CHAPTER FIVE
NO CENTRAL BANKER HAS HAD to adapt his views more under the public eye than Ben Bernanke, the chairman of the Federal Reserve Board. In February 2004, in a speech to the Eastern Economic Association, Bernanke, then a governor of the Federal Reserve Board, spoke of the “Great Moderation,” the observation that the fluctuations of output and inflation in industrial countries had come down steadily since the mid-1980s. Because the Holy Grail of economic management is strong, steady growth, without booms, busts, or high inflation, this trend suggested that something was working.
Bernanke considered three possible explanations: first, that we might have just been lucky, with the world economy experiencing fewer accidents such as war and oil-price increases over this period. Second, that economies had changed, for example as corporations developed systems to acquire sales information more quickly and to translate it more continuously into production and inventory decisions. Such improvements could explain how economies had been able to avoid the more dramatic inventory buildups and production cutbacks that had characterized previous recessions. Third, as a result of advances in our economic understanding, central bankers, many of them former academic economists, understood better how monetary policy affected economic output.
Because he is a careful economist, in addition to being a very good one, Bernanke suggested that there was merit in all three explanations. However, he stressed the view that monetary policy had become much better. Unlike the policy makers in the 1960s and 1970s, who operated with rudimentary and often incorrect beliefs about economic relationships, today's central bankers, he felt, understood far better how the economy works. Bernanke is, if anything, more cautious and nuanced than the typical policy maker, but the overall tone of his speech was triumphant: the policy levers for managing a modern economy were well understood, which was why we already had milder recessions. The implication, perhaps unintended, was that with steady progress, we could do away with recessions altogether.
By September 2008, however, Bernanke, now chairman of the Federal Reserve Board, having realized the limits of monetary policy, was pleading for help from Congress, arguing that “despite the efforts of the Federal Reserve, the Treasury, and other agencies, global financial markets remain under extraordinary stress. Action by Congress is urgently required to stabilize the situation and avert what otherwise could be very serious consequences for our financial markets and for our economy.”1 In short, monetary policy was not working, and only a bailout of the financial system by Congress could stabilize the economy and avert a depression. Where had the Fed gone wrong?
With the benefit of hindsight, it appears that the Federal Reserve made two mistakes. First, the jobless recovery from the recession of 2001 induced the Fed to keep interest rates extremely low for a sustained period. A lot of excesses were building up, in the rest of the world as well as in the United States, but theory and politics conveniently came together to keep the Fed on hold. Second, the Fed actively encouraged the financial markets to believe it would follow an asymmetric policy: it would not lean against a potential unsustainable rise in asset prices, but it would remain ready to pick up the pieces if a bubble burst. Both these implied promises did considerable damage, because in attempting to stimulate sluggish job creation, they set off an orgy of financial risk taking.
Unfortunately, because the Fed's actions were consistent both with its mandate and with the prevailing academic orthodoxy, it has not been forced to rethink its policies. Moreover, in an environment of high and persistent unemployment, the political pressure on it to persist with such policies will make change very difficult.
The Federal Reserve has a mandate from Congress to promote a healthy economy. This means maintaining maximum sustainable employment and stable prices. Also, it has been entrusted since its founding in 1913 (in the wake of the Banker's Panic of 1907) with helping to ensure the stability of the financial system.2
In the past, economists believed that the components of the healthy-economy mandate—the goals of maximum sustainable employment (high growth) and stable prices (low inflation)—were incompatible over the long run, because high growth might require high inflation. Implicit in this trade-off (known as the Phillips curve after William Phillips, who found such a relationship in the U.K. economy between 1861 and 1957) was the belief that you could fool all of the people all of the time. Injecting more inflation would lead people to believe they were getting paid more for the goods they produced and to work harder—thus expanding output—not realizing that everything else was becoming costlier at the same time.
In the late 1960s and early 1970s, even as data suggested the Phillips-curve relationship between inflation and unemployment was breaking down, the “rational expectations” revolution started taking hold of monetary economics. It explained why the Phillips-curve relationship was theoretically untenable. The essential idea was that the public understood the objectives of policy makers and the frameworks they operated with, so they would not cooperate by being fooled. If the central bank had a policy of inducing high inflation, producers would rationally expect that all prices would go up and would not exert more effort when they saw the prices of their own products go up. Rather, they would understand that the additional dollar they earned was actually worth less in terms of its ability to purchase goods and services. The long-run level of employment of the economy would be determined by factors like the business climate, incentives to innovate, and the ability of firms to hire or lay off workers easily, not by inflation.
