Presentation to the Annual AEA/ASSA Conference
San Francisco, CA
By Janet L. Yellen, President and CEO, Federal Reserve Bank of San Francisco
For delivery on January 4, 2009, 10:15 AM Pacific, 1:15 PM Eastern
Download PDF Version (118KB)
Comments on "The Revival of Fiscal Policy"
I'm delighted to have the opportunity to comment on the papers of my distinguished colleagues, Martin Feldstein, John Taylor, and Alan Auerbach at a session devoted to a topic of such critical policy importance. With the U.S. economy in a year-old recession, marked by a particularly sharp contraction in recent months, and ensnared in ongoing financial market distress, fiscal policy intervention is certainly needed. It is important that it be informed by the kind of cogent thinking and analysis that these papers embody.
I agree with Marty that the current downturn is likely to be far longer and deeper than the "garden-variety" recession in which GDP bounces back quickly. As Marty points out, a defining characteristic of this downturn is its cause. Typically, recessions occur when monetary policy is tightened to subdue the inflationary pressures that emerge during a boom. This time, the cause was the eruption of a severe financial crisis. Cross-country evidence suggests that, following such an event, GDP remains subdued for an extended period. And consistent with this evidence, many forecasters expect this to be one of the longest and deepest recessions since the Great Depression. Indeed, the crisis is ongoing. Risk-aversion in financial markets remains exceptionally high; deleveraging is widespread; the markets for most private asset-backed securities are dysfunctional; financial institutions, both large and small, have failed; and the economic downturn is causing delinquencies to rise, threatening further financial distress; households and businesses face an ongoing credit crunch; and housing and financial wealth has plunged. Marty points out, and I agree, that the likely impact on consumer spending of the decline in wealth thus far—one of a number of factors weighing on this sector—is, on its own, quite substantial. And house prices are continuing to slide.
The Federal Reserve has responded aggressively since the onset of the crisis with both interest rate reductions and a variety of new, specialized liquidity facilities to improve the functioning of impaired financial markets. At the FOMC's last meeting in December, the Committee reduced its target for the federal funds rate to a range of 0 to ¼ percent, effectively hitting the "zero bound" for the first time since the Great Depression. With nominal interest rates now essentially zero, we have reached the limits of what conventional monetary policy, working through short-term interest rates, can do to stimulate the economy. A special concern in the current circumstances is that a decline in inflationary expectations could cause real interest rates to rise. The potential for real interest rates to rise as inflation declines creates a significant downside risk because, with an extended period of abnormally high unemployment in the forecast, it is increasingly likely that inflation will fall to undesirably low levels. This is an additional consideration that, to my mind, militates in favor of strong policy responses.
I certainly agree with John Taylor that the zero bound doesn't mean that monetary policy has run out of options. At another session this afternoon, I will discuss in detail a range of non-conventional measures, using the Fed's "balance sheet," that will be a prime focus for monetary policy going forward. The Fed has already announced or undertaken, and can expand, programs designed to improve the functioning of credit markets. These programs complement the actions of the Treasury, the FDIC, and the Fed to inject capital into the banking system, to stabilize systemically important financial institutions, to expand deposit insurance, and to temporarily guarantee the senior debt of banking organizations. A clear lesson of history is that a sine qua non for sustained economic recovery following a financial crisis is a thoroughgoing repair of the financial system. The necessary steps are occurring now, but restoration of the financial system to full health will be a long, drawn-out process.
