August 19, 2015

Unconventional Lessons for Unwinding Unconventional Policy, I: The Volcker Disinflation By Jon Faust

The Fed is (data dependently) on the cusp of attempting to engineer something unprecedented—a relatively benign liftoff of short-term interest rates after an extended period with those rates near their lower bound. Many analysts have been carefully analyzing past tightening episodes hoping to understand the likely Fed behavior, and I am regularly asked what I take as the lessons from, e.g., 2003-2005 and 1994. Here is what history teaches: the FOMC will do what it usually does after 6 years near the lower bound for nominal rates and when its balance sheet is over $4 trillion.

The current situation is far from the range of historical experienced, and there will be no simple lessons from history. But there may be more subtle lessons in both the successes and failures of the past. This is the first in a series—not sure how long a series—of posts about unconventional history lessons for the unwinding of unconventional policy.

The most recent example of the Fed engineering anything resembling an unprecedented transition is the Volcker disinflation.[1] After coming into office in 1979 Chair Volcker almost immediately set about a severe tightening of monetary policy that saw inflation fall from over 10 percent to about 5 percent by August of 1982. At that point, the Latin American debt crisis threatened the survival of many large U.S. banks and soon became the focus of policy.

Consider the nitty gritty of how the Fed managed policy. The Fed adopted new operating procedures for the transition, a nonborrowed reserves targeting procedure directed at hitting monetary growth goals stated each year in the form of cones for various measures of money. During the period, each Friday’s announcement of the money supply numbers took on the market-moving importance that is now reserved for nonfarm payroll announcements.

How did this new procedure work out? Inflation came down and has stayed down. Let’s call that success. Volcker is widely, and rightfully in my view, heralded as a hero for his leadership in achieving this objective.

The policy framework, however, was a mess. Because the behavior of the money supply bore no close relation to Volcker and the Fed’s goals, the money supply was redefined several times and the cones were moved mid-year. If we view the monetary cones as the goalposts of policy, the Fed both re-designed the football and moved the goal posts and still didn’t get the ball through the uprights.

The figure below from an article by noted historian Bob Hetzel[2] illustrates this point, and Hetzel’s extended note to the table makes clear several additional machinations:

Hetzel’s note to Table: In order to display the data available contemporaneously, M1 is taken from the first Board of Governors statistical release H.6 showing complete monthly figures for a given year. In 1980, M1-B is used. In 1981, shift-adjusted M1-B is used. This series adjusts other checkable deposits for shifts from nondemand deposit sources. The discontinuity after 1981 arises from the discontinuance of the shift adjustment. After October 1982, the target range for M1 was replaced by a “monitoring” range. The dual ranges for M1 in 1983 reflect the rebasing of the M1 monitoring range in July 1983.

At the time, the Fed was excoriated both by economists[3] and Congress. I was a research assistant[4] at the Kansas City Fed during this period and was tasked with reading all of Volcker’s testimony on monetary policy; being young and naive, I was shocked at the level of discourse. People complain about the tone in Congress today, but the tone of the monetary policy debate was at least as inflamed in the early 1980s.

But few today remember that the Fed’s operating procedure and communications framework were, by any reasonable assessment, a complete hash. Volcker brought down inflation, and missing the cone for M1-b adjusted is relegated to the footnotes of history.

What was the key to Volcker’s success? The key is probably that everyone inside and outside the Fed was pretty clear on the objective the Fed was pursuing: bring down inflation. Everyone could also see that the actual policies followed—regardless of the framework under which they were implemented—were broadly consistent with that prime objective. The rest was sideshow.

For liftoff, I think the lesson is clear. The prime objective will be to make short-term rates go up. Most informed economists agree that the Fed has the tools to do this. That is, some combination of the array of tools available to the Fed can get the job done. And the details? Will the federal funds rate stay near the middle of the announced target range? Will the ranges have to be widened to accomplish this goal? Will the parameters of the overnight reverse repurchase agreement facility have to be adjusted? I don’t know the answers to any of these questions, but if the Volcker case is any indicator, the operational details of implementing this unique transition may at times look quite messy. In the famous modal scenario, however, rates will go up, and the rest will largely be forgotten.

