The National Debt versus the Deficit
But first we must distinguish between the gross national debt and the outstanding national debt, both of which the government regularly reports. The gross national debt includes the holdings of the various federal trust funds for Social Security, Medicare, and several smaller government programs. These programs have run surpluses that are in essence loaned to cover other government expenditures. Consequently, the figure to which most economists and commentators refer is the outstanding national debt, net of the trust funds (i.e., the debt held by the public, including the Federal Reserve). Changes in neither measure of the national debt match annual deficits or surpluses, nor do the changes in either measure match each other.
Let me start with the debt held by the public. When the U.S. government borrows money to cover ordinary expenditures, the full amount appears in annual outlays, the annual deficit, and the annual increase in the outstanding national debt. But when the U.S. government borrows money to make direct loans to private parties, usually only the net present value of those loans is counted as an outlay and part of the deficit. Under the Federal Credit Reform Act of 1990 (applied retroactively in all official budget statistics), the net present value of each loan (and loan guarantee) is the estimated amount of the subsidy (or cost to the government, exclusive of administrative costs). The subsidy is analogous to the loan loss reserves that banks set aside when they make loans. The outstanding national debt, in contrast, increases by the full amount of the loan. The national debt can therefore increase by more than the deficit when the U.S. government makes direct loans and by less than the deficit when the loans are repaid.
This accounting practice today is affecting how the TARP (Troubled Asset Relief Program) bailouts are reported. The Congressional Budget Office puts the TARP's contribution to the fiscal year 2009 deficit at $184 billion, even though the TARP is expected to add an additional $461 billion to the outstanding national debt during the same fiscal year. That means President's Obama's reported budget, coming in at nearly 28 percent of GDP (the highest since World War II), significantly understates the increase in federal spending on a pure cash-flow basis.
The same thing happened last September when the Treasury initiated its Supplementary Financing Program, in which it eventually borrowed half a trillion dollars solely for the purpose of reloaning it to the Fed, primarily to finance the Fed's currency swaps with foreign central banks. None of that money was booked as part of federal outlays, and so it could not appear in the annual deficit. Yet however much remains outstanding (currently down to $200 billion) will be included whenever the national debt is reported.
The additional difference between the outstanding national debt, held by the public, and the gross national debt is of course the total amount in the federal trust funds. Since the trust funds can rise or fall independent of the U.S. government's annual outlays, the increase in the gross debt can be more or less than the increase in the outstanding debt. The easiest way to comprehend this relationship is to keep in mind that when all the trust funds are totally exhausted, the two ways of measuring the national debt will perfectly coincide. (Currently, under intermediate assumptions about future variables, the Medicare trust fund will be empty in 2017 and the Social Security trust fund will be empty in 2037).
A technical point: Many mistakenly believe that, when the Johnson Administration moved Social Security from"off-budget" to"on-budget," it made a major difference. But because both"off-budget" and"on-budget" sets of government outlays and receipts are regularly reported and unified, the change was of little significance, even from an accounting perspective. Indeed, in 1983 Social Security was technically moved"off-budget" again. For details about these virtually meaningless convolutions, go here.