Earlier this month I was lucky enough to attend the Federal Reserve's much-anticipated conference on "Monetary Policy Strategy, Tools, and Communication Practices." This was the "research" component of a series of Fed Listens events kicked off this February and wrapped-up last week. The aim of the series was to allow Fed officials "to hear perspectives from representatives of business and industry, labor leaders, community and economic development officials, academics, nonprofit organization executives, and others" concerning how the Fed might more perfectly fulfill its mandate.
To call both the Chicago conference and the Fed Listens series unusual would be an understatement: traditionally, the Board of Governors tends to be a somewhat insular institution, if not one plagued by groupthink. So far as I'm aware it has never put so much effort into hearing what Fed outsiders — and particularly non-academic outsiders — have to say about how it should do its job. Even if it makes for little if any discernible change, this display of increased Fed openness to outsider opinions is at least a step in a good direction.
Insiders and Outsiders
As befit its purpose, the Chicago program consisted mainly of technical research presentations and commentaries by academics, with Fed officials presiding.* The audience of just over 160 persons was, in contrast, divided roughly equally between current and former Fed staff and officials and others — including at least a dozen journalists as well as representatives of several Washington-area think tanks. Cato had the good fortune to be represented twice over, by myself and also by CMFA Managing Director and V.P. in charge of Working Miracles Lydia Mashburn.
Even so, it wasn't easy for us, or any of the outsiders in the audience, to take part in the formal proceedings. The Q&A sessions tended to be short. And of four seating sections available — two surrounding a central aisle and two others adjacent to the left and right walls — the central ones, most visible from the podium, were reserved for Fed staff. Yours truly, always a fast study when it comes to Q&A, never managed to stick his oar in that day. Fortunately there were plenty of opportunities for discussion with Fed officials between the sessions, so anyone could get his or her two-cents in one way or another.
Off the Table
If anything disappointed me, it was not so much the discussions as the program itself. The problem wasn't a lack of interesting presentations and papers: there were plenty of both, covering topics ranging from the meaning of maximum employment to the bearing of global considerations and financial instability concerns on monetary policy. The problem was that some important alternative monetary policy strategies and tools that might have been given serious attention were either
ignored excluded from the program altogether or given very short-shrift.
To be sure, some potential alternatives were never meant to be up for discussion. As Rich Clarida explained at a pre-Chicago Fed Listens event,
The Congress charged the Federal Reserve with achieving a dual mandate –maximum employment and price stability — and this review will take this mandate as given. We will also take as given that a 2 percent rate of inflation in the price index for personal consumption expenditures is the operational goal most consistent with our price-stability mandate. While we believe that our existing strategy, tools, and communications practices have generally served the public well, we are eager to evaluate ways they might be improved. That said, based on the experience of other central banks that have undertaken similar reviews, our review is more likely to produce evolution, not a revolution, in the way we conduct monetary policy.
While one can understand the Fed's desire to avoid a revolution, it's not easy to see why there should have been no discussion at all of alternatives to its present "floor type" operating system: certainly no one believes that the new system is the only one compatible with the Fed's dual mandate, let alone with it's 2-percent inflation target. (Some, indeed, have wondered whether it may even be incompatible with that target!) Yet the program offered no opportunity at all for consideration of modifications to, let alone replacements for, this most important of all Fed tools — or for discussion of the related question of the proper long-run size of the Fed's balance sheet.
Admittedly there had been hints from the start that the Fed's operating system would be as off-limits at Chicago as its mandate and 2 percent inflation target. When the Fed announced its planned conference last November, Bloomberg's Rich Miller reported that "The review is not expected to include a look at how the Fed mechanically controls short-term interest rates, which is being discussed internally." Still I held out the hope that Miller's claim, which was reported nowhere else, was wrong: after all, I thought, if a review of the Fed's "tools" doesn't include a discussion of how it controls short-term rates, that doesn't leave very much.**
In the event what it left was an evaluation of alternative "unconventional" monetary policy devices, including quantitative easing and negative interest rates, led off by a nice survey of those by Eric Sims and Jing Cynthia Wu. The Fed's methods of conducting "conventional" monetary policy, however unconventional they may be, were kept entirely off limits. Indeed, much to my surprise and disappointment, there wasn't even any formal discussion of possible improvements to the Fed's established operating system, including Jane Ihrig's and David Andolfatto's Standing Repo Facility plan.
Selling NGDP Short
If I was surprised to see proposals like Jane and David's given no attention at Chicago, I was stunned by Lars Svensson's offhand dismissal of nominal GDP targeting in the conference's session devoted to Monetary Policy Strategies. NGDP targeting is, after all, among the most prominent and widely-discussed non-revolutionary alternative monetary policy strategies out there just now, with numerous serious proponents both inside and outside of the Federal Reserve System. And the subject of Svensson's session was alternatives to having the Fed aim at all times at 2-percent inflation. NGDP targeting certainly fits that bill.
In his presentation Svennson nevertheless waved NGDP targeting aside, declaring it (but not price-level targeting, temporary price-level targeting, or average inflation targeting) inconsistent with the Fed's mandate. In his paper Svensson does discuss NGDP targeting, but only in an appendix, to which he consigns it after declaring it to have "some obvious disadvantages and no advantages compared with flexible inflation targeting or price-level targeting."
Now, NGDP targeting certainly has its drawbacks, including the low frequency of NGDP data and the fact that it tends to be revised after first being reported. But if inconsistency with the Fed's mandate is one of them, then I dare say that all of the strategies Svensson considered are just as inconsistent with that mandate. So, for that matter, is the Fed's current practice. After all (as Svensson himself acknowledges in his paper), that practice has resulted in a price level which, far from being stable, resembles a random walk with drift. If the Fed has been able to convince Congress that that's "promoting…stable prices," why should it have any trouble justifying an NGDP-targeting strategy that's equally if not more consistent with modest long-run inflation?
