Rethinking the Institutional Architecture of Monetary Policy in Light of the December 2026 FOMC Decision
Meeting Frequency, Discretionary Interest-Rate Adjustments, Strategic Signalling and Artificial Intelligence: A Critical Assessment of Conventional Monetary-Policy Practice under Radical Uncertainty
Farid Novin
Revised edition · 10 November 2026
Abstract
The Federal Open Market Committee approaches its meeting of 8–9 December 2026 with the evidence pointing in one direction and the institutional conventions pointing in another. The October consumer price report, released on the morning of 10 November, showed core inflation rising by 0.3 percent for the third consecutive month, above what economists had expected; the October employment report showed payroll growth of 165,000, almost three times the consensus; and Governor Christopher Waller, speaking in London hours before the inflation release, indicated that a 25-basis-point increase was his baseline and that a 50-basis-point increase should be considered if core inflation printed at or above 0.3 percent. It did.
This paper uses that conjuncture as a live test of an argument about the architecture of monetary policy: that the frequency of meetings, the size of rate adjustments and the manner of communication are instruments to be calibrated to the informational content of events, not administrative conventions to be defended for their own sake. It reassesses the original argument against the new evidence, examines the Federal Reserve’s own announced changes to its meeting calendar, treats Governor Waller’s speech as a case study in conditional commitment and strategic signalling, considers artificial intelligence both as an analytical instrument and as a macroeconomic shock, and closes with a Bayesian game-theoretic assessment of the December decision and a revised institutional architecture.
I. Introduction: A Live Test of the Argument
The institutional design of monetary policy is usually discussed in the abstract, as though calendars, increments and communication protocols were neutral technical choices. They are not. Each of them embeds an assumption about how quickly the economy changes, how reliable the available information is, and how markets will interpret what a central bank does or fails to do. In calm periods these assumptions are rarely tested. In periods of genuine uncertainty they are tested in public, and the quality of the institution is revealed by whether its conventions help or hinder the decision in front of it.
The present moment is such a test. The Federal Reserve raised its target range by 25 basis points on 16 September, the first increase since 2023, bringing the range to 3.75–4.00 percent. It then held the range on 28 October by a vote of ten to two, with Governor Lisa Cook and Cleveland Federal Reserve Bank President Beth Hammack dissenting in favour of a further 25-basis-point increase. Within two weeks, the data have moved decisively against the view that the pause could be prolonged. The question for December is no longer chiefly whether to raise the policy rate. Futures markets had priced that as close to certain even before the inflation release. The question is how large the increase should be, how it should be communicated, and what it will reveal about the institution taking it.
That question touches every element of the original argument. A debate about whether the committee should meet six, eight or fifty-two times a year turns out to be inseparable from a debate about whether it can recognise, in real time, the difference between noise and a change of regime. A debate about whether 25 basis points is a law of nature or a habit turns out to be inseparable from a debate about whether a habit can carry the weight of credibility when credibility is what is at stake. And a debate about whether artificial intelligence can improve monetary policy turns out to be inseparable from the fact that artificial intelligence is itself one of the forces now moving the economy.
The historical usefulness of institutional conventions should not be confused with economic necessity. The appropriate objective is to maximise the quality, timeliness and adaptability of monetary-policy decisions, not to maximise the number of meetings or to impose uniformity on the size of adjustments. The remainder of this paper tests that proposition against the evidence of November 2026.
II. The Evidence Base on 10 November 2026
A. Prices: Broadening, Not Receding
The consumer price index rose by 0.4 percent in October after an increase of 0.6 percent in September. Excluding food and energy, prices rose by 0.3 percent, matching the pace of August and September, against a consensus expectation of 0.2 percent. The composition matters as much as the headline. Shelter, which rose by 0.3 percent, accounted for more than 30 percent of the total increase. Recreation rose by 0.8 percent, motor-vehicle insurance by 0.7 percent and medical care by 0.3 percent, while used vehicles fell by 1.2 percent, which is the kind of offset that, had it not occurred, would have made the core reading higher still. Energy prices rose by 0.7 percent following an increase of 5.2 percent in September, with gasoline up by 1.0 percent.
