Could the Fed Go Broke?

Imagine the Fed as a regular bank. It has a.s.sets-loans to commercial banks and Treasury bonds; and liabilities-commercial bank reserve deposits and the currency in your wallet. The Fed earns interest on its a.s.sets and pays some interest on reserves but those $20 bills in your wallet are an interest-free loan to the Fed. This produces a big profit for the Fed, called seignorage, seignorage, which it hands over to the Treasury. This isn't chump change. In 2009, it gave the Treasury $47 billion. This means every taxpayer has an interest in how the Fed manages its balance sheet. which it hands over to the Treasury. This isn't chump change. In 2009, it gave the Treasury $47 billion. This means every taxpayer has an interest in how the Fed manages its balance sheet.

During the financial crisis, the Fed shifted from safe Treasurys to riskier things like mortgage-backed securities, loans to banks and AIG, old Bear Stearns a.s.sets, commercial paper, and so on. Could this strategy cause the Fed to go broke? Almost certainly not. The Fed does not carry its bonds at market value, so fluctuations in their prices don't affect its profit. If some of its loans default and the collateral is inadequate, it would record a loss. But the losses would have to be ridiculously large to wipe out its profit, never mind its substantial capital. Anyway, the Fed could have no capital at all and still do its job, although having to ask Congress for more capital wouldn't be good for its independence and our confidence in it.

Into the Weeds Banks are acutely vulnerable to panic. Most of their money is tied up in loans. They keep cash on hand to repay some depositors, but not enough to repay all of their depositors.

In 1873 Walter Bagehot, an early editor of The Economist, The Economist, wrote in wrote in Lombard Street, Lombard Street, "A panic grows by what it feeds on . . . [It is] a species of neuralgia." In such a panic, investors abandon any kind of risky a.s.set and demand the safest, most liquid thing: cash, or its closest subst.i.tute, Treasury bills. Only a central bank can create more cash. Bagehot recommended it lend against any good collateral, at a penalty interest rate to discourage frivolous borrowing. "A panic grows by what it feeds on . . . [It is] a species of neuralgia." In such a panic, investors abandon any kind of risky a.s.set and demand the safest, most liquid thing: cash, or its closest subst.i.tute, Treasury bills. Only a central bank can create more cash. Bagehot recommended it lend against any good collateral, at a penalty interest rate to discourage frivolous borrowing.

This is the principle behind the loans the Fed makes from its discount window discount window (although there's no actual window). It lends to commercial banks, accepting as collateral loans, securities, and other a.s.sets discounted from face value. Loans from the Fed to banks are charged the discount rate. To encourage banks to first borrow from each other before borrowing from the Fed, the discount rate, since 2003, has been set above the Federal Funds rate target. Occasionally, the actual Fed funds rate spikes above its target if, for example, banks suddenly have to meet a cl.u.s.ter of payments and all try to borrow at once. On those occasions, the discount window is a useful safety valve for them. (although there's no actual window). It lends to commercial banks, accepting as collateral loans, securities, and other a.s.sets discounted from face value. Loans from the Fed to banks are charged the discount rate. To encourage banks to first borrow from each other before borrowing from the Fed, the discount rate, since 2003, has been set above the Federal Funds rate target. Occasionally, the actual Fed funds rate spikes above its target if, for example, banks suddenly have to meet a cl.u.s.ter of payments and all try to borrow at once. On those occasions, the discount window is a useful safety valve for them.

Over the years, the discount window has dwindled in importance as banks found other sources of funds at home and abroad. Banks used it when the payment system didn't operate properly, as on 9/11, or if they were short of required reserves. Most notoriously, troubled banks used it when no one else would lend to them, as Continental Illinois did in 1984. It eventually failed. This gives the discount window a stigma, and healthy banks avoided it at all costs. Getting around that stigma is why during the recent crisis the Fed made loans from the discount window via auctions: They were cheaper, and more anonymous.

The Federal Reserve Act says the Fed can only lend to banks, thrifts, and credit unions. That made sense when banks dominated the economy. But as Chapter Fourteen will show, they don't anymore. Between 1980 and 2007, their share of credit (excluding federal borrowing) shrank to 23 percent from 50 percent. Mutual funds, investment banks, finance companies, hedge funds, and the like supply the rest. Instead of deposits, they fund themselves by issuing bonds and short-term money market paper and by borrowing from banks. When the crisis. .h.i.t, lenders wouldn't refinance their paper or their loans. They faced the equivalent of a nineteenth-century bank run, but with no lender of last resort.

In the 1930s, Congress inserted a loophole in the Federal Reserve Act. It said that with a supermajority vote of governors, the Fed could lend to companies other than banks-temporarily. The Fed was reluctant to use Section 13(3) as this loophole is called. It often has had to fight pressure, usually successfully, to lend to politically favored const.i.tuencies from farmers to homeowners. Such loans help some people at the expense of others, and expose the taxpayer to loss if the loans go bad. That sort of lending should be left to elected politicians. And so 13(3) sat largely unused-that is, until 2008.