Less clear, and here I part ways with John, are the lessons from Japan's recent experience with "quantitative easing." In his paper for this session, John asserts that the BOJ's quantitative easing strategy worked well, while fiscal policy was ineffective. My interpretation of the evidence is exactly the opposite. The BOJ targeted an extraordinarily high level of excess reserves in the banking system, in the hope that a flood of such reserves might stimulate additional bank lending. The BOJ's quantitative easing policy probably had some beneficial effect, but mainly because it symbolized the BOJ's commitment to combat deflation by keeping interest rates at zero for an extended time. The expectation of an extended period of zero short-term interest rates was probably instrumental in lowering longer-term nominal rates. But the expansion of excess reserves to extraordinary levels appears, on its own, to have had very little impact on financial conditions. The Fed's "balance sheet" strategy has a different motivation. We are focused on pursuing carefully tailored programs to remedy specific financial market dysfunctions. For example, our suite of programs is designed to enhance liquidity in financial markets, to improve the availability of term funding in the money market, to restore the functioning of the commercial paper market, and to stimulate the issuance of new asset-backed securities to support lending to consumers and small businesses.
I am sanguine that the Fed's new programs will be helpful in restoring credit flows. But many of the new approaches are experimental, and there is a great deal of uncertainty concerning their likely effects. Even with vigorous Fed action to restore credit flows, an extended period of economic weakness is likely. This explains why, in the statement following its December meeting, the FOMC noted that it anticipates that an exceptionally low level of the federal funds rate will be appropriate for some time.
For all of these reasons, I support Marty's conclusion that there is an exceptionally strong case for substantial fiscal stimulus over the next few years. In ordinary circumstances, there are good reasons why monetary, rather than fiscal policy, should be used for stabilization purposes. But these are exceptional circumstances, and fiscal policy can help get the economy going.
Given that a substantial fiscal stimulus package is now under consideration, it is important for economists to debate the form it should take and raise questions about the likely effectiveness of particular fiscal strategies. With respect to optimal fiscal design, I have heard more alliterative mantras during the past year than at any time in my professional career! A year ago, the mantra was "Timely, targeted, and temporary." More recently, I've heard "Speedy, substantial, and sustained." In recent testimony, John argued for fiscal policies that are "Permanent, pervasive, and predictable." And the award for the longest, most detailed, and least alliterative list goes to the International Monetary Fund whose staff this week recommended a package that is "large, lasting, diversified, contingent, collective, and sustainable."
Conceptually, I would note that fiscal policy can play two logically distinct roles in the present crisis. The first, which is the main topic of the papers in this session and my subsequent comments, is to support aggregate demand. I will discuss some issues that should be considered in formulating a stimulus package for this purpose. The second, which is crucially important in the current crisis, is to address the maladies that now afflict the financial sector. Indeed, fiscal resources have already been deployed to bolster the capital of the banking system, to stabilize systemically important institutions, and to support programs designed to assist homeowners and mitigate foreclosure. Fiscal policy can play a further supportive role in unclogging credit market flows. For example, the new Term Asset-Backed Securities Loan Facility (TALF) is a joint program between the Federal Reserve and the Treasury, using TARP funds, and is designed to improve the flow of credit to households and businesses. The Fed cannot conduct such programs on its own, since taxpayer funds are at risk. As I noted earlier, the resolution of financial sector difficulties is essential for a return to sustained growth following a financial crisis. This conclusion is endorsed by the IMF in its recent position paper.
Let me now turn to some design issues that relate to the use of fiscal policy for stabilization purposes. First, as both John and Alan highlight, automatic stabilizers have an important role to play, and this role could be expanded. These stabilizers have the desirable property of being predictable and automatic—agents know they're coming, and they inevitably provide a boost just when it's needed. Using aggregate CBO data on the cyclical contribution to the budget deficit, John argues that these stabilizers are as effective as ever; it is hard, though, to square this finding with Alan's analysis showing that the automatic stabilizers have become smaller over time, as marginal tax rates have declined.
In the current context, however, I don't think that existing automatic stabilizers—of whatever size—are likely to provide a sufficient boost to aggregate demand. Moreover, both John and Alan make the important point that the predictable fiscal policy response to changing economic conditions does not rely solely on automatic stabilizers. Alan highlights the specific examples of countercyclical fiscal policy in the 2000s, as well as the predictable component over a long period of time in discretionary investment incentives. John notes, but does not emphasize, that the structural component of the budget deficit—reflecting specific, discretionary changes in legislation—has been highly responsive over the past decade or longer to changes in the output gap.
