|Economic Outlook logo 130|
President-elect Barack Obama faces the most difficult economic challenge confronting an incoming American president since the election of Franklin Delano Roosevelt seventy-six years ago in 1932. When he assumes office on January 20, Obama will need to act decisively with heretofore unprecedented fiscal policy steps, in conjunction with measures by the Federal Reserve to increase the money supply and lower long-term interest rates. All of this must be done to help contain and reverse the accelerating global slowdown by halting the rapidly deepening American recession. We can only hope that other national leaders and central banks will follow suit.
Adverse Feedback Loop Unleashed
The dismaying collapse of global housing and financial markets, evident in stock market losses of 40-60 percent and plunging real estate values that have occurred largely over the last six months, has erased over $30 trillion--one-third of global wealth. Little wonder, then, that the global economy has begun to contract sharply as the United States, Europe, Japan, and much of the rest of Asia have entered or are about to enter sharp and prolonged recessions.
A dreaded adverse feedback loop has manifested Fed chairman Ben Bernanke's fear of a self-reinforcing cycle in the United States of financial collapse followed by a sharp economic downturn and subsequent further financial collapse, and that pattern has gone global--suddenly and viciously. Even China has not been spared, as its projected growth rates for 2009 have dropped in the span of a month from 9 percent to below 5 percent, prompting the heretofore confident Chinese government to unveil, on November 7, a nominally massive fiscal stimulus package equivalent to about 17 percent of GDP over two years (about $2.4 trillion scaled to U.S. GDP). As is often the case, the Chinese exaggerate. When measured in terms of initiatives proposed in addition to those already in the pipeline, the Chinese stimulus package collapses to a far more modest 0.2 percent of GDP in 2009 and perhaps 2 percent of GDP in 2010--higher than had been scheduled, but far short of the 8.5 percent of GDP per year suggested by the initial announcement.
The tendency to announce huge headline numbers for fiscal stimulus or financial rescue packages has become widespread and is probably counterproductive. Exaggerated numbers, not accompanied by prompt action, only reinforce negative sentiments about an uncontrollable collapse in financial markets and economic activity. Just since early October, when the $700 billion U.S. Troubled Assets Relief Program (TARP) was launched, rescue plans have been announced in Europe, Russia, and Asia with a total value of another $2.1 trillion. The Fed's balance sheet has expanded from $800 billion to more than $2 trillion without boosting the money supply because bank deleveraging has collapsed the money multiplier, the banking system's ability to generate liquidity worth several times the value of its reserves or deposits. Meanwhile, the global financial meltdown has continued--even accelerated--as economic growth slows more sharply, signaling another round of downward pressure on financial and housing markets. While the headline numbers are large, the rescue packages will take time to implement, even partially, and more time to implement effectively without entailing massive waste that we can ill afford at this time. Beyond that, the size of these costs is modest relative to the scope of massive wealth losses resulting from a collapse of housing and equity values.
Fortunately, late in 2008, we are not entering the third year of a global depression, as was the case after the election of FDR in November 1932. But we are at a critical point at which insufficient action--either rooted in denial that a crisis exists or in wasted time resulting from poorly conceived measures with misleadingly large numbers attached to them--could land us in a global depression. So, too, could political bickering over the steps that need to be undertaken.
The situation is made even more tenuous by the delay in implementing necessary U.S. policy measures that is tied to the presidential transition. Roosevelt's initiatives--a boost in the price of gold (to produce reflation) and a capital injection into the banking system (like that already begun under the TARP)--were delayed until March 1933, when he was inaugurated president. Fortunately, the transition is shorter now with Inauguration Day on January 20, and, as noted, some capital injections to the banking system are already underway under the TARP program.
We must hope that on January 20 Obama will be ready to announce drastic stimulus measures. In the meantime, we must look to the Federal Reserve to provide further support for the economy by moving from interest-rate targeting to quantitative monetary easing as suggested on November 10 by--among others--Minneapolis Federal Reserve Bank president Gary Stern. I will return to these themes after surveying the evidence of an alarming collapse in the U.S. and global economies.
A sudden cessation of credit availability is disastrous for U.S. economic growth, driven as it has been by credit-financed consumption. Even in previous postwar recessions, on average, consumption growth has actually contributed a positive 0.7 percentage points to overall growth, thereby helping to mitigate the recessions' intensity. When a credit crunch occurs, however, consumption collapses and unemployment soars.
The last severe American credit crunch occurred after March 1980 when then-president Jimmy Carter imposed blanket credit controls on the U.S. economy, making credit virtually unavailable at any price. The economy collapsed almost instantly, much as it has done in the current credit crunch that began in mid-September. In March 1980, U.S. employment rose by 112,000. Then, just two months later, employment fell by 431,000. During four months, from March 1980 to July 1980, the unemployment rate, usually a slow-moving, lagging indicator, rose from 6.3 percent to 7.8 percent. From February 1980 to May 1980, the ISM Index of manufacturing activity plunged from 50.2 (indicating neither expansion nor contraction of manufacturing) to 29.4--a record low.