Three features of this pattern bear on the institutional argument. First, a core rate of 0.3 percent a month, sustained for a quarter, is a rate somewhat above 3.6 percent when annualised, and it is the three-month pattern, not the single observation, that carries the information. Second, the strength is not confined to a single volatile category. When shelter, services such as insurance and medical care, and discretionary categories such as recreation move together, the more natural interpretation is breadth, and breadth is what distinguishes a persistent disturbance from a temporary one. Third, analysts have attributed core inflation to the combined influence of tariffs, energy prices and demand related to artificial intelligence, which means the sources of pressure are heterogeneous and are unlikely to fade on a common schedule.
B. The Labour Market and Real Activity: Strong, with a Distributional Qualification
Nonfarm payrolls rose by 165,000 in October, compared with a consensus near 65,000 and an average monthly gain of about 48,000 over the preceding year. Revisions added 84,000 jobs to the August and September totals. The unemployment rate fell to 4.1 percent from 4.2 percent, and it has stayed in a narrow band between 4.1 and 4.3 percent since March. The decline occurred even though 421,000 people entered the labour force, and the participation rate edged up to 61.9 percent. Average hourly earnings rose by 0.4 percent in the month, the largest monthly gain since January, and by 3.1 percent over twelve months. Third-quarter gross domestic product grew at an annualised rate of 3.4 percent, and the Federal Reserve Bank of Atlanta’s tracking estimate for the fourth quarter stood near 3.5 percent.
The distributional qualification is that wage growth of 3.1 percent remains below consumer-price inflation, so that real hourly earnings have been falling for months. This helps explain the apparent paradox of a strong labour market and widespread public discontent about the cost of living. It also bears on the policy debate, because it weakens the argument that wage pressure is the transmission channel for persistent inflation and strengthens the argument that the pressure comes from supply-side shocks, administered prices and investment-driven demand. A central bank facing inflation of that character has a harder task, because the instruments it controls act most directly on a channel that is not the dominant one.
A further point concerns what constitutes a strong employment number. The Committee’s own October statement observed that job gains have kept pace with the workforce. When the growth of labour supply is slow, the number of jobs required merely to hold unemployment constant is lower than it was in earlier years, and a payroll gain of 165,000 therefore represents more tightening of the labour market than the same figure would have implied in a period of rapid labour-force growth. The estimate of that breakeven pace is itself uncertain, which is precisely the kind of unobservable quantity discussed in the original paper.
C. The Policy Stance, the October Hold and Market Pricing
The Committee’s October statement said that inflation remains elevated and added a short declaration that the Committee will deliver price stability. Chair Kevin Warsh described the deliberative process as one of interrogating reality, complicated by high levels of investment, a steady labour market and elevated inflation, and said that financial and credit conditions had begun to tighten for some businesses and households following the September increase. He also denied that the proximity of the 3 November midterm elections had played any role in the decision to hold.
That denial deserves analytical attention rather than either acceptance or dismissal. A hold six days before a national election is, in the language of signalling theory, a pooling action: it is what an institution would do if it were purely data-driven and cautious, and also what an institution would do if it were quietly sensitive to the electoral calendar. Observers cannot distinguish the two from the action alone, and they therefore infer from the subsequent pattern of behaviour. If the Committee now acts decisively and in proportion to the data, the inference of political caution recedes. If it hesitates again, the inference strengthens. The December decision carries, in this sense, a reputational burden that goes beyond its direct macroeconomic effect.
Market reaction to the morning’s events was immediate. Before the inflation release, Governor Waller’s speech had lifted the futures-implied probability of a December increase of 25 basis points toward certainty. After the release, the implied probability of a 50-basis-point increase rose from roughly one percent to a range of about 10 to 14 percent, depending on the time of observation and the source. The ten-year Treasury yield rose by more than ten basis points to approximately 5 percent, a level it had also reached earlier in the month and which in October represented the highest yields in roughly twenty-four years. The two-year yield rose to about 4.69 percent. Equities fell and the dollar strengthened. Commentators at major asset managers now describe a further increase in 2027 as a plausible, though not certain, continuation.