Hence, private agents rationally expect that in times of economic weakness, fiscal policy will respond and that expectation itself serves as a kind of automatic stabilizer in its own right, reassuring households and businesses that the government will do everything in its power to prevent a long period of intense economic distress. Agents have almost surely, at least implicitly, built some such expectation about discretionary fiscal policy into their current behavior, even if they are highly uncertain about the details of the response. Perhaps it would be desirable in the longer run to make the policy rule more explicit and predictable—and, as Alan emphasizes, to devote more analysis to understanding the effects of different policy rules. But I don't think now is the time to change the implicit, if imprecise, rules and disappoint agents' expectations about the discretionary fiscal response to the present crisis.
A second design principle that must be kept in mind, given the huge, long-term budget challenges facing the country, is that current responses need to be consistent with long-term fiscal discipline. Conceptually, the choices about the long-term size of government—with revenues that are consistent with the desired level of spending—and the choices about the optimal short-run stimulus are quite separate. But long-run expectations about fiscal sustainability do matter, even in the short run. If agents don't believe that tax cuts and/or spending increases are temporary, and are concerned that policy responses threaten the sustainability of the long-run run budget picture, then long-term real interest rates are likely to rise in response, undercutting, conceivably even overwhelming, the short-term stimulus. In the 1990s, the term "Rubinomics" came to summarize this possibility (in reverse): that "contractionary" measures to improve long-run fiscal balance could provide net, short-term stimulus through this channel.
In the present context, this logic creates potential tradeoffs between short-run and long-run considerations. In particular, the permanent income hypothesis implies that permanent tax cuts affect spending in the short run by more than temporary tax cuts; but if these tax cuts are not consistent with long-run spending realities, then they are inconsistent with fiscal sustainability and are not credible. For this reason, I am a bit skeptical about Marty's suggestion that we implement broad-based, permanent tax cuts. Especially given future obligations for social security and Medicare, voters and legislators might resist the sizeable reductions in other spending that would be required to keep taxes permanently lower than current legislation implies.
Of course, a concern for fiscal discipline is one reason why temporary tax rebates seemed desirable to many observers last year. The rebates were explicitly targeted at lower-income individuals who were more likely to be credit constrained and thus responsive to temporary as well as permanent changes in income. Some micro evidence suggests that the rebates were effective in boosting spending. On the basis of aggregate data, however, both Marty and John agree that the tax rebates were a bust—they failed to have enough of an impact on consumer spending to make a dent in the aggregate spending data. There are difficult econometric issues here—for example, credit constraints on households were tightening around the same time the checks were arriving, which could have obscured the true effects. Nevertheless, the rebate checks at best delayed the most severe period of the current recession, and at worst were ineffective.
How about investment incentives? In contrast to temporary tax rebates, economic theory implies that temporary investment incentives should be even more effective in stimulating aggregate demand than permanent incentives. Hence, they should work well in terms of providing large short-term bang for the buck with little damage to long-run fiscal balance. Unfortunately, Alan's work suggests that, while this logic might hold, it isn't the whole story. Investment incentives are clearly countercyclical—when the economy slows, investment incentives rise and firms contemplating capital investment know this. So the anticipation effects could cause undesirable fluctuations in investment and economic activity.
In theory, as Marty suggests, temporary increases in government spending—or, perhaps, changes in the timing of government spending—fit the bill well. Such spending has the advantage of hitting aggregate demand directly. There is uncertainty in the academic literature about the multiplier, although I think that the multiplier is likely to be larger than usual now because private spending is less likely than usual to be crowded out. Of course, it is unclear how quickly temporary spending programs can be ramped up and, as Marty suggests, there could be bottlenecks of various sorts.
