The Cost-Benefit Challenge

In its latest Room for Debate feature, The New York Times poses the question "Should the Fed Risk Inflation to Spur Growth?" Befitting a balanced panel of blogging experts, Mark Thoma (Economist's View) says "yes," John Cochrane (The Grumpy Economist) says "no," and Edward Harrison (Credit Writedowns) says something like "irrelevant question, it's going to do neither."

The whole discussion, naturally, is about differing assessments of the costs and benefits of additional monetary stimulus. Not surprisingly, this was also a theme disclosed in the just released minutes of the July 31–August 1 meeting of the Federal Open Market Committee:

Participants also exchanged views on the likely benefits and costs of a new large-scale asset purchase program. Many participants expected that such a program could provide additional support for the economic recovery both by putting downward pressure on longer-term interest rates and by contributing to easier financial conditions more broadly. In addition, some participants noted that a new program might boost business merits of purchases of Treasury securities relative to agency MBS. However, others questioned the possible efficacy of such a program under present circumstances, and a couple suggested that the effects on economic activity might be transitory. In reviewing the costs that such a program might entail, some participants expressed concerns about the effects of additional asset purchases on trading conditions in markets related to Treasury securities and agency MBS, but others agreed with the staff's analysis showing substantial capacity for additional purchases without disrupting market functioning. Several worried that additional purchases might alter the process of normalizing the Federal Reserve's balance sheet when the time came to begin removing accommodation. A few participants were concerned that an extended period of accommodation or an additional large-scale asset purchase program could increase the risks to financial stability or lead to a rise in longer-term inflation expectations...

The questions about the costs and benefits of any particular policy intervention are abundant, and for virtually every potential pro there is a potential con. Here is my personal, certainly incomplete list of pros/cons or benefits/costs associated with another round of large-scale asset purchases:

Pro:	Lower interest rates (and perhaps a lower dollar) will on balance spur spending.
Con:	The expectation of low interest rates for a longer period of time will reduce the urgency to borrow and spend.
Pro:	Expanded asset purchases and lower rates will preserve needed liquidity in financial markets.
Con:	Expanded asset purchases and lower rates will create or exacerbate financial market distortions.
Pro:	More monetary stimulus reduces the probability of an undesirable disinflation in the near term.
Con:	More monetary stimulus increases the probability of undesirable inflationary pressures in the longer term.
Pro:	Lower Treasury and MBS rates will induce an appetite for risk taking that is needed to get productive resources "off the sidelines."
Con:	Lower Treasury and MBS rates will induce an appetite for risk taking that sets us up for the next bubble.
Pro:	Monetary policy is the only channel of support for the economy, absent new fiscal policies.
Con:	Monetary policy support is relieving the pressure to make needed fiscal reforms that would be much more effective than monetary stimulus.
Pro:	With additional monetary stimulus, GDP growth will be higher and unemployment lower than they would otherwise be, and outcomes may be more consistent with the FOMC's mandate to promote maximum employment.
Con:	With additional monetary stimulus, the exit from monetary stimulus once the economy improves will be more difficult than it would otherwise be, and outcomes may be inconsistent with the FOMC's mandate to achieve price stability.
Pro:	The performance of the economy has not been consistent with the FOMC's mandated objectives.
Con:	The economy is slowly moving in the direction of the FOMC's mandated objectives, and the Fed should "keep its powder dry" in case of further deterioration of the economy.

Many, if not all, of these benefits and costs are familiar, and there has been no shortage of opinions advanced. To some extent, it is inevitable that the weighting of these costs and benefits will be to a large degree judgmental. How one weights the risks associated with continued high rates of unemployment versus the risks of imbalances that may arise from low interest rates, for example, is a subjective thing.

Nonetheless, it would be helpful to frame these subjective judgments with a background of some hard evidence. For example, how much employment can we gain for a given quantity of asset purchases (or any other monetary policy option on the table)? What are the likely—or existing—distortions created by low interest rates, and what do we think are the tangible costs associated with those distortions? And so on.

So, here is my challenge question: No matter what your opinion about what should be done, and whether you arrive at a conclusion by casual observation, econometric studies, or historical evidence, what do you think is the best evidence concerning any or all of these costs and benefits?

OK, then. Bloggers blog, commentators comment.

Update: Jon Hilsenrath sizes up some of the costs of a new bond-buying program.

By Dave Altig, executive vice president and research director at the Atlanta Fed

 

The (Unfortunately?) Consistent Record of the Recovery: Duy and Thoma Respond

Mark Thoma, always generous in linking and reposting our musings here at macroblog, took a look at my last post and read a sense of helpless resignation:

David Altig of the Atlanta Fed argues that "the majority of FOMC participants don't seem to think that the unemployment rate will improve that quickly, but "it is not at all obvious that the pace of the recovery is inconsistent with the FOMC's view of achieving its dual mandate." It sounds as though the Fed has given up -- we've done all that we can, there's nothing more we can do, so we won't even try -- and we're not about to risk even the tiniest bit of inflation to find out if we are wrong (and this is despite assurances from Bernanke and others that the Fed is not out of bullets)...

Tim Duy, whose observations in fact motivated my post, had his own response to my comments:

Altig's calculations make the important assumption that the labor force participation rate holds at 63.7%. This effectively assumes that none of the decline in the labor force participation is cyclical. Instead, it is all structural...

