Goldman has been on a roll this week. After losing all credibility (or whatever they had) with the markets, the objective media and Main Street, but not their clients, who were the ones losing the most for interacting with the squid, yet refuse to take their business elsewhere for fear of being locked out from the market monopolist with the greatest amount of inventory (yes, economies of scale when compounded with not so subtle forced liquidations of key competitors end up in monopolistic outcomes), now their economic team is taking a gamble with its own reputation (this is the team that won the best big bank economic team aware for 2009). In a note distributed to clients, entitled "What's the Right Measure of US Government Debt?" Andrew Tilton and Alec Phillips try to present the case that contrary to what you may have heard, the $12.8 trillion of US debt is not really worth losing sleep over. In fact the next time Goldman needs a bailout and the resultant $2-20 trillion of new debt are added to the make the 2s30s at about 100%, that should not be a source of concern either.
We present the "essay" in its entirety.
What's the Right Measure of US Government Debt?
Investors prone to worry about the sustainability of government debt may feel increased anxiety given the wide variation in estimates of government liabilities. In today’s comment we list some of the major reasons why figures differ and organize them in a common framework.
When reviewing debt figures it is important to distinguish between a) federal and state/local debt, b) gross vs. net debt, c) present vs. future obligations, d) on- vs. off-balance sheet commitments, and e) stronger vs. weaker commitments. For those focused on the federal government’s near-term financial condition, federal net debt of about 50% of GDP (including net GSE liabilities) may be an appropriate measure; for those wishing to compare government debt sustainability across countries, gross general government debt to GDP of 70% is probably a fairer comparison. Unfunded liabilities of 200%+ of GDP are extremely worrisome, but serve mainly to make plain that fiscal policy must and will change.
As many economies and markets recover from the worst of the financial crisis, investors have turned their attention to the creditworthiness of governments. The fiscal support needed to contain the financial damage and kick-start growth has been considerable, and has naturally raised questions about the sustainability of government borrowing in many countries.
One source of anxiety for many investors has been skepticism regarding the commonly used government budget statistics, which exclude a variety of off-balance sheet liabilities. This skepticism has in turn led to a plethora of estimates about the “true” amount of government debt, from federal debt held by the public (54% of GDP in 2009) to the present value of total government liabilities (several times GDP). In today’s comment we list some of the major reasons why estimates differ, and organize them in a common framework. Though still incomplete, the table below may serve to provide some context for the various estimates of debt. (Note that some figures are our estimates, and in cases where 2009 figures are unavailable, we have used 2008 data.)
A good deal of the confusion stems from the various forms of governmental obligations that exist. We can distinguish among these in several ways:
1. Federal vs. state/local. Although some analyses focus only on state and local obligations as a share of state GDP in evaluating municipal debt sustainability, or on federal debt/GDP in evaluating federal fiscal sustainability, it is more instructive to combine the federal, state, and local debt burdens as a share of GDP, given that the revenue used to pay interest on this debt ultimately comes from the same source – the income generated by the economy. For example, while the current federal debt/GDP ratio is generally viewed as sustainable, it would appear less sustainable if there were other claims, such as taxes to fund state/local debt, on the same income. A second justification for a combined “general government” approach is to net out claims between state and federal governments; states hold Treasury and GSE securities that comprise some of the gross federal debt in the table below, but net out once the balance sheets of governments at all levels are consolidated.
2. Gross vs. net debt. This distinction used to be important mainly at the state/local level, where governments have in aggregate typically held financial assets roughly equal to and sometimes exceeding credit market liabilities. At the federal level this distinction has been much less important, as the Treasury has typically held little by way of financial assets. However, recent financial market intervention has resulted in additional assets and liabilities coming onto the federal balance sheet, either officially or unofficially. First, the Treasury financed its TARP program through debt issuance, but used nearly all of the proceeds to purchase financial assets with considerable value. Second, the Treasury significantly strengthened the implicit federal guarantee of GSE liabilities when it put them into conservatorship in 2008, effectively increasing the amount of federally backed debt. However, the GSEs hold nearly an equivalent amount of assets. Thus the net effect of financial stabilization programs on debt levels is considerably less significant than the change in gross debt would imply.