Together, these considerations reinforce the IMF's recent suggestion that a diversified package of policies should be deployed. That is, within the class of fiscally sustainable policies, there is considerable uncertainty about the effects of any individual policy. So we should try a lot of things, including, as Marty suggests, government spending.
Of course, there are many things the government could spend money on and this brings me to a third design principle: Policy should aim to promote spending on goods and services with higher rather than lower social value. After all, the goal of policy is to raise welfare. And spending money on projects that aren't viewed as having much social value—such as digging holes and filling them in again—adds a lot less to social welfare than available alternatives, even if such low-value activities can succeed in increasing aggregate demand. Tax cuts do have the advantage of giving individuals the choice about how to spend their money. But, as I've already noted, it is unclear whether fiscally sustainable tax cuts will have sufficient stimulative effect.
In this context, as Marty's argument implies, there are clear macroeconomic gains from direct government spending. In cost-benefit terms, the aggregate demand externalities increase the benefits of government spending relative to the costs. Hence, programs that already look worthwhile ought to be implemented; and some programs that might not otherwise pass a cost-benefit test could also be justified given the considerable slack in the economy. There are real dangers, however, in pushing this logic too far and Japan's experience, to my mind, illustrates them. John suggested that discretionary fiscal policy made no meaningful contribution to Japan's recovery. I disagree with that characterization. Japan's decision to raise its consumption tax and cut public spending in 1997 aborted an incipient recovery then in progress; and the fiscal stimulus put in place in 1998 and 1999 appears to have prompted a rebound. Most studies find that fiscal measures did have a positive stimulative effect, although estimates of the fiscal multipliers vary. Even so, I have the impression that John's view is widely held in Japan. I suspect that fiscal policy was both unpopular, and widely viewed as ineffective, in Japan because so much of it involved spending on "bridges to nowhere"—public works projects of dubious social value. Such spending appears to have undermined popular support for fiscal stimulus.
If ever, in my professional career, there was a time for active, discretionary fiscal stimulus, it is now. Although our economy is resilient and has bounced back quickly from downturns in the past, the financial and economic firestorm we face today poses a serious risk of an extended period of stagnation—a very grim outcome. Such stagnation would intensify financial market strains, exacerbating the problems that triggered the downturn. It's worth pulling out all the stops to ensure those outcomes don't occur.
# # #
1. The views expressed here are my own and do not necessarily represent those of my colleagues in the Federal Reserve System. I would like to thank the staff of the San Francisco Fed's Economic Research Department, and John Fernald and John Judd in particular, for outstanding support in developing these remarks.
2. See, for example, Carmen Reinhart and Kenneth Rogoff, "Is the 2007 U.S. Subprime Crisis So Different?" American Economic Review Papers and Proceedings 98(2), May 2008, or International Monetary Fund, World Economic Outlook, October 2008, Chapter 4. (http://www.imf.org/external/pubs/ft/weo/2008/02/pdf/text.pdf).
3. Lawrence Summers, November 2008, quoted at http://blogs.wsj.com/economics/2008/11/17/summerss-new-view-on-stimulus-sustained-not-temporary/ Downloaded December 31, 2008.
4. John Taylor, "The State of the Economy and Principles for Fiscal Stimulus," Testimony before the Committee on the Budget, United States Senate, November 19, 2008. Downloaded December 31, 2008.
5. Antonio Spilimbergo, Steve Symansky, Olivier Blanchard, and Carlo Cottarelli, "Fiscal Policy for the Crisis," IMF Staff Position Note, December 29, 2008, SPN/08/01.
6. See Spilimbergo et al., 2008.
7. Christian Broda and Jonathan Parker (2008), "The Impact of the 2008 Tax Rebates on Consumer Spending: Preliminary Evidence," mimeo University of Chicago, GSB, July 2008.
8. This is consistent with the view expressed in Spilimbergo et al. (2008, page 25).