D. What the Data Cannot Tell Us
One feature of the October release deserves particular emphasis because it connects directly to the paper’s treatment of measurement under uncertainty. The Bureau of Labor Statistics did not publish an annual inflation rate for October, because the federal government shutdown of October 2025 prevented the collection of the data required to construct the comparison. The household-survey comparisons in the employment report are affected in the same way. At the moment when the Committee most needs a clean reading of the inflation trend, the standard twelve-month measure is unavailable.
Analysts have improvised. One calculation that assumes the October 2025 index lay midway between its September and November values puts annual headline inflation at about 4 percent and annual core inflation at about 2.8 percent. Such figures must be treated as illustrative, not as official statistics, but they permit a rough assessment of the real policy rate. With the midpoint of the target range at 3.875 percent, the real policy rate measured against headline inflation is approximately zero, and against core inflation it is approximately one percentage point. In an economy growing at more than 3 percent with unemployment at 4.1 percent, a policy rate that offers no real return to savers is difficult to describe as restrictive, which is the substance of the case for continued tightening.
III. Meeting Frequency: Surveillance, Deliberation and Intervention
The original argument rested on a three-part distinction: continuous surveillance of the economy, periodic deliberation about the accumulated evidence, and intervention in the form of a change in the policy instrument. These activities need not follow the same timetable. A central bank can monitor continuously and deliberate often while changing its instrument rarely, and it can respond to a genuine disturbance without waiting for a date that was fixed before the disturbance occurred.
On 28 October the Federal Reserve announced a change in its own calendar that engages this distinction directly. Chair Warsh stated that eight meetings a year remain the right number, but that beginning in 2027 one of them will be devoted exclusively to broad questions of monetary-policy strategy, with neither a vote on the federal funds rate nor a policy statement anticipated, and that in 2028 two meetings will be so devoted, leaving six for setting policy. He said it was not yet obvious what, if anything, would be published after such meetings. Some analysts expect the first such meeting to fall in October 2027, between two meetings that include quarterly projections, which would produce a gap of roughly twelve weeks between opportunities to change the policy rate, though with two sets of inflation reports in the interval.
A. What the Warsh Reform Gets Right
The reform accepts, in practice, the central proposition of the original paper: that deliberation and intervention are different activities and that the committee’s time is not best used if every meeting is organised around a vote. A meeting set aside for thinking about frameworks, longer horizons and the structure of the institution’s own reaction function is a meeting at which the pressure to produce a numerical decision cannot distort the analysis. The chair’s own formulation, that policy should be assessed at regular and information-rich intervals while room is also reserved for strategic deliberation, is a restatement of the same separation.
B. Where It Differs from the Original Argument
The reform nonetheless moves in a direction the original paper did not advocate, and the difference is instructive. The original proposed to increase the frequency of deliberation, including weekly review sessions supported by continuous surveillance, while leaving the frequency of intervention to be governed by the information received. The Federal Reserve is instead reducing the number of meetings at which intervention is possible and leaving the remaining deliberative capacity to be decided by the calendar. A fixed date on which the committee has committed not to decide is itself a calendar-based rule. It substitutes one administrative convention for another, and it does so at the very moment when the evidence suggests that the pace of events may warrant more, not less, flexibility.
The institution retains the authority to act between meetings. But that authority has been exercised only rarely, most recently in early 2020 at the onset of the pandemic, and markets are accordingly likely to interpret any unscheduled action as a signal of crisis. The more a central bank concentrates its scheduled decision points, the more informational weight each unscheduled move carries, and the larger the cost of using it for anything short of an emergency. This is the central design problem: a calendar with fewer decision dates requires a better-defined protocol for acting between them, and no such protocol has been announced.
