Managing the Federal Debt


In Washington, on Wall Street, and in foreign capitals, all eyes are on mounting government debt, particularly America's. The scope of American borrowing, and the ease of the government's access to credit, offer useful signals about the strength of our economy and about the future direction of global public finances — especially during this time of economic uncertainty.

But while observers of our debt all track the same trends and the same figures, the most knowledgeable analysts disagree about just what the data reveal. To some, the market for United States Treasury securities — the means by which the government sells debt to the public so it can spend money in excess of tax revenue — appears to be buoyed by insatiable demand. To others, it seems like a bubble ready to collapse. The yields on Treasury bills and notes (or the interest rates the government must pay to people who buy securities) sit at historic lows. But America's public finances (its debt, deficits, and financing needs) hardly differ from those of Spain, which faces the very real prospect of losing access to credit. Some analysts argue that America risks a second Great Depression by borrowing too little, and therefore allowing demand to wither. And yet others argue that we are borrowing too much, thereby risking a currency crisis.

How can the same evidence lead to such wildly divergent conclusions? What exactly are the prospects for America's access to credit in the years to come? And what do the data say about the U.S. dollar's position as the global currency of choice? To answer these questions, we must first think through just what the national debt really is; how it is sold, and to whom; and how our debt looks today in historical perspective.


Simply put, the government borrows to fund public spending in excess of tax revenue and to cover the interest expenses associated with past borrowing. The Treasury borrows money on the government's behalf by auctioning debt to the public.

That debt takes three major forms. The first consists of Treasury bills (often called T-bills), which are securities that mature in less than one year. They do not earn interest along the way, but are sold below their face value — and so are essentially guaranteed to yield a modest return to the buyer. (Indeed, T-bills are often taken to be the least risky generally available investment option in the world.)

The second form of debt involves Treasury notes and bonds, which are securities that take longer than a year to mature (notes generally take up to ten years, bonds usually take 30 years), but that pay off some set percentage of their face value in interest every six months. The interest rate depends on demand (since the securities are sold at auction), and therefore acts as a measure of the appeal and value of federal debt, as well as an indicator of the state of the American and global economies. The higher the demand for these securities, the lower the interest rate the Treasury is forced to pay to buyers. Today's very low rates therefore suggest demand is quite strong.

The third form of U.S. government debt is called Treasury Inflation Protected Securities (TIPS), which are basically bonds that are indexed for inflation. The interest rate at which they pay off is constant, but the principal amount against which interest payments are calculated is adjusted based on the consumer price index, thus protecting the holder against the risk that high inflation will reduce the real value of his investment. Because of this added protection, TIPS tend to pay lower interest rates than notes and bonds.

The goal of the Treasury Department's Office of Debt Management is to issue the mix of these securities that best raises the funds needed to support government spending, while also minimizing the government's long-term borrowing costs. The mix therefore changes as demand changes, and in response to economic conditions. At the end of fiscal year 2009, bills accounted for 28.5% of outstanding Treasury securities, bonds and notes accounted for slightly less than 64%, and TIPS accounted for 7.9%.

Treasury auctions occur regularly based on a fixed schedule; there are now usually two auctions per week, on Mondays and Thursdays. As each auction approaches, the Treasury provides details about the securities being offered and the total amount of money the government is looking to raise. Participants in the auction have the option of bidding on either a competitive or non-competitive basis. Competitive bidders provide the Treasury with the quantity of securities they are interested in buying and the minimum yield they are willing to accept. Non-competitive bidders agree to accept the winning yield, whatever it may be, in exchange for assurance that their full order of securities will be filled. Progressively higher competitive bids are accepted until the Treasury sells all of the securities at auction, so that demand ends up setting the yield, and therefore the price of government debt.

There are three kinds of bidders at these Treasury auctions: primary dealers, other direct bidders, and indirect bidders. Primary dealers are the 18 financial institutions that have a direct trading relationship with the Federal Reserve Bank of New York, including large American firms like Bank of America, Goldman Sachs, and Citigroup, as well as some foreign firms (like Credit Suisse, Deutsche Bank, and the Japanese firm Mizuho Securities). As a condition of their direct trading access, these dealers are required to make bids at Treasury auctions; they are also consulted by the Treasury and the Federal Reserve regarding debt issuance. These firms generally sell the securities they purchase at auction to other parties, and between them are responsible for distributing most American government debt to both institutional and individual buyers around the world.