Today's economy is suffering not from an exogenously imposed credit crunch, but from a credit crunch tied to massive deleveraging by the banking system. Since mid-September, when the collapse of Lehman Brothers intensified the deleveraging process, the U.S. economy has been in free fall, much as it was in 1980. Employment, which had been falling gradually until the summer and was initially reported to have dropped by 73,000 in August, was revised downward sharply in August and September to produce a payroll drop over the three months ending in October totaling 651,000 workers. The unemployment rate has gone from 5.7 percent in July to 6.5 percent in October and, as in 1980, is headed sharply higher--probably to well over 7 percent by the end of the year. The ISM survey has dropped sharply from a break-even reading of 49.9 in August to a highly contractionary 38.9 in October. All of these measures are headed still lower.
The drop in consumption spending as credit has dried up for most households has been even more dramatic. Real consumption spending fell at a 3.1 percent annual rate during the third quarter--the weakest showing since the spring of 1980 after the panic that ensued following Carter's imposition of credit controls. Consumer spending is slowing even more rapidly in the fourth quarter. During October, U.S. auto sales dropped to a 10.6 million unit annual rate, corresponding to a 50 percent annualized decline over the three months ending in October. The Commerce Department's initial estimate reported that October retail sales fell by 2.8 percent, the largest decline since the department reinstituted this statistical series in 1992. Although part of that severe drop in October was due to sharply lower gasoline prices, the annual growth rate of nominal retail sales was -7.2 percent in the three months ending in October, portending annualized consumption growth probably in the vicinity of -5 or -6 percent during the fourth quarter, enough to subtract nearly four percentage points from overall growth. Sharply lower investment spending will subtract more, even if there are modest offsets from net exports and government spending.
The dramatic slowdown underway in the U.S. economy is problematic for three reasons. First, its suddenness and intensity will feed back negatively onto U.S. financial markets, causing the need for still more deleveraging by U.S. banks and reduced availability of credit. Second, the rapid drop in U.S. spending growth has been accompanied by a sharply intensified global slowdown. The significantly lower global growth will, along with a stronger dollar, steeply reduce the heretofore substantial positive contribution to U.S. growth from exports. An already noted sharp weakening of Chinese growth has dried up export markets for export-sensitive Asian producers. The collapse in oil prices attendant on the collapse in U.S. growth has similarly dried up financial flows from oil producers, especially Russia, into emerging markets such as Eastern Europe. European banks are heavy lenders to Eastern Europe and therefore deeply exposed to a collapse in those economies and to a sharp rise in their inability to service their debt to Western European banks. Another country, Iceland, that had been sharply boosted by a global financial bubble-boom has virtually defaulted on all its debt, while economic activity there has ground to a near-complete halt.
The third and perhaps most sinister problem emerging from another round of America's adverse feedback loop is a sharp rise in the risk of global deflation. Highly indebted households and businesses are already devastated by the lack of available credit. If deflation emerges--and some forward-looking market measures of price moves, coupled with a virtual collapse in global commodity prices, have already hinted at a sharp rise in deflationary expectations--the real burden of debt rises rapidly along with the demand for cash. A deflationary spiral must be avoided at all costs since such a development would virtually ensure a global depression.
U.S. Policy Measures
While the United States cannot singlehandedly stave off a global financial and economic crisis, decisive measures to break the adverse feedback loop described above constitute a necessary, if not sufficient, condition to contain the damage by reducing the risk of deflation. Between now and January 20, the Federal Reserve can step up efforts to contain the demand collapse that is weakening the U.S. and global economies. Interest-rate targeting has gone about as far as it can go. Although the Fed's official federal funds target is 1 percent, the effective fed funds rate on most days is merely a quarter of a percent. The reason is that the cash trapped in the banking system can only earn about ten basis points on short-term Treasury bills because households and firms are so frightened by possible insolvencies in the financial system that they seek the absolute safety of direct short-term claims on the U.S. government. Cash-stuffed banks choose to lend their funds to the Federal Reserve at a slightly higher rate--twenty-five to thirty basis points--than is currently available in the fed funds market.
The Fed needs to undertake quantitative easing, whereby it prints money and expands its balance sheet by direct purchases of longer-term Treasuries or mortgage-backed securities. The result of such quantitative easing would be to push mortgage rates down toward 4 percent from the current level of 6 percent. This would help to ease stresses in the mortgage market and thereby help to contain what is likely to be a reacceleration of the fall in house prices. Further, lower borrowing costs would help, at the margin, to stimulate spending by households that are not in acute financial distress. A broad-based quantitative Fed easing that adds directly to the liquidity of households and firms, while lowering key mortgage interest rates, would provide some systemic relief and thereby reduce the pressure for "special case" bailouts like those sought by insurance companies and automobile manufacturers.