This view eliminated the incompatibility in the long run between the economic goals of low inflation and maximum sustainable employment. According to the new orthodoxy, by keeping inflation low and thus eliminating all the uncertainty and distortions associated with high and variable inflation, central bankers would give the economy its best chance of achieving its potential growth rate and thus maximum sustainable employment. However, there is still a short-run trade-off between growth and inflation, stemming from the notion that every economy has a potential growth rate—an inbuilt maximum safe speed. Make the economy go any faster, and wages and inflation start accelerating because demand exceeds productive capacity; slow it down, and wages and inflation start falling. When the potential growth rate is reached, the economy is effectively at maximum sustainable employment—all the unemployed are either fully occupied in searching for appropriate jobs or are unemployable—and any effort to further accelerate growth will only increase competition and wages for employed workers, and thus inflation. So the ideal central-bank policy is to keep the economy perpetually at its potential growth rate.
Unfortunately, no one really knows what the potential growth rate is, though they have reasonable guesses. And this rate can change if the structure of the economy changes—for example, if the industries that are dominant in the economy change. The best indicator for a central banker is inflation. A rise in the inflation rate indicates that the economy is exceeding the speed limit; if the inflation rate is falling, the economy can benefit from more stimulus. Of course, because monetary policy operates with lags—raise interest rates now, and the effects are felt in the economy only many months later—a central bank that waits until it stares inflation in the eye before withdrawing stimulus has waited too long. Therefore, central banks attempt to project what their policies will do (typically over a two- to three-year horizon) and adopt policies that will keep future inflation close to a target and thus maintain growth close to potential.
In the years before the crisis, central bankers and academia thus converged on variants of targeting inflation as their primary objective. Of course, they also had to consider the objective of financial stability. According to conventional wisdom, central bankers had only one instrument with which to carry out monetary policy—the short-term interest rate—and they could not target more than one objective with it. Concerns about financial stability would complicate and make less intellectually rigorous the process of setting monetary policy. Financial stability was left to be tackled through “prudential” measures like capital requirements and relegated to the less glamorous supervisory and regulatory arms of central banks.
The Fed conducts monetary policy largely through the short-term interest rate (the overnight federal funds rate). It sets this rate by intervening in the inter-bank market for reserve money. Through this rate, the Fed hopes to influence the long-term interest rate. According to the most commonly held economic view, investors in the market see the long-term interest rate (say the ten-year Treasury bond rate) as being a function of the sequence of the short-term interest rates that are expected to prevail over time. So if the short-term interest rate is expected to remain low over the next ten years, the long-term interest rate will be low, whereas if the rate is expected to be low only for the next two months and then climb to a higher plateau, the long-term interest rate will be high. This reasoning is known as the expectations hypothesis. By holding down the short-term rate, especially if the market believes it will be held low for a sustained period, the Fed can influence expectations of the future short-term rate and hence the long-term interest rate.
Long-term interest rates are extremely important in the economy. A lower long-term interest rate increases the value of long-term assets such as equity, bonds, and houses because dividends, interest payments, and the services provided by the house are discounted at a lower rate. It thus increases household wealth and, consequently, spending. A low long-term interest rate also makes it less attractive for households to save and more attractive to consume, thus again contributing to demand. Finally, long-term interest rates determine the profitability of real investment: lower long-term interest rates make today's value of future profits higher, giving corporations more incentive to invest as well as greater ability to borrow.
The short-term interest rate may also have direct effects on economic activity. Many borrowing rates are tied to short-term interest rates: for example, the interest payment on an adjustable-rate mortgage falls if the Fed cuts interest rates, leaving more money for households to spend. Through a low policy rate, the Fed may also signal to the market that it intends to keep liquidity conditions—that is, the ability to borrow—easy over the foreseeable future. Banks and finance companies then have the incentive to make illiquid term loans, confident that they can refinance from the market.
What I have outlined is the conventional view of how monetary policy works. Let us now see how the Fed responded to the dot-com bust and the recession in 2001, and what the conventional view may have missed out.
After the crash in the NASDAQ index in 2000–2001 and the recession that followed, the Federal Reserve tried to offset the collapse in investment by cutting short-term interest rates steadily. From a level of 6½ percent in January 2001, interest rates were brought down to 1 percent by June 2003. Such a low level, unprecedented in the post-1971 era of floating exchange rates, sent a strong signal to the economy. House purchases picked up as more people found they could afford the lower mortgage payments. Increased housing demand encouraged more home construction, which was already being given a boost by the low interest rates at which developers could borrow.
Output growth, riding on productivity growth, was strong, but jobs were really what the public and politicians wanted. Growth by itself did not put food on the table, pay bills, or reduce anxiety for those who were unemployed and seeing their benefits running out, or for those who feared for their jobs. Unfortunately, as we have seen, job growth simply did not pick up. Industrial and service companies continued pruning workers, and the new jobs in construction did not offset job losses elsewhere. Unemployment peaked only in June 2003, long after output growth had resumed and the recession was officially over.