Other direct bidders, meanwhile, are the more than 800 financial institutions beyond the primary dealers that use the Treasury auction system to place bids with the Fed without the involvement of an intermediary institution. They do not have the same close relationship with the Treasury, but their participation — by providing a broad and diverse array of buyers at auction — helps to tie securities prices to real-world economic conditions.

Finally, indirect bidders generally fall into two broad categories: investment funds (like money-market funds and hedge funds) that do not have a direct trading relationship with the Fed, and foreign institutional investors (principally central banks abroad).

According to Treasury data compiled from auctions that took place between 2001 and 2009, brokers (generally primary dealers) accounted for 55% of all notes and bonds acquired at auction and 61% of all bills. The next largest acquisitor category was foreign investors, who purchased 18% of all bonds and notes and 9% of all bills at auction. Investment funds followed, buying 11% of notes and bonds and just under 10% of bills. The Federal Reserve System itself is a non-competitive bidder at auctions, acquiring 12% of notes and bonds and 16% of bills from 2001 to 2009. Its participation, however, is generally limited to rolling over maturing obligations into new securities that it uses to help conduct its monetary policy. (The Fed uses Treasury securities as collateral when it borrows from and lends to large financial institutions, which it does in order to manage the money supply.)


The distribution of buyers of debt at Treasury auctions does not necessarily tell us who actually owns these securities over the long term, since many buyers eventually sell their securities to other investors. Indeed, even though primary dealers are the largest buyers at auction, they generally sell the securities they purchase in secondary markets; as a result, they end up holding very little federal debt. At the end of March, for instance, brokers and dealers owned only 1.6% of all Treasury securities outstanding.

The largest holders of Treasury securities by far are foreign institutions, which own nearly 50% of all Treasury securities outstanding (despite acquiring less than 20% at auction). The next largest holder of debt is the Federal Reserve, with 9.4% of all Treasury securities, followed by individual Americans, who own 9% of the total; mutual funds, which own 7.8%; and state and local governments, which own about 6.5%.

Foreigners hold about $4 trillion in Treasury securities today, and roughly $2.9 trillion of that amount is owned by so-called foreign official investors: central banks and finance ministries that build reserves of foreign currencies to support their exchange rates or monetary systems (as discussed below). Most of them use the U.S. dollar for this purpose, which makes it the global reserve currency of choice. At the end of 2009, the dollar accounted for 62% of declared official foreign-exchange reserves, up from 59% in 1995. It is striking to note that, even though the euro was introduced in 1999 to create a second global reserve currency, the euro share of foreign-exchange reserves at the end of 2009 — 27% — was roughly equal to the percentage of reserves in 1995 denominated in the old currencies of those European countries that eventually adopted the euro (such as the Deutschemark and French franc).

For governments, having foreign-exchange reserves can serve a number of useful purposes. Some countries, especially those whose economies depend heavily on trade, seek to maintain fixed exchange rates (by which the values of their currencies are pegged to the value of a particular foreign currency, most often the dollar, or some set mix of foreign currencies). If a country pegs its currency to the dollar, for instance, its trading relationship with the United States — and its relationships with other countries that peg their currencies to the dollar — become far easier, since buyers and sellers don't constantly have to readjust prices to deal with fluctuating exchange rates.

In order to keep its own currency's value so strictly tied to that of another country's, however, a nation's central bank must be able to control the price of its own currency on the open market — by buying large amounts of its own currency when the price falls too low, and selling off when the price goes too high in relation to the currency it seeks to match. In order to constantly maintain such balance, the central bank needs large reserves of foreign currency, so that it has the flexibility to buy up a great deal of its own money as needed, under the demands of real-time trading.

