Or so claims a recent headline. But that’s not what I actually said, as I try to clarify below.
In a recent paper I examined off-balance-sheet liabilities of the U.S. federal government. I concluded that these came to $70 trillion as of 2012, or 6 times the size of net federal debt held by the public. These off-balance-sheet liabilities take the form of implicit and explicit guarantees and commitments, but should not be construed as comparable to debt obligations of the U.S. Treasury.
The first decision you have to make in deciding how much the government really “owes” is how to treat government trust funds such as Social Security. These represent sums that the U.S. Treasury has promised to pay to the trust funds. They are counted as an asset of the trust funds and a liability of the U.S. Treasury. One perspective is that these represent obligations that one part of the U.S. government– the Treasury– has promised to pay to another part– the Social Security Administration. If you view this as money that the federal government just owes to itself, you might choose to exclude it from your measure of government debt. Leaving out the sums that the Treasury owes to the various government trust funds leads to the measure known as the net federal debt, or debt owed to the public. This came to $11.3 trillion as of the end of fiscal year 2012.
On the other hand, some argue that these trust funds should not be viewed as money the government owes to itself, but instead represent an obligation of the Treasury to deliver real resources to current and future retirees. To meet these obligations the Treasury will need to raise taxes or cut spending, just as it will need to do to honor the explicit Treasury debt owed to the public. If this is your perspective, you might want to include the sums that the Treasury owes to the various government trust funds, and look at the gross debt rather than the net debt. Gross federal debt came to $16 trillion as of the end of last fiscal year.
However, if you pursue that line of reasoning, you would ask not what is the Treasury’s accounting entry for the Social Security Trust Fund, which is nothing more than the accumulated amount that Social Security in the past has contributed to the Treasury in the way of an excess of past Social Security taxes over past Social Security benefits paid out. Instead, you would want to look at how much money would be needed in order to provide the benefits scheduled to be received by current program participants, defined to be anyone currently 15 years or older, that is, anyone currently potentially paying into or receiving benefits from Social Security. You can then compare that with the sums expected to be received from future Social Security taxes. Such calculations involve a number of guesses, and can be quite sensitive to assumptions about items like the interest rate and the economic growth rate. What I did was just use the numbers reported by the Social Security trustees report (Table IV.B.7). Their report calculates the excess of the present value of expenditures over revenues for current program participants to be not the $2.6 trillion obligation officially acknowledged in the form of the Treasury debt owed to the Social Security trust fund, but instead to be $26.5 trillion.
Similar calculations from the trustees reports for Medicare report Medicare’s net unfunded liabilities for current program participants to be $27.6 trillion. For more details see Table 4 and the accompanying discussion in my paper.
Here’s how my paper summarizes the relevance of these numbers for U.S. fiscal challenges:
The federal government might well choose to reduce payments to beneficiaries relative to those anticipated under the program’s current practice, or might increase future payroll taxes. But these are of course the same options that the government would consider in figuring out how to honor its official on-balance-sheet liabilities as well. The political difficulties that the government might face in making changes to the public’s perceived Social Security obligations should reasonably be regarded as an important influence on the government’s ability to honor its on-balance-sheet liabilities. For this reason it seems entirely appropriate to include these implicit commitments in an accounting of the federal government’s combined off-balance-sheet liabilities.
My paper also looked at a number of other off-balance-sheet obligations. One important category is assistance to housing. With Fannie Mae and Freddie Mac currently in conservatorship, it is useful to reflect on the size of their liabilities in addition to other federal housing-related agencies. The government-sponsored enterprises have issued not just sizable direct debt of their own (which is not included in the $16 trillion gross federal debt figure noted above), but have also issued guarantees on huge numbers of private mortgages. Adding together all federal off-balance-sheet liabilities associated with housing, I come up with a figure of $7.5 trillion, by itself alone 2/3 the size of the net federal debt. And here is how my paper suggests those numbers should be interpreted:
It should be recognized that such liabilities do not have the same status as the direct debt obligations of the Treasury itself. For one thing, there are some offsetting assets, namely the mortgages held outright. The value of the mortgages would never fall to zero, so that using the notional exposure is a significant overstatement of the conceivable net outlays that would ever be required from the federal government to honor these commitments.
Instead, what these numbers do accurately convey is the huge footprint of off-balance sheet items on the housing sector, along with their enormous growth over the last several decades.
My paper also discusses a number of other significant off-balance sheet liabilities of the U.S. government. For example, FDIC deposit guarantees in 2012 came to $7.4 trillion, which I summarize this way:
the stresses of the most recent financial crisis were not enough to cause these guarantees to result in direct cash outflows from the U.S. Treasury, and the program seems to have worked in this instance as intended.
My paper nonetheless also notes that earlier related FSLIC guarantees did result in a significant U.S. fiscal burden, and that the recent deposit guarantees by the Irish government contributed to a near fiscal crisis for that country.
Off-balance-sheet obligations of the U.S. government are quite substantial– 6 times the size of Treasury debt held by the public. But that does not mean that the real debt of the U.S. government is $70 trillion. Off-balance sheet commitments are of a fundamentally different character than those officially on the balance sheet.
Even so, off-balance-sheet commitments can be quite important. Here’s what I think is the correct conclusion to take away from these numbers:
…the budget impact associated with an aging population and other challenges could turn out to have much more significant fiscal consequences than even the mountain of on-balance-sheet debt already accumulated.