I will admit that I have floated this argument in many circles, and one common refrain is that today is different due to social media, the round-the-clock news coverage, and the hyper-critical Congress. Perhaps these are first-order considerations. Perhaps because of twitter, the Fed must not only achieve the objective, but also achieve it gracefully. If so, we may be doomed.

But I think not. To fill out the modal scenario, the FOMC will emphasize that raising the level of short-term rates is the prime objective and that it has the tools to achieve this. Various adjustments to the particular manipulation of tools will be made, presumably amid a firestorm of blog posts and nasty congressional hearings. And rates will rise, and we will all turn to the next subject.

A note of caution: there are many ways that events in the world could derail this modal scenario, just as the debt crisis replaced disinflation on center stage for Volcker. But that is another subject.


1.Disinflation obviously had many precedents, but this event is probably as close as we can come to unprecedented. In many ways, what is now called the Great Inflation was a unique worldwide phenomenon, and there were certainly unique features to ending it. [back]

2.Robert Hetzel, Monetary policy in the early 1980s. Economic Review, Federal Reserve Bank, March/April 1986. [back]

3.For example, in “A Map for the Road from Dunkirk,” [New York Times, March 21, 1982, as quoted by Hetzel] Nordhaus argued,

The first, step [of a new economic policy] would be to bring down the curtain on the disastrous monetarist experiment of the last two years. The Federal Reserve should be directed to cease and desist its mechanical monetary targeting and to set monetary policy with an eye to inflation and unemployment… At the same time, the Fed should overhaul its operating procedures. The techniques of emphasizing supply of bank reserves rather than interest rates since October 1979 has produced greater volatility of both interest rates and the money supply.

Note to the New York Times headline writers: if I have my history right, a map to the road from Dunkirk was not the issue; perhaps “A chart for sailing from Dunkirk.”[back]

4.With Bryon Higgins (a responsible adult), I threw my naive 2 cents worth in on the topic of redefining money: Bryon Higgins and Jon Faust, NOW’s and Super NOW’s: Implications For Defining and Measuring Money, Economic Review, Federal Reserve Bank of Kansas City, January 1983.[back]

Read more/comment
March 19, 2015

Dots … By Jon Faust

Just before the release of the FOMC’s Survey of Economic Projections (SEP) in December, I posted a piece saying that the dots would reveal nothing and arguing that, by design, the dot plot is not likely to help us understand policy. My main critique is that the dots convey the range of opinions, but shed no light on how the differences will be resolved. Shortly after 2 pm yesterday, Ross Margolies, a steadfast (and succinct) supporter of the CFE, emailed: I thought the dots meant nothing.

I suspect that reporting a 3-month policy projection each September would lead to an annual feeding frenzy of short sightedness.

Good point. My earlier comments were wrong in neglecting an important case: the dots are informative when they are tightly clustered. For example, over the 3-year history of the dots, essentially everyone on the FOMC expected rates to stay at zero at least through the first year reported in the SEP, and the SEP very effectively conveyed that unity.


Read more/comment
March 16, 2015

Happy Anniversary Chair Yellen! By Jon Faust

Post-FOMC Press Conference, March 2014

ANNE SAPHIR: Then once you do wind down the bond buying program, could you tell us how long of a gap we might expect before the rate hikes do begin?

CHAIR YELLEN:… [S]omething on the order of around six months or that type of thing. But, you know, it depends. What the statement is saying is, it depends what conditions are like. We need to see where the labor market is, how close are we to our full employment goal-that will be a complicated assessment not just based on a single statistic… Inflation matters here, too, and our general principle tries to capture that notion. If we have a substantial shortfall in inflation, if inflation is persistently running below our 2 percent objective, that is a very good reason to hold the funds rate at its present range for longer.

The first anniversary of Chair Yellen’s initial press conference is at hand and it’s a good time to review the past year. In doing so, I’ll mainly follow the norm in the press of focusing on developments regarding the date interest rates will lift off; but I’ll turn to more important issues toward the end.