Having large foreign-exchange holdings can also help a nation protect its currency against speculative attack. If speculators try to drive down the price of the nation's currency on the open market by selling it short, the central bank can sell off its foreign holdings and buy up its own currency, pushing back against the short-sellers by increasing demand and therefore boosting the currency's value. The larger the stock of foreign-exchange reserves a country has, moreover, the less inclined investors will be to launch speculative attacks on its currency in the first place (since they know such efforts are unlikely to succeed, and therefore likely to be quite costly). Thus large reserves of foreign currency — often dollars — can be a powerful deterrent.

Who has built such reserves? As of April, the largest foreign holders of Treasury securities were the People's Republic of China ($900 billion, or 11% of the entire Treasury market), Japan ($796 billion, or 9.5%), the combined OPEC nations ($239 billion, or 3.5%), Brazil ($164 billion, or 2%), and Russia ($113 billion, or 1.4%). In addition, a combined $474 billion (nearly 6% of Treasury securities outstanding) were held in accounts located in the United Kingdom and in Caribbean and other North American banking centers (the Bahamas, Bermuda, the Cayman Islands, the Netherlands Antilles, and Panama). While the holdings of some hedge funds and other private investors are included in the U.K. and Caribbean figures, some foreign official purchases on the secondary market are also likely included in this total, as foreign central banks often acquire Treasury securities from dealers in London and offshore jurisdictions.

The U.S. Treasury's reliance on financing from the foreign official sector carries risks that are difficult to quantify, because of the political considerations involved. From a pure market perspective, concentrated foreign Treasury holdings would not be problematic: Countries could succeed in using their Treasury holdings to inflict financial pain on the United States — flooding the market with American debt, and thereby driving up interest rates and disrupting our economy — only by suffering large losses on their own holdings. But this normal balance of risks is altered by the potential for non-market, or politically motivated, decisions. China, for instance, would suffer enormous economic harm if it ever used the American debt it owns to hurt our economy; given the degree to which China's economic growth is derived from its trade surplus with the United States, destroying Americans' ability to buy Chinese goods would be highly counterproductive. Yet in the heat of a strategic confrontation — not to mention a war — China's leaders might decide that such pain is worthwhile in the short run; the damage they could do would certainly be significant. Moreover, it would be naïve to think that the opinions of major creditor nations are not given special weight when American economic policies (and even other issues like human-rights abuses abroad) come up during bilateral discussions and other diplomatic activity. Without a doubt, buying large amounts of American debt gives these countries unique political and strategic, as well as economic, advantages.

Though these foreign holdings of American debt remain very large, they have actually declined slightly as a share of total securities outstanding — to 48% in April 2010 from 50% in April 2009, according to the Federal Reserve. This decline was not the result of any selling of securities, or even of a failure to add to existing portfolios; foreign holdings of Treasury securities actually increased by 16% over the 12 months ending in March. But that growth was insufficient to keep the foreign share of the Treasury market constant, because the total amount of Treasury debt outstanding increased by 21% in the course of that period, and a greater share of it was purchased domestically than had been the case in recent years.

Why did domestic purchases of Treasury securities increase so significantly? A large part of the increase involved massive purchases by the Federal Reserve, whose holdings of Treasury debt grew by an extraordinary 58% over the 12 months in question. In late 2008, many financial institutions were unable or unwilling to lend money, but the typical approach to jumpstarting such lending — reducing interest rates to lower banks' borrowing costs and so encourage them to lend — was not an option, as the Fed's rates were already near zero. The Fed therefore sought to give financial institutions cash in exchange for securities they owned, launching a policy of "quantitative easing" — by which the Fed purchased Treasury notes from financial institutions (as well as the debt and mortgage securities issued by government-sponsored enterprises, or GSEs, like Fannie Mae and Freddie Mac) using money essentially created out of thin air.

While the acquisition of GSE securities accounted for a much larger share of the quantitative-easing purchases than did Treasury securities, the Fed still added $284 billion to its Treasury portfolio during the 12 months ending in March 2010 — equal to about 20% of net new Treasury issuance. Other domestic buyers also increased their purchases of federal debt in an effort to find safe investments in the wake of the financial crisis. Although the Treasury holdings of mutual funds declined during this period (driven primarily by the overall contraction in the money-market mutual-fund industry), the Treasury holdings of private pension funds grew by 59%. The holdings of individual households grew by 46%. But by far the largest increase was in the holdings of commercial banks, which grew by 115%.

