Signs for September
- At Jackson Hole, the deluge of Fedspeak will continue to emphasize a data dependent approach where officials remain unsure about the extent and timing of the further easing cycle but, for the moment, lean towards a September cut.
- Look for any changes in Powell’s language on the labor market (“solid”) and the stance of policy (“well positioned”) relative to the July meeting for a more definitive lean.
- Core CPE forecasts are tracking at or slightly above expectations from June, while the unemployment rate is still a bit below forecasts but has left-tail concerns that are hard to shake.
- The Fed’s communication of its reaction function has been similar to optimal control for much of the past year and that will remain their default approach to communicating around both their various baselines and differing risk assessments. The baseline is that sluggish growth is necessary to keep inflation returning to target but that is a tricky path to tread.
The Long-run Strategy
- The Fed’s statement on longer-run run goals and strategy is likely to be amended in a modestly hawkish way, removing some of the asymmetries which were put in place (the flexible average inflation target and, less guaranteed, the emphasis use employment short-falls vs deviations) to respond to the less relevant asymmetry of the zero lower bound and the different inflation regime we are now in versus that prevailing from 2009-20.
- Changes to the SEP + Dot Plot and other communications could but any eventual tweaks are likely to be modest in scope and impact.
- It would be good for the Fed to take a more robust approach to supply shocks given the experiences of the past 5y but the framework review is likely to make only more modest shifts. Some new language here is possible. Approaches such as a wider or higher inflation target band or nominal GDP or income targeting, would add robustness but seem beyond the scope of this review.
On neither side of the mandate is there anything close to universal agreement but the slowdown in labor demand growth, despite weak supply-side trends, seems to be winning on the margin over the inflation mandate for much of the FOMC in the very near term. This makes a September cut the most likely outcome but, with the data currently in hand, it seems a bit less likely than current market pricing. However, market pricing seems to reflect a small chance of a 50bps cut on the basis of more appreciable labor market weakness in the August data and thus a bit lower all-in odds of a cut period than shown on WIRP, or similar, due to the left tailed skew.
Powell’s speech Friday morning is likely to center on introducing the new consensus statement as well as include a discussion and analysis of the past five years’ experiences, lessons, and mistakes. This is Powell’s valedictory Jackson Hole conference as well as the framework review so we should expect a bit more of a retrospective speech than normal.
During Powell’s remarks I will be looking for changes in the baseline characterization of the labor market (“solid”), the stance of policy (“well positioned”) as well as a more in depth or concerning tone around the likely negative benchmark revisions to the non-farm payrolls data (roughly -600k expected as a preliminary announcement in early September) which would almost surely be suggestive of greater odds of a September cut. Hawkish pushback in the form of increasing inflation concern, particularly regarding services inflation (also going to be revised a bit higher in late September) and inflation expectations risks.
As a bit of an aside, it seems worth noting that 20% odds of a mild recession roughly equilibrate with a 2.5% core PCE inflation forecast over the medium-term in simple optimal control-like policy frameworks. Given the Fed’s greater descriptive use of this type of language in recent months[1] this seems worth highlighting as inflation begins its tariff-driven move higher from an already too high base. I ultimately doubt the Fed would allow quite so much of a persistent inflation overshoot given that many hawkish members of the Committee would worry about inflation, but this highlights just how modest recession odds can substantially shift acceptable levels of trend inflation as a justification for a roughly neutral policy stance.
Taken together, this seems likely to setup a September meeting where the concerns and baselines of the June meeting are largely repeated. It remains challenging for the Fed to be comfortable with any further easing in the labor market, despite their baselines’ seeing the unemployment rate move up to 4.5% in Q4, given the recessionary concerns which any labor market easing, especially on above neutral levels of slack, raise. They do want to prevent reheating but that is a fine line to tread, especially with aggressive rhetoric from the Trump administration (even if they are causing the downside shocks). Barring an extremely hot August print, certainly possible, the 3.1% median and 3.4-3.5% upper-end 2025 core PCE forecasts are likely to remain observationally equivalent at this point. The possible reacceleration in some services inflation measures (more from Gerard here) is more troubling in this regard than tariff-driven inflation the Fed wants to look through unless inflation expectations start to deanchor. That means much of the SEP baseline will roll forward with a further easing presumption but not guarantee. If the Fed does end up cutting in September, it certainly seems possible that the dovish dissents in July are followed by a hawkish one or two in September.