The basic message from Chair Yellen last March was that we were steadily approaching the time when an initial increase in the federal funds rate would be appropriate, that the precise timing of lift off was likely to move forward or back in time depending on incoming information regarding employment and inflation, and that (more…)

Read more/comment
February 27, 2015

Two big questions By CFEGuru

By Bob Barbera and Jon Faust

Simple plots of recent GDP and inflation data highlight two profoundly important questions facing monetary policymakers in the United States. GDP is at a level several percentage points below reasonable estimates of its pre-crisis trajectory (Fig. 1):[1] Will we ever regain that lost output? Inflation had run well below the Fed’s objective for the two years prior to the recent plunge (Fig. 2):[2] Will inflation bounce back, all the way back, to the Fed’s target? (more…)

Read more/comment
February 16, 2015

Jobs, inflation, and growth in 2015 By Bob Barbera

Recent readings for the U.S. economy are filled with contradictions. Non-farm payroll gains were quite strong. Unemployment fell to 5.7% from 6.1%, last September. Nonetheless, real GDP growth was soft, up only 2.6%. And retail sales were quite weak. Reconciling robust gains for employee hours and a low jobless rate with tepid increases for real GDP and soft nominal spending gains is very difficult.

The most recent Survey of Professional Forecasters suggests that forecasters generally reconciled these contradictions by doing the following:

Assume a nearly instantaneous sharp slowdown for job growth.
Project a modest additional decline for unemployment.
Assume a continuation of unprecedented weakness for labor productivity growth.
Project a modest pace of advance for real GDP.
Assume almost no change for core inflation.

Specifically, for 2015 the recent survey has:

Non-farm payroll gains average 238,000/month. This compares to the 336,000 average monthly gains of the past 3 months, the 282,000 average monthly gains of the past 6 months, and the 267,000 average monthly gains of the past 12 months.

Unemployment falls to 5.2% in Q4:2015 By sharply curtailing the pace of job growth, over the four quarters of 2015, forecasters avoided writing down projections that took unemployment below 5%. They also avoided projecting a rebound for the participation rate.

Labor productivity rises by a tepid 0.8% Notwithstanding the slower pace for job growth, that growth would still suggest a gain for employee hours of roughly 2%. A traditional projection for labor productivity, say a gain of 1.8%, would require forecasters to embrace hefty gains for real GDP. Instead, labor productivity gains are consigned to rise by less than 1%.

Real GDP climbs by 2.8% Embracing evidence of soft nominal consumer spending trajectories, and acknowledging the likely drag on output that trade will deliver, real GDP growth is projected to approximate the pace registered in Q4:2014, a pace well short of the gains witnessed in Q2:2014 and Q3:2014.

Core inflation is projected to remain steady, rising 1.7% in 2015. All forecasters know that gasoline prices have plunged and that headline inflation will fall for a few months, simply a consequence of petroleum product price declines. Beyond that, however, the consensus forecast looks for little change, anticipating a 1.3% rise in core inflation in Q1:2015, followed by a 1.8% rate of inflation for the remainder of the year.

What is the most striking aspect of the consensus view? Almost to a person, forecasters felt compelled to sharply rein in jobs growth. The current recovery reminds us that one should never doubt dire predictions, but where is the evidence for a sharp deceleration? Should this really be the modal forecast? Plunging petroleum prices, history makes fairly clear, act as a stimulus in the U.S. It is true that the rise for the dollar and the fade for economic growth in Europe, Japan and China all combine to suggest that trade will be a drag on real growth in 2015. That said, netting out the effects of rest-of-world weakness against a halving of the U.S. consumer energy bill, probably suggests a small positive change in U.S. economic momentum.

I suspect that the projected, sharp job market retrenchment is not due to signs of weakness. Instead, I suspect it is because of the forecasting havoc that continued strong job growth would bring to the rest of the forecast.

Some simple arithmetic

Suppose job growth does not slow, and instead climbs at the average pace in place over the past 3 months? Suppose further that labor force participation is stable, as it was over the four quarters of 2014. The jobless rate, in such circumstances falls to 4.1%. In addition, total hours worked will have climbed at a 2.7% pace. With all that, one would be forced to forecast a rise for real GDP of nearly 4%, even if one embraces the notion of a tepid, less than 1% climb for labor productivity.