The banks' overall balance sheets remained roughly constant during this period, so their move to government debt came at the expense of other investments — specifically an 18% reduction in corporate bond holdings and a 17% reduction in bank loans. The pronounced move toward Treasury holdings reflected a desire by banks to shed private-sector credit risk and to seek greater safety in the wake of the financial crisis. Given that the percentage of loans in default in the banking system reached record levels in 2009, banks had good reason to be wary of more private-sector credit exposure.

The banks had an additional motive for buying federal debt. As a result of the credit crisis and the flagging economy, the rate at which banks lend to one another hovered near 0.25% for much of 2009, while ten-year Treasury notes yielded 3.5%. This meant that banks could borrow from other banks cheaply, and then use that money to buy Treasury securities, which would earn them more than 3% in net interest income. The ability to obtain such wide margins without the assumption of much credit risk obviously increased the appeal of Treasury securities.

But it has also created concerns about interest-rate risk, as profit margins could narrow — or even turn into losses — if short-term interest rates for interbank borrowing were to shoot upward unexpectedly. The government has therefore strongly encouraged banks to refrain from this practice in recent months, and the banks have begun to heed the Fed's warnings.


The Treasury decides how much debt to issue, and in what form, as part of its "quarterly refunding" process. Each quarter, the Treasury assesses its available funds, its expected funding needs (based on how much money government agencies are expected to spend in excess of tax revenues in the forthcoming quarter), and the amount of outstanding debt or interest that will be coming due during the quarter and will need to be repaid. After constructing its own estimate of cash needs (with input from the Office of Management and Budget), the Treasury also seeks advice from primary dealers — as well as from some of the other big private-sector buyers of government debt — inquiring about their expectations for economic growth, the federal deficit, and the kinds of Treasury securities they would find most attractive. Then, in February, May, August, and November of each year, the Treasury formally announces the amount of net debt it expects to issue in the subsequent two quarters. It also provides a rough estimate of the types of securities it expects to issue, and offers a brief summary of its overall debt-management policy.

The Treasury's most recent announcements paint a grim picture of the government's debt obligations and prospects. At the end of March 2010, total outstanding federal-government debt held by the public was $8.3 trillion (the rest of the familiar $13 trillion national-debt figure is money the government owes itself — to pay for obligations like Social Security or Medicare, for instance). OMB estimates that this public-debt figure will reach $9.2 trillion by the end of this fiscal year, which would be roughly equal to 63% of gross domestic product. In 2011, OMB expects the federal government to run an additional $1.4 trillion budget deficit, which would push the debt-to-GDP ratio to 69%. The last time the federal debt represented a larger share of the economy was in 1950, when it stood at 80% of GDP as a result of the debt incurred during World War II.

Over the second half of 2010 and the first three quarters of 2011, the federal government is therefore likely to issue about $1.9 trillion of net new debt, roughly 10% of GDP over that period. It is important to remember that this is a net figure, which measures only incremental additions to the stock of outstanding debt, not the funds used to pay back maturing debt.

But the Treasury also has to pay for debt that comes due during this period. And when those maturing obligations are added to the total, the magnitude of gross scheduled Treasury debt issuance is positively gargantuan. In fiscal year 2009, the federal government ran a $1.4 trillion budget deficit, equal to a post-war record of 9.9% of GDP. But over the same one-year period, the U.S. Treasury issued $8.9 trillion in debt, equal to a staggering 62% of GDP. Some of this total represents money used to acquire financial assets as part of the Troubled Asset Relief Program (TARP), but most of it simply demonstrates the Treasury's reliance on bills and shorter duration notes, which must be refinanced frequently, requiring greater net debt issuance. (A $1 billion debt financed with a 91-day bill, for instance, must be refinanced four times in a single year.) Although net issuance for all of fiscal year 2010 is likely to be about $1.36 trillion, the gross Treasury issuance will likely exceed $7 trillion, or 48% of GDP.