The Fed has never before inserted itself so deeply into life-and-death decisions about who would get money and under what conditions. Many in and outside of Congress recoiled, especially when the companies who were saved reported ma.s.sive profits and handed out fat bonuses. The Fed became party to the bankruptcy proceedings of companies whose loans it bought. It even became owner of a hotel chain.

As the crisis pa.s.sed, the Fed closed most of its special lending facilities. Yet, having once become lender of last resort to the entire financial system, the Fed will be expected to do so again. "Regardless of whether that window is officially opened or closed, the market now a.s.sumes that it will be opened if necessary," said Vikram Pandit, head of Citigroup.

The Bottom Line * The Federal Reserve has made its name managing the economy through monetary policy, but its parents had a different career in mind: to act as lender of last resort when banks ran out of cash. The Fed is uniquely suited to the job because it can simply create whatever money it needs to lend, primarily through loans from its discount window, and withdraw the money from existence when the loans are repaid.* During the financial crisis the Fed dusted off a loophole to lend not just to banks but to a wide a.s.sortment of companies. In so doing it may have saved the country from another Depression, but it also awakened politicians to its formidable power.

Chapter Twelve.

The Elephant in the Economy What the Government Giveth and Taketh Away IN 2009, during the rancorous fight over health care, an agitated voter told his Congressman to "keep your government hands off my Medicare." The next year, a man angry about a tax audit crashed his plane into an Internal Revenue Service building in Austin, Texas, killing an employee.

Fiscal policy arouses strong feelings among people because it affects the shape of society. It provides services the private sector can't-like national defense-or won't-like national parks-at a publicly acceptable price. It sends checks to the unlucky, the elderly, the sick, and the poor, and pays for them by taxing the more affluent, the young, and the employed, affecting their incentives to earn and to invest. People's demands on government are unlimited but what they are willing to pay in taxes sure isn't. Reconciling what they want with what they'll pay for is a never-ending balancing act. The government's solutions may not be efficient for the economy or fair for people, but simply what's politically palatable.

The federal government leaves a pretty big footprint on the economy. From 1970 to 2007, it collected the equivalent of 18 percent of gross domestic product (GDP) in taxes, and spent about 21 percent of GDP. The difference between those figures results in a budget deficit (more on that in the next chapter). And yet they barely hint at the breadth of things on which the government spends, the impact of how it taxes, and the mind-numbingly complex way Congress and the President deal with both.

What the Government Giveth You can divide federal government spending into three main categories.1. Interest Paid on the Debts Taken on Since the American Revolution. Interest Paid on the Debts Taken on Since the American Revolution. For most of the last decade this type of expenditure was an afterthought, averaging 8 percent of spending and 1.6 percent of GDP. As the national debt climbs in coming years, it will become a much bigger presence in the budget, consuming some 4 percent of GDP by 2020. There's not much politicians can do about this category. For most of the last decade this type of expenditure was an afterthought, averaging 8 percent of spending and 1.6 percent of GDP. As the national debt climbs in coming years, it will become a much bigger presence in the budget, consuming some 4 percent of GDP by 2020. There's not much politicians can do about this category.2. Discretionary Spending. Discretionary Spending. Name a federal activity and odds are it is discretionary, from the Peace Corps and the federal courts, to national defense, the largest, topping $700 billion a year. Congress must appropriate funds every year for discretionary activities. No appropriation, no activity. For the last decade, discretionary spending has averaged about 38 percent of total spending. This is the category of spending over which politicians have the most control. Name a federal activity and odds are it is discretionary, from the Peace Corps and the federal courts, to national defense, the largest, topping $700 billion a year. Congress must appropriate funds every year for discretionary activities. No appropriation, no activity. For the last decade, discretionary spending has averaged about 38 percent of total spending. This is the category of spending over which politicians have the most control.3. Mandatory Spending. Mandatory Spending. Also called Also called ent.i.tlements, ent.i.tlements, mandatory spending comprises 60 percent of federal spending. Mandatory spending doesn't require an annual appropriation: it is already set in law. For example, Social Security benefits are dictated by the terms of the Social Security Act. Congress often amends these laws, but if does nothing, the spending continues on autopilot. mandatory spending comprises 60 percent of federal spending. Mandatory spending doesn't require an annual appropriation: it is already set in law. For example, Social Security benefits are dictated by the terms of the Social Security Act. Congress often amends these laws, but if does nothing, the spending continues on autopilot.

Initially, growing ent.i.tlement spending reflected our natural urge to expand the safety net as we become richer. In the future, though, it will be fueled by demographics and medical inflation.

