Gross vs net government debt -- a tedious topic with important practical consequences

19 Feb 2011 by Jim Fickett.

The main fallacy in common thinking about government debt comes from including assets, but not liabilities, of government branches other than the Treasury. Gross government debt, which includes intragovernmental debt, provides the best simple measure of long-term solvency. Net debt, which includes only debt held by the public, is more important for understanding markets in the short term.

I'm in the process of writing a longish post on whether a Japanese debt crisis, predicted for many years, might now be a near-term danger. To talk about sovereign debt one must be clear on how to measure it. This is particularly an issue for Japan. Gross government debt, which means all treasury obligations, is about 200% of GDP. But net government debt, which subtracts out government assets, is only about half that.

To be a little clearer on the definitions involved, we can take a look at the US Social Security Administration. The SSA took in more from taxes than it provided in benefits for many years, and the excess was stored away in two trust funds (Old-Age and Survivors Insurance, and Disability Insurance). What does it mean that the SSA has money saved in trust funds? SSA has given the cash to the Treasury in exchange for “special issue” securities, and Congress has spent the money. This is directly analogous to you or I buying a Treasury bond – the money is immediately spent, but we hope to be paid back due to the Treasury's power to tax.

The securities issued by the Treasury to the SSA are counted in gross debt, but not in net debt (usually called, in the US, debt “held by the public”), because the debt issued by the Treasury is offset by the asset held by the SSA.

The arguments about whether gross or net is the better measure, like the arguments about whether the government is measuring inflation correctly, never will be settled. But we can make some progress if, instead of the overly simplistic “which is better?”, we ask the more nuanced, “for what purposes is gross debt the better measure, and for what purposes is net debt more helpful?”

The Congressional Budget Office explains why net debt is the more important measure for understanding immediate impacts of budget deficits on the wider economy:

Gross federal debt comprises both debt held by the public and debt issued to various accounts of the federal government, including the major trust funds in the budget (such as those for Social Security). Because the debt issued to those accounts is intragovernmental, it has no direct, immediate impact on the economy. Instead, it simply represents credits to the various government accounts that can be redeemed as necessary to authorize payments for benefits or other expenses.

That is, in the short term, it is only changes in net debt that come to the open market, affect interest rates, and perhaps crowd out other investment.

Some argue that gross debt does not matter at all, because the securities issued by the Treasury to the SSA trust funds simply represent “money that the government owes itself”. This is a serious error. If you look only at the Treasury balance sheet, the debt is there, and will, through additional taxes, have to be paid some day. If you widen the picture to include the SSA balance sheet, then you must include not only the SSA assets – the trust fund balances – but the SSA liabilities as well – the present value of future benefit obligations – which are larger than the assets. In other words, if you look at the Treasury and SSA balance sheets in aggregate, the assets and liabilities at the SSA net to a negative, and actually worsen the overall picture.

The Financial Times made exactly this point regarding Japan's debt, in an article last year:

Optimists insist that it is wrong to fixate on Japan's gross debt, since in net terms the state only owes the equivalent of slightly more than 100 per cent of GDP. After all, the lower net number reflects the fact that the state has the equivalent of more than Y100,000bn in foreign exchange reserves and also holds a big chunk of its own debt.

It is unclear, however, how much help other state assets would be in extremis. Social security funds, for example, hold well over a third of total assets - but their future obligations are even bigger.

In the case of Japan, one other asset often mentioned, supposedly offsetting government debt, is foreign exchange reserves. The logical fallacy here is exactly the same as that for the US Social Security Administration. As explained well by Michael Pettis, foreign exchange reserves are acquired by issuing local currency bonds. Barring exchange rate fluctuations, the assets and liabilities of the government agency handling such reserves net to zero. There is no net asset with which to reduce Treasury debt.

Where does this leave us with the gross debt concept? Gross debt looks only at the Treasury balance sheet. It would be better, if it were possible, to look at the balance sheet of the whole government, perhaps including as well some closely related enterprises backed by the government, like Fannie and Freddie in the US. My guess, given the nature of politics and government, is that few government branches or entities have a balance sheet that actually nets to a positive amount. If that is true then even gross government debt, which looks only at the Treasury, underestimates the total the government owes. Gross debt is, then, a simplification and possibly an underestimate, but still the best simple estimate we have of what the government owes.

And where does that leave us with respect to its importance? If it is only changes in net debt that affect the markets, why worry about how much is really owed? The answer to that is that that net debt is affected by gross debt in the long term. Eventually the piper must be paid. The SSA again provides an example. The trust funds are, with current projections, about at their peak. That means that while, in past years, the SSA passed a lot of cash to the Treasury, reducing the amount that had to be raised by tax or borrowed, now the reverse is true – in the future more will have to be raised by tax or borrowed, pressuring net debt, in order to pay back the trust funds.