While most analysts of budget sustainability correctly focus on the net debt figure, gross issuance is still important, because it exercises a powerful influence over the interest rates the government must pay on its debt. The more often a debtor is forced to return to his creditor to roll over his outstanding obligation, the more opportunities the creditor has to renegotiate the terms of the loan (or to refuse to renew the borrowing entirely). The amount of debt subject to such renegotiation therefore determines how swiftly the interest rate paid on the government's outstanding debt can change — and a sudden spike in interest rates caused by creditor unease, or other market shocks, would affect the cost of servicing all newly issued debt. Gross debt issuance is therefore hugely consequential, because it represents the amount of our debt exposed to potentially unfavorable fluctuations in financing conditions.

This uncertainty makes it more difficult not only to manage but even to forecast the federal government's deficit. The 2012 deficit will consist of the primary deficit (that is, direct government spending in 2012 in excess of tax revenues), plus the cost of paying interest on the debt we owe at the end of 2011. If most of the debt were scheduled to mature in 2013 and beyond, the total interest expense could be estimated with relative certainty, since the rate paid on the bulk of it would already be known. But with so much debt maturing between now and the end of 2012, the expected future interest expense becomes less certain, since the effective interest rate can change at every refinancing.

This rollover problem is particularly egregious for the United States, because our debt in recent years has generally been issued in shorter-term securities than have the debts of many other developed nations. At the end of 2009, the average maturity of the outstanding U.S. Treasury debt was 55 months (just over four and a half years). By comparison, the United Kingdom — which faces a deficit situation similar to ours — has an average maturity on its debt of 12 years. Federal policymakers are clearly aware of this problem; after relying on bills to finance much of the debt growth of late 2008 and 2009, the Treasury is now working to extend the maturity of the public debt. The average maturity of newly issued Treasury securities in 2010 is 77 months (over 6 years), which will extend the average maturity of our outstanding debt to 59 months by this winter. The portion of debt funded by bills is expected to drop to 19%, with a long-term target of 12%. The Treasury estimates that this financing shift will reduce the portion of debt maturing in 2011 to 29% of the total outstanding debt. This is an improvement, but only a modest one: It means the Treasury will still have to issue 19% of GDP in debt in fiscal year 2011 just to roll over existing obligations, before accounting for whatever new deficits will also need to be funded.


Government policymakers aren't the only ones who need to worry about borrowing terms. Increases in the interest rates the federal government must pay to borrow money would have painful repercussions for private budgets, too, because they would affect the rates that private borrowers must pay on their own loans.

Indeed, discussions of the national debt too often overlook the fact that, in recent decades, private indebtedness in America has been growing as quickly as public indebtedness. Even as the federal debt reaches and exceeds post-war records as a share of the economy, it remains well within historical bounds as a share of the total U.S. credit market. An $8.3 trillion federal debt accounts for just 16% of the total value of all American credit-market obligations outstanding — that is, of all public and private debt in America. This is actually slightly less than the federal debt's average share of the combined stock of public and private debt since 1970 (16.2%).

The stability of this ratio in the face of burgeoning federal borrowing is a reflection of the dramatic increase in the private-sector debt-to-GDP ratio. In 1970, total non-federal debt outstanding in America was equal to 125% of GDP. Today, it is roughly 300% of GDP. The astonishing growth of household debt (from 44% to 93% of GDP) and financial-sector debt (from 12% to 103% of GDP) means that a much larger share of national income has already been pledged to service debt obligations than has generally been the case in American history. This means that the growing cost of federal borrowing — driven by increases in the putatively risk-free rate from which all other borrowing rates, public and private, are measured (including rates on things like mortgages, credit cards, and student loans) — would lead to the eating up of a much greater share of national income than an analysis of the federal budget alone would suggest. This makes proper management of the federal debt all the more important, and the staggering growth of that debt all the more worrisome.