Stimulative fiscal policy will be necessary to complement the demand stimulus that comes from quantitative easing. Tax rebates have already proven to be ineffective because nervous households, whose saving is far below desired levels, just tend to either save the rebates or use them to pay down debt. Spending measures such as public works (outlays to improve roads and bridges) do not provide the prompt income boost currently needed by distressed households and firms. Cost-effective implementation of such measures requires time, while hasty implementation entails waste of scarce resources.
The best available fiscal policy measure would be a sharp reduction in the payroll tax, which would boost household disposable income while giving firms an incentive to retain more workers on their payrolls. Total annual collections from households and firms of payroll tax levies total about $625 billion, about 7 percent of disposable personal income. A payroll tax is labeled as the primary means to finance Social Security and Medicare benefits, but those benefits are financed out of government revenues and would, of course, continue to be provided at their full level. The payroll tax is a poorly designed fiscal measure because it acts as a tax on employing labor and, in times of falling demand, a tax on retaining labor. The payroll tax is the primary tax paid by more than 60 percent of American households and so constitutes a marginal disincentive to further work.
If the payroll tax (of which households pay half directly) were suspended--say, for a year or eighteen months--households would experience an immediate 3.5 percent increase in disposable income that they could employ to sustain consumption and pay down debts. Since the payroll tax is regressive, falling more heavily on lower income households, its repeal would be progressive, while transferring a substantial increase in disposable income to the low-income households who are likely to need it most and therefore likely to spend most of it.
For firms, a reduction in their payroll tax payments would reduce their incentive to lay off workers by reducing the cost of keeping workers on the payroll. In effect, firms would be prompted to shift more toward labor as a factor of production because of a reduction in the tax on employment of labor that the payroll tax entails.
A payroll tax holiday is a radical measure. Opponents will claim that it constitutes a threat to maintaining Social Security and Medicare benefits. That claim would be unfortunate and untrue. The federal government is obligated to pay retirement and medical benefits whether it finances them out of a payroll tax, an income tax, or by additional borrowing, which would be the case in current circumstances. The sharp rise in disposable income that would result from a payroll tax holiday would constitute a far more effective fiscal stimulus than many of the other measures currently under consideration.
Those who claim that sharp increases in federal borrowing and the national debt would be ill-advised at the present time, when the economy is weakening while deflation threatens, have failed to study Japan's history in the 1990s. Japan ran annual deficits as high as 8-9 percent of GDP, the equivalent of over a trillion dollar deficit in the United States, while interest rates on government debt in that nation continued to fall even as total government debt relative to GDP rose above 100 percent--well above current projections for the U.S. debt-to-GDP ratio.
If a combination of quantitative monetary easing by the Fed and a payroll tax holiday that injects an additional $625 billion a year into the disposable cash flow of households and firms is undertaken promptly, the United States stands a fair chance of avoiding another devastating round of the adverse feedback loop, whereby a sharp economic slowdown intensifies the credit crisis. Such measures will be criticized as too radical and too potentially inflationary to be considered. We can only hope that, taken together, such measures will, at least, be reflationary so that we avoid a downward growth spiral that brings with it an abrupt slippage into deflation. That outcome, which would sharply increase the real burden of already excessive debt in the United States and worldwide, needs to be avoided at all costs. Should the aggressive monetary and fiscal measures entailed by quantitative easing and payroll tax cuts prove too stimulative, they can easily be terminated or reversed. The history of the Great Depression and Japan's experience with deflation in the 1990s show that, while it is not pleasant to deal with a bout of inflation, it is less costly than coping with a persistent and intensifying bout of deflation.
The Nonintervention Option
There is, I should point out, a tradition that would argue for a noninterventionist solution to the global financial and economic meltdown. The classical economists who argued against John Maynard Keynes's deflationary liquidity-trap dilemma and the aggressive fiscal measures he recommended were led by Arthur C. Pigou, who articulated the "Pigou effect" or "real balances effect." If the price level falls far enough, Pigou postulated, then the real value of cash balances would rise by enough to make consumers feel better off, and, therefore, they would spend more. Beyond that, sharply lower prices of current goods and services would attract more spending at the expense of future goods--saving--by consumers whose wealth had been enhanced by a sharp, deflation-induced rise in the value of their money balances. The scale and substitution effects for more spending on current goods both operate in the same direction.