.Ent.i.tlements deserve a closer look because they dominate federal spending and, left alone, will overshadow everything else. There were no ent.i.tlements until the creation of Social Security in 1935. It was joined by Medicare and Medicaid in 1965. These three account for most of the growth in mandatory spending. Other ent.i.tlements, like veterans' pensions and food stamps, have grown much more slowly.

Initially, growing ent.i.tlements spending reflected our natural urge to expand the safety net as the United States became richer. Since 1990, the proportion of the U.S. population enrolled in Medicaid has risen from 10 percent to 15 percent and Barack Obama's health reforms are expected to raise it further. In the future, though, ent.i.tlements will be fueled by demographics and medical inflation. As the U.S. population ages, the bill for Medicare and Social Security benefits grows. Meanwhile, science continually comes up with new, more expensive ways to treat our ailments. And because most Americans have health insurance, they tend to consume more health care than if they paid the full price. If left unchecked, these three ent.i.tlements will go from 10 percent of GDP to a crushing 18 percent by 2050.

Social Security was originally designed to be financed by payroll taxes. This worked when the taxes paid by workers far outstripped benefits collected by retirees. The extra money went into a trust fund. As the population ages, however, the number of workers per retiree is falling and payroll taxes now barely cover benefits. True, the trust fund still has $2.5 trillion in it. But that's economically meaningless: The money consists entirely of federal government IOUs. To repay those IOUs, the federal government has to come up with the money elsewhere either by borrowing from the public or raising taxes, just as if the trust funds didn't exist. Medicare is partly funded by a payroll tax and by premiums paid by beneficiaries, but these don't cover the program's costs now, and the gap will only grow.

Every year the trustees of Social Security and Medicare report on the gap between future revenue and benefits. Expressed in 2009 dollars, they put that "unfunded obligation" at $104 trillion over all coming years, or 8 percent of all future GDP. That estimate depends a lot on a.s.sumptions of how long people live, how fast their wages grow, the rate of health care inflation, and future interest rates. Moreover, Social Security's contribution to that gap is relatively stable. Health care is the real time bomb.

What the Government Taketh Away Spending money is the fun part. Raising the taxes to pay for it is what makes people squeal, and Americans have a long history of squealing. In the 1790s, a tax on whiskey provoked a rebellion against the administration of George Washington. Back then, the government got most of its revenue from taxing items that were easy to find, such as imports and liquor. As government has grown, it found other things to tax: wages, investment income, profits, capital gains, gasoline. Its biggest revenue raisers are the personal income tax, payroll tax, and the corporate income tax.

Taxes provoke no end of argument between Republicans and Democrats, conservatives and liberals, economists and politicians. One flashpoint is how to share the tax burden. Most Republicans and Democrats agree the tax system should be progressive, which means the rich should pay higher tax rates than the middle cla.s.s, and the poor should pay little or no tax. They differ, however, on how far to go: Should the rich pay double the tax rate as the middle cla.s.s? Triple? Should the poor get refundable tax credits even if they pay no tax? Should payroll and sales taxes, which consume more of the income of the poor than the rich, be progressive?

Another flashpoint is over exemptions, deductions, exclusions, and credits. Though pitched as tax cuts, these are all types of tax expenditures, expenditures, a targeted boost delivered through taxes instead of spending. They range from credits for disabled coal miners to deferred capital gains tax on electrical transmission lines. Some are sensible, such as the child credit and the charitable deduction, while others encourage bad behavior. For instance, the deduction for employer-provided health care encourages wasteful health spending, while the deduction for mortgage interest encourages people, in particular the wealthy, to take out bigger mortgages. a targeted boost delivered through taxes instead of spending. They range from credits for disabled coal miners to deferred capital gains tax on electrical transmission lines. Some are sensible, such as the child credit and the charitable deduction, while others encourage bad behavior. For instance, the deduction for employer-provided health care encourages wasteful health spending, while the deduction for mortgage interest encourages people, in particular the wealthy, to take out bigger mortgages.

Besides raising money, taxes change behavior. The behavioral response to taxes, though, is often exaggerated.

.A final flashpoint is over how taxes affect behavior. In 1990, George H. W. Bush slapped a luxury tax on yachts, private planes, expensive cars, and the like. Sales of yachts promptly collapsed with the help of a recession and the tax was repealed in 1993.

In the 1970s and 1980s, supply-side supply-side conservatives, like economist Arthur Laffer and Congressman Jack Kemp, claimed that if taxes on wages and investment income were cut, workers would supply so much more labor and investors so much more capital that tax revenue would actually go up. Even mainstream Republican economists didn't buy it. It was politically irresistible, though, enabling politicians to promise both lower taxes and reduced deficits with a straight face. conservatives, like economist Arthur Laffer and Congressman Jack Kemp, claimed that if taxes on wages and investment income were cut, workers would supply so much more labor and investors so much more capital that tax revenue would actually go up. Even mainstream Republican economists didn't buy it. It was politically irresistible, though, enabling politicians to promise both lower taxes and reduced deficits with a straight face.