With inflation low and unemployment high, the Fed's healthy-economy mandate suggested it should keep interest rates low. Indeed, given the level of unemployment and the consequent slack in the economy, Fed officials, including Ben Bernanke, openly worried about the possibility of deflation, even in mid-2003, when quarterly GDP growth was around 3 percent.3 The Federal Reserve seemed to be influenced by the recent experience of Japan, which had faced prolonged price deflation and slow growth in the 1990s as a result of the collapse of its real estate bubble. But this concern was misplaced: unlike Japan, the United States in 2001 had not experienced a debt crisis, only a meltdown of the overvalued tech stocks. A debt crisis could have caused a downward spiral of bankruptcies, job losses, and fire sales that might have triggered a deflation. But the effects of a stock meltdown were, and historically have been, much milder.4 Consumer price inflation in the United States never fell below 1 percent over this period, despite downward pressure from low-cost imports (we do not, of course, know what it would have been without the easy Fed policy); and more important, future expectations of inflation were firmly above 1 percent and nearer 2 percent, the Fed's unofficial target. Indeed, the disinflationary pressures at that time may well have arisen because foreign competition was forcing U.S. producers to become more productive as well as to keep wage increases limited, rather than because demand was excessively low.
By mid-2003, almost every measure of economic activity other than inflation and unemployment was picking up strongly. Demand in the United States was strong: the United States’ trade deficit, a measure of the demand in the United States that was being satisfied from abroad, was widening rapidly. Indeed, one reason that the pace of U.S. job growth was especially slow in manufacturing may have been that countries outside the United States, like China and Japan, were resisting the appreciation of their currencies against the weakening dollar, thus ensuring that their exports continued to be competitive in the U.S. marketplace. The Fed was now effectively adding stimulus to a world economy that was growing strongly, with jobs being created elsewhere but not in the United States. Commodity prices around the world started a steady rise, suggesting that worldwide economic slack was decreasing. If the Federal Reserve, the world's central banker in all but name, had been focused on sustainable world growth, it should have been tightening monetary policy by raising interest rates. But its mandate covered only the United States.
John Taylor of Stanford University has pointed out that even measured against what is known as the Taylor rule (an empirical characterization of past Federal Reserve interest-rate policy, which sees the short-term policy rate as a function of the inflation rate and the gap between the output the economy is capable of and what it actually produces), the Fed should have started raising interest rates by early 2002.5 But it continued to reduce rates until as late as June 2003. In a speech in 2010 at the American Economic Association's annual meetings, Ben Bernanke defended Fed policy, saying that if inflation was properly measured, the Fed had not departed from the Taylor rule during this period. In truth, the problem was that output growth had not resulted in job growth. And the Fed was focused not on output, as the Taylor rule would suggest, but on jobs.6
When the Fed finally started to raise rates in June 2004, it was extremely fearful of killing off a nascent jobs recovery. So it took pains to accompany its rate hikes with announcements that interest rates would be low for “a considerable period” and would rise slowly at a “measured pace”—namely, 25 basis points at every scheduled meeting of the board. This strategy clearly helped keep long-term interest rates low, but not because expectations of future short-term rates came down, as the expectations hypothesis would suggest. Instead, the risk premium on long-term government bonds—the additional spread that the market demands to take the risk of bond prices fluctuating—fell even as the Fed raised short-term interest rates, with the result that long-term interest rates fell and bond prices rose.7 Indeed, a generally low premium for risk ensured that the prices of all risky or long-term assets, including housing, rose, even as the Fed raised rates slowly. The Fed's policy seemed to be working because it made risk more tolerable!
The Fed did worry about the deteriorating quality of lending and made some supervisory noises over time. But with its foot pressed firmly on the interest-rate accelerator, the supervisory measures were ineffective. Ultimately, it was probably also Fed actions that brought the party to an end. Higher short-term interest rates raised the payments on adjustable-rate mortgages as well as prospective payments on mortgages with rate resets. With demand for housing falling off (and Fannie and Freddie eventually tempering their purchases of mortgage-backed securities), house prices stopped rising. As a result, over-extended borrowers found it hard to refinance before the initially low “teaser” rates on their mortgages expired, and households began to default on payments. Foreclosures caused house-price declines, and the whole momentum of the boom was reversed. The Fed, as we now know, intervened too late. The bubble had inflated enormously, and the ensuing bust has been extremely painful.