There are really two separate thoughts in these comments. So let me take them in turn, but first recap what I said in the previous post (or at least what I meant to say):

• Stepping back and looking at the data, I am drawn to the conclusion that the U.S. economy looks like it has settled into a pattern of something like 2 percent GDP growth with net job creation somewhere around 150,000 payroll jobs per month.


• The unemployment projections published in the FOMC participants.' June Summary of Economic Projections (SEP) would, under certain assumptions, be consistent with annual job growth averaging the 150,000 per month pace we have seen over the past year and a half.

The under certain assumptions caveat is obviously important. To Duy's point, my mapping of the apparent employment trend to the SEP unemployment forecasts does assume a constant labor force participation rate. Multiple macroblog posts this year have offered skepticism about exactly that assumption (here and here, for example), and any rise in the labor force participation rate will require faster job growth to get the same unemployment rate outcomes. But as far as I know—I think as far as anyone knows—participation will rise only if we get that faster job growth in the first place. My only point was that the SEP unemployment rate submissions in June are not obviously out of line with what appears to be the current trend in job creation.

In fact, I cannot tell you what assumptions underlie the unemployment rate (and growth) projections in the SEP. I can tell you only that these are the outcomes the individual participants view as consistent with "appropriate monetary policy." And that brings us to Mark Thoma's concern that the very slow progress toward higher growth and lower unemployment in the SEP implies that the Fed has "given up, "done all that we can," and "won't even try."

I definitively do not want to leave an impression that this view is implied by the SEP. As I noted in my original post (and duly noted in both the Thoma and Duy responses), the definitions of appropriate monetary policy that condition the individual SEP contributions are not spelled out. Lacking that information, one should not infer that monetary policy is assumed by any one individual as fixed or without influence.

Could it possibly be that an unemployment rate at 7.5 percent and GDP growth of 2.8 percent in 2013 (the more optimistic forecasts in the majority, or the "central tendency" range in the June SEP) are consistent with monetary policy having a nontrivial positive impact on the economy?

Of course monetary policy does not operate in a vacuum. As our boss, Atlanta Fed President Dennis Lockhart, said in a speech yesterday:

Monetary policy can exert a powerful positive influence on an economy, but as Chairman Bernanke has pointed out, monetary policy is not a panacea.

I'm not really aware of any models matched to real-world data that suggest monetary policy actions can (at acceptable cost) quickly and completely overcome all of the shocks and headwinds that may present themselves.

You may believe otherwise—that is, you may believe that, for current circumstances, monetary policy is a panacea. Or, less dramatically, you may believe that more monetary stimulus would surely yield something better than what was implied in the June SEP. Fair enough. But you should not believe that lackluster numbers in the SEP tell you anything about individual FOMC participant's views on the efficacy, desirability, or likelihood of further monetary actions, one way or the other.

By Dave Altig, executive vice president and research director at the Atlanta Fed

What’s to be done?

It's always hard to please everyone. Sometimes it's hard to please anyone. You probably don't need a lot of convincing on this point, but if you desire one more case study look no further than the past week's commentary on monetary policy, starting with Wonkblog's rather negative performance review (post by Brad Plumer) of the Fed's recent policy decisions:

"Right now, unemployment is falling more slowly than the Fed expected when it issued its forecasts back in April... When the Fed published its forecasts, it expected more jobs reports like April's, which initially showed the economy adding 115,000 jobs new jobs. But that hasn't happened… Which means the Fed's own numbers prove the Fed is failing to meet its dual mandate of keeping unemployment and inflation low. (Inflation is below the central bank's target right now; unemployment is not.)"

Plumer favorably references an earlier item from the Peterson Institute's Joe Gagnon, bearing the damning title "The Fed Shirks Its Duty":

"On June 20, 2012, the Federal Reserve System's Federal Open Market Committee extinguished the last shred of doubt as to whether it intends to achieve its mandated objectives."

Carnegie Mellon's Alan Meltzer similarly wonders "What's Wrong With the Federal Reserve?" But his lament, published earlier this week in the op-ed pages of the Wall Street Journal, doesn't exactly mesh with the Gagnon-Plumer school of thought.

"One of the Fed's big mistakes is excessive attention to the short term, over which it has little influence...

"The problem with the short term is that data reported today are subject to revision, or reflect only transitory changes. The better economic data last winter are one of many examples. Would the reported improvement in the economy persist? We didn't learn the answer until weaker data reported this spring. Is the slowdown persistent or temporary? We can only guess.

"Executing monetary-policy changes in response to transitory data is a mistake...

"Today's economic problems are serious, but the Fed can't do much about them if these problems are not monetary. Very expansive monetary policies did help during the crisis of 2008–09, but they're not what is needed now…"

I don't see a dispute here about the fact in the first half of the year the U.S. economy has grown considerably slower than most people—including those in the Fed—thought it would. As usual, the dispute comes down to how to interpret those facts and what to do about them.

Material differences of opinion about how to interpret the current economic environment was the focal point of a speech given today by Atlanta Fed President Dennis Lockhart, in Jackson, Mississippi. Acknowledging the divergent views represented by the Plumer, Gagnon, and Meltzer views, President Lockhart offers his own:

"The question that the members of the FOMC confront is whether there is more that can be done to address the related challenges of slower GDP growth and tepid job creation. So, to wind up, let me give you my take on the key questions underlying a decision to bring on more monetary stimulus.

