Timothy Geithner
Timothy Geithner; drawing by John Springs

By now there are few people who do not acknowledge that the major American financial institutions and the markets they dominate turn out to have served the country badly in recent years. The surface evidence of this failure is the enormous losses—more than $4 trillion on the latest estimate from the International Monetary Fund—that banks and other lenders have suffered on their mortgage-related investments, together with the consequent need for the taxpayers to put up still larger sums in direct subsidies and guarantees to keep these firms from failing. With nearly 9 percent of the labor force now unemployed and still more joining their ranks, industrial production off by 13 percent compared to a year ago, and most companies’ profits either falling rapidly or morphing into losses, it is also evident that the financial failure has imposed huge economic costs.

The government has moved aggressively, and on several fronts, to stanch the immediate damage. The Federal Reserve has not only eased monetary policy to the point of near-zero short-term interest rates but created a profusion of new programs to extend credit to banks as well as other lenders. Congress, at the Obama administration’s behest, has enacted nearly $800 billion of new spending and tax cuts aimed at stimulating business and consumer spending. First the Bush administration and now Mr. Obama’s have experimented with one new plan after another to rescue lenders and reliquify collapsed credit markets.

But despite the universal agreement that no one wants any more such failures once this one has passed, there is a troubling lack of attention to reforms that might prevent such a crisis from recurring. By now everyone realizes that excessive risk-taking, systematic mispricing of assets, and, often, plain reckless behavior (not to mention some instances of criminality, although to date surprisingly few of these have come to light) helped cause the current mess. At the same time, most people recognize both that parts of the American economy have been capable of dynamic growth and that the US financial markets have had a part in promoting that growth. The result is a reluctance to consider changes to the current system. Substantial interference with financial markets, it is said, amounts in the end to centrally planned allocation of an economy’s investment process, and will result in technological stagnation and wasted resources. Milder attempts at regulation will either prove ineffective—the private sector can afford better lawyers than the government can—or at best merely lead financial institutions to relocate to more lightly regulated jurisdictions like the Cayman Islands.

As in past financial declines, what is sorely missing in this discussion is attention to what function the financial system is supposed to perform in the economy and how well it has been doing it. Today attention is mostly focused on banks’ and other investors’ losses from buying mortgage-backed securities at inflated prices. What is neglected is the consequence: if the prices of the securities were too high, this meant that the underlying mortgage rates were too low, and so too many houses were built, and too many Americans bought them. In just the same way, when the 1990s stock market boom crashed, everyone talked about investors’ losses on their telecom stocks, not the fact that if the stocks’ prices were too high, the cost of capital to the firms that issued the stocks was too low, and so communications companies laid millions of miles of fiber-optic cable that nobody ended up using.

In both instances, the cost was not just financial losses but wasted real resources. True, over longer periods of time the American financial system has seemed reasonably effective at allocating resources rather than wasting them. The economy’s pace of technological advance and growth in production since World War II suggest that the banks and the stock and bond markets can steer investment capital reasonably effectively to firms that will use it productively, often including start-ups trying out a wholly new idea—Microsoft, or Google, or, earlier on, Apple. But the effectiveness of the economy’s mechanism for allocating capital should be a matter for serious quantitative evaluation, not a matter of faith.

Moreover, to ask just how efficient a financial system is in allocating capital leads naturally to the question of the price that is paid for such efficiency. In recent years the financial industry has accounted for an unusually large share of all profits earned in the US economy. The share of the “finance” sector in total corporate profits rose from 10 percent on average from the 1950s through the 1980s, to 22 percent in the 1990s, and an astonishing 34 percent in the first half of this decade.1

Those profits accruing to the financial sector are part of what the economy pays for the mechanism that allocates its investment capital (as well as providing other services, like checking accounts and savings deposits). But even a stripped-down version of the cost of running the financial system includes not just the profits that financial firms earn but also the salaries, the office rents, the travel budgets, the advertising fees, and all of the other expenses they pay. The finance industry’s share of US wages and salaries has likewise been rising, from 3 percent in the early 1950s to 7 percent in the current decade.2 An important question—which no one seems interested in addressing—is what fraction of the economy’s total returns to productively invested capital is absorbed up front by the financial industry as the costs of allocating that capital.