With the benefit of hindsight, it easy to suggest that the Fed made mistakes even in the traditional conduct of monetary policy: for instance, it may well have overestimated the risk of deflation. In some ways, though, the threat of deflation seemed to be a low-probability red herring, put forth to explain why the Fed kept rates on hold. The true problem was unemployment, which made raising rates politically impossible. In the past, when economic growth and job recovery coincided, this was not an issue. With jobless recoveries though, growth and jobs became somewhat divorced. The Fed would have to be on hold for a long time if it wanted to see jobs reappear.
For those who believe the Federal Reserve is too independent, the notion that it is subject to political pressure may seem unthinkable. Yet, as Fed governors admit in private, pressure is applied all the time by Congress. Powerful politicians, in off-the-record conversations with Fed governors, frequently make veiled and not-so-veiled threats to scrutinize Federal Reserve activities and reduce its independence unless the Fed complies with their wishes. Although typically they stay off monetary policy, their desires are not hard to read when unemployment is high. Ironically, the Federal Reserve's desire to remain independent is the lever with which Congress makes it compliant.
At this time, however, the Fed needed little convincing to keep interest rates low. The prevailing orthodoxy suggested that Fed policy makers should worry only if inflation was getting out of hand. And it was not: inflation in the prices of goods (like cars and milk) and services (like haircuts and laundry) was quiescent, and indeed, if anything, the Fed feared deflation. So the Fed was free to focus on the second part of its mandate, full employment. Yet even while the Fed attempted to convince unwilling corporations to invest through ultralow interest rates, the prices of financial assets and housing were skyrocketing. But the orthodoxy suggested asset prices could be ignored.
Rapidly rising asset prices should have sounded alarm bells. They were driven by a number of forces other than the traditional ones: increased risk taking, more foreign money looking for debt claims, and expanding credit.
Low short-term interest rates pushed investors to take more risk, for a number of reasons.8 Some institutions, like insurance companies and pension funds, had contracted long-term liabilities. At the low interest rates available for safe assets, they had no hope of meeting those liabilities. Rather than falling short for sure, they preferred to move into longer-term riskier bonds, such as mortgage-backed securities, that paid higher interest rates. In addition, as long-term interest rates fell and the value of stocks, bonds, and housing rose, households felt wealthier and may have felt the confidence to take more risks. Some of these choices may have been irrational. As my colleague Richard Thaler has argued, when gamblers win money, they take more risks, because they treat their earlier winnings as “house” money—not their own—and therefore less important if lost. Whatever the reason, with investors more willing to take risks, the risk premium on all manner of assets came down.
One effect of the search for yield was that money moved out of the United States into other countries, especially into the high-yielding bonds, stocks, and government securities in developing countries. But many of these countries were fearful of losing out in the race to supply goods to the U.S. market. Their central banks intervened to hold down the value of their currency by buying the U.S. dollars that were flowing into their countries from the domestic private entities that had acquired them and reinvesting these dollars in short-term U.S. government bonds and agency bonds.9 Thus, even as the Fed pushed dollars out, central banks in developing countries pushed them back in. In a number of industrial countries, private entities recycled the dollar inflows: German banks and Japanese insurance companies bought seemingly safe U.S. mortgage-backed securities with the dollars their customers deposited. The money leaving the United States looking for riskier assets around the world thus came back to the United States, looking for seemingly safe but higher-yielding debt-like securities. In some ways, Federal Reserve policy was turning the United States into a gigantic hedge fund, investing in risky assets around the world and financed by debt issued to the world.10
Credit also expanded. Rising asset prices themselves gave households and firms the collateral with which to borrow—a channel that Chairman Bernanke himself had pointed out when he was a professor at Princeton University.11 Indeed, much of the financing of low-income borrowers was predicated on house prices rising and borrowers refinancing once the low teaser rates ran out. Thus a higher house price, rather than increased income, was the means through which borrowers would keep themselves current on payments.
In addition, the promise that liquidity would be plentiful over the foreseeable future meant that bankers were willing to make longer-term illiquid, and hence risky, loans.12 But with firms unwilling to invest, banks went looking for deals that would create demand for loans.13 One option was for private equity investors to acquire firms, relying on banks to finance the deals. Banks, in turn, packaged the loans they made—creating collateralized loan obligations (CLOs) —and sold debt securities against them, thus obtaining the funds to make yet more loans. The result was that larger and larger leveraged acquisitions were proposed to satisfy the seemingly insatiable investor hunger for debt claims. As the rush to lend increased, lending standards declined rapidly: the classic signs of the frenzy were “covenant-lite” loans, bereft of the traditional covenants banks put in to trigger repayment if the borrower's condition deteriorated, and pay-in-kind bonds, schemes by which borrowers who could not pay interest simply issued more bonds. As I argued earlier, the recent crisis was not caused only by lending to the poor!