"I think the output gap—the amount of slack in the economy—is neither as sizeable as the high-end estimates, nor is it zero. If there were no slack at all, 8.2 percent unemployment would represent full employment. If this were so, the economy would have undergone profound structural change over the last five years. As I weigh the findings of research by Federal Reserve economists and others, I do not think a compelling case has yet been made that structural adjustment has played a dominant role in slowing growth and progress against unemployment.

"If, on the other hand, slack in the economy were close to the high estimates, we should have seen more and more persistent downward pressure on prices and wages than has, in fact, been the case. Deciding on the extent of the output gap is not straightforward. I believe the truth is in the gray middle.

"On the risk associated with the balance sheet: in my judgment, some further use of the balance sheet to promote continued recovery and/or financial stability brings with it manageable risks. I think reversal of the cumulative balance sheet scale and maturity structure can be accomplished in an orderly manner. But the step of additional balance sheet expansion should be undertaken very judiciously. Such a step would take us further into uncharted territory.

"On the likely effectiveness of further monetary stimulus—a policy that would necessarily be brought to bear at least in part through credit channels—I think we should have modest expectations about what further action can accomplish. I do not think this means monetary policy is impotent or has reached its limit. But I don't see more quantitative easing or similar policy action as a miracle cure, especially absent fixes in policy areas outside the central bank's purview."

And to the dimming forecasts:

"So, as one policymaker, here's my situation: my support for the current stance of policy rests on a forecast that sees a step-up of output and employment growth by year-end and into 2013. If the economy continues on the track indicated by the most recent incoming data and information, that forecast will become untenable, as will the policy premises underlying it."

Plumer and Gagnon argue we are already at that point. Meltzer believes otherwise. Lockhart is weighing both possibilities. That approach pleases neither camp, but it's the right thing to do.

By Dave Altig, executive vice president and research director at the Atlanta Fed

The armchair Fed historian

I enjoy researching economic history—some of my work with Steve Quinn on early central banking is here, here, and here.

The only problem with historical research, though, is that it tends to involve some real work—long hours spent with dusty archival volumes, consumption of lots of coffee and antihistamines, and a steady hand on the digital camera.

Just recently—and somewhat belatedly—I became aware of a Google application (thanks to Benjamin Guilbert's blog) that lets would-be economic historians breeze over the rough stuff and do some interesting research from their own computer keyboards. The application is called Ngrams, and here's how it works.

Basically, Ngrams counts occurrences of words in books that have been scanned by Google into its Google Books database. It then plots out the frequencies of these words as annual time series. These plots can then be used to measure how interest in a topic varies over time—to construct "cultural histories."

There are some limitations to this technique, mostly related to unavoidable issues in Google's database. For example, the dataset I chose to work with covers only English-language publications and stops in mid-2009. (See the Ngrams website for more detailed information.)

The six charts below represent a first attempt to use Ngrams to delve into the cultural history of the Federal Reserve.

Question 1: How popular is the Federal Reserve as a discussion topic compared with other central banks?

• Search terms: Bank of England, Federal Reserve, Reichsbank, Bundesbank, Bank of Japan
• Time period: 1900–2008

(enlarge)

• My interpretation: almost from its beginning in 1913, the Federal Reserve has been the primary focus of English-language writing on central banks.

Question: 2 The Fed was founded as a means to counteract banking panics. What has been the impact of the Fed on the discussion of panics?

• Search term: bank panic
• Time period: 1866–2008

(enlarge)

• My interpretation: that bank panics were widely discussed in the wake of three National Banking Era panics in 1873, 1893, and 1907, no surprise. Interest in bank panics peaked following the widespread bank failures of the early 1930s. This topic became less popular after World War II, but interest reawakened with the numerous savings and loan failures of the 1980s and early 1990s.

Question 3: One of the early policy goals of the Fed was to improve the efficiency of the check payment system. When did use of checks become the norm for ordinary Americans?

• Search terms: pay envelope, pay check, paycheck
• Time period: 1900–2008

(enlarge)

• My interpretation: in 1920, most people did not have checking accounts and were paid in envelopes stuffed with cash. By 1960, most households had checking accounts and were paid by "pay check," later contracted to "paycheck."

Question 4. What has been the impact of the Fed on people's concerns about inflation and unemployment?

• Search terms: unemployment, inflation
• Time period: 1900–2008

(enlarge)

• My interpretation: interest in unemployment shot up during the Great Depression, fell back in the postwar years, but resurged in the 1970s. Discussion of unemployment then falls steadily to the end of the sample in 2008. Inflation was rarely discussed until the United States left the gold standard in 1933. Interest in inflation remained below unemployment until inflation began to accelerate in the 1970s. Since about 1980, interest in these two topics has been almost identical.

Question 5: In the mind of the public, which policy goal should the Fed be most concerned with: price stability, financial stability, or employment?

• Search terms: price stability, financial stability, Phillips curve (as an imperfect proxy for "employment"; note that the original article by William Phillips appeared in 1958)
• Time period: 1900–2008

(enlarge)

• My interpretation: financial stability was paramount until after the 1951 Treasury-Fed Accord. Price stability then takes center stage until the turn of the 21st century but by 2008 had converged with financial stability. Interest in the Phillips curve seems to have peaked in the early 1980s.

Question 6: What has been the impact of two "big ideas" on monetary policy, proposed by Robert E. Lucas (1976) and John B. Taylor (1993)?