As it happens, tax revenue fell after Ronald Reagan's and George W. Bush's tax cuts and it rose after Bill Clinton's tax increases. A lot of this, though, had nothing to do with changes to tax rates, but the health of the economy. In a boom, people and corporations earn more so they pay more tax. In a recession, they pay less.

Supply-siders have a point, even though they exaggerate it. High tax rates do discourage work and encourage tax avoidance. For example, a high income tax wouldn't discourage a traveling salesman from working enough to buy a home, but it might discourage him from working enough to install a swimming pool. In a 2009 study, economists Emmanuel Saez, Joel Slemrod, and Seth Giertz concluded raising rates on the rich by 1 percent may cause them to report 0.1 percent to 0.4 percent less taxable income. Thus, it is not just the total tax take that matters, but how it is levied. All else equal, it is better to tax things we want less of-higher gasoline taxes would discourage driving, carbon emissions, and imported oil. Raising taxes on dividends and capital gains, on the other hand, would discourage investment that makes workers more productive. But if taxes on dividends are abolished, accountants will find a way for wealthy clients to convert their wages to dividends.

Into the Weeds Reconciling this mult.i.tude of competing priorities for spending and taxes is the job of the federal budget. In parliamentary systems, like Britain's and Canada's, the Prime Minister draws up a budget and Parliament pa.s.ses it. It's like a steak: a solid cut of meat that changes little between the cow and the dinner plate. The United States' budget is more like sausage, a mixture of ground meat from different parts of the animal stuffed into a misshapen skin.

Let's take a look inside the sausage factory. Under the Const.i.tution, the president can only propose spending and taxes; Congress has final say, subject to the president's veto. In its first century of existence, the United States had no federal budget. Individual agencies would request money and Congress would appropriate funds for them one at a time. In 1921, the president began formulating a single budget with the creation of what is now called the Office of Management and Budget (OMB). Congress clawed back some of its influence in 1974 after a standoff with Richard Nixon, establishing its own budget process and the Congressional Budget Office as its nonpartisan scorekeeper.

Even during the best of times, Congress routinely ignores requests by the president to kill pet programs or tax breaks.

.Though the fiscal year begins October 1, the budget process begins a year and a half earlier when federal agencies submit their budget requests to the OMB. No later than the first Monday in February, the OMB submits the president's budget to Congress. Besides laying out thousands of individual proposals and cost estimates, the budget is also a political doc.u.ment that lays out the president's domestic agenda. How much he actually gets depends a lot on his approval rating-Congress grants a popular president more of what he wants-and his party's control of Congress. But even during the best of times, Congress routinely ignores requests by the president to kill pet programs or tax breaks.

After receiving the president's budget, Congress starts its own process. The Senate and House budget committees pa.s.s a budget resolution that sets out spending and revenue totals to which all other tax, program, and appropriations bills should conform. The resolution isn't a law, and can't be vetoed by the president. Both chambers are supposed to pa.s.s the resolution by April 15, though they routinely miss this target. And at least four times since 1998 Congress couldn't agree on a resolution at all.

After the resolution pa.s.ses, and even if it doesn't, individual committees get to work. Tax proposals are handled by the Finance Committee in the Senate and the Committee on Ways and Means in the House. Mandatory spending proposals are a.s.signed to the relevant authorizing committee-Medicare to the Finance and Ways and Means committees, food stamps to the agriculture committees, and student loans to the education and labor committees.

Discretionary spending is the purview of the House and Senate appropriations committees. Each has 12 subcommittees whose chairmen are dubbed "cardinals," dealing with a particular part of the budget. This is the favored time for individual lawmakers to earmark money in an agency's budgets for a special projects or const.i.tuents, such as a bridge to nowhere in Alaska or swine manure research. Earmarks have grown from a relatively harmless way for a legislator to promote his state or district to a vehicle for vote-buying and even corruption. But they are economically insignificant. They seldom exceed 1 percent of federal spending, and they only result in the reallocation of money that would be appropriated anyway.

Getting discretionary spending bills into law is essential because the federal government can't spend money that Congress hasn't appropriated. Since Congress routinely misses the October 1 deadline to pa.s.s all 12 of its appropriations bills, it usually has to pa.s.s a continuing resolution continuing resolution to fund the government in the interim. Some years, Congress and the president deadlock and, with no authority to spend money, the government shuts down, most famously in 1995-1996. Emergency functions, such as national defense and air traffic control, can continue, but employees get IOUs instead of paychecks until the deadlock ends. to fund the government in the interim. Some years, Congress and the president deadlock and, with no authority to spend money, the government shuts down, most famously in 1995-1996. Emergency functions, such as national defense and air traffic control, can continue, but employees get IOUs instead of paychecks until the deadlock ends.

