访谈联邦储备银行主席伯南克(第二部分)

Interview with Ben S. Bernanke*

ASSET BUBBLES AND MONETARY POLICY

Rolnick: Some claim they can identify a speculative bubble in a variety of markets. The dot-com sector was viewed as an asset bubble, and then some said we have had a speculative bubble in housing. How confident can we be in identifying such bubbles, and if we can be confident, what's the role of the central bank in dealing with these events?

Bernanke: I think it's extraordinarily difficult for the central bank to know in advance or even after the fact whether or not there's been a bubble in an asset price. The mere fact that an asset price has gone up and come back down again doesn't mean that there was a bubble in the technical sense that the price movement was completely divorced from fundamentals. Moreover, if a bubble does exist, there is no guarantee that an attempt to “pop” it won't lead to violent and undesired adjustments in both markets and the economy. The central bank should focus the use of its single macroeconomic instrument, the short-term interest rate, on price and output stability. It is rarely, if ever, advisable for the central bank to use its interest rate instrument to try to target or control asset price movements, thereby implicitly imposing its view of the proper level of asset prices on financial markets. History has shown us clearly that that type of policy has more often than not led not only to a large decline in asset prices but also to a large decline in the general economy.

I do think there are several useful things that central banks can do about potential mispricing in asset markets. First, I think that many distortions in asset prices have arisen historically because of various kinds of structural regulatory problems in the underlying markets. For example, research on historical episodes suggests that large asset price increases are sometimes preceded by credit booms. In many cases, this pattern results from the fact that the country in question deregulated its banking system, giving banks extra powers, but did not enhance the supervisory structure adequately at the same time. The result is that institutions have an incentive to make economically bad investments, to take advantage of the “put” provided by the government safety net.

So you could have a situation where a badly managed deregulation of some financial market, such as in the case of our own savings and loan crisis or in episodes during the 1980s in Japan and Scandinavia, has the potential to create a one-way bet that generates a destabilizing move in asset prices. It's extremely important for central banks, or for financial supervisory agencies in those countries that have them, to ensure that the underlying microeconomic regulatory structure is such that moral hazard and misalignment of incentives are not pervasive in the system. More often than not, such moral hazard problems are the source of asset prices becoming disconnected from fundamentals. So good oversight of the banking system and of the broader financial system is one very important way in which central banks or other agencies can prevent these kinds of problems.

Secondly, of course, the central bank has a very important obligation to maintain the stability of the underlying institutions. To the extent that there are large movements in asset prices that

threaten the stability or functioning of exchanges or other institutions, the central bank may have to play a role to try to stabilize those institutions, as the Fed did in October 1987 for example.

Ultimately though, I think the main objective of the central bank is to provide macroeconomic stability, and it's neither its comparative advantage nor its objective to try to second-guess the asset price decisions of investors in asset markets.

Rolnick: You've made an interesting point that some of these so-called bubbles may be the result of bad policy.

Bernanke: Yes, that was a theme of my very first speech as a governor. I suggested that, from a policy perspective, there are two ways to approach bubbles: One is interest rate policy, the other is micro-regulatory policy. Micro-regulatory policy is the much better approach, in my view.

THE GREAT DEPRESSION AND MORAL HAZARD

Rolnick: In 1930, the Bank of United States failed. Milton Friedman and Anna Schwartz in their work on the Great Depression argue that the failure of this bank was the pivotal point in what became the worst economic decline in U.S. history. They suggest that if the Fed had rescued this bank, the Great Depression might only have been a short, albeit severe, recession. Some of us have argued that Friedman and Schwartz provide the rationale for the policy that today is known as “too big to fail”—that there are some institutions that are so big that we can't afford to let them fail because of the systemic impact on the rest of the economy. The downside of that policy, of course, is the moral hazard problem. How do you interpret this history?

Bernanke: As a historical matter, whether that particular bank failure was pivotal or was an isolated case is quite controversial. A lot of good research has questioned the Friedman-Schwartz conclusion that that particular episode was the kickoff of the U.S. banking crisis. But certainly, the whole experience of the 1930s clearly shows some of the threats associated with large-scale and pervasive bank failures, including collapse of the money supply, disruptions of credit extension and other banking services, and blows to confidence.

The experience of the Depression led to the creation of deposit insurance, which stabilized banking but created a potential moral hazard. So we required yet another countervailing force, which is the supervisory and regulatory function of the central bank and other bank regulators. The tension between providing stability and dealing with moral hazard problems that arise from guarantees is pervasive in economic policy.

In the case of too big to fail, the question has been raised whether or not the failure of very large banks, because of the many links those banks have with other counterparties, might threaten the stability of the system. I think, generally, we shouldn't have a fixed doctrine; we should have a case-by-case analysis. The default option is that the bank should be allowed to fail. But even in those cases where the central bank and the deposit insurance corporation decide to prevent the failure of a large institution in the interest of systemic stability, there's a great deal to be done to minimize the moral hazard implications. For example, the shareholders can be forced to take losses, the management can be replaced, and other actions can be taken that greatly reduce the incentives to take unwarranted risks.