The 2020 Framework and What Went Wrong?
The Fed’s current statement on longer-run goals and monetary policy strategy very much reads like the culmination of the post-GFC cycle document that it is (see here). The Fed characterized the changes made in 2020 to the statement, announced at the virtual Jackson Hole conference that year, as being driven the lessons of the past decade (here). These included:
- The adoption of the “broad-based and inclusive” maximum employment goal. The labor market reaction function was switched to the asymmetric use of “assessments of the shortfalls of employment from its maximum level” rather than the symmetric “deviations” in prior versions. The first of these has frequently been a target of criticism from outside the Fed system and did lead to some shifts in the Fed’s rhetoric but the topline unemployment rate has always been the key indicator, which with most others of more narrow categories tend to comove. The second change was far more dovishly impactful.
- The second change was moving to the flexible average inflation target, where “following periods when inflation has been running persistently below 2 percent, appropriate monetary policy will likely aim to achieve inflation moderately above 2 percent for some time.” This, more than the above I think, formalized the lagging policy response that made for a more dramatic catch-up period in 2022-23.
- Third, the above changes and the structural backdrop assumed that “the federal funds rate consistent with maximum employment and price stability over the longer run has declined relative to its historical average. Therefore, the federal funds rate is likely to be constrained by its effective lower bound more frequently than in the past.” This obviously has failed to place as the world of persistent demand shocks, modest fiscal stimulus, and dis- or de-flationary supply-side developments were all undone by covid. How much the Fed thinks the world will gradually revert to the former one remains an open question with a wide range of views, best summed up by the distribution of the long-run dot.
Some of these challenges still seem to apply, particularly the observation that neutral rates are lower than they have been historically, but the at the time structurally dovish implications and interpretations of them seem unlikely to carry forward to nearly the same extent that they did in 2020. It is also the case the neutral rates are likely somewhat above the level the Fed assumed in 2018-20 and trend inflation has always switched to above target rather than below.
What the 2020 framework failed to include was a sense of imagination around the shocks which could impact the economy. While covid was unprecedented, in many ways the event itself and the economic policy responses to it were different in scale not kind. The 2020 framework review and the reaction function it led to have been well discussed (see views from former Vice Chair Clarida, the BIS’ Carstens, and a conference at the Brookings Institution). There were certainly some failures of imagination on the part of Fed officials but perhaps most concretely, they “got aggregate supply wrong” to quote from Clarida’s above linked speech. The framework focused on the challenges of the zero lower bound and financial system dominated shocks in a world where supply-side developments had generally been procyclical (sustaining the cycle) and, on net, deflationary.
The Fed’s subsequent policy followed the lead of the framework and consensus forecasters, who as a group including the Fed did not appreciate the snapback in growth or boom in inflation that was to come, in locking itself into to dovish of a policy approach. But it remains up for debate if the reaction function and policy framework or the too anchored forecasts were more the greater cause of the avoidable part, which is likely only a small part of the initial overshoot but more of the latter longer-lived inflation, of the inflation overshoot.
Formalizing the 2020 framework as “maximum employment subject to an inflation constraint centered around 2%” remains a close to optimal statement of policy intent but lacks enough detail to be operationalized smoothly. While the unemployment rate is often a good summary statistic in this regard, the Fed’s consistent pessimism on the ultimate non-inflationary level of the unemployment (due to cyclically supportive positive hysteresis in labor supply). I would also add that this interpretation seems more in the qualitative spirit of the dual mandate than a symmetric employment mandate does. Its failure to think about how maximum employment and inflation centered around 2% should be linked is a challenge in a world where the two sides of the mandate have risks at cross purposes.
Where Will This Leave the Fed?
The joint asymmetries of the old framework are largely likely to be jettisoned. Effectively implementing flexible average (sort of unspecified but in way which locked in reactive policy to ensure some eventual overshooting took place) inflation targeting would have been challenging under benign circumstances and events were far from benign. The average will go, and we will move towards the more standard flexible average inflation target which implicitly focuses on 6-24m inflation forecasts as the primary inflationary input to monetary policy decision making.