Rising asset prices would not be a problem if markets were well behaved and kept asset prices tied to fundamentals. In the case of housing, prices should be a function of interest rates, local demographics, household incomes, and local zoning regulations constraining the supply of housing.
Unfortunately, asset price growth can be self-reinforcing. For instance, higher house prices give existing homeowners home equity that they can borrow against to make the down payment for better houses, leading to a rise in prices for those houses as well. And a history of house price growth can lead naive new home buyers to swallow their real estate agent's sales pitch and put their money down expecting the price appreciation to continue. Indeed, for a while such expectations may be logical, because there are many more existing homeowners with enough home equity to move up.
In most markets, savvy investors can take a contrarian position when prices depart too much from fundamental value. In the housing market (as well as in the market to take firms private), few opportunities exist for investors to take a short position—that is, sell houses they do not have so as to make a killing when prices fall. This typically means that the optimistic, who buy housing, tend to have undue influence.14 So house prices, and more generally, asset prices, can rise excessively, and their reacquaintance with reality can be brutal indeed.
Central bankers argue that they really should not be in the business of figuring out when asset prices are too high: after all, do they really know that much more than market participants? This not a silly argument. Many markets work well by themselves, and introducing the whims and fancies of the central bank governor into the way prices are determined could create more problems than it solves. But history warns that markets such as housing, which are driven by bank lending, are different: not only are they very thin (relatively few house sales determine the value of housing for the whole country), but they also do not allow for investors to take short positions. Prices in these markets can run away from fundamentals. And the adverse spirals associated with house-price busts can be very damaging indeed : as prices fall, lending vanishes, and people cannot repay their mortgages; thus foreclosures increase, and prices drop further.
The key warning signal of unsustainable growth in asset prices is an accompanying growth in credit.15 Before the crash of 1929, the warning signal was the growth in margin loans against shares even as stock prices increased. Before the most recent recession, alarm bells should have sounded in every central bank meeting as a boom in real estate lending accompanied house price growth, and lending to private equity grew with ever-higher transaction prices. Indeed, credit growth has historically been one of the factors determining how central bankers set policy interest rates; but in recent years, academics have persuaded many of them that such behavior is archaic. To their credit (no pun intended), the European Central Bank and some developing-country central banks, like the Reserve Bank of India, have continued to pay attention to credit growth in determining their monetary policy.
Rapid credit growth was deemed of importance in the past partly because it was thought to presage inflation and partly because it reflected a possible deterioration in the quality of credit. Academics argued that the links between credit growth and inflation were tenuous (here they were right) and that credit problems were a historic curiosity in industrial countries because of improvements in bank management and supervision (here they were obviously wrong).
A second argument central bankers offer is that in the midst of a frenzy, when investors expect double-digit rates of price growth, raising rates by a fraction of a percent is ineffective.16 There are two difficulties with this argument. First, the key issue is expectations. If the central bank can convince investors that it is serious about fighting asset-price inflation—in the same way as it convinces them it will fight goods-price inflation—expectations about price growth can deflate fast, especially in the early stages of a bubble. Put differently, small changes in the central bank interest rate can affect expectations about price growth considerably. The fact that asset prices are growing at double-digit rates does not mean that policy rates have to be raised commensurately. Second, bubbles develop based on a kind of “greater fool” theory—that even if an asset is already trading at an inflated price, someone will be willing to buy it at an even more inflated price. By signaling that it will tighten liquidity conditions, and thus constrain financing and trading, the central bank can signal to investors that there will be fewer fools out there with the capacity to buy, making it more difficult for the bubble to grow.
Indeed, instead of discouraging the development of bubbles, the Fed encouraged it through an implicit commitment, which might have done far more damage than any other Fed action. This commitment, the so-called “Greenspan put,” essentially said that the Fed could not really tell when asset prices were building up into a bubble, and so instead the Fed would ignore asset prices but stand ready to pick up the pieces when the bubble burst. To understand why this commitment was made, we need to go back to 1996.
In late 1996, the Fed chairman, Alan Greenspan, an astute and experienced (though somewhat ideological) economist, became concerned about the high level of the stock market. In a famously brave speech at the American Enterprise Institute, he wondered whether the central bank should also worry when the prices of real estate, equities, and other earning assets were rising rapidly. And in the opaque language that he had perfected, he came as close as a central banker can to saying he thought stocks were overvalued:
But how do we know when irrational exuberance has unduly escalated asset values, which then become subject to unexpected and prolonged contractions as they have in Japan over the past decade?…We as central bankers need not be concerned if a collapsing financial asset bubble does not threaten to impair the real economy, its production, jobs, and price stability.…But we should not underestimate or become complacent about the complexity of the interactions of asset markets and the economy. Thus, evaluating shifts in balance sheets generally, and in asset prices particularly, must be an integral part of the development of monetary policy.17
In his autobiography, Greenspan admits wondering whether the market would understand what he was getting at.18 It did—and ignored him! The stock market opened substantially lower the next day but regained its losses in a day. And it was right to ignore him, because the Fed did not follow up Greenspan's concern with an increase in interest rates, even though he had hinted at such action in his speech. Greenspan never explained why he did not act: quite possibly his hand was stayed by the furious reaction he engendered when market participants realized he was trying to talk the market down.