Rolnick: And uninsured depositors?

Bernanke: I would include in the previous sentence that uninsured depositors should often take some losses as well. So one could try to maintain the functioning of the bank and preserve most of its depository liabilities while still addressing the moral hazard issue. There's a trade-off there, obviously, a trade-off between ensuring the stability of the system and making sure there's enough cost felt by the shareholders, managers and uninsured depositors that the ex ante incentives to take unwarranted risks are minimized.

Rolnick: What do you think of the idea that bank regulators and the federal government should make a commitment that uninsured depositors in large banks will always face some risk? Under such a commitment, markets would price uninsured deposits accordingly, and the price signal could be used by regulators to help evaluate bank risk.

Bernanke: A number of people have suggested that we rely on uninsured, subordinated debt to provide a market signal about the quality of the bank. I think that's a very interesting idea, and I note that market signals play some role in current bank supervision and would continue to do so under the proposed reforms to the Basel capital standards. I would be reluctant to put all of our eggs in that basket, however, because banks are inherently somewhat opaque institutions, and it's not evident that the market signal would be completely informative about the state of the financial institution. So I would supplement market signals with capital requirements and various forms of direct supervision.

Rolnick: Of course, if I'm a really safe bank, I have an incentive to be more transparent.

Bernanke: That's true. These market devices are potentially helpful, and as I mentioned they are in fact part of the Basel II proposal. They should be viewed as complementary to other forms of supervision.

LESSONS FROM THE GREAT DEPRESSION

Rolnick: I would like to come back to the research on the Great Depression of the 1930s and its causes. My view is that the profession has made a lot of progress in understanding this period in our economic history. But it still bothers me that we are far from an agreement as to what caused the Great Depression. We have some understanding of why it may have lasted for 10 years, but even that is controversial. You're clearly one of the leading experts in this area. What's your view? What are the lessons we should learn from the Great Depression? What do you as a governor of the Federal Reserve System take away from that calamity?

Bernanke: The Depression was a very complicated event, a global event with both political and economic causes. For those reasons I'm sure we'll never have a complete understanding or complete agreement about its sources. I do think though that the degree of consensus on the cause of the Depression is probably greater now than it has ever been. And that consensus includes not necessarily an exclusive role but a very important role for both monetary and financial factors.

The collapse of the money supply that was engendered by the combination of mistaken monetary policies in the United States and a few other countries, and the transmission of those policies throughout the world by a gold standard that in many ways had a bias toward deflation built into it, seems clearly to have been the major single cause of the Depression. The evidence for this view—for example, the fact that whether a country adhered to the gold standard after 1931 or not explains much of the difference in cross-country experience—is very strong.

Simultaneously, the world was also afflicted with a great deal of financial instability. Banking crises were not unique to the United States. They were experienced in many other countries, as were exchange rate crises, stock market crashes and the like. Many of these had their roots in the political and economic instability of the 1920s. Many scholars have traced how financial instability can engender bad outcomes in the real economy.

So in terms of the causes, again, I think there's not a full consensus and there never will be, but I think at this moment there is a fairly broad view that these monetary factors and perhaps to a lesser extent financial factors were crucial.

The lessons one learns from that are, first, that monetary stability and price stability are incredibly important for generating overall macroeconomic stability, and second, that financial stability also is of major importance. Because financial instability can generate real instability, maintaining a sound and stable financial system should be a high priority for the central bank and the government.

Those are the two broadest lessons I would draw from it. I think it's important to understand that the deepest cause of the Depression, as [Massachusetts Institute of Technology economic historian] Peter Temin once said, was World War I and the Peace of Versailles that followed it. These left the world not only in an economically fragmented and unbalanced situation but also in a highly contentious political situation. Powerful political forces prevented the kinds of economic cooperation that might have prevented some of the problems that later occurred. So there was a confluence of political and economic factors that made the Depression as severe as it was. But again, if I had to narrow it to two main lessons, I would stress the crucial importance of both monetary and financial stability.

OUTSOURCING CONTROVERSY AND
ECONOMIC LITERACY

Rolnick: The outsourcing of jobs to India, China and elsewhere has become a huge controversy in the United States. Most economists would say such job flows are a normal and healthy phenomenon. And as you indicated in a recent speech, outsourcing abroad accounts for a very small fraction of total job loss in the United States each year—perhaps a bit more than 1 percent.

Does this controversy indicate that the economics profession has failed to educate the public about international trade in particular and economics more generally? What can be done to improve economic literacy?