Multiple appropriations bills are routinely merged into a single omnibus bill omnibus bill either to speed things up or to force through provisions that might not pa.s.s on their own. If more money is needed after the fiscal year begins, Congress pa.s.ses a either to speed things up or to force through provisions that might not pa.s.s on their own. If more money is needed after the fiscal year begins, Congress pa.s.ses a supplemental bill. supplemental bill.

If individual spending and tax bills don't total up to what the budget resolution envisioned, they are, in theory, forced to conform through a process called reconciliation. reconciliation. Reconciliation has since evolved into a vehicle for major legislative changes. Unlike most bills, reconciliation bills cannot be filibustered in the Senate so they can pa.s.s with 51 instead of 60 votes, and debate is limited to 20 hours. This makes it attractive for contentious legislation, such as Bush's tax cuts and parts of Obama's health care overhaul. Reconciliation, however, has its limits-its Byrd rule forbids amendments that are Reconciliation has since evolved into a vehicle for major legislative changes. Unlike most bills, reconciliation bills cannot be filibustered in the Senate so they can pa.s.s with 51 instead of 60 votes, and debate is limited to 20 hours. This makes it attractive for contentious legislation, such as Bush's tax cuts and parts of Obama's health care overhaul. Reconciliation, however, has its limits-its Byrd rule forbids amendments that are nongermane, nongermane, meaning they have nothing to do with the budget. Generally, it can't be used to widen the deficit, although Bush broke that tradition with his tax cuts. There's typically one reconciliation bill every one to two years. meaning they have nothing to do with the budget. Generally, it can't be used to widen the deficit, although Bush broke that tradition with his tax cuts. There's typically one reconciliation bill every one to two years.

In budget battles, hyperbole and partisan exaggeration are the weapons of choice. Thank goodness for the Congressional Budget Office.

.In budget battles, Congressional leaders and presidents regularly indulge in hyperbole and partisan exaggeration. Thank goodness for the Congressional Budget Office (CBO). Though appointed by Congressional leaders, the CBO director is nonpartisan and doesn't endorse bills. By evaluating their impact and cost, though, he can make or break them.

The CBO gets things wrong, sometimes spectacularly. In the late 1990s, it repeatedly underestimated future surpluses. Starting in 2001, it made the opposite mistake as deficits replaced surpluses, only partly because of Bush's tax cuts. But its errors are unbiased. They result from misjudgments or mistaken a.s.sumptions, not a presumption that a policy is good or bad.

The CBO shares its watchdog role with Congress' Joint Committee on Taxation (JCT). Although its chief of staff is a partisan appointment, the committee's role and its staff are nonpartisan. The JCT a.n.a.lyzes and helps write tax legislation. The CBO uses its revenue estimates to score estimate the cost of legislative proposals.

The budget process in the United States makes deficit reduction hard because touching any spending program or tax break arouses opposition from concentrated const.i.tuents and allied legislators. To overcome this, Congress sometimes tries to impose rules on itself that make it hard to run deficits, much as Odysseus had himself tied to his mast to resist the call of the sirens. One such rule is Pay-As-You-Go (Paygo), which requires that new tax cuts or mandatory programs not increase the deficit and must be paid for with higher taxes or lower spending elsewhere in the budget. Paygo, however, has loopholes. Major deficit reductions such as in 1990 have usually required bipartisan deals, often made at summits with just a handful of decision makers.

Hitting Closer to Home State and local governments together spend about as much as the federal government. State and federal budgets are similar, but there are some important differences.* States spend less on benefit checks and more on delivering services, like college education, prisons, and road maintenance. Thus, they employ more people.* States get a third of their general revenue from the federal government, mostly for shared programs like Medicaid. For the rest, they depend heavily on both sales and income taxes.

In all but four states, the fiscal year begins on July 1. More than half of states have a one-year budget cycle while the remaining states mostly have a two-year cycle. They rarely budget beyond two years, which makes for myopic decision making. Like Congress, state legislatures have final say over the budget and four (notably California) need a supermajority to pa.s.s a budget. But governors have more say than presidents. In 44 states, governors can veto individual items rather than an entire spending bill, whereas the Supreme Court struck down a similar presidential line-item veto. Governors also have more power to change spending or taxes without the legislature's permission if the budget goes off course.

Every state except Vermont is required to balance its budget. If, part way through the fiscal year, a deficit materializes, many states require the governor or legislature to eliminate it before the year ends. States can borrow for capital projects such as prisons and highways.