• Search terms: Lucas critique, Taylor rule
• Time period: 1970–2008

(enlarge)

• My interpretation: in his 1976 paper, Lucas argued that there were limits on the usefulness of statistical relationships (the Phillips curve in particular) in monetary policymaking. Partly in response to the Lucas critique, Taylor in 1993 proposed that central banks follow a simple rule in setting short-term interest rates. Interestingly, discussion of the Lucas critique peaked around the time of the publication of Taylor's paper. Interest in the Taylor rule was still growing at the end of the sample in 2008.

You may or may not agree with the choice of search terms or the interpretations of the search results, but you are welcome to conduct your own historical research with the same application—all from the comfort of your armchair, no digital camera required. We'll have more of these cultural histories to share in later posts.

By Will Roberds, research economist and senior policy adviser at the Atlanta Fed

The relative expansion of central banks’ balance sheets

Dave Altig's recent macroblog post on policy actions that affected the Fed's balance sheet made me wonder about how changes to the Fed's balance sheet since the financial crisis compared with other central banks.

Relative to before the financial crisis, the Federal Reserve's asset holdings are currently about 3.3 times larger. Initially, the source of that increase was the collateral associated with various temporary lending facilities that the Fed used to address the financial panic. Those assets were then replaced on net by purchases under the first large-scale asset purchase program in 2009. Then in late 2010, asset holdings increased further as a result of a second large-scale asset purchase program.

Of course, size isn't everything. While it might be tempting to try and interpret the change in the size of the central bank's balance sheet as a summary statistic of the degree of monetary policy accommodation, as Dave Altig's post points out, that interpretation is not so straightforward. Increasing the size of the balance sheet is not the only thing a central bank can do to ease monetary policy when short-term interest rates are very low. For example, in late 2011 the Fed began a maturity extension program that changed the composition of the assets on the balance sheet, but this program did not materially alter the size of the balance sheet.

With this caveat in mind, the following chart compares the proportionate changes in the size of asset holdings of five central banks over the period from the first quarter of 2007 through the first quarter of 2012: the Federal Reserve (FR), the Bank of England (BE), the European Central Bank (ECB), the Bank of Canada (BC), and the Bank of Japan (BJ).

One take-away from the chart is the large variation from country to country. Here are some observations:

1. Bank of England: Through mid-2011, the proportionate increase in the Bank of England's asset holdings was roughly similar to the Fed's. But then the Bank of England began a second round of large scale asset purchases that sharply increased the size of its balance sheet. By the first quarter of 2012, the Bank of England's asset holdings were about 4.2 times as large as they were before the financial crisis.


2. European Central Bank: Through mid-2011, the ECB's asset holdings were about 1.7 times their precrisis level. But the sharp increase in the ECB's longer-term lending programs in recent months has resulted in a large increase in the size of ECB's balance sheet. By the first quarter of 2012, the ECB's asset holdings were about 2.5 times what they were before the financial crisis.


3. Bank of Canada: In 2009, the Bank of Canada's asset holdings had increased to about 1.6 times their precrisis level—similar to the ECB's increase. But as liquidity pressures in Canadian financial markets eased, the Bank of Canada's asset holdings declined in 2010. By the first quarter of 2012, the Bank of Canada's asset holdings were around 1.3 times the precrisis level. (Note that the Bank of Canada's asset data are through February 2012.)


4. Bank of Japan: The balance sheet of the Bank of Japan did not increase materially during the financial crisis, but has increased somewhat over the last year. By the first quarter of 2012 the Bank of Japan's asset holdings were about 1.2 times the pre-crisis level.

While size isn't everything, it is something. A large expansion in a central bank's balance sheets can create broad policy risks. This study by researchers at the St. Louis Fed suggests that large-scale balance sheet increases are a viable monetary policy tool, provided the public believes the increase will be appropriately reversed (citing the experience of Nordic countries in the early 1990s) or that the reserves created by the expansion will remain within the banking system (citing changes to bank settlement systems in the United Kingdom and New Zealand in the mid-2000s). New York Fed President Bill Dudley touched on some risks in an interview on CNBC today:

"...We've expanded our balance sheet a lot over the last few years. And additional actions do have costs, and so we have to weight them relative to the benefits...

"One set of cost is the extent we expand our balance sheet or we sell short-dated treasury securities and buy long-dated treasury securities, we have more risk, in terms of our portfolio, interest rate risks...

"The second issue, of course, is if we expand our balance sheet, we could create anxiety among some people that this might actually sow the seeds for future inflation. I don't think expansion of the balance sheet, in any way, compromises the Fed's ability to keep inflation in check over the longer term. But it doesn't matter just what I think. If people in the market think that expansion of the balance sheet could cause future inflation, we have to take those expectations into consideration as a potential cost of monetary policy."

