.

In the best case, climbing interest payments would eat up revenues and force the federal government to cut back on the actual operations and programs. The result would be exactly opposite the one desired by Obama and company, which is real expansion of government. In the worst case, the Federal Reserve would pick up the tab, if it could scrape together a voting majority with the stomach for wading into a political firestorm. But that is another deus ex machina — of dubious durability — and not a surefire way of getting around the debt limit.

I am through with this subject. Comments are closed.

UPDATE (06/08/14)

I note, very belatedly, that Friedman later amended his post to add this:

A friend who knows much more law than I do writes:

It turns on, on further research, that Congress anticipated and prevented the very trick you have devised. Public Law 104-121, section 107(a), prohibits redemption of Social Security trust fund securities prior to maturity for any purpose other than the payment of benefits or administrative expenses.So it’s still true that the debt limit cannot block social security payments, at least until the trust fund runs out. But my multi-trillion dollar get out of jail free card has been cancelled.

Curses, foiled again.

Friedman later wrote a post that is properly focused on the ability of the federal government to continue paying SS benefits, regardless of the debt ceiling, as long as the trust fund is sufficiently large. The trustees expect the fund to be exhausted in 2033.