Balancing a state budget is a harrowing experience because while spending grows steadily, revenue gets whip-sawed by economic ups and downs. When a recession drives down revenue, states have to raise taxes or cut spending, making the recession worse. To escape the straitjacket, states often resort to gimmicky moves, such as skipping pension contributions or counting the proceeds of a bond issue as revenue. That is sort of like getting a home equity loan and counting the new cash as salary.

Still, the balanced budget requirement helps keep state debts low. Total state and local debt in 2009 was just $2.4 trillion, with state debt accounting for $1 trillion. A lot of that state debt was backed by earmarked fees like highway tolls. Debt backed by the full faith and credit of the states was only about $400 billion in 2008 or just 3 percent of GDP. States do, however, have large unfunded gaps in their employee retiree plans: $1 trillion as of 2008, the Pew Center on the States estimated in a 2010 report. Since state economies are less diverse than the country's as a whole and they can't print money, states are at greater risk of default than the federal government, although none has done so since Arkansas in 1933.

The Bottom Line * The federal government is a gigantic player in the economy and it will get bigger in coming years as government services expand, the population ages, and interest on the national debt mounts.* Federal spending comes in three varieties:1. Interest on the debt.2. Discretionary spending.3. Mandatory spending.* Tax revenue comes mainly from personal and corporate income and payroll taxes. Compared to other countries, the United States relies relatively little on consumption taxes such as on gasoline or a value-added tax.* Every year the president proposes a budget; Congress accepts some of it but ignores a lot as it pa.s.ses the appropriations, tax, and mandatory program laws.* Unlike the federal government, states must balance their budgets each year, which makes for profligacy in good times and wrenching austerity in bad times.

Chapter Thirteen.

Good Debt, Bad Debt How Government Borrowing Can Save or Destroy an Economy GEORGE PAPANDREOU ran for election in 2009 promising to reinvigorate Greece's recession-gripped economy by raising public salaries, investing in infrastructure, and helping small business. Shortly after becoming prime minister he discovered the budget deficit had exploded to 13 percent of Greek gross domestic product (GDP). Investors fled the country's bonds, driving their interest rates up to punishing levels. Papandreou was soon slashing salaries and raising taxes in a desperate effort to stave off default.

Government borrowing is like Ritalin. At the right dosage it can jolt a lethargic economy out of recession. Overdosing, as Greece discovered, can bring on seizure.

Borrowing money isn't necessarily bad. In fact, the U.S. national debt arguably helped bind the country together at its birth. In a famous compromise with Thomas Jefferson, Alexander Hamilton persuaded Congress to a.s.sume the colonies' debts in return for moving the capital from Philadelphia to present-day Washington. For most of the next century and a half, the federal government was small and conservatively run. The budget was in surplus more often than deficit; the national debt in 1860 was lower than in 1791. (A deficit deficit occurs when the government's revenue falls short of spending in a particular year. The occurs when the government's revenue falls short of spending in a particular year. The debt debt is the sum of all deficits ever run.) is the sum of all deficits ever run.) All that changed in the 1930s. In five of every six years since, the government has run a deficit. Although a family that keeps borrowing money every year to pay its bills would eventually have its credit cards canceled and file for bankruptcy, countries are different. Since their economies grow over time, they can keep borrowing as long as total debt doesn't get out of line with GDP. In theory the government can tax the entire GDP if necessary to repay the money-something no individual or company can do.

Three's a Crowd There's nothing wrong with the government borrowing to finance an investment, such as a highway, that pays off long into the future. That way, future taxpayers help pay for something that also benefits them. But, deficits now mostly finance things like social programs that only benefit today's citizens. Yet, future taxpayers will have to repay those deficits with interest. with interest.

Furthermore, deficits may chip away at long-term economic growth. To understand how they do this damage, imagine a watering hole on the African savannah with just enough water to support a pride of lions and a herd of zebras. Then one day, a bunch of elephants move in. Soon, lions and zebras are dying of thirst. Like the watering hole, the pool of savings from which businesses and households borrow is finite. When government deficits start drawing on that pool, the three-way compet.i.tion for money pushes up long-term interest rates and crowds out private investment-perhaps a family decides not to buy a house or a business decides not to expand. That hurts future growth.

Elephants don't crowd out lions and zebras when they're drinking from a lake instead of a watering hole. Similarly, deficits are less likely to crowd out private investment when the pool of savings is global instead of local. In the last decade, the world has had more savings than it knew what to do with, and the United States kept borrowing without pushing up long-term rates. Still, even this doesn't let deficits off the hook; when the United States borrows by selling bonds to foreigners, the influx of foreign money might lift the dollar, penalizing U.S. exporters.

These nefarious effects are hard to pin down but they're real. In 2004, before becoming Barack Obama's budget director, Peter Orszag wrote a study with William Gale that found a deficit of 3.5 percent of GDP (roughly its average since 1982), year in and year out, raises interest rates by 0.4 to 0.7 percentage points and shrinks the economy by 1 percent to 2 percent.