John Robertson, vice president and senior economist in the Atlanta Fed's research department

The three faces of postcrisis monetary policy

The latest edition of the San Francisco Fed's Economic Letter (written by Michael Bauer)has a nice review of the different channels through which the Fed's Large Scale Asset Purchase (LSAP) programs—QE, or quantitative easing more popularly—are thought to work:

"Central bank LSAPs potentially may affect interest rates through at least three channels. Notably, all three channels can broadly affect longer-term interest rates, extending beyond those securities that the central bank announces it will purchase:

• A portfolio balance channel, because the supply of long-maturity bonds available to private investors is reduced. The reduced supply of longer-term securities targeted by the Fed lowers the amount of interest rate risk in investor portfolios. That in turn decreases the risk premium that they require to hold both the targeted securities and other assets of similar duration. Longer-term interest rates are lowered across the board as a result. Gagnon et al (2011) emphasize this channel for QE1.
• A signaling channel, which arises when the Fed's announcements are interpreted as signals of its intent to hold down short-term interest rates further into the future. Bauer and Rudebusch (2011) argue that this channel played an important role for QE1.
• A market functioning channel, because QE1 provided relief when conditions in financial markets were dire, liquidity very low, and panic widespread. The Fed's intervention calmed investor fears. Thus, the intervention substantially supported a range of asset prices, including MBS and corporate bonds, lowering their yields."

The article references include links to the Gagnon et al. paper and the Bauer and Rudebusch paper, but none to any studies addressing the "market functioning channel." So I'll provide one: "Did the Federal Reserve's MBS Purchase Program Lower Mortgage Rates?" by Diana Hancock and Wayne Passmore, both senior staff members for the Federal Reserve of Board of Governors. According to Hancock and Passmore, the market functioning channel is key to appreciating the impact of QE1:

"We use empirical pricing models for MBS yields in the secondary mortgage market and for mortgage rates paid by homeowners in the primary mortgage market to measure how distorted mortgage markets were prior to the Federal Reserve's intervention, and the course of market risk premiums during the restoration to normal market functioning...

"We argue that this return to normal pricing occurred because the Federal Reserve's announcement signaled a strong and credible government backing for mortgage markets in particular and for the financial system more generally...

"More specifically, we estimate that the Federal Reserve's MBS purchase program over the course of 16 months reestablished normal market pricing in the MBS market and resulted in lower mortgage rates of roughly 100 to 150 basis points for purchasing houses. Most of the decline in mortgage rates occurred between the announcement of the program, on November 25, 2008, and the implementation of the program in the first quarter of 2009. After this point, both mortgage rates and risk premiums remained relatively stable until the end of the Federal Reserve MBS purchase program."

Hancock and Passmore note that the portfolio balance channel may have played a role after the completion of the QE1 purchases once market functioning had normalized, but the biggest bang was that renormalization itself.

Bauer's observations align with Hancock and Passmore's conclusions:

"QE1 had very pronounced effects on interest rates. The key announcements led to decreases of close to one percentage point. The announcements not only lowered yields on targeted Treasury securities and MBS, but also on corporate bonds...

"The two other programs, QE2 and MEP [maturity extension program], also affected yields of securities that were not targeted for Fed purchases... Generally though, QE2 and MEP affected interest rates much less than QE1 did. One reason is that bond market functioning had largely returned to normal. In addition, expectations of future short-term interest rates were already very low when these programs were announced, leaving little room for further signaling effects. Finally, QE2 and MEP were smaller than QE1."

Earlier this week, in a speech delivered in Tokyo at the Institute of Regulation and Risk, Federal Reserve Bank of Atlanta President Dennis Lockhart provided his view on this evidence:

"In my view, these [the QE1] purchase programs played an important role in the transition away from the emergency lending facilities created earlier in the crisis. The emergency credit facilities worked well to stem the downward spiral of the immediate post-Lehman period. Financial markets began the process of repair during the first half of 2009 but were still suffering from relatively serious liquidity pressures. The QE1 operation sustained the liquidity support that had been previously provided by lending through the emergency facilities.

"Because asset purchases largely replaced emergency loans made during the crisis, the net increase in the Fed's balance sheet was relatively modest. In this sense, the quantitative easing label is misleading. The intent and effect of the policy was not to inject a new and sizable quantity of reserves into the economy. Rather, the effect was to sustain liquidity in still struggling and fragile financial markets, particularly those related to residential real estate. For that reason, I prefer the term ‘credit easing' to describe this policy action."

However, the smaller impact of QE2 leads Lockhart to a different conclusion regarding the largest contribution of that program:

"I view QE2 differently. The FOMC [Federal Open Market Committee] formally announced QE2 in November 2010, with its decision to purchase $600 billion in longer-term Treasury securities. However, the policy was signaled in an important speech from Federal Reserve Chairman Ben Bernanke in August of that year. The circumstances at the time were dominated by a falling trend in measured inflation, weakening inflation expectations, and rising probabilities of outright deflation. Each of these developments was effectively reversed as the expectations for QE2 took root, expectations that were ultimately validated by FOMC action.

"Unlike QE1, QE2 did materially expand the size of the Federal Reserve's balance sheet. In my view, this distinction is important. The intent and effect of the two rounds of asset purchases were different. QE1 served to maintain liquidity at a time when financial markets were exceptionally unsettled. In contrast, QE2 was a more traditional monetary action to preserve price stability."

In a sense, this places the effects of QE2 in the signaling channel category, albeit with an emphasis on inflation expectations rather than interest rates directly.

Bauer's article also covers post-QE2 policy—the maturity extension program (MEP, or "Operation Twist") and the insertion of specific calendar dates (currently at least late 2014) to provide forward guidance on the period of time that the FOMC anticipates that the federal funds rate will remain at exceptionally low levels. Lockhart also describes these policies in terms of the "signaling channel," though in these cases with interest rate effects front and center:

"In terms of intent and effect, I think of the explicit forward guidance and the MEP in similar terms. We have entered a phase of the recovery in which sustained monetary accommodation is warranted in order to preserve and advance what is still modest progress on employment and economic growth. Importantly, this modest progress is occurring in the context of what, for me, is acceptable performance with respect to our price stability mandate. Actions that reinforce the maintenance of policy accommodation are appropriate. It is through that lens that I view the MEP and explicit forward guidance on policy rates."