C. The Present Cycle as a Test
The practical stakes are visible in the current sequence. Some market economists anticipate a December increase, a pause in January, and a further increase in March if inflation does not improve. Under a calendar with a decision-free meeting in autumn 2027, a tightening cycle that extended into that year would be subject to a longer interval between adjustments than the data might justify. The adequacy of the architecture will therefore be judged not in a calm period but in precisely the kind of environment now emerging, in which inflation is above target, real activity is strong and the appropriate pace of adjustment is genuinely uncertain.
D. The October Meeting and the Informational Value of Meetings
The October meeting illustrates the original argument about the value of a meeting in a different way. Its principal output was not a change in the policy rate but a revelation of disagreement, in the form of two dissents in favour of tightening, together with the chair’s statement that the meeting schedule would change. A meeting at which the rate is unchanged can nonetheless be highly informative if it alters the distribution of expectations about future decisions. Conversely, the information carried by the dissents was largely about timing, since both dissenters favoured the same 25-basis-point increase the committee was expected to approve at its next opportunity. A framework in which the committee could have moved earlier on the basis of the accumulating evidence would have converted that disagreement about timing into a decision, rather than leaving it as a signal that markets had to interpret.
IV. Artificial Intelligence, Real-Time Data and the Limits of Measurement
The original paper proposed that artificial intelligence be used as an analytical and early-warning instrument, and not as an autonomous authority. The events of November 2026 sharpen that proposition in two ways, one that tests the instrument and one that complicates its object.
A. The Instrument: Why the Shutdown Gap Matters
An artificial-intelligence surveillance system can process payments data, credit flows, commodity prices and survey responses at a speed no committee could match. But it cannot supply observations that were never collected. The absence of an October 2025 price index, which has removed the standard annual comparison from the October 2026 release, is a reminder that the value of any analytical system depends on the integrity and continuity of its inputs. A model trained on a series with a hole in it must either interpolate, which imports an assumption, or ignore the gap, which imports a bias. The paper’s earlier observation that real-time data are not necessarily accurate data therefore has an immediate empirical illustration. Research from the Federal Reserve Bank of Richmond has examined how data revisions and measurement error complicate even systematic policy decisions, and the Bank for International Settlements has emphasised that monetary policy is ultimately an exercise of judgment under uncertainty.
The institutional implication is that central banks should maintain independently constructed, privately sourced and rigorously documented alternative measures of price and labour-market conditions, so that a disruption to official statistics does not leave the committee reasoning from a gap. This is a task to which machine-learning methods are well suited, provided their outputs are subject to independent validation and are presented to policymakers with explicit statements of uncertainty.
B. The Object: Artificial Intelligence as a Macroeconomic Shock
The complication is that artificial intelligence is no longer merely a tool available to the central bank. It is a force acting on the economy the central bank is trying to read. The Committee itself has described capital investment as robust and productivity growth as strong, and market analysts have identified demand linked to artificial intelligence as one of three shocks, alongside tariffs and energy prices, shaping core inflation. Strategists at a major asset manager have meanwhile referred to wider jitters about the sector as a risk that sits alongside the case for further tightening.
This has three consequences for monetary-policy design. First, an investment boom raises aggregate demand in the short run and may raise productive capacity in the longer run, so that the neutral rate of interest, which cannot be observed directly, may be higher than historical estimates suggest. A model that assumes a stable neutral rate will then understate the degree of tightening required, and the error will not be visible in the model’s own diagnostics. Second, the concentration of the investment boom in a small number of assets creates the possibility of a nonlinear financial adjustment, which is exactly the category of exceptional disturbance for which the original paper argued that conventional frameworks need to be reconsidered rapidly. Third, any artificial-intelligence system trained on the historical relationship between investment, productivity and inflation is, by construction, trained on a regime that may now be ending.
The conclusion is a strengthening of the original position. Artificial intelligence should assist human deliberation by processing more information and identifying emerging deviations. It cannot resolve the deepest uncertainty, which concerns whether the underlying economic relationships have shifted. That question requires institutional judgment, and the more the economy is transformed by the same technology that supplies the analytical tools, the more important it is that final authority remains with accountable policymakers.