Consider that, from 1951 to 1980, total U.S. borrowing averaged about 12% of GDP per year. Between 1980 and the first quarter of 2008, borrowing increased by two-thirds, to more than 20% of GDP per year. Debt grew about one and a half times faster than GDP during this period, resulting in the explosive growth of the aforementioned indebtedness ratios. But over the past year and a half, in the wake of the credit crisis of 2008, this trend has reversed itself sharply in the private sector. In the first quarter of 2009, total economy-wide borrowing was negative — meaning more money was being repaid than borrowed — for the first time since the Federal Reserve began tracking such data in 1946. It has remained negative in the four quarters since. The universe of outstanding debt is shrinking, as negative borrowing (that is, the repayment of debt) necessarily implies negative lending (the selling of assets to raise cash with which to pay back loans), since total borrowing equals total lending in any given period. Such selling off of assets can cause an economy-wide decline in prices — in other words, deflation.

In this context, federal borrowing of 10% of GDP per year may not seem too alarming. Indeed, many argue it is necessary to head off a disastrous deflationary spiral and that the supportive role of debt-financed public expenditure is all that has kept the United States from falling into an economic depression. But however healthy the short-term effects, the government's debt issuance must still abide by the basic arithmetic of flow-of-funds accounting. This means that the money government spends must come from somewhere else in the economy.

Since 1951, the public sector has run an average deficit (nearly all of it federal) of roughly 2.4% of GDP, and the household sector has run an average deficit of about 3.2% of GDP. These combined deficits have been financed by an average surplus of 4.5% of GDP in the non-household private sector (both commercial and financial businesses) and an average surplus of 1.1% of GDP in the external sector (largely foreign trade and foreign investment). This means that the savings of businesses, which come largely from retained earnings or undistributed profits, have financed both the mortgage and consumer borrowing of households — and more than half of the public-sector deficit — these past six decades. The remaining half of the public-sector deficit, meanwhile, was financed by the external sector.

These data paint a picture of a precarious economic situation. In the near term, public borrowing needs will be four times larger than they have averaged over the past 50 years, and will remain twice the historical average for years to come. Financing these deficits would require the public sector to drain large amounts of capital from the private sector at a time when private-sector growth is essential. Swiftly reducing these deficits, meanwhile, might prevent the government from siphoning off the private capital stock, but would require sharply curtailing government spending and borrowing, which might be just as dangerous to the economy right now. Which risky path American policymakers take will depend on how they understand the implications of their two options — that is, how they understand the role of public debt in our economy.


This brings us back to the two different ways of reading today's economic data. In the past two years, the federal government has been able to sell its debt on extremely favorable terms (in part because commercial borrowing declined dramatically, leaving investors with fewer options and so making federal debt more attractive). But can its string of luck continue? The answer depends on one's understanding of what exactly is happening to our economy in the aftermath of the financial crisis of 2008. This is why there has been such intense disagreement among experts about the future of the federal debt: The data can accommodate two stories.

The first is that the American economy suffers from excess capacity resulting from a temporary decline in aggregate demand. In this view, the large private-sector financial surplus of the moment (that is, the fact that businesses are saving and repaying more than they are spending and borrowing) reflects a weakness in consumer spending and business investment in the real economy. Deficit-financed government spending — putting money into the hands of citizens to spur private consumption, or making direct government purchases — is all that is preventing the economy from falling into a deflationary spiral caused by the liquidation of assets (which is the necessary counterpart to the private sector's effort to reduce outstanding debt).

The second view is that the private-sector financial surplus is driven by consumer and producer expectations about the future tax implications of government borrowing. Named after the classical economist David Ricardo and popularized by Harvard economist Robert Barro, this "Ricardian" theory (in its various forms) argues that the private sector responds to increases in public-sector borrowing by increasing its estimates of future taxation, therefore lowering its estimates of future disposable income and reducing current spending. Under this theory, the private-sector surpluses come from the increased savings caused by these revised expectations. Indeed, if one is ever likely to see Ricardian effects manifest themselves, it would be in the wake of a sudden 50% increase in the debt-to-GDP ratio. And, as with the first view, the data above fit this story perfectly.

This debate is also mirrored in the dispute about the impact of debt-financed stimulus over the past two years. If private-sector surpluses are a response not only to reduced demand, but also to public-sector debt growth, then borrowing huge sums of money to fund a stimulus bill could well end up making things worse, or at least fail to make them much better. The government's financial position could end up deteriorating without a corresponding boost to the economy.