3. Present vs. future obligations. Any discussion of fiscal sustainability would be incomplete without a discussion of future obligations, particularly those related to retirement-related programs. As shown in the table below, governments hold significant assets in pension funds (in the case of state/local) and intragovernmental trust funds (in the case of federal). These assets, along with revenue that continues to be collected through dedicated payroll taxes, are in most cases sufficient to pay current obligations related to these programs. However, these programs are far from balance over the long term; pension plans for retired government employees are underfunded by 33% of GDP according to estimates from the Pew Center on the States and the Government Accountability Office, and federal entitlement programs face an unfunded liability of at least 129% of GDP (based on 75-year deficits calculated by the Social Security and Medicare trustees) if current policy remains unchanged.
4. On vs. off-balance sheet. Although on-balance sheet commitments should obviously be counted toward debt totals, off-balance sheet commitments can be more subjective. First, there is no agreed-upon amount for many of these obligations, such as the pension or entitlement shortfall noted above. Second, it isn’t clear where to draw the line—for instance, most observers count some form of pension underfunding, and while many analysts left out GSE obligations before the entities were put into conservatorship, we suspect more are inclined to include them today.
5. Strong vs. weak commitments. How off-balance sheet commitments are treated depends in large part on the degree to which governments are (or are perceived to be) bound to make good on certain obligations. For instance, there is little dispute that debt service holds the top spot in the priority of claims on governmental revenue, particularly at the federal level. Pension obligations come only slightly below; there have been few episodes of state and local governments reneging on pension obligations, and in some states they hold as high if not a higher legal position than debt service. On the other hand, entitlement programs represent a softer commitment. Medicare Part A, which is financed through dedicated payroll taxes placed in a trust fund, represents a fairly strong commitment to future retirees. The rest of Medicare (Parts B and D), which bases eligibility on Medicare Part A but is financed out of general revenues and premiums from current enrollees, is more of a hybrid, with not quite as strong a commitment to leave benefits unchanged. Medicaid, by contrast, is financed entirely out of general revenues, and only a fraction of the taxpayers who pay for the program will use it.
What’s the appropriate measure to use? It depends on the purpose, but we would highlight three main ways of looking at the numbers:
1. For assessing the federal government’s current financial position. Here, we would focus on net on-balance sheet government debt, plus net debt related to GSE/financial sector commitments: these are the obligations that are likely to be relevant in the very near-term, less the liquid assets available to pay them. This particular measure is not too concerning, at just under 50% of GDP in 2009. Of course, the budget situation is deteriorating rapidly: we expect the current-year federal deficit to exceed 10% of GDP, and see deficits still in the mid-single-digit range through the end of the decade.
2. Debt sustainability, especially for international comparisons. At the international level, gross general government debt (70% of GDP in the US) is probably the most widely used measure. This includes state and local obligations, some of which ultimately could be backed by the federal government in a crisis. It does not net out assets, since the liquidity and quality of government assets is difficult to compare across countries. Ideally, we’d like to compare net general government debt, including the likely cost of contingent liabilities, but in practice this is very difficult.
While we’re on the subject, it’s worth pointing out that comparisons of the debt of certain US states with other sovereign nations are apples-to-oranges. The economic situation of states is clearly affected by federal government policies and debt, and so at the least, one should apply a pro rata share of federal government debt if making any kind of international comparison.
3. To gauge the magnitude of decisions facing policymakers. Here, total net debt of all types, including unfunded entitlements, is probably the best measure—simply because it makes plain that the current combination of government tax and spending policies is unsustainable and will require tough decisions soon. Higher taxes are likely to be part of the solution, but reform to entitlement spending also is inevitable. For exactly this reason, we don’t think that debt/GDP figures of 200%+ are a good representation of the United States’ current position. They are based on an assumption that policy doesn’t change, but we know it must.
Andrew Tilton / Alec Phillips