A "Mini Revolt"
Nor was I the only one disappointed by Svensson's treatment of NGDP targeting. In dismissing that option Svensson offered to answer questions about it during Q&A. Well, he not only got questions about NGDP targeting — he found himself fending-off what Bloomberg called a "Mini Revolt." The troublemakers, far from being among the gathering's true outsiders, were all insiders of sorts: Notre Dame's Eric Sims, an author of one of the program's papers; Evan Koenig, the Dallas Fed's principal monetary policy advisor, and Jim Bullard, President of the St. Louis Fed and a sitting member of the FOMC. Svensson's commentator, Bank of Canada Advisor Sharon Kozicki, also chimed in with a kind word about nominal GDP targeting. Bullard spoke for all of them and, I dare say, for more than a few other audience members, in calling Svensson's spurning of NGDP targeting "off base."
It was in order to take part in this revolt that I tried hardest to get my own flailing arm recognized. But as luck would have it session moderator (and Peterson Institute President) Adam Posen, having considerately chosen to look for questions from the wings rather than the center rows, determined, in a rare display of right-wing bias, to start with that side of the hall!
I had hoped to ask Professor Svensson whether, in assessing the merits of NGDP targeting, it made sense to treat temporary deviations of actual from target inflation as a "loss," that is, as something to be regretted. You see, in his paper appraising the various alternatives to strict inflation targeting, Svensson assumes that the Fed's statutory mandate
must be interpreted as full employment and stable prices being two separate goals, so there are two target variables, namely employment and prices (or inflation). In contrast, targeting nominal GDP implies that there is only one target variable, namely nominal GDP.
To which I say, "phooey"! As I've already pointed out, once one grants, as Svensson does, that the only monetary policies worth considering all allow for occasional deviations of the inflation rate (not to mention the price level) from its long run target, there's no reason to insist on treating such deviations as equally and necessarily unfortunate. Instead, one can and should distinguish those deviations that actually go along with keeping real output and employment close to their optimal levels from deviations that don't. If, for example, a policy achieves optimal employment by allowing for temporarily above-target inflation, that above-target inflation shouldn't be reckoned a loss at all. Put another way, and notwithstanding certain central bankers' preference for what some call a "balanced" approach, there's no reason for supposing that the Fed's mandate calls for assigning equal weights to the inflation and output (or employment) components of Svensson's central bank "loss function." Instead, in comparing alternative policies, all of which result in the same (say, 2 percent) long-run inflation rate or price-level trend, preference should be given to the one that minimizes the "real" short-run losses. Concern about those losses is, after all, why we don't insist on strict price-level or inflation-rate targeting.
Price Stability is a Means, not an End
More fundamentally, as Svensson recognizes, and as good central bankers readily acknowledge, price stability, however interpreted, is desirable not as an end in itself, but a means for avoiding unwanted changes in real economic variables — the only variables on which peoples' well-being depends — including output and employment. The whole point of nominal GDP targeting is to allow such occasional price level and inflation-rate movements as are needed to achieve precisely that outcome. Nominal GDP targeting is thus able to avoid, without special pleading, such absurdities as monetary tightening in response to adverse supply shocks.
Admittedly Svensson sees things differently. For him it's nominal GDP targeting itself that can lead to perverse outcomes. "In particular," he writes, "for a given level of nominal GDP, less than full employment (with resulting lower GDP) would be fine as long as prices would be appropriately higher." But unlike my adverse supply-shock scenario Svensson's begs the question — two questions, actually. First, if the Fed has kept nominal demand stable, allowing prices to rise "appropriately" in light of that objective, what, if not an adverse supply shock, is causing the drop in real GDP? Second, how would the other alternatives favored by Svensson help? Consider ordinary (symmetric) price level targeting. As Svensson himself observes, a central bank committed to that strategy
cannot “look through” supply shocks that temporarily drive up inflation, but must commit to tightening policy in order to reverse the effects of the shock on the price level. This reversal could be gradual and responsive to real-side conditions, as indeed the theory of flexible price-level targeting suggests. Nevertheless it implies a possibly painful tightening even as the supply shock depresses employment and output.
That hardly sounds like an improvement to me; and I'm inclined to think that any loss function that says otherwise ought to be taken with a large pinch of salt.
An Ongoing Conversation?
Whatever the merits of Svensson's paper, and of the rest of the papers and comments presented at Chicago, those merits shouldn't be confused with those of the conference itself. The point of the conference was not to settle anything, but to convince us outsiders that "the Fed Listens" to them. And listen it did, for several months. The real test will be whether it keeps on listening.
*The exceptions were two panels featuring community leaders, one on "What Does Full Employment Look Like for Your Community or Constituency?" and the other on "Transmission of Monetary Policy to the Economy: Beyond the Headlines."
*[Added 6/26/19 at 2:22P] After I published this Ellen Meade, Senior Adviser at the Federal Reserve Board's Division of Monetary Affairs and Vice Chair Clarida's Special Adviser, wrote to me, explaining that the floor vs. corridor and balance sheet issues were not included in the Chicago program because the FOMC had already reached a decision regarding these matters this January (https://www.federalreserve.gov/newsevents/pressreleases/monetary20190130c.htm). Ms. Meade is of course correct; and I might well have understood that announcement to have implied that public input would not be sought concerning the already-resolved questions. I have only myself to blame for having hoped the topics might come up anyway!