V. The Twenty-Five-Basis-Point Convention Under Test
The original paper argued that 25 basis points is an operational convention and not an economic law, and that the appropriate size of an adjustment should depend on the magnitude and persistence of the disturbance, the responsiveness of demand and inflation, the existing degree of restraint, the condition of the financial system and the distribution of risks. The December decision tests this proposition, because the live choice is between the convention and a departure from it.
A. The Case for 50 Basis Points
The argument for a larger move is built on five observations. The first is the persistence and breadth of core inflation, which has run at 0.3 percent for three consecutive months and exceeded expectations in the latest. The second is the strength of the real economy: growth above 3 percent, falling unemployment, rising participation and accelerating wages in October. The third is the level of the real policy rate, which on the best available approximation is close to zero against headline inflation, a stance difficult to reconcile with a restrictive intent. The fourth is the asymmetry of credibility: a central bank that allows inflation to run above its objective for close to six years, as the chair himself acknowledged, has accumulated a reputational deficit, and expectations are at greatest risk of de-anchoring when the public concludes that the institution lacks resolve. The fifth is Governor Waller’s own reasoning, that hiking only 25 basis points would put credibility at risk, which identifies the size of the action, and not merely its direction, as the carrier of the signal.
B. The Case for 25 Basis Points
The argument for the conventional move is no less serious. First, the policy rate has already begun to bite on the margin: the chair has reported tighter financial and credit conditions for some businesses and households, and long-term yields have risen to levels not seen in decades. A sizeable part of the tightening is being delivered by markets, not by the policy rate. Second, policy operates with long and variable lags, and Brainard’s classic result that uncertainty about the effect of an instrument can justify attenuating its use applies with particular force when the real economy is being reshaped by an investment boom of uncertain scale. Third, energy prices, which rose by 5.2 percent in September and by a further 0.7 percent in October, are the most volatile component of the index, and a policy response calibrated to the combined pattern risks overreacting to a supply impulse that may partly reverse. Fourth, the October dissenters themselves asked for 25 basis points, not 50, so that a 50-basis-point decision would move beyond the position of the committee’s most hawkish voters and risk a visible split. Fifth, the Committee will receive wholesale-price data, further employment data and projections before the meeting, and an increment that preserves optionality for January and March has option value that a larger move forgoes.
C. Evaluating the Two Cases
The two cases are not symmetric in the information they rely on. The case for 50 basis points depends heavily on the trajectory of the next several data releases: if wholesale prices and the measure of personal consumption expenditures preferred by the Committee confirm the strength seen in the consumer price index, the three-month pattern will have acquired additional corroboration and the marginal cost of a larger move in terms of regret will fall. If those releases soften, the case for the conventional increment is restored. This is precisely the situation in which the original paper’s principle applies: the appropriate size of the adjustment is endogenous to the information that arrives between now and the meeting, and the institution should preserve the capacity to select either.
A related question is whether the choice set should extend beyond the two familiar alternatives. The original paper argued that an adjustment of, say, 35 or 40 basis points could in principle be more appropriate than either 25 or 50. This remains true as a matter of economic logic, and the administered-rate framework, which is set by the Board of Governors, does not technically require that the target range move in multiples of a quarter point. But the paper’s own analysis of communication cautions against treating the freedom to choose as a reason to exercise it. In a period when the institution’s credibility is the variable under scrutiny, an unconventional increment risks being read as evidence of unresolved disagreement, a compromise between camps, and not as a calibrated judgment. The appropriate lesson is therefore not that the Federal Reserve should adopt an unusual number in December, but that its deliberations should treat the numerical menu as a variable and not as a constraint, and that any departure from convention should be accompanied by an explanation of the reasoning that makes it intelligible.