Further, the latest financial crisis is a sharp reminder that the simple operating expense of running the financial system—including profits of financial firms—is not the only cost if this system also exposes the economy at large to episodic losses in production and incomes, and to the need for taxpayer subsidies. Today those losses are mounting, and so are the subsidies. Many US banks, including some of the largest ones, are now insolvent. The bank rescue plans offered to date by both the Bush and the Obama administrations amount to ever more expensive fig leafs for avoiding concrete recognition of this sad development.

In the same effort, the Financial Accounting Standards Board—the independent organization designated by the SEC to set accounting standards—acting at the strong urging of Congress, recently changed its rules to allow banks more latitude to claim that assets on their balance sheets are worth more than what anyone is willing to pay for them. (Next time you apply for a loan, try mentioning FAS 157-4 and telling your banker that you should be allowed to calculate your net worth with your house priced not at what comparable houses are selling for now but at what you paid for it and what you hope you’ll get for it if you hold on to it for some years. The banker will laugh, even while the bank applies just such standards to its own balance sheet.)

Another fundamental issue that the current discussion has overlooked almost entirely is the distinction between the losses to banks and other lenders that reflect genuine losses of wealth to the economy, and other losses that don’t. When the value of your house falls, that’s a loss of wealth to the economy as a whole. If you keep paying your mortgage, you bear the loss yourself: your net worth is diminished by the amount of the decline in the home’s price. If you default on your loan, then someone else—maybe the bank that lent you the money, maybe some investor to whom the bank sold the loan—also bears part of the loss. If the government steps in and reimburses the bank, or the investor, the taxpayers will bear part of the loss as well. But however this loss is divided, what is inescapable is that someone, somewhere, will bear it. What much of today’s debate is about is how these losses should be divided among homeowners, banks, loan-purchasing investors, and the taxpayers. But the loss must be borne by someone, and America’s economy is poorer because it has occurred.

By contrast, suppose you and your neighbor have bet on whether today’s peak temperature would exceed fifty degrees. One of you was right, the other wrong. One of you won, the other lost, and the amount the winner won is identical to what the loser lost. There is no loss of wealth to the economy, merely a transfer of wealth from the loser to the winner. Many of the huge losses that American financial institutions have sustained in the current crisis are of this second kind. None of them was betting on the weather, but they were taking positions that amounted to placing bets on outcomes that represented no change in wealth to the economy as a whole. And with regard to these positions, for every loser featured in the latest newspaper story about banks posting losses and turning to the government for bailouts there is also, somewhere, a winner.

The most telling example, and the most important in accounting for today’s financial crisis, is the market for credit default swaps. A CDS is, in effect, a bet on whether a specific company will default on its debt. This may sound like a form of insurance that also helps spread actual losses of wealth. If a business goes bankrupt, the loss of what used to be its value as a going concern is borne not just by its stockholders but by its creditors too. If some of those creditors have bought a CDS to protect themselves, the covered portion of their loss is borne by whoever issued the swap.


But what makes credit default swaps like betting on the temperature is that, in the case of many if not most of these contracts, the volume of swaps outstanding far exceeds the amount of debt the specified company owes. Most of these swaps therefore have nothing to do with allocating genuine losses of wealth. Instead, they are creating additional losses for whoever bet incorrectly, exactly matched by gains for the corresponding winners. And, ironically, if those firms that bet incorrectly fail to pay what they owe—as would have happened if the government had not bailed out the insurance company AIG—the consequences might impose billions of dollars’ worth of economic costs that would not have occurred otherwise.

This fundamental distinction, between sharing in losses to the economy and simply being on the losing side of a bet, should surely matter for today’s immediate question of which insolvent institutions to rescue and which to let fail. The same distinction also has implications for how to reform the regulation of our financial markets once the current crisis is past. For example, there is a clear case for barring institutions that might be eligible for government bailouts—including not just banks but insurance companies like AIG—from making such bets in the future. It is hard to see why they should be able to count on taxpayers’ money if they have bet the wrong way. But here as well, no one seems to be paying attention.

Barack Obama
Barack Obama; drawing by John Springs

Why has there been so little discussion of fundamental issues like this distinction among losses? Why is so little said about the trade-off between the goal of allocating the economy’s capital efficiently and the need to shrink the enormous costs of the financial industry in doing so? One obvious reason is political. There is a long arc from Roosevelt’s acceptance of a useful role for government institutions and government regulation to the conviction of Reagan and Thatcher that the government is never the solution but actually the problem. A second, closely related reason is ideological: the faith, personified by Alan Greenspan with his early dedication to the writings of Ayn Rand and his staunch opposition to regulations while chairman of the Federal Reserve, that private, profit-driven economic activity is self-regulating and, when necessary, self-correcting.