On the labor market, I am less sure how the Fed will tweak the framework. The asymmetric “shortfalls” language is likely to revert to 2012’s “deviations” around maximum employment. Shortfalls seems more consistent with the statutory language of the dual mandate in a narrow sense but with upside inflation risks far more salient now than in 2020 this reversion helps anchor medium-term dynamics some (although it does not seem likely the Fed formally addresses the inflation dependence of its assessments of labor market slack; see more below).
The more hawkishly inclined, and somewhat inconsistently perhaps the White House and a number of Chair hopefuls as well, could argue that the Fed’s pursuit of “broad-based and inclusive” view of full employment was another reason it failed to respond to incipient inflation in time. There may be some truth to this, but the prior cycle had showed the Fed that preemptive tightening on the basis inflationary assumptions stemming from view on the inflationary level of unemployment rate were not correct and labor supply dynamics were notable disinflationary there. It is admittedly unclear if the current gradual easing in slack off of the overheated conditions in the labor market, particularly with the bullwhip in immigration flows, will allow for the labor market to see such positive supply-side outcomes that attenuate medium-term inflationary trends.
What Maybe Should Change but Isn’t Guaranteed?
The Fed should make clear under what circumstances it will use different types of balance sheet policy (QE/QT) and temporary lending facilities going forward. This is likely to be somewhat less relevant going forward given the distance from the zero lower bound but as we saw in spring ’23 with the SVB blowup and response (and the BoE experienced with the mini budget crisis) balance sheet policy outside of the usual rate supportive QE has remained a feature of monetary policy. I expect that these largely successful and short-lived experiences have made central bankers more willing to use lending and market functioning QE going forward even as long-term rate driven QE will remain in the backseat for some time to come.
Increasing the emphasis on nominal rather than real + inflation growth paths would make the Fed’s framework more robust to changing short-term supply side conditions, particularly to one-off supply-side shocks such as tariffs or oil price moves. These approaches have been widely discussed since the GFC. While I think there is substantial merit in them, the fact they were not considered during the much larger review in 2019-20 suggests that they will likely not be a part of the framework revisions during this round either. Employ America (here) makes an interesting dovish case for nominal consumption and income targets as a way to assess the likely persistence of inflation because “when the business cycle and labor markets are fragile, we typically do not see inflation coincide with robust labor income or consumer spending growth. When they are stronger, as they were in 2022, the presence of inflation can take on firmer persistence. Focusing on nominal labor income and consumer spending, instead of just GDP in totality, also avoids penalizing investment-led GDP growth.” David Beckworth has long been an advocate of nominal GDP targeting for similar reasons (here).
The inflation-dependent context of maximum employment’s stability seems unlikely to be touched on officially. This point is often implicitly made by Chair Powell in the post-meeting press conferences (“without price stability, we cannot achieve the long periods of strong labor market conditions that benefit all Americans”) but it has not been formally codified by the Fed noting that its assessment of maximum employment cannot be an absolute aim but rather has to take place in the context of labor markets which are consistent with inflation at target. This is made more complicated by the different post-covid inflation regime and the Fed’s own history of under-estimating the economy’s ability to continue seeing solid employment outcomes that are non-inflationary (the late 1990s and the 2015-19 era when positive labor hysteresis allowed lower non-inflation unemployment rate and higher prime-age LFPR levels than seen in the prior cycle). There may be less ability to sustain tighter intensive margins of slack (urate, quits) but with non-recessionary cyclical stability there is little reason to think that more extensive margins of slack couldn’t keep improving, but historically extensive margins lag intensive ones, making this at best a challenging balancing act for the Fed when inflation is likely to be above target for 6 or more years in a row.
To quote Powell at the July press conference, “if you saw that the risks to the two goals were moving into balance, if they were fully in balance, that would imply that you should move toward a more—a more neutral stance of policy” and “the economy is in—is in, you know, good shape, but it’s an unusual situation where you have risks to both your employment mandate and your inflation. That’s the nature of a supply shock. And it’s probably not surprising that there would be differences and different perspectives on that as well as different views of where the neutral rate is, so they—different views of how tight policy is.” ↑