It is also worth noting how much the debate has already shifted. A governor who was among the Committee’s advocates of easing in the previous cycle is now describing a 25-basis-point increase as his baseline and a 50-basis-point increase as a serious option. That migration is, in itself, informative about how the weight of evidence has changed, and about how quickly the committee’s centre of gravity can move when the data do not cooperate with earlier expectations.
VI. Signalling and Credibility: Waller’s Conditional Commitment
Governor Waller’s speech of 10 November is a near-perfect specimen of the strategic signalling the original paper analysed, and it repays close reading. He did three things in a single address. He announced a baseline: even if October inflation improved, as he expected it would, he leaned toward a 25-basis-point increase in December. He announced a conditional rule: if core inflation came in around 0.3 percent or higher, a stronger move would be needed and a 50-basis-point increase should be considered. And he attached a statement about consequences to the failure to act, saying that hiking only 25 basis points in that case would put credibility at risk, which is a risk he said he was not eager to take.
A. Conditional Commitment as a Self-Binding Device
The significance of the speech lies in its timing. It was delivered before the data were known, and it specified in advance the datum that would trigger a larger response. That is the structure of a conditional commitment. By naming the threshold ex ante, the speaker raised the reputational cost, for himself and for the institution, of a response inconsistent with the stated rule once the data were released. The data then arrived exactly at the threshold, which converted a hypothetical into an obligation. The futures market registered both stages: the probability of a 25-basis-point increase moved from a range of roughly 76 to 85 percent to close to certainty during the speech, and the probability of 50 basis points rose from about one percent to ten to fourteen percent after the release.
This is a more refined form of signalling than the original paper’s example of a newly appointed governor who must demonstrate resolve through a costly action. Here the cost is not a surprise but a pre-commitment, and the credibility effect operates through the expected consequences of a failure to follow through. The device works only if the speaker’s reputation for consistency is at stake, and only if the audience believes that other members of the committee share the stakes. It also has a risk. A conditional rule announced by one member narrows the choice set for the whole committee, and if the committee then chooses the conventional increment, it will have contradicted a published signal and may be read as having been unable to build consensus.
B. The Distinction Between Credibility and Predictability
The episode also illustrates the paper’s distinction between predictability and credibility. Markets were able to forecast the direction of the December decision with considerable confidence, and the conditional rule made the magnitude partly predictable as well. Predictability of this kind is valuable. But it would be a mistake to treat it as a substitute for credibility, which depends on whether the institution’s actions are seen to follow from its objectives when they are costly. A central bank that always does what the market expects is predictable. A central bank that does what is required even when the market expected otherwise, and whose action is subsequently vindicated, is credible. The two coincide only if the market’s expectations are correctly formed, and the experience of 2021 and 2022, when the characterisation of inflation as transitory was contradicted by subsequent developments, shows how fragile that coincidence can be.
VII. Communication under a Minimalist Chair
The current Federal Reserve is conducting an experiment in minimalist communication. Chair Warsh has shortened the policy statement and the press conference and has said, in the plainest terms, that he does not believe in forward guidance absent exigent circumstances or a crisis. His approach is coherent as a reaction to a period in which elaborate guidance became a source of difficulty, as when public characterisations of inflation were overtaken by events. Its virtue is that it avoids binding the institution to a path that the data may not support.
Its cost, however, is the central point of the original paper’s discussion of the distinction between transparent communication and continuous tactical signalling. When the centre says little, the periphery speaks more. In the week of the most consequential data release of the quarter, it was a governor in London who supplied the market with the most operationally useful information about the December decision. Such speeches are in effect unofficial monetary-policy announcements, and the more the chair withholds, the more weight individual remarks must carry. The institution thereby acquires exactly the problem that minimalism was designed to avoid, which is a market interpreting every utterance as a policy signal, but without the discipline of a coordinated message.
The empirical literature on central-bank communication reaches a similar conclusion from a different direction. The tone of Federal Reserve statements contains information that helps explain monetary surprises and anticipate later decisions, and a recent review of communication with the public emphasises its growing importance for policy effectiveness while questioning how much of it reaches ordinary households. Taken together, these findings suggest that the choice is not between communication and silence. It is between communication that is designed, coherent and conditional, and communication that is improvised, fragmented and discovered by the market in the speeches of individual members. The first can be done with very few words. The second cannot be avoided by speaking less at the centre.