Instead, the Fed watched while stock prices continued rising in the dot-com boom, as companies without earnings or even revenues sold shares at astronomical prices based on the number of “eyeballs” they attracted to their websites. The Fed even cut rates following the Russian debt default in 1998 and the collapse of the hedge fund Long-Term Capital Management, and raised interest rates mildly starting only in 1999.
When the stock market eventually crashed in 2000, the dramatic initial response by the Fed ensured that the recession was mild even if job growth was tepid. In a 2002 speech at Jackson Hole, Alan Greenspan now argued that although the Federal Reserve could not recognize or prevent an asset-price boom, it could “mitigate the fallout when it occurs and, hopefully, ease the transition to the next expansion.”19 This speech seemed to be a post facto rationalization of why Greenspan had not acted more forcefully on his prescient 1996 intuition: he was now saying the Fed should not intervene when it thought asset prices were too high but that it could recognize a bust when it happened and would pick up the pieces.
The logic was not only strangely asymmetrical—why is the bottom easier to recognize than the top?—but also positively dangerous. It fueled the flames of asset-price inflation by telling Wall Street and banks across the country that the Fed would not raise interest rates to curb asset prices, and that if matters went terribly wrong, it would step in to prop prices up. The commitment to put a floor under asset prices was dubbed the “Greenspan put.” It told traders and bankers that if they gambled, the Fed would not limit their gains, but if their bets turned sour, the Fed would limit the consequences. All they had to ensure was that they bet on the same thing, for if they bet alone, they would not pose a systemic threat.
Equally important, the willingness to flood the market with liquidity in the event of a severe downturn sent a clear message to bankers: “Don't bother storing cash or marketable assets for a rainy day; we will be there to help you.” Not only did the Fed reduce the profitability of taking precautions, but it implicitly encouraged bankers to borrow short-term while making long-term loans, confident the Fed would be there if funding dried up. Leverage built up throughout the system.
For a long time, central banks justified not focusing on asset prices by arguing that if Alan Greenspan had acted on his intuition in 1996, he would have snuffed out a boom that, despite the slump in 2000, took the stock market and U.S. household wealth to unprecedented heights. On March 2, 2009, though, the S&P 500 closed at 700, below its level of 744.38 on the day in 1996 when Alan Greenspan made his fateful speech. Of course, to date it has regained substantial ground, but perhaps Greenspan could have averted thirteen years of lost returns if indeed he had backed his words with action on interest rates. Whether the political system would have allowed him to do so is, of course, another matter.
The recent recession has started some rethinking on the objectives of monetary policy, though even as I write, the Fed is keeping interest rates at rock-bottom levels because unemployment is high, even while all manner of asset prices are rising. The saving grace today is that credit growth is still tepid, and it is unlikely that we will have another housing boom while memories of the last one are still fresh. But the financial sector is, if anything, innovative, even in the ways it gets into trouble!
I said earlier that academics and central bankers had converged on the view that there is no incompatibility between the objectives of seeking maximum growth and keeping inflation low in the long run. But there does seem to be some incompatibility between the monetary policies that encourage real investment and growth—maintaining predictably low interest rates over a sustained period and expressing a willingness to flood the market with liquidity when it is tight—and the monetary policies that discourage the coordinated one-way bets by financial market participants that have proved so damaging—pursuing unpredictable policies with no assurance of liquidity support.
The argument that monetary policy has no role in leaning against asset-price bubbles is both timid and self-serving, and it takes the Fed out of a key role it can play in assuring financial stability. Of course the Fed should proceed cautiously and lean against an incipient bubble only when there is substantial evidence that it exists, tempered by the knowledge the fears of a bubble could be baseless. To resign the role of party pooper, however, is to buy political acceptability at great risk to the economy.
More controversial is whether the Fed should cut policy interest rates to rock bottom in order to revive the economy. Although such an action seems costless, it imposes an enormous cost on savers and offers an enormous windfall to debtors, especially banks. Because it is a relatively hidden transfer, it elicits little comment or protest, especially as well-off savers tend to keep their heads down at times of crisis. But it is a transfer nevertheless, amounting to hundreds of billions of dollars a year. Moreover, it offers a one-way bet to bankers: plunge the system into trouble, and they will get a great deal on interest rates. Finally, it is not clear that ultralow nominal interest rates (around 0 percent) offer a significantly greater incentive for firms to invest than merely low interest rates (2 to 3 percent), but the difference in risk taking between ultralow and low interest rates could be enormous.