Lockhart's remarks provide his perspective on three somewhat distinct policy challenges—market dysfunction, disinflationary pressures, and a need to sustain monetary policy accommodation—that motivate his support for the three major policy initiatives of the postcrisis period:

"Let me summarize this brief tour of postcrisis monetary policy. I view the sequence of nontraditional monetary policy actions as tailored responses to the particular needs of the economy and financial system at the time they were implemented. My conclusion is that by and large policy actions have been appropriate to the diagnosis of circumstances at the time. And in my assessment they have worked pretty well."

In this light, President Lockhart delivers his policy punch line:

"I have reframed to some extent the original question of what more can be done around the point that policy actions must be matched to circumstances. The challenge policymakers face is judging appropriateness of a tool for circumstances. As popular as it might be in some quarters to rule out further LSAPs (QE3, as it is known), I do not think this option can be taken off the table. QE3 will work under the right circumstances. But I don't believe such circumstances prevail at this time."

By Dave Altig, executive vice president and research director at the Atlanta Fed

 

Is inflation targeting really dead?

Harvard's Jeffrey Frankel (hat tip, Mark Thoma) is the latest econ-blogger to cast an admiring gaze in the direction of nominal gross domestic product (GDP) targeting. Frankel's post is titled "The Death of Inflation Targeting," and the demise apparently includes the notion of "flexible targeting." The obituary is somewhat ironic in that at least some of us believe that the U.S. central bank has recently taken a big step in the direction of institutionalizing flexible inflation targeting. Frankel, nonetheless, makes a case for nominal GDP targeting:

"One candidate to succeed IT [inflation targeting] as the preferred nominal monetary-policy anchor has lately received some enthusiastic support in the economic blogosphere: nominal GDP targeting. The idea is not new. It had been a candidate to succeed money-supply targeting in the 1980's, since it did not share the latter's vulnerability to so-called velocity shocks.

"Nominal GDP targeting was not adopted then, but now it is back. Its fans point out that, unlike IT, it would not cause excessive tightening in response to adverse supply shocks. Nominal GDP targeting stabilizes demand—the most that can be asked of monetary policy. An adverse supply shock is automatically divided equally between inflation and real GDP, which is pretty much what a central bank with discretion would do anyway."

That's certainly true, but a nominal GDP target is consistent with a stable inflation or price-level objective only if potential GDP growth is itself stable. Perhaps the argument is that plausible variations in potential GDP are not large enough or persistent enough to be of much concern. But that notion just begs the core question of whether the current output gap is big or small. At least for me, uncertainty about where GDP is relative to its potential remains the key to whether policy should be more or less aggressive.

In another recent blog item (also with a pointer from Mark Thoma), Simon Wren-Lewis offers the opinion that acknowledging uncertainty about size of the output gap actually argues in favor of being "less cautious" about taking an aggressive policy course. The basic idea is familiar. It is a simple matter to raise rates should the Fed overestimate the magnitude of the output gap. But with the short-term policy rates already at zero, it is not so easy to go in the opposite direction should we underestimate the gap.

No argument there. As I pointed out in a May 3 macroblog item, Atlanta Fed President Dennis Lockhart has said the same thing. But, as I argued in that post, this point of view is only half the story. Though I agree that the costs are asymmetric to the downside with respect to the FOMC's employment and growth mandate, they look to me to be asymmetric to the upside with respect to the price stability mandate. And I view with some suspicion the claim that we know how to easily manage policy that turns out to be too aggressive after the fact.

My issues are not merely academic. In an important paper published a decade ago, Anasthsios Orphanides made this assertion:

"Despite the best of intentions, the activist management of the economy during the 1960s and 1970s did not deliver the desired macroeconomic outcomes. Following a brief period of success in achieving reasonable price stability with full employment, starting with the end of 1965 and continuing through the 1970s, the small upward drift in prices that so concerned Burns several years earlier gave way to the Great Inflation. Amazingly, during much of this period, specifically from February 1970 to January 1977, Arthur Burns, who so opposed policies fostering inflation, served as Chairman of the Federal Reserve. How then is this macroeconomic policy failure to be explained? And how can such failures be avoided in the future?...

"The likely policy lapse leading to the Great Inflation …can be simply identified. It was due to the overconfidence with which policymakers believed they could ascertain in real-time the current state of the economy relative to its potential. The willingness to recognize the limitations of our knowledge and lower our stabilization objectives accordingly would be essential if we are to avert such policy disasters in the future."

With this historical observation in hand, it seems a short leap to turn Wren-Lewis's thought experiment on its head. Arguably, the last several years have demonstrated that nonconventional policy actions have been quite successful at short-circuiting the disinflationary spirals that pose the central downside risk when interest rates are near zero. (If you can tolerate a little math, a good exposition of both theory and evidence is provided by Roger Farmer.)

On the opposite side of the ledger, we know little about the conditions that would cause the Fed to lose credibility with respect to its commitment to its inflation goals, and very little about the triggers that would cause inflation expectations to become unanchored. Thus, I think it not difficult to construct a plausible argument about the risks of being wrong about the output gap that is exact opposite of the Wren-Lewis conclusion.