So who is right? Should we seek to borrow more today, or less? Ever since the economist John Maynard Keynes put forward his brilliant exposition on the "paradox of thrift" (which argues that, if everyone in the economy were to reduce spending and increase saving in recessionary times, demand would drop so sharply that the economy would collapse and overall savings would actually decline), there has been a tendency among economists to view calls for debt reduction as misplaced moralizing. And indeed, the risk of insufficient deficit spending today could be very real. At the very least, we might examine our current situation in the light of historical precedent: Between 1936 and 1938, the government pursued a contractionary fiscal and monetary policy that closed a 5.4%-of-GDP budget deficit within two years. The result was a fall in output and a spike in unemployment — contributing to the Great Depression's "second dip." Today's policymakers certainly have reason to fear repeating this mistake.

But economists on the other side of the argument are motivated by concerns just as real. Their worry is not that the United States would default on its debt: The government borrows in a currency that it prints, and it is difficult to conceive of a situation in which it would be more advantageous for the United States to renounce obligations than to print whatever amount of dollars would be necessary to meet them. The real problem is that bond-market investors are not oblivious to this flexibility. When it appears likely that a country will print money to inflate away unsustainable debt burdens, interest rates rise to incorporate an inflation risk premium — thus increasing the burden on the government and on private borrowers. The danger, then, is that excessive borrowing will bring investors' hunger for Treasury securities to an end, causing a spike in interest rates that could crush the American economy and send it into a debt spiral we would find very difficult to escape.

This latter scenario cannot be dismissed when we consider the sheer scale of recent and projected federal borrowing. OMB estimates that the debt-to-GDP ratio will reach 77% in 2020. But using the same policy baseline and more pessimistic assumptions about interest rates and GDP growth, the International Monetary Fund estimates that the debt-to-GDP ratio will actually reach 106% in 2020. Service payments on that debt would likely consume more than a quarter of expected tax revenue. Needless to say, diverting one in four tax dollars to interest payments would not be conducive to financing the explosion in Social Security, Medicare, and Medicaid spending over this same period. And a significant worsening of the terms on which the Treasury is able to borrow would squeeze the nation even further.

Those who deny the reality of this danger point to the persistent investor demand for American debt even in the midst of sharply deteriorating American public finances. But the source of that persistent demand is perhaps best explained by George Soros's quip, paraphrasing Winston Churchill, that "the dollar is the worst currency except for all the alternatives." The strength of the Treasury market reflects the weakness of other countries' fiscal positions, and the power of inertia and familiarity. These factors have allowed the Treasury to borrow seemingly without limit in order to absorb the extraordinary financial shock the economy has suffered.

But the weakness of other currencies is a shaky edifice on which to build confidence in the Treasury's ability to fund itself at low rates in the future. It is important to remember that today's deficit and debt projections, grim though they are, are based on the assumption that this edifice will hold. But if investors come to see America's fiscal position as just as weak as "all the alternatives," we may see a shift in interest rates and relative pr ices that policymakers are now failing to account for.

The leftist intellectual Irving Howe once wrote that the liberal imagination is often constrained by what he called the "provincialism of the immediate" — the presumption that things as they exist today will continue on forever. That presumption is certainly evident in our public finances, as discussions about debt sustainability tend to focus on the burden imposed on "our children and grandchildren" while ignoring the potential for abrupt shifts in investor sentiment today. But the history of modern financial markets is characterized by sudden breaks and swings that cannot be reliably forecast. The current spike in demand for Treasury securities is itself an example of such a change, as it was precipitated by the sudden flow of money out of other investment options in the wake of the financial crisis. Today's historically low yields on Treasury securities should not inure American policymakers to unsustainable debt growth, or make them overconfident of their ability to anticipate and pre-empt a crisis.

The Treasury market's status as a safe haven is not an immutable feature of economic life: It is a function of institutional credibility that took generations to build, but that would take just a fraction of that time to destroy. Were Treasury securities to lose their status as the global reserve asset of choice to gold, other commodities, or a different currency, the consequences for the American economy would be disastrous. Unlikely as such a scenario might seem at the moment, today's fiscal policies unquestionably increase the probability of its coming to pass.

Jason Thomas is a research and teaching fellow in the department of finance at the George Washington University School of Business.