What’s the best way to measure the national debt?

A recent meeting of the Obama administration's deficit reduction commission raised an important question: We all want to reduce the national debt, but how do we measure the debt we're trying to reduce?

One of the witnesses at the fiscal commission's hearing last week was University of Maryland economist Carmen Reinhart, whose book on financial and debt crises with Harvard economist Kenneth Rogoff has become something of a must-read (or at least, must-cite) among the commentariat. In their book, This Time Is Different: Eight Centuries of Financial Folly, Reinhart and Rogoff focus on "gross debt," which, according to the International Monetary Fund (IMF),

consists of all liabilities that require payment or payments of interest and/or principal by the debtor to the creditor at a date or dates in the future. This includes debt liabilities in the form of SDRs, currency and deposits, debt securities, loans, insurance, pensions and standardized guarantee schemes, and other accounts payable.

This measure differs substantially from the oft-cited $8.5 trillion publicly held debt. The IMF measure includes intragovernmental debt, in particular the Social Security and Medicare trust funds. As a result, the measured level of debt is actually $13 trillion, or around 88 percent of GDP. Reinhart and Rogoff have warned that 90 percent of GDP tends to be a crossover point where gross debt affects interest rates, economic growth, and other factors (although the threat can occur at much lower levels for countries with serial histories of default).

But some have countered that the best measure is simple publicly held debt, by which standard the debt-to-GDP ratio is "only" 58 percent. Since only this debt is frequently rolled over in international credit markets, it provides a more accurate measure of how willing international lenders must be to support our spending. While the Social Security and Medicare trust funds are rolled over as well, the government does so by borrowing from itself.

But consistency is important: many of those who argue against including intragovernmental borrowing in measures of public debt also argue that the $2.5 trillion trust fund protects Social Security through 2037, backed by the full faith and credit of the U.S. government (a credit unbroken since the abrogation of the gold clause in the 1930s). But the trust funds cannot be an asset to Social Security but a debt to no one; if the Social Security "crisis" is put off for decades by the existence of the government bonds in the trust fund, then the debt owed to Social Security—which will need to be repaid beginning this year—should likewise be counted.

That said, I personally don't consider trust fund securities debt quite the same as publicly held debt, since the government can control when and whether trust fund debt is repaid by altering the Social Security benefit formula. Raise the retirement age enough or reduce benefits enough and the debt will never come due. And, as noted above, trust fund debt is not rolled over in the same way as publicly held debt. So I agree with those who say that intragovernmental debt is different than publicly held debt, although these folks should then also acknowledge that Social Security's trust fund isn't exactly like a private-sector fund.

But here's what I think matters: Reinhart and Rogoff's statistical analysis showed a relationship between high gross debt and undesirable financial and economic outcomes. But I would be surprised if Reinhart and Rogoff wouldn't reach similar conclusions if they focused only on publicly held debt—since gross debt and publicly held debt are almost surely correlated—except they would likely show trouble beginning at debt-to-GDP ratios below 90 percent.

More importantly, one thing we can't do is mix and match measures. For instance, the Economic Policy Institute's Monique Morrissey writes (in an otherwise not-bad paper on Social Security and the budget) that

a cross-country comparison by economists Carmen Reinhart of the University of Maryland and KennethRogof of Harvard University found no evidence that debt-to-GDP ratios below 90% had any impact on economic growth.

But Morrissey uses this finding to rebut a recommendation by the Peterson-Pew Commission that the country stabilize debt-to-GDP ratios at around 60 percent by 2020. The problem here is that the Peterson-Pew Commission was focusing only on publicly held debt, while Reinhart-Rogoff focuses on gross debt. Morrisey's interpretation suggests that we still have another 30 percentage points of debt relative to GDP to give, while gross debt shows we're already at the level where Reinhart and Rogoff begin to see trouble.

In other words, the precise definition of public debt we focus on matters less than the broad conclusion that, however measured, we're amassing too much of it.