The proposal to hold strategy meetings without a statement raises the same issue. If the chair does not anticipate releasing a statement after such meetings, the market will form its own view of what was discussed, and the vacuum will be filled by leaks, speeches and conjecture. A brief, carefully worded account of the questions the Committee examined, without any indication of decisions, would preserve the intellectual freedom of the meeting while reducing the risk that its silence is read as a signal.
VIII. A Bayesian Game-Theoretic Assessment of the December Decision
A. The Hypotheses
The decision problem can be organised around three competing hypotheses about the inflation process. The first holds that the recent strength is predominantly a supply-side phenomenon, driven by tariffs and by energy prices that have been disrupted by the confrontation in Persian Gulf, and that it will fade as those impulses are absorbed. The second holds that inflation is broadening and becoming self-sustaining, supported by strong demand from the investment boom, a resilient labour market and the passage of tariff costs through to a widening range of prices. The third holds that the economy has entered a different regime, with higher productive investment, a higher neutral rate and a permanently higher inflation floor than the Committee’s models assume.
These hypotheses carry different implications for the appropriate policy response. Under the first, modest tightening followed by patience is optimal, and a large move risks an unnecessary loss of output. Under the second, a faster response is warranted because the cost of delay compounds. Under the third, the appropriate response is not merely a larger move but a revision of the framework in which the move is made.
B. The Updating
A reasonable prior, held before the autumn data, would have assigned the greatest weight to the first hypothesis and a smaller weight to the second, with a modest residual for the third. The evidence since then has been informative in a specific way. A single month of strong core inflation would have been well explained by all three hypotheses and would have moved beliefs little. A third consecutive month at 0.3 percent, accompanied by breadth across shelter, insurance, medical care and recreation, is considerably more probable under the second and third hypotheses than under the first, because a supply shock confined to particular goods would not normally generate such a uniform pattern in services. The employment report is similarly diagnostic: a payroll gain well above the consensus, falling unemployment and accelerating wages are more likely under strong demand than under a pure supply disturbance. The strength of third-quarter growth and the investment boom tilt the balance further.
My own subjective assessment, which should be understood as a calibrated judgment and not as a statistical estimate, is that the weight on the first hypothesis has declined from a majority to a minority position, that the second hypothesis now carries the largest share, and that the third has risen from a residual to a meaningful minority. The qualification is that the evidence in favour of the first has not disappeared: energy volatility remains substantial, the used-vehicle category has been falling, and the absence of an official annual figure means that the three-month annualised measure carries more of the interpretive burden than it ordinarily would. A responsible posterior therefore remains wide, and the appropriate policy is one that is robust to an error in either direction.
C. The Strategic Environment
The decision is not made in isolation. At least four sets of actors respond to it. The first is the Committee itself, which contains a hawkish bloc that dissented in October, a centre that has migrated toward larger increases, and a chair who prefers minimal communication and appears reluctant to pre-commit. The second is the financial market, which anticipates the decision and reprices long-term rates in advance, so that part of the tightening effect is delivered before the policy rate moves. The third is the executive branch, which has for years argued that the policy rate should be much lower and which has placed the Federal Reserve under legal and political pressure. The fourth is the general public, whose inflation expectations are formed by experience of prices more than by any speech, and whose tolerance for a prolonged erosion of real earnings is limited.
The interaction among these actors creates a sequential game with incomplete information. The Committee wishes to establish that it will respond to evidence regardless of political circumstances. The market wishes to infer whether the Committee’s resolve is real. The executive branch can raise the cost of tightening but cannot change the arithmetic of inflation. In such a game, the principal value of a decisive action is that it separates committees that are willing to bear political cost from those that are not. If the Committee increases by 25 basis points and accompanies the decision with a clear conditional statement of its readiness to do more, it separates itself from a committee that merely hesitates, while preserving its option to respond to subsequent data. If it increases by 50 basis points it makes the separation sharper, but it also accepts a larger risk of being seen to have overreacted if subsequent data soften.