More damaging still is the Fed's ongoing attempt to prop up housing prices, both indirectly through low interest rates and directly by lending into the housing market. Although such support is justified as a way to allow the bubble to deflate slowly, it contributes to prolonged delays in adjustment in the housing market. Instead of homeowners and lenders biting the bullet on losses and moving on, they have the incentive to wait and see. But so long as there is a prospect for further adjustment, buyers, too, stay out of the market. And unless the over-supply in the housing market is cleared out, builders have little incentive to resume construction. The Fed could be not only delaying the recovery of the housing market but also reinforcing the sense that it will not get in the way of price increases but will prevent price falls. The Greenspan put is quickly becoming the Bernanke put.
In sum, the Fed's conduct of monetary policy between 2002 and 2005, while roundly criticized by all but central bankers and monetary economists (with notable exceptions), had two important limitations. First, it was fixated on the high and persistent unemployment rate and did its best to bring it down by trying to encourage investment. It signaled that it would keep rates low for a sustained period and offered the Greenspan put if firms were still not convinced. Critics should recognize that this fixation was in full accord with its mandate and, more important, that there would have been political hell to pay if it had raised interest rates much earlier than it did. This policy, however, may have had a greater effect on credit growth and asset prices than on job creation outside the real estate industry: corporations were still working away the excesses of the dot-com boom.
Second, the dominant academic orthodoxy indicated that so long as inflation was quiescent, central bankers had nothing to worry about. Indeed, to worry was to destroy the purity of the theoretical system that had been built, for that would admit of multiple objectives and lead to market confusion. Instead, central bankers should keep their eyes fixed on inflation (or the lack thereof) and let bank supervisors worry about risk taking. Unfortunately, the supervisors had been muzzled, this time on the ideological grounds that they would do more harm than good by restraining the private sector.
The bottom line is that the debate over monetary policy, which was once thought settled, will have to be reopened again. Among the most pressing issues are the trade-offs between policies intended to generate investment and employment and policies intended to ensure financial stability. Asset-price inflation will have to enter the policy debate. Moreover, the Fed will have to consider whether it is setting policy only for the United States, or, in reality, for a much larger global economy. Much needs to be done.
This is as good a point as any to try to understand the failings of academic economists in the macroeconomic sphere. Many commentators have gone overboard in poking fun at economists’ models, deriding them as oversimplified. Others wonder about the excessive mathematical complexity of some modeling, and yet others combine the criticisms by arguing that human behavior is too complex to be captured by mathematical models.
The most realistic model would be one that details all individuals and their whimsical behavior, and all institutions, but it would be hopelessly complex and of little value in analysis. The whole point of economic modeling is to create useful simplifications of the economy that allow us to analyze what might happen under varying policies and conditions. The test then is whether the model is a useful simplification or an oversimplification.
Many past macroeconomic models had a single representative agent making all decisions. The representative-agent models were easy to work with and did offer useful predictions about policy, but they took for granted the plumbing underlying the industrial economy—the financial claims, the transactions, the incentive structures, the firms, the banks, the markets, the regulations, and so on. So long as these mechanisms worked well, the models were a useful simplification. And during much of the “Great Moderation” that Bernanke referred to, the plumbing worked well and served as a good basis for abstract reasoning.
But as soon as the plumbing broke down, the models were an oversimplification. Indeed, the models themselves may have hastened the plumbing's breakdown: with the Fed focused on what interest rates would do to output rather than to financial risk taking (few models had a financial sector embedded in them, let alone banks), financial risk taking went unchecked.
In a haunting parallel to Robert Lucas's famous critique of Keynesian models, in which he argued that those models would break down because modelers did not account for how the economy would react to policies that attempted to exploit past correlations in the data, modeling that took the plumbing for granted ensured the breakdown of the plumbing. In coming years, macroeconomic modeling must incorporate more of the plumbing, which has been studied elsewhere in economics.