Bernanke: As I argued in my recent speech, there is an overwhelming case that trade increases economic welfare. With respect to jobs, it is true that trade promotes structural change that displaces some jobs, but trade creates many opportunities for increased employment as well, including high-wage employment. While recently many people have been concerned specifically about the outsourcing of business services, few are aware that the United States runs a healthy trade surplus in services—that is, there is considerably more (and higher-value) “insourcing” to the United States than there is outsourcing from the United States abroad. Employment of Americans by foreign-owned firms, such as foreign automobile manufacturers, is a major source of domestic employment as well.

With respect to economic literacy, I have no magic bullet; responsible economists just have to keep getting the word out. Specifically with respect to trade, though, I think there is a legitimate issue arising from the fact that you can't buy insurance against losing your job to new foreign competition, so that workers who are displaced by trade bear most of the associated costs, rather than society at large. We should consider expanded support and retraining for workers displaced for any reason. I think if people were a bit less fearful of the impact of change on their own financial well-being they might be more amenable to arguments that trade is highly beneficial to the economy as a whole.

JAPAN'S ECONOMY

Rolnick: Beginning in about 1992, Japan lost a decade of growth. Was poor monetary policy the cause of Japan's stalled economy? If not, what was the cause?

And last year, by contrast, growth in Japan was rapid-per capita GDP grew by over 4 percent. Does better monetary policy account for this dramatic improvement?

Bernanke: Japan's lost decade had a number of causes, including the severe difficulties of its financial system and structural rigidities in its economy (except for its strong export sector). Excessively cautious monetary policy did play a role in the lost decade, however, because it did not do all that it could have done to arrest and reverse the deflation. Recently, monetary policy in Japan has been more proactive. Together with a long-delayed strengthening of the Japanese banking system, I think that monetary policy has played a role in the economic improvement we have seen in Japan. Deflation seems to be moderating. It is certainly too early to declare victory, however, and I hope that expansionary monetary policy, financial reform and structural reform will continue to be vigorously pursued in Japan.

THE ROLE OF ACADEMIC RESEARCH

Rolnick: What role does academic research have in an institution like the Fed? The System employs hundreds of economists. They do a variety of work, from forecasting the economy to developing theoretical models of how the economy works. What do you see as the role of research at the Board and at the district banks? And how should policymakers in the Federal Reserve System use this research?

Bernanke: I think research plays a critical role in the Federal Reserve System for at least two distinct reasons. One is that we need to have highly trained and highly competent economists to do our analysis, and given that we must compete in an academic job market, we're able to attract the best and the brightest only by allowing them some freedom to supplement their day-to-day analyses with the pursuit of individual research objectives. So part of it is simply attracting the best talent.

The more important reason is that the research itself provides an important long-run perspective on the issues that we face on a day-to-day basis. Speaking for myself, most of what I know about monetary policy has come out of academic research. There is a great deal of excellent work out there, both theoretical and empirical, that addresses fundamental practical issues of policymaking. Having done academic research myself, I think I can separate the more useful from the less useful. The best work has had a profound effect on how we think about policy.

In short, a strong research department ensures that you'll have high-quality economists to address the issues you're facing on a day-to-day basis with the best, most current tools. Moreover, the research that is produced allows policymakers to take a broader perspective on the issues that we face. The presence of research departments in Reserve banks is also very important. First, because all the presidents participate in FOMC discussions, they need staff support and guidance on making decisions about monetary policy; likewise, research provides important support in other policy areas such as bank supervision and consumer compliance. Also, historically, having separate research departments in the Reserve banks has allowed for some heterogeneity of opinion and viewpoint. St. Louis, for example, with its traditional monetarist perspective, has for decades provided a different viewpoint than others in the System. So the distribution of research groups across the country has provided some antidote to potential groupthink in the viewpoints expressed across the FOMC table.

Rolnick: The counterargument, of course, is that people on the outside are looking in and seeing economists disagreeing and debating on a variety of key issues. Policymakers need answers but economists can't seem to agree. So the outsider's view is: How can you operate in such a controversial environment?

Bernanke: Well, economics has many substantive areas of knowledge where there is agreement but also contains areas of controversy. That's inescapable. You're much better off knowing what the controversies are, what the issues are and what the arguments are on both sides than you are operating in a vacuum or in ignorance.

Economics is a very difficult subject. I've compared it to trying to learn how to repair a car when the engine is running. The economy is always changing, our knowledge of it is very incomplete, and our ability to predict it is not impressive. Nevertheless, I think that having good data, good statistics—and the United States generally has better macroeconomic statistics than most countries—and having good economists to interpret those data and present the policy alternatives, has a substantially beneficial effect on policymaking in the United States, not only in monetary policy but in other areas as well. I think in the end good economic policy research makes a very big difference to the welfare of the average person.

Rolnick: Thank you, Gov. Bernanke.