Trusting the Fund: Can we rely on the government's Social Security nest egg?

Update 13-Mar-2013: A newer and much improved post on this topic can be found here.  I recommend you read it instead.Trust me!

There's a certain camp in the Social Security debate that's convinced the massive and growing Social Security trust fund is utterly worthless. The payroll tax surplus, they say, has long been eating holes in the pockets of lawmakers, who have been unable to resist spending it. In return, the trust fund has received nothing but worthless government "IOU's". Can the trust fund doubters possibly be right?

A mere layman myself, with no formal training in the arcane arts of economics, accounting, or politics, I have nevertheless discovered that clear thought and simple arithmetic go a long way toward cutting through the thick fog of trust fund controversy. Let's think our way through it together.

For purposes of this discussion, I suggest four possible trust fund management scenarios. As we evaluate these four, I stipulate that total government spending be the same regardless of which scenario is chosen. As always, if government spending exceeds revenue, then the government must borrow the difference and thereby increase the national debt. And as always, government borrowing cannot exceed the "debt ceiling" authorized by Congress.

Scenario 1 - Payroll tax receipts go into the government's general fund, and the Treasury issues to the Social Security trust fund U.S. government bonds in the amount of any payroll tax surplus.

Scenario 2 - Surplus payroll tax receipts go directly to the Social Security trust fund. The trust fund then invests this surplus by buying U.S. government bonds in the financial markets.

Scenario 3 - Surplus payroll tax receipts go directly to the Social Security trust fund. The trust fund invests this surplus by buying non-U.S. government securities in the financial markets. Such securities could be equities, foreign government bonds, whatever.

Scenario 4 - Surplus payroll tax receipts go directly to the Social Security trust fund. Every night a government bureaucrat named Fred takes the day's receipts home and hides them under his mattress.

Scenario 1 is what actually happens. Payroll tax receipts intended for Social Security go to the government's general fund and are, as the critics claim, spent on all manner of government programs. Critics complain that since the compensating bonds issued to trust fund are merely transactions between accounts within the government (according to one description, they are "accounting devices"), they amount to little more than "IOU's" from the government to itself.

It is also true that the trust fund does not consist of real resources at the government's disposal to pay future Social Security claims. Remember: the money has already been spent. The trust fund bonds are claims against the government, but of course Social Security is itself a government program. What shall we make of this state of affairs? Has someone pulled a fast one? Has the payroll tax surplus been pilfered—spent and long gone, with nothing to show for it? That is what we hope to answer here.

Scenario 2 has the trust fund investing the surplus by active participation in the financial markets. The trust fund becomes a big buyer of government bonds. On the other hand, compared to Scenario 1 the government must sell a lot more bonds to the markets to finance its deficit spending. The two effects cancel exactly, and the final situation is just as with Scenario 1: the trust fund is "invested" in U.S. government bonds. Although the market transactions make it appear less—at least superficially—that the government is writing IOU's to itself, essentially nothing is different.

Scenario 3 has the trust fund invested in non-U.S. government securities. While economists debate on whether government participation in the (equity, for example) markets on such a massive scale is a good thing or a bad thing, at least the trust fund now consists of real resources that can be sold as needed to pay Social Security claims. Is this a big improvement? Consider:
First—just to give us some numbers to work with—we provide the 1998 CBO projection of U.S. government debt for the year 2008. All numbers are billions:

Table 1.	1998 CBO projected debt for 2008
Debt held by the public	3,251
Debt held by the Social Security trust funds	2,277
Debt held by other government accounts	1,617
Total federal debt	7,145

Since under Scenarios 3 and 4 the trust fund does not invest in government bonds, and since we have stipulated that government spending (and therefore borrowing) is the same under all scenarios, our Scenario 3 2008 debt projection would look like this:

Table 2.	Projected debt under Scenario 3
Debt held by the public	5,528
Debt held by other government accounts	1,617
Total federal debt	7,145

The total indebtedness of the government remains unchanged. The only difference is that a higher proportion of that debt is now held by the public, since the trust fund is out of the business of loaning money to the government.

Critics of the trust fund as it is now managed, such as Penn State accounting professor J. Edward Ketz, say that we "make the mistake of believing that Social Security has $2 trillion of assets without examining and realizing that these assets predominately consist of receivables from the general fund. These receivables aren't collectible unless additional taxes are imposed on the populace." Professor Ketz goes on to imply that these "additional taxes" are tantamount to paying for Social Security twice! What strange reasoning this is, coming from a highly educated expert in his field.

The first response to Professor Ketz is to say that the government must ultimately make good on all of the national debt, not just the part represented by the trust fund. And as we have seen in Tables 1 and 2, the size of that debt remains unchanged regardless of whether or not the trust fund holds government bonds. So in contrasting Scenarios 1 and 3, does it really make a difference whether taxes paid in 2020 (to pull a date out of the air) are "imposed on the populace" to pay for Social Security benefits instead of to pay for, say, the war in Iraq that we couldn't afford, and for which we had to borrow money in 2003, 2004, 2005, and beyond? In terms of the actual tax burden imposed on the public, it makes no difference at all.

The point is worth reiterating: Since the government is currently running deficits, the government must borrow money to conduct its affairs—to the tune of some $400 billion this year—and the Iraq war is just one prominent example of a government program financed by borrowing. Does it really matter, come 2020, whether that money was borrowed from the Social Security trust fund, or from John Q. Public? Or from the government of China, for that matter? No matter who the lender, the money must be repaid.

So when we first begin to tap the trust fund in 2018 or thereabouts, it seems logically just as correct—actually more so—to say we need to raise revenue to pay off war debts as to say we need it to pay Social Security benefits. As the war (and other) debts are paid, real assets will flow into the trust fund. Alas, Professor Ketz, the populace will pay for Social Security but once after all.

Simply put, what Professor Ketz has apparently just discovered is something a few of us have already noticed: the government has been borrowing money. A lot of it. And for a very long time. The mistake made by Professor Ketz and others is to reach the conclusion that that borrowing somehow constitutes a looting of Social Security, or that the government's general fiscal problems (which are significant) are specifically or exclusively problems with Social Security. In our most horrific nightmares we might imagine the government defaulting on its debt, but there is no logical reason to assume that the debt held by the trust fund is more at risk than the debt held by the public.

Let's take one more stab at Table 2. Yes, total government debt is the same as in Table 1, but now we have a Social Security trust fund with $2.3 trillion in real assets. Better, no? No. That's because Table 1's 2.3 trillion of "worthless IOU's" (ie., debt held by the trust fund) have in Table 2 been replaced by $2.3 trillion of debt to the public, and the public must be paid. As a mental exercise, let's pay the public. Let's sell our "real" assets in the trust fund, and pay off a corresponding amount of public debt. Since we still must ultimately pay Social Security benefits, we find ourselves back, once again, to the situation described by Table 1. That is, debt held by the public has been reduced, but since we still are obligated to pay Social Security benefits, debt held by the trust fund has increased. The two tables, it seems, are equivalent. That's just another way of saying that it doesn't much matter whether the government borrows from the trust fund or from the public. Either way, the debt must be repaid

Where does that leave us? It leaves us with a Social Security system that is funded far into the future. The trust fund is alive and well. Don't let anybody tell you otherwise.

---

Update Aug 16, 2010: An interesting assessment by New York Times columnist and Nobel Prize winning economist Paul Krugman.
Copyright (C) 2005 James Michael Brennan, All Rights Reserved