D. Scenarios and Judgments
Three scenarios organise the range of plausible outcomes. In the first, which I judge the most probable, the Committee raises the target range by 25 basis points on 9 December, accompanies the decision with a statement and projections indicating that further increases are likely if inflation does not improve, and keeps open the possibility of a larger move at a subsequent meeting. In the second, the Committee raises the range by 50 basis points, most likely if the wholesale-price and consumption-expenditure data reinforce the consumer-price pattern and if the November employment report is again strong. In the third, an adverse financial event, perhaps connected with a repricing of technology-sector valuations or a disorderly rise in long-term yields, leads the Committee to moderate its response or pause, a scenario of lower probability but substantial consequence.
My assessment is that a 25-basis-point increase is the most likely outcome, with a probability in the neighbourhood of three in four, that a 50-basis-point increase has a probability somewhat higher than the current market-implied figure but below one in four, and that the remaining probability is distributed across other outcomes. These are judgments conditional on the information available on 10 November, and they should be revised as the intervening data are released. I would assign greater weight to a larger move than futures markets do chiefly because of the explicit conditionality of Governor Waller’s statement, which the data have now satisfied, and because the stated cost of a failure to follow through is expressed in the vocabulary of credibility, which is the variable on which the institution’s reputation currently depends.
E. The Normative Judgment
The more difficult question is what the Committee ought to do, as distinct from what it is likely to do. Here the argument must be stated carefully. The evidence supports a December increase and supports the view that the risks to the inflation outlook are on the upside. It does not by itself establish that 50 basis points is superior to 25. What it does establish is that the choice should be made on the basis of the evidence and not on the basis of the convention. If the Committee concludes, after the intervening releases, that inflation is broadening and the neutral rate is higher than assumed, a larger increase is defensible and, given the real policy rate, may be preferable. If it concludes that the strength of October is partly transitory, the conventional increment, together with a firm commitment to respond to further evidence, is more prudent. The institutional failure to avoid is the choice of a number because it is customary, or because it is unthreatening, rather than because it is justified.
IX. A Revised Institutional Architecture
The analysis supports a more flexible and more explicit institutional architecture than the one the Federal Reserve is presently adopting. Seven elements follow.
1. Continuous Surveillance with Redundant Data Sources
The central bank should maintain a permanent surveillance system that draws on official statistics, privately sourced high-frequency data and independent alternative measures. The experience of the missing October 2025 price index shows that reliance on a single official series is a vulnerability. Artificial-intelligence methods should be used to detect deviations from established patterns and to construct alternative estimates, with documented uncertainty and independent validation.
2. Deliberation Scheduled by Information, Not Only by the Calendar
Regular deliberation should continue, and the committee should be able to convene more frequently during periods of heightened uncertainty. The establishment of strategy-focused meetings is a constructive recognition that deliberation and intervention are different activities. But the frequency of review should not be fixed by calendar alone, and the committee should define in advance the conditions under which it will convene outside the schedule.
3. A Defined Intermeeting Protocol
Because the number of decision meetings is being reduced, the authority to act between meetings must be made credible, and its credibility depends on being predictable in form. The Committee should publish a short statement of the circumstances in which it would consider an unscheduled decision, such as a sharp deterioration in financial conditions, a sudden change in inflation expectations or a major supply disruption, so that an intermeeting action is not automatically interpreted as an emergency and its signalling content is correspondingly better understood.
4. Rate Adjustments Calibrated to the Disturbance
Conventional increments should remain reference points, not constraints. The December deliberation illustrates why. The relevant question is whether the adjustment is consistent with the assessment of the economy, the transmission mechanism and the likely market response, and the explanation of the choice should be as prominent as the choice itself. Where the Committee departs from convention, whether upward or downward, it should explain