The danger is that monetary economists will try to wish away the links between monetary policy, risk taking, and asset-price bubbles. Bernanke came close to doing so in his 2010 speech to the American Economic Association, where he argued that it was not the Fed's defective monetary policy—which he considered entirely appropriate, given the Fed's views on inflation—but its inadequate supervision that helped trigger the crisis. He concluded: “Although the most rapid price increases occurred when short-term interest rates were at their lowest levels, the magnitude of house price gains seems too large to be readily explained by the stance of monetary policy alone. Moreover, cross-country evidence shows no significant relationship between monetary policy and the pace of house price increases.”20
Of course, no one claims that the Fed alone was responsible for the housing debacle. Government policies favoring low-income housing, as well as private-sector mistakes, contributed significantly. But to suggest that it had no role is disingenuous. Indeed, a detailed study published in the Federal Reserve Bank of St. Louis Review in 2008 presents evidence that “monetary policy has significant effects on housing investment and house prices and that easy monetary policy designed to stave off the perceived risks of deflation in 2002–2004 has contributed to a boom in the housing market in 2004 and 2005.”21
Moreover, there is no reason why there should be a strict relationship across countries between monetary policy and the rate of house-price growth over any common period of time: the rate of price growth might depend on a variety of factors that are specific to each country, including how high house prices already are.22 The broader point is that monetary economists need to take note (as they are now doing) of the other channels through which monetary policy might have effects.
As developing countries cut back on demand following their crises in the 1990s, and as industrial-country corporations worked off their excess investment following the dot-com bust, the world's exporters searched once again for countries that would reliably spend more than they produced. The United States, which was already pushing to encourage household consumption to appease those left behind by growth, had added reasons to infuse substantial fiscal and monetary stimulus in response to the downturn: the jobless nature of the recovery and the weak U.S. safety net. In addition to a substantial fiscal stimulus that pushed a government budget that was temporarily in surplus into large fiscal deficits, the Fed kept its foot pressed on the monetary accelerator, even while giving all sorts of assurances to the markets on its willingness to maintain easy monetary conditions and to step in to provide liquidity in case the financial markets had problems. These assurances had the desired effect of leading to an explosion of lending, which unfortunately continued expanding and deteriorating in quality even after the Fed started tightening. For an unsustainable while, though, the United States provided the demand the rest of the world needed.
The U.S. political system is acutely sensitive to job growth because of the economy's weak safety nets. The short duration of unemployment benefits in the United States, as well as the substantially higher costs of health care for those who do not have jobs, were not excessively painful when recessions were short: they gave laid-off workers strong incentives to find new jobs even while U.S. businesses created them. But if recessions are likely to be more prolonged than in the past, the system has to change, if only because the old social contract—short-duration benefits in return for short recessions—is breaking down.
One reason is simply moral. No modern economy should force workers who lose their jobs to make such painful decisions as choosing which of their children to protect with medical insurance. Not only is this situation barbaric, it is also unsustainable, for those who lose out economically have every incentive to use political means to regain what they have lost. While a democratic system eventually responds, the response can be unpredictable, adding to worker uncertainty. There is a strong case for strengthening the U.S. safety net in ways that will not hamper the flexibility of the economy greatly.
Another problem with a weak safety net is that the United States tends to overreact, and other nations underreact, to downturns. Because every country knows that the politically vulnerable United States has to respond with expansionary policies and that some U.S. demand will spill over to the rest of the world, their incentive to change the structure of their economy, or their policies in downturns, is commensurately less.
But perhaps the most important problem is that the ad hoc policies the United States is forced into do enormous damage to the long-term health of the economy, both directly and through their effects on the financial sector. One could argue that discretionary fiscal and monetary policy in the midst of a downturn gives the United States the ability to calibrate its response to the severity of the downturn. But fiscal policy undertaken at the point of a gun is rarely as dispassionate or as well thought out as one might like. Yes, Congress could simply extend unemployment benefits, as it has done in the current recession. But politicians often want to do more. And the public's anxiety gives them the license to bring out all their pet projects, all the favors to special interests, and all the schemes their ideological leanings and political connections predispose them to.
Similarly, as we have seen, the Federal Reserve, though ostensibly independent, has a very difficult task. It is extremely hard to ensure rapid job growth in an integrated, innovative economy where firms use recessions to refocus on becoming more productive or to strengthen their global supply chains, shifting jobs elsewhere. Moreover, the new technologies employed in hiring allow firms the luxury of waiting to fill positions. The sustained easy monetary policy that is maintained while jobs are still scarce has the effect of increasing risk taking and inflating asset-price bubbles, which again weaken the fabric of the economy over the longer term. If the United States cannot tolerate longer bouts of unemployment, but those bouts are here to stay, we risk going from bubble to bubble as the Federal Reserve is pressured to do the impossible and create jobs where none are forthcoming.
It is now time to turn to vulnerabilities in the financial sector to see why the fault lines came together to make banks take the risks they did. I focus on two issues. First, why did mortgage lending go berserk (which is the subject of the next chapter)? Second, why did the banks take on so much default and liquidity risk (which is the subject of Chapter 7)?