How can we invent a climb for real GDP of nearly 4%, amid tame nominal spending trajectories and stable core inflation? That’s simple. We can’t.

Suppose instead that we temper our enthusiasm for job strength and we embrace the trajectory in place over the past 6 months. And, once again, let’s assume an extension of a stable participation rate. In this case, Unemployment, Q4:2015, averages 4.5%. And we still need to contrive a story that allows for real GDP growth of around 3.5%, amid soft spending trajectories and core inflation a bit below 2%. Again, a very tough story to cobble together.

This daunting accounting, I suspect, helps explain the willingness to envision a violent job growth downshift. Accepting recent jobs trajectories boxes you into other storylines that you are not ready to tell. Jettison the jobs strength and it is much easier to roll out consensus views for much of the rest of your forecast.

Thinking inside the box

Is there any way to keep the jobs strength and tell a plausible economy wide story while remaining boxed in by the confusing configuration of recent data? One way to do so requires us to make two leaps. Bet on a major fall for core inflation. And assert that labor force participation, at long, long last, is now on the rise. Second things first. Over the 7 years through Q4:2013, the participation rate fell at a 0.8% per year pace. What happens if it rises by 0.8% in 2015? Put the rate at 63.3% and the jobless rate averages 5% in Q4:2015, even with monthly job gains running at 330,000 throughout the year. Throttle back the job growth assumption to the six month trajectory and what happens? A year of 280,000 per month job gains ends with the jobless rate at 5.3%. What about output growth? January’s import price data may give us the key. Core prices fell sharply, by 0.6%, month-on-month. They were down 1.1% year-on-year. Moreover the data show a clear acceleration on the downside. This is all too easy to explain. There is always an echo effect in other prices, when energy prices plunge. We also have a 15% rise for the U.S. dollar versus a very wide basket of U.S. trading partners’ currencies. Lastly all indications suggest that inflation in China is in sharp retreat. This suggests prices of goods sent from China could fade this year. Suppose the consensus assumption about core inflation is wrong? If core inflation falls to 0.7%, we have a percentage point of additional real growth, for any given level of nominal spending growth. We can, therefore, assert the following. Real GDP growth of 3.5% to 4% is achievable without a material acceleration for spending, if core inflation fades in 2015. Furthermore, in this scenario, the jobless rate can end the year a bit above 5%, notwithstanding strong gains for monthly payrolls. We simply need to embrace the notion of a bit of bounce back in the participation rate, after 7 very lean years. What about tepid productivity? We get labor productivity up to 1.5% in a world of 280,000 per month job gains and 4% real GDP. Not an impressive number. But one that does not conjure up bleak stories of continued malaise in the aftermath of the Great Recession.

The CFE forecast

In broad terms, the story just given explains what our forecasting team here at the CFE submitted to the Survey of Professional Forecasters. Let’s be clear, like the other forecasters, we are sometimes right in key features of our forecast, always wrong somewhere, and occasionally wrong in general. One of the most important reasons for making a forecast, however, is that it forces one to devise a story that most plausibly reconciles all the available data. With the current confusing constellation of data, this is really making a story that is least implausible. Most forecasters seem to have concentrated the implausible part mainly on job growth. We have spread it around a bit more in a way that we think is worth considering.

Read more/comment
December 26, 2014

Capital controls for Russia? By CFEGuru

“The main lesson from international experience is that controls on capital outflows can work—but only if they are associated with a credible policy plan addressing the underlying cause of the confidence crisis.”

Olivier Jeanne, of the Center and Peterson Institute, has an interesting op-ed in the Dec. 23 Financial Times arguing that capital controls may be useful policy tool for Russia at present, so long as they are used only to provide breathing space to begin putting in place more fundamental policy changes.

Read more/comment
December 18, 2014

Patience and prices By Bob Barbera

Consensus expectations were off regarding the November estimate for U.S. CPI inflation released yesterday. The 0.3% fall for headline inflation was a larger drop than estimated by 82 of the 84 economists who ventured forth with an opinion in Bloomberg’s survey. No-one offered up a forecast of a greater fall than 0.3%. We suspect that consensus expectations for inflation in early 2015 are similarly off the mark. Back-of-the-envelope calculations suggest that for a relatively wide range of oil price scenarios, over the next several months, the U.S. headline inflation rate will plunge.