When Deficits Help The wisdom of deficits changes during recessions. When people lose their jobs, bonuses, or hours and corporate profits fall, they pay less tax. That deepens deficits but softens the blow to private spending. Meanwhile, benefits for the poor, such as those who are on Medicaid, and the unemployed, automatically rise.

Businesses and consumers usually trim their own borrowing during recessions out of caution for the future. With less compet.i.tion for saving, deficits are less likely to push up long-term interest rates as we saw in 2009 when the deficit hit 10 percent of GDP but bond yields fell.

The biggest drawback of fiscal stimulus is that it is usually redundant. Central bankers are much better at managing the business cycle.

.That sort of increase in the deficit is pretty much automatic, and it usually reverses during recovery. Sometimes, though, governments give the economy an added jolt with fiscal stimulus: an extra tax cut or a dollop of spending. This is usually a bad idea. It can take months, even years, for money targeted at roads, sewers, and power lines to clear the necessary approvals. Tax cuts may be saved rather than spent. Just one-third of the $96 billion in tax rebates distributed as part of a fiscal stimulus in 2008 were spent in the first year, according to a 2009 study by Claudia Sahm, Matthew Shapiro, and Joel Slemrod. If the economy really needs more stimulus, it's better to leave it to the Fed. It can lower interest rates on an hour's notice, and raise them just as quickly when the need has pa.s.sed, without reelection clouding its judgment.

However, fiscal stimulus is a good idea under one special case: when the Fed has already lowered short-term interest rates to, or close to, zero. Then, fiscal policy may be the only means left to pull the economy out of recession. This happens rarely; one of those few times was in 2009, which is why most economists backed a major federal stimulus. Raising taxes and cutting spending to reduce deficits is also riskier in such a situation, because the Fed can't compensate by lowering rates.

Different types of stimulus have different multiplier effects. The Congressional Budget Office thinks a dollar spent by the federal government on a jet fighter or a park ranger will ultimately generate an estimated $1 to $2.50 of total activity, whereas a dollar of tax cuts will generate between $0.50 and $1.70. But quant.i.ty does not equal quality. Building a bridge to somewhere and a bridge to nowhere will both create jobs. That doesn't mean they're equally useful.

Debt Traps and Debt Crises Most of the time, debt does its damage gradually, like termites in the attic. As lenders demand higher interest rates, the deficits slowly suffocate private investment and a growing share of national income goes to paying interest on the debt. Year after year this nibbles away at the economy's foundations. Sometimes, though, it is like a fire that races through the house. Investors suddenly decide not to lend at all. Interest rates skyrocket, the currency collapses, and economic activity implodes. It's like the conversation between two characters in Ernest Hemingway's The Sun Also Rises. The Sun Also Rises. "How did you go bankrupt?" one asks. "Two ways," the other replies. "Gradually and then suddenly." "How did you go bankrupt?" one asks. "Two ways," the other replies. "Gradually and then suddenly."

Investor confidence is crucial. An otherwise bearable debt becomes intolerable if interest rates rise sharply.

.Unfortunately, it is hard to know in advance whether a crisis will be gradual or sudden. A key danger sign is a high and rising debt-to-GDP ratio. Investor confidence is crucial. An otherwise bearable debt becomes intolerable if investors suddenly demand sharply higher interest rates. A tipping point occurs when interest rates climb above a country's nominal growth rate. At that point, the debt-to-GDP ratio will automatically rise unless the country runs a budget surplus, excluding interest. For example, if a country's nominal GDP grows 4 percent and it pays 6 percent interest on its debt, it needs an annual surplus, excluding interest, of 2 percent of GDP to keep the debt steady as a share of GDP.

Walter Wriston, the legendary chief executive of Citicorp, was ridiculed for saying countries don't go bankrupt, after Latin American loans almost wrecked his bank. But he was right: A creditor can't drag a deadbeat country through bankruptcy court and grab its a.s.sets. Countries may choose to default because they can't pay or because they don't want to. So lenders must worry about a country's ability and willingness to pay.

These differences explain why some countries can go longer without a debt crisis than others. At the end of World War II, the United States' debt reached 120 percent of GDP, and Britain's 200 percent; neither experienced a crisis. Nor has j.a.pan, even though its debt at the time of this writing exceeds 200 percent of GDP. By contrast, Mexico's debt was only 35 percent of GDP when a crisis struck in 1994.

In general, investors give a longer leash to countries with a long history of paying their debts like the United States, Canada, and Britain. Such countries are often allowed the luxury of borrowing in their own currency, which insulates them from one of the major causes of debt crises: the inability to repay foreign currency debt. By contrast, investors give shorter rope to countries that regularly stiff them. Argentina, for example, has defaulted three times since 1980 alone, most recently in 2001.