The economists George Akerlof and Robert Shiller suggest a third reason. In their view, the problem is also intellectual—a systematic failure of thinking on the part of their fellow economists. Taking the title of their new book from a phrase famously used by John Maynard Keynes, Akerlof and Shiller argue that what is missing in the worldview of today’s economists is sufficient attention to “animal spirits,” by which they mean the psychological and even irrational elements that figure importantly in so many other familiar aspects of personal choices and personal behavior, and that, they believe, pervade economic behavior too.

Akerlof and Shiller identify five distinct elements in what they call “animal spirits”: confidence, or the lack of it; concern for fairness, that is, for how people think they and others should behave—for example, that a hardware store shouldn’t raise the price of snow shovels after a blizzard despite the increased demand; corruption and other tendencies toward antisocial behavior; “money illusion,” meaning susceptibility to being misled by purely nominal price movements that, because of inflation or deflation, do not correspond to real values; and reliance on “stories”—for example, inspirational accounts of how the Internet led to a “new era” of productivity. The omission of these five aspects of “animal spirits,” they argue, blocks conventional economics from either understanding today’s crisis or providing useful ideas for dealing with it.

Most of Animal Spirits consists of surveys of disparate research programs in which either Akerlof or Shiller or both have participated—in many cases, led the way—demonstrating that the conventional economic models fail to fit the facts of one or another aspect of observable economic behavior. The topics they cover include (among others) joblessness, saving, the volatile prices of financial assets and of real estate, and the prevalence of poverty among African-Americans. Akerlof, a professor at Berkeley who deservedly won the Nobel Prize in 2001, and Shiller, a Yale professor who in time should do likewise, are both outstanding economists, and the work they review in each of these areas of research is of high quality. In this respect, their book is similar to recent offerings by other scholars that have also presented for a broader audience many of the findings of what has come to be known as “behavioral economics.”3

Akerlof and Shiller, however, want Animal Spirits to do more than summarize distinct findings and their separate implications for policy. They see the limitations of today’s conventional macroeconomics as systemic: “the theories economists typically put forth about how the economy works are too simplistic. That means we should fire the weather forecaster.” And “firing the forecaster means giving up the myth that capitalism is purely good.”

In an intuitively appealing illustration, Akerlof and Shiller suggest how to think about what’s included and what’s not in conventional economic theory, and what they intend their book to accomplish:

Picture a square divided into four boxes, denoting motives that are economic or noneconomic and responses that are rational or irrational. The current model fills only the upper left-hand box; it answers the question: How does the economy behave if people only have economic motives, and if they respond to them rationally? But that leads immediately to three more questions, corresponding to the three blank boxes…. We believe that the answers to the most important questions regarding how the macroeconomy behaves and what we ought to do when it misbehaves lie largely (though not exclusively) within those three blank boxes.

One of the inevitable difficulties with this kind of argument is that it depends so much on just how words are used. What is the difference between an economic and a noneconomic motive? If I buy a new car because I like its styling and good gas mileage, that’s presumably an economic motive. What if it’s a hybrid and part of my reason for buying it is that I value the “story” of my doing my bit to help slow global warming? If a business owner provides scholarships for his employees’ children, is his motive to treat them fairly for fairness’ sake or to foster their loyalty and thereby improve their productivity?

The answer, in the end, is that an “economic” motive is whatever economists include in their theories of how people behave. And since different economists are always proposing different theories, what constitutes an “economic” motive can differ from one theory, and one economist, to another. Since at any particular time there are dominant theories, there is some coherence to what people would understand as an “economic” motive; and so the point Akerlof and Shiller are trying to make here is far from empty. But it is more elusive than they suggest. The distinction between what’s “rational” and what’s not is, if anything, even more fraught.

Sometimes, moreover, Akerlof and Shiller’s substantive arguments fall victim to a similar circularity. The element of animal spirits on which they place the most emphasis in their account of the current crisis is confidence. It is, they say, “the first and most crucial of our animal spirits.” Is it rational or not for me to put my money in a bank in which I have confidence? Or to buy a stock if I have confidence in the company’s business prospects? According to Akerlof and Shiller, this kind of behavior is, by definition, irrational since “confidence” for them is a kind of faith, not a matter of rational analysis: “The very term confidence…impl[ies] behavior that goes beyond a rational approach to decision making.” (Similarly, when confidence is explained in terms of trust, “the very meaning of trust is that we go beyond the rational.”)