If we embed today’s energy futures prices into consumer product prices, the drop in first quarter 2015 CPI inflation could be breathtaking.

A plunge for headline inflation is already baked in the cake for December. The E.I.A. surveys of gasoline prices at the pump December-to-date are down around 70 cents, year-on-year. That is almost double the 32 cent year-on-year fall registered in November. Based on the energy component alone, the CPI headline looks set up to fall around 0.7% for the month, which would take year-on-year headline inflation down to 0.3%. And it appears there is more to come. If we embed today’s energy futures prices into consumer product prices, the drop in first quarter 2015 CPI inflation could be breathtaking. For example, near-contract futures prices for gasoline are trading down $1.15, year-on-year. (more…)

Read more/comment
December 16, 2014

Of dots and (considerable) periods By Jon Faust

In a recent post, I argued that the Bernanke and Yellen Fed’s have been striving for a ‘no tea leaves’ approach to policy communications. An astute JHU student responded, ‘How about those dots?’ Good students can be annoying that way.

There is a sound reason for publishing the dot plot, but we should not expect these ‘if you were Czar’ policy paths to be of much value in clarifying policy intentions of the FOMC.

At the outset, let me remove some suspense by sharing what Wednesday’s new dot plot will clarify about the likely path of policy: nothing.

What is the dot plot? The FOMC publishes a Survey of Economic Projections (SEP) quarterly, giving the 19 FOMC participants’ projections for real activity and inflation.[1] Along with these projections, each participant submits his/her view of where the federal funds rate will be at the end of the current year and end of the next couple years. These year-end rates for the participants are then turned into the dot plot—the October plot is below. (more…)

Read more/comment
December 6, 2014

What do markets expect the Fed to do? By Jon Faust

Matt Raskin, along with several co-authors (Richard Crump, Emanuel Moench, William O’Boyle, Carlo Rosa, and Lisa Stowe), has published a series on measuring policy interest rate expecations on the NY Fed’s Liberty Street Economic Blog.

These blog posts provide an excellent discussion of how to interpret market and survey-based indicators of interest rate expectations. For example, the authors note, “Market prices provide timely information on policy expectations. But as we emphasized in our previous post, they can deviate from investors’ expectations of the most likely path because they embed risk premiums…”

By the way, Matt Raskin recently returned to Hopkins to complete his PhD after leaving a few years ago to work at the New York Fed. He is now an Assistant Vice President in the Markets Group, and I can attest that Matt’s analysis played an important role in policy discussions on a number of occasions over my recent time at the Fed.

Good work by Matt and his co-authors.

Read more/comment
November 21, 2014

Today’s CPI release: If you just squint, you’ll see … By Jon Faust

By Bob Barbera and Jon Faust

The FOMC minutes released yesterday and today’s CPI data release underscore a remarkable shift. Over the past five years, the state of the labor market has dominated monetary policy discussions, but for the first time since the crisis inflation is now taking center stage. All through the recovery, of course, inflation hawks have warned that inflation would soon demand our attention. But probably neither hawks nor doves predicted that excessively low inflation would be our concern as unemployment moved closer to normal. And yet today’s CPI data and the FOMC commentary reported yesterday remind us that, in the short run, inflation is probably headed lower.

Today and in coming months, analysts at the Fed and elsewhere will be parsing the data and squinting extra hard to see signs that inflation will, without additional policy measures, move back up to desired levels. The difficulty, of course, is that factors such as the falling price of oil and of other commodities and the rising value of the dollar are putting downward pressure on inflation.1 A centerpiece of inflation analysis in situations like this is examining the behavior of sub-components of inflation that are less likely to be affected by transitory forces. These indices, one hopes, will give a clearer sense of where a shadowy beast known as underlying inflation may be headed. The baseline view of many analysts is that underlying inflation is headed slowly back to the Fed’s objective. This may be the right baseline, but recent data have, in our view, significantly eroded the confidence we can have in this baseline view.


Read more/comment