Countries escape debt through one of five ways. The most painless is to grow its way out: economic growth generates revenue to shrink the deficit and brings down the debt-to-GDP ratio.

Another way of controlling debt is austerity: painful cuts in spending or increases in taxes. This brings down interest rates, which also cuts the deficit. This is how rich countries like Ireland and Denmark in the 1980s and Canada in the 1990s escaped their debt traps and it is how the United States turned things around in the early 1990s. Another way is a bailout, when another country or the International Monetary Fund (IMF) comes to the rescue as the United States did for Mexico in 1994-although austerity is often a condition of such rescues.

Then, for countries like the United States that borrow in their own currencies, there's inflation, which reduces the real value of existing debts. As we saw in Chapter Five, though, creating inflation is easier said than done. And it may not help: If investors smell inflation, they'll charge higher interest rates or refuse to lend. The government may then have to twist the central bank's arm into keeping interest rates low, or force private citizens and banks to buy its bonds at artificially low rates.

The fifth route out of a debt trap is to default.

Could It Happen Here?

The United States' deficits are chewing away at the rafters, but are they about to burn the house down? It doesn't seem likely. Investors' faith in U.S. debt is b.u.t.tressed by history, culture, and law. When the Continental Congress couldn't pay revolutionary soldiers their wages, they gave them IOUs instead. Many veterans sold their IOUs to speculators. Alexander Hamilton persuaded the new republic to pay off all the IOUs, including those held by speculators, to establish the country's creditworthiness.

The only time the United States did anything like defaulting was its decision in 1934 not to honor previous promises to repay in gold. At the time of this writing, the debt is a manageable 60 percent of GDP and interest rates are no higher than GDP growth. The United States still has the "exorbitant privilege" of borrowing in the world's most popular currency. As we saw in Chapter One, its long-term economic outlook is brighter that most countries' thanks to higher population growth.

Crises used to be restricted to emerging countries. Not anymore.

.But there are reasons to worry. It used to be that debt would loom large only once promises to retiring baby boomers came due. No more. The Great Recession left gaping structural deficits that are already driving up debt as a share of GDP, unless there's an economic boom or big changes to policies. Crises used to be restricted to emerging countries. Not anymore. Wealthy Iceland needed the IMF's help in 2008 to avoid default, and Greece had to be bailed out by the IMF and the rest of Europe in 2010. When U.S. debt was last this high, in the early 1950s, it owed it mostly to its own citizens. Today, the United States owes half to other countries, who may be quicker to flee if they smell trouble. (To be sure, most of those foreign lenders are central banks with few alternatives for their money.) Keeping investors comfortable doesn't require the budget to be balanced overnight or that we pay all the debt back. It means ensuring that when the economy is healthy again, the debt will stop rising. That, however, means painful and divisive choices between raising taxes and shrinking government.

We can't count on a windfall from falling interest rates because they're already low. Discretionary spending is too small to trim the deficit much unless our military commitments shrink dramatically. Ent.i.tlements will have to bear the brunt, through less generous benefits, cost control, reduced eligibility for government programs, and higher payroll taxes.

Taxes will also likely have to rise to reduce the deficit. One popular proposal is to impose new taxes on consumption rather than income, with the idea of encouraging Americans to save. A higher gasoline tax, for example, would also reduce carbon emissions. Another possibility is a value-added tax, or VAT. A VAT is charged throughout the production process of a product. For example, a baker might pay $0.05 VAT on flour and collect $0.25 in VAT on bread that he sells to the consumer. After deducting the tax he paid, he submits the tax he's collected to the government: $0.20 in this case. Another way to raise revenue is to trim tax breaks, which add up to $1 trillion more per year and make the tax code complicated and inefficient.

It's hard to believe but politicians have made these sorts of tough decisions before. When Social Security teetered on insolvency in 1983, they raised payroll taxes and the retirement age. In the 1990s, they cut Medicare provider rates and changed welfare from matching state funding to block grants. George H. W. Bush in 1990 and Bill Clinton in 1993 raised taxes.

Unfortunately, these episodes may have taught politicians the wrong lesson. Bush lost re-election in 1992. In 1993, Marjorie Margolies-Mezvinsky, a first-term Democratic Congresswoman from a conservative suburban Philadelphia district, cast the deciding vote for Clinton's tax increase. It helped turn the budget around but she wasn't around to take credit: In 1994, voters threw her out.

Today, Congress is its most polarized since the 1920s. It is increasingly difficult for Democrats to countenance cuts to ent.i.tlements and even harder for Republicans to raise taxes. If events in Greece and in the United States are a guide, it will take a wake-up call from the markets, in the form of rising long-term interest rates, to get politicians to act.