This tendency to argue by definition also sometimes affects questions beyond whether something is to be classified as economic or noneconomic, or rational or irrational. One of the reasons confidence is so important in Akerlof and Shiller’s list of animal spirits is that when people have too much of it, they behave in ways that cause the economy to become overheated. But this argument, too, turns out to be a matter of definition: “The term overheated economy, as we shall use it, refers to a situation in which confidence has gone beyond normal bounds….”

Concerns like these aside, the broader question is whether Akerlof and Shiller succeed in making more of the different kinds of research they report on in Animal Spirits than the sum of the book’s disparate parts. The answer is yes in some respects, no in others.

They succeed in demonstrating both the narrowness of mainstream macroeconomic thinking in recent decades and the stranglehold that this thinking has placed on the economics profession’s ability either to explain phenomena like today’s crisis or to advance potential solutions. For example, economics today is largely taught using mathematical models to describe outcomes under different conditions. And for purposes of macroeconomics—the study of the economy as a whole—most of the standard models do not admit the possibility of unemployment. The reason is not that no one knows unemployment exists. Rather, no one has figured out how to allow for it within the confines of sufficiently simple mathematics; and faced with the choice between excluding unemployment and sacrificing analytical simplicity, most macroeconomists have opted for the former.

To point to another example even more central to what is happening currently, the standard macroeconomic analysis today also does not acknowledge the existence of banking or other kinds of borrowing and lending. Instructive models of credit markets are certainly available.4 But incorporating them within the standard macroeconomic models would place too much strain on mathematical simplicity.

Akerlof and Shiller succeed, too, in demonstrating that conventional macroeconomic analyses often fail because they omit not just readily observable facts like unemployment and institutions such as credit markets but also harder-to-document behavioral patterns that fall within the authors’ notion of “animal spirits.” Confidence plainly matters, and so does the absence of it. When the public mood swings from exuberance to anxiety, or even fear, the effect on asset prices as well as on economic activity outside the financial sector can be large.

As they argue, these effects can plausibly be larger than the fluctuations attributable to more concrete factors, such as monetary policy or oil prices, that economists more typically incorporate in their analysis. Money illusion, by which people are influenced by purely nominal price movements, is also clearly important to some aspects of how the economy in aggregate behaves. (As they show, concerns for fairness, tendencies toward corruption, and reliance on “stories” also influence some aspects of economic behavior, but whether they are significant for the aggregate economy remains unclear.) “It is necessary,” they argue,

to incorporate animal spirits into macroeconomic theory in order to know how the economy really works. In this respect the macroeconomics of the past thirty years has gone in the wrong direction.

The effort to “clean up macroeconomics and make it more scientific,” to impose “research structure and discipline,” has proved disastrously confining.

But what then? Is there more to Animal Spirits than a list of important influences for economists to try to take into account? Akerlof and Shiller believe they have proposed a new model for macroeconomic analysis. Comparing their work to Keynes’s discussion of animal spirits from the 1930s, they write:

This book…describes how the economy really works…. With the advantage of over seventy years of research in the social sciences, we can develop the role of animal spirits in macroeconomics in a way that the early Keynesians could not. And because we acknowledge the importance of animal spirits, and accord them a central place in our theory rather than sweep them under the rug, this theory is not vulnerable to attack.

There are two problems here. One is simply the familiar tendency to overstate, especially when writing about one’s own work. It is hard to believe that Akerlof and Shiller really see their ideas as invulnerable to attack. Similarly, after listing at the outset the subjects about which they will summarize the research that they and other scholars have done—e.g., “Why do economies fall into depression?”—they declare that “animal spirits provide an easy answer to each of these questions.” The answers they summarize may be right or not (I vote yes) and they may be well argued or not, but they are not “easy.”

The more important question is whether what Akerlof and Shiller have offered in Animal Spirits amounts to “a theory” in the sense that it could stand in place of the current theories that they criticize for being based entirely on rational responses to economic motives. There is a difference between a series of ideas about different aspects of economic behavior and an integrated account of macroeconomic fluctuations. Akerlof and Shiller are surely on the right track in pointing to elements that are missing from today’s conventional models, and in arguing that incorporating them into mainstream macroeconomic analysis would help. But they have neither done this nor shown others how to. Hence their goal of replacing what macroeconomists teach their students is likely to be disappointed, at least for now.