There are some serious problems with Pigou's analysis. First, and perhaps most important, it ignores the existence of nominal contracts. With labor agreements calling for the payment of a certain level of dollar wages, a drop in the price level boosts real, inflation-adjusted wages, thereby raising the real cost of labor to producers--who then seek to cut employment--and exacerbating the reduction in employment already underway in a disinflationary or deflationary environment. Of course, a prompt fall in money wages would eliminate the negative substitution effect operating against employing labor as real wages rise. But, in fact, as Keynes and many others have observed, wages are sticky downward. Workers and unions strongly resist wage cuts with the force of existing labor contracts, which often call for rising money wages over a period of years. Such contracts made sense as a way to maintain the real purchasing power of wages in a period of rising prices, but they are highly destructive of employment in a prolonged period of falling prices. Most contracts simply do not envision a period of persistently falling prices, like that required for the Pigou effect to cause aggregate demand to rise.
In addition to labor contracts, most contracts between borrowers and lenders are written in nominal (current dollar) terms. Mortgages are an excellent example. A ten-year fixed mortgage specifies a fixed nominal interest rate. At, say, a 6 percent fixed interest rate, the "real" cost of borrowing is 4 percent, given a 2 percent inflation rate that reduces the real value of the payments specified by a fixed-rate mortgage.
Since a mortgage is tied to a particular dwelling, its "real" cost is more accurately the nominal mortgage rate minus the annual rate of increase of the price of the house or property that is financed by the mortgage. During the housing boom, real estate boosters were fond of pointing out that a 6 percent mortgage interest rate on a property rising by 10 percent a year is a -4 percent real cost of borrowing. The real burden of debt is reduced by inflation and reduced even more by inflation that is higher than had been expected when the mortgage specifying a fixed nominal rate was written. As house prices accelerate upward while mortgage brokers snap up mortgages and package them into derivative securities acquired by risk-hungry investors, whose purchases of mortgage-backed securities keep mortgage rates from rising, the "real" cost of financing a home falls. As a consequence, more buyers rush in to push up the price of housing to unsustainable, unaffordable levels when the bubble bursts.
When the 10 percent housing inflation rate rapidly turns into a 10 percent housing deflation rate (as it did starting in 2005), the real cost of a 6 percent fixed-rate mortgage flips from -4 percent to 16 percent. The homeowner still owes 6 percent a year on the mortgage contract tied to an underlying asset that is losing value at 10 percent a year--far more costly than when the asset was rising in value at 10 percent annually. It is little wonder that millions of homeowners with large mortgages whose payments are fixed in current dollars want to renegotiate their mortgages.
If the overall price level--not just the price of houses--were falling at 2 percent annually, the real burden of debt would rise for all households and firms. Six percent nominal interest rates that were contracted for based on an expected 2 percent inflation rate (that is, an expected real borrowing cost of 4 percent) would, given 2 percent deflation, become 8 percent (six nominal plus two from deflation) in real terms--double the expected real cost of borrowing.
For a fully indexed economy with fully flexible prices, deflation is not so problematic. But with price rigidities, deflation can wreak havoc. With a moderate deflation rate of 2 percent, a typical consumer would see his real wage rising by 2 percent and so could, perhaps, afford the higher real cost of borrowing. But suppose falling prices boost real wages and result in a layoff for a consumer with a nominally fixed 6 percent mortgage? He has no income and a mortgage with a suddenly much higher real interest rate. He defaults.
Nominal contracts and sticky prices and wages make the persistent deflation required for the Pigou effect to operate highly problematic. This is another way of saying something more obvious. Since deflation is an unusual phenomenon, most contracts do not incorporate it into their terms. Long-term nominal contracts for wages or mortgage payment in fixed-dollar terms--or rising dollar rates, as with a multiyear labor contract that accounts for typical inflation--all have embedded in them an implicit, expected inflation rate. That is why central banks aim for low and stable inflation levels.
Even without nominal rigidities tied to contracts, a persistent and accelerating negative inflation rate can be dangerous in a postbubble economy. In theory, there is a price level that will clear a market--goods, labor, bonds or money, and foreign exchange. But the move to a lower price level can be highly disruptive because of the existence of cash--a riskless non-interest bearing claim on the government. As deflation accelerates, the demand for cash increases since the deflation rate itself becomes the riskless real return on holding cash. Higher deflation means a larger demand for cash and more selling of real assets in order to add to cash balances. That selling, in turn, intensifies the rate of deflation, which accelerates the rise in the real cost of borrowing and employing labor.
Avoid Deflation at All Costs
Deflation is a dynamically unstable, self-reinforcing process that must not be allowed to take hold, not only because of the huge dislocations it creates in labor and credit markets, but also because it can accelerate rapidly to a pace that no feasible recontracting can accommodate. The overshooting of deflation to the downside on the way to what may ultimately be Pigou's equilibrium price level in a frictionless, static world is a bungee jump that no economy heavily reliant on nominal contracts with no provision for deflation could withstand. Let us hope that Bernanke and Obama reach the same conclusion very soon.
John H. Makin (firstname.lastname@example.org) is a visiting scholar at AEI.