I end up about where I did in my previous post. Flexible inflation targeting, implemented in such a way that the 2 percent long-run inflation target rate exerts an observable gravitational pull over the medium term, feels about right to me. Despite what Frankel seems to believe, I think that idea is far from dead.

By Dave Altig, executive vice president and research director at the Atlanta Fed

Symmetric goals, asymmetric risks

Mark Thoma has been hanging out with my boss or at least was at the same conference, where Thoma had a chance to try out his reporter chops:

"I just got out of a press conference with Dennis Lockhart (Atlanta Fed president) and Charles Evans (Chicago Fed president). I can't say there was any real news, but I did manage to ask a question... I asked whether the 2% inflation target was truly symmetric."

Thoma got an answer, though seemingly not one that left him totally convinced:

"Both insisted that the target is symmetric. However, Lockhart said that we know much more about the effects of inflation than deflation, and that preventing deflation was therefore job number one (which doesn't really answer the question). He didn't explain what he is so afraid of if inflation goes up...

"Evans, while explicitly agreeing the target was symmetric, made comments that indicated that it may not be. He said the Fed has not done a very good job of communicating its tolerance around the 2 percent target, both up and down, and they need to improve. But if the target is really symmetric, simply saying that (along with the tolerable range) is all that is required. Talking separately about tolerance for over and under-shooting isn't needed."

Evans' and Lockhart's statements stand on their own, but we've collected some information via the Atlanta Fed's Business Inflation Expectations survey that helps me think about the Thoma question. The chart below plots the answers, collected in the February and April surveys, to the following query: "Projecting ahead, to the best of your ability, please assign a percent likelihood to the following changes to unit costs per year over the next five to 10 years."

The question focuses on unit labor costs in order to elicit responses about what businesses may actually be planning for, as opposed to their guesses about a more abstract concept of overall inflation. The question focuses on expectations five to ten years into the future because the inflation goal of the Federal Open Market Committee (FOMC) is explicitly a long-run objective.

The obvious pattern in these survey responses is their asymmetry to the upside. The most probable outcome, according to the respondents, is that long-term costs will rise in a range that includes the FOMC's long-run inflation objective. But they also put an almost 50 percent probability on annual outcomes higher than 3 percent. Less than 20 percent probability is placed on costs rising at rates of less than 1 percent.

This picture is one of asymmetric risks to the inflation outlook, and as such it is an important element in thinking through policy choices. Symmetry in the sense of having an equal distaste for misses on either side of an objective does not necessarily imply symmetry with respect to the risks of meeting that objective.

To begin with, the Fed does have a dual mandate. Disinflation or, in the extreme, deflation has the potential to be problematic for growth and employment when interest rates are very low. The reason, if you buy the analysis, is that, with no room for rates to move lower, a decline in inflation raises the real cost of borrowing. A higher cost of borrowing restrains spending, creating an additional drag on economic activity in already tough circumstances.

The reverse argument has, of course, been made for temporarily tolerating inflation that is somewhat higher than the long-run objective. But even if you aren't quite sold on the wisdom of that approach—and I'll get to that in a bit—it is clear that, with policy rates near zero, misses to the downside on inflation bring risks to the Fed's growth mandate that are not implied by misses to the upside. Hence President Lockhart's comment regarding the importance of preventing deflation.

So why not respond to this asymmetric risk to growth by taking a chance on higher inflation? That question takes us back to the chart above. Taken at face value, the probabilities reported by our survey respondents suggest that the FOMC has been pretty successful in convincing folks that very low rates of inflation or deflation will not be allowed to set in. Perhaps this conviction is not surprising given the relatively aggressive responses of the committee to the disinflation scares of 2003 and 2010.

But the response to asymmetric risks to growth at low inflation rates may have had the effect of inducing asymmetric risks to the upside with respect to Fed's price stability mandate. Again taken at face value, the results of our business inflation expectations survey definitely imply a one-sided bet by businesses on how the FOMC might miss on its inflation objective. That could well explain why one would be so concerned if inflation rises. Just as there are asymmetric risks associated with below-objective inflation when it comes to the Fed's growth and employment mandate, there are asymmetric risks associated with above-objective inflation when it comes to the price stability mandate.

By Dave Altig, executive vice president and research director at the Atlanta Fed

Taking a deeper dive into the definition of inflation

Sometimes our familiarity with a topic gets in the way of our understanding of it. I often think that about inflation. It's widely known that inflation is a condition of rising prices, or what my favorite macroeconomics text defines more precisely in this way: "Inflation is an increase in the overall level of prices in the economy." And while this idea is serviceable for some purposes, it's pretty inadequate if you try to think about how a central bank goes about the business of controlling inflation.

Here's an example. Friday's retail price report for March revealed that prices rose 3.5 percent from February and averaged 3.7 percent (annualized) over the first three months of the year. I think it's pretty clear we have seen an "increase in the overall level of prices in the economy" this year. But how should the Federal Reserve respond to this rise?

I'm not going to offer an answer that question. I'm a policy adviser, not a policy maker. But we need to dig a little deeper into the textbook to get a better understanding of the inflationary process and how a central bank contributes to that process before we offer up an opinion on the matter.