And because what they have here is a set of examples of how “animal spirits” matter for specific aspects of economic behavior, not a coherently worked out theory of how the macroeconomy behaves, it is not surprising that Animal Spirits is also thin on suggestions for what to do about the current crisis. Akerlof and Shiller mostly restrict their recommendations to how economists should think, not what policymakers should do. The one concrete proposal they offer is that the government should have a target for the expansion of credit—that is, for the amount of borrowing and lending in the economy, presumably defined in some measurable way.

This idea has considerable merit (some years ago I wrote a series of papers advocating it myself5). But it is hard to see how it would help address the problems our economy faces today. Everyone knows that a principal objective of the Treasury’s bank bailouts, the Federal Reserve’s numerous new facilities for credit, the government’s takeover of Fannie Mae and Freddie Mac, as well as other actions, is to restore the functioning of the credit markets so that borrowers can again obtain financing. It is not clear how stating a specific target for credit expansion would help further that goal. It is also not clear that the proposal follows, in any direct way, from the importance of animal spirits in influencing economic behavior.

What else? The “world of animal spirits gives the government an opportunity to step in,” Akerlof and Shiller write. But in what ways? Apart from a target for credit growth, they mostly leave it at that. “This book cannot give the detailed answers” to current policy questions, they write. This is a pity. Because they are so convincing in their critique of modern macroeconomics, and because they point to the current crisis as a prime example of why this matters, readers will understandably want to know what policies they would recommend.

In his recent book The Subprime Solution, Shiller offered a series of proposals, recommending, for example, a new version of the Home Owners’ Loan Corporation. which operated from 1933 to 1936 and provided generous subsidies for home mortgages, insisting, however, that they be paid off by steady monthly payments. He also advocated expanded financial disclosure requirements, greater availability of financial data, and a new government agency like the Consumer Product Safety Commission that could protect consumers by informing them about the safety of financial products.

Shiller made more far-reaching suggestions as well, proposing markets in which families could insure the risk inherent in owning their house6 or earning their livelihood, and even a new way of expressing prices in a unit of account that would adjust for inflation or deflation, so as to help people overcome money illusion. Here he anticipated the discussion of “animal spirits” in his book with Akerlof:

The subprime crisis was essentially psychological in origin…. Denying the importance of psychology and other social sciences for financial theory would be analogous to physicists denying the importance of friction in the application of Newtonian mechanics.

But The Subprime Solution was (by intent) mostly focused on mortgages and the fallout from the subprime collapse. In Animal Spirits, Akerlof and Shiller make a much broader argument, but they say little about its concrete implications for macroeconomic policies: How much and what kind of fiscal stimulus is desirable now? What should be the role for monetary policy now that short-term interest rates are near zero? What changes should be made in financial regulation once the crisis is past?

Some of the proposals for regulatory reform now being put forward by others appear at least to connect to parts of Akerlof and Shiller’s discussion. For example, they call attention to the potential instability inherent in the “new shadow banking system”—that is, institutions such as investment banks and even the off-balance-sheet entities sponsored by the banks themselves that now carry out many of the lending functions of banks but are not regulated as such and whose obligations are not ordinarily insured. They observe, correctly, that “there can be a ‘run’ on these institutions just as there can be a run on traditional banks.” They are right; what happened last year to Bear Stearns and Lehman was, in effect, a run on a nonbank.

But what, then, should be done to prevent more such runs? Some ideas now under discussion include stronger accounting rules and wider capital requirements, so that hedge funds and insurance companies and other nonbanks too would have to hold capital against their risk positions. There is also support for empowering the government to take “prompt corrective action” to force nonbanks to address problems in their balance sheets, or if necessary take them into receivership, just as it already can with banks. A more controversial idea is to reinstitute some form of the Glass-Steagall Act that, until Congress repealed it in 1999, barred commercial banks from also doing investment banking. The Obama administration has proposed a new financial “super-regulator.” How do animal spirits bear specifically on any of these issues? Akerlof and Shiller do not say.

But their silence on these and other policy issues and their lack of a full-blown “theory” to replace the strait-jacketed macroeconomics of the past few decades do not undercut the force of Akerlof and Shiller’s central argument. Their harsh judgment of current mainstream macroeconomic thinking is true, and they are right about the importance of what they call “animal spirits” to many of its crippling shortcomings. Animal Spirits provides an agenda for new thinking, and one well worth pursuing.

April 30, 2009

This Issue

May 28, 2009