The inadequacy of the textbook definition of inflation is a topic that's been bugging me for a long time. It's silent about the cause of the rise in prices and therefore does not make any distinction about whether the price increase is a corollary of the policies of the central bank or from another source that a central bank can only nominally influence. This distinction is an important one. (For what it's worth, this 1997 article I wrote explains how I think the term "inflation" has come to be synonymous with "price increase." And here's why I think a deeper appreciation of what constitutes "an increase in the overall level of prices" affects how one thinks about the inflationary experience in real time.)

Here's how economist Irving Fisher put it in his 1913 article "The Monetary Side of The Cost of Living Problem":

"If it can be shown, for instance, that today the good things of this world are becoming scarce on the one hand while money and its substitutes are not becoming plentiful it would be reasonable to conclude that the fault lies with goods and not with money. If, on the contrary, it can be shown that money and its substitutes are becoming plentiful and that goods are not becoming scarce, it is reasonable to conclude that the fault lies chiefly on the monetary side."

Are all price increases inflation? Should central banks use their tools to fight against any rise in costs? The answers to these questions have important implications for how we measure inflation in the short run versus the long run, the horizon over which a central bank ought to be held accountable for achieving price stability, and how monetary policies should be calibrated in the face of rising prices.

In that spirit, the Atlanta Fed has developed an animated video that provides additional perspective on the issue of inflation—specifically, what sorts of price increases are part of the inflationary process that is under the control of the central bank, and which are not. This video is the first in a new series that covers economic issues and the work of the Fed. The video is part of the Atlanta Fed's new feature called "The Fed Explained," which includes a range of new and existing content. We hope these videos stimulate conversation on a range of topics important to the Federal Reserve. The format is targeted to the general public as well as to students and teachers, and we hope it provides an engaging supplement to the study of the Federal Reserve.

Let us know.

Mike Bryan, vice president and senior economist at the Atlanta Fed

Why we debate

It's been a while since we featured one of my favorite charts—a "bubble graph" comparing average monthly job changes during this recovery with average changes during the previous recovery, sector by sector.

(enlarge)

If you try, it isn't too hard to see in this chart a picture of a labor market that is very close to "normalized," excepting a few sectors that are experiencing longer-term structural issues. First, most sectors—that is, most of the bubbles in the chart—lie above the horizontal zero axis, meaning that they are now in positive growth territory for this recovery. Second, most sector bubbles are aligning along the 45-degree line, meaning jobs in these areas are expanding (or in the case of the information sector, contracting) at about the same pace as they were before the "Great Recession." Third, the exceptions are exactly what we would expect—employment in the construction, financial activities, and government sectors continues to fall, and the manufacturing sector (a job-shedder for quite some time) is growing slightly.

For the skeptics, I below offer a familiar chart, which traces the level of total employment pre- and post-December 2009, compared with the average path of pre- and post-recession employment for the previous five downturns:

(enlarge)

We are now more than 16 quarters past the beginning of the recession that began in the fourth quarter of 2007, and total employment is still 4 percent lower than it was at the beginning of the downturn. In the previous five recessions by the time 16 quarters had passed, employment had increased by about 6 percent. Even in the worst case, indicated by the lower edge of the gray shaded area, employment growth was flat—and that observation is qualified by the fact that the recovery from the 1980 recession was interrupted by the 1981–82 recession.

This unhappy comparison is not driven by the construction, financial activities, and government sectors. In the area of professional and business services, which has logged the largest average monthly employment gains in the current recovery, the number of jobs still sits 2.7 percent below the level at the outset of the last recession, as the chart below shows.

(enlarge)

Total private-sector jobs in education and health services, which never actually contracted during the recession, nonetheless remain abnormally low in historical context.

(enlarge)

In these charts lies the crux of some very basic disagreements about the appropriate course of policy. The last three graphs draw a clear picture of labor markets that are underperforming by historical standards—a position that I take to be the conventional wisdom. An argument against following that conventional wisdom centers on the question of whether historical standards represent the appropriate yardstick today. In other words, is the correct reference point the level of employment or the pace of improvement in the labor market from a permanently lower level? For the proponents of the latter view, the bubble chart might very well look like a return to normal, despite the fact that employment has not returned to prerecession levels.

One way to adjudicate the debate, in theory, is to rely on the trajectory of inflation. If there remains a significant amount of slack in labor markets, as the conventional interpretation of things suggest, there ought to be consistent downward pressure on prices. The case for consistent downward pressure on prices is not so obvious. Measured inflation appears to moving in the direction of the Federal Open Market Committee's 2 percent long-term objective.

(enlarge)

Also, the Atlanta Fed's own monthly survey of business inflation expectations, which surveys a panel of businesses from our Reserve Bank district, indicates that this inflation number (shown in our March release from earlier this week) is in line with what private-sector decision makers anticipate:

"Survey respondents indicated that, on average, they expect unit costs to rise 2.0 percent over the next 12 months. That number is up from 1.9 percent in February and comparable to recent year-ahead inflation forecasts of private economists. Firms also reported that their unit costs had risen 1.8 percent compared to this time last year, which is unchanged from their assessment in February. Inflation uncertainty, as measured by the average respondent's variance, declined from 2.8 percent in February to 2.4 percent in March, the lowest variance since the survey was launched in October 2011."

Does that settle it? Not quite. There may not be much evidence of building disinflationary pressure, but neither is there building evidence of an inflationary push that you would expect to see if the economy were bumping up against capacity constraints. Obviously, the story isn't over yet.

By Dave Altig, senior vice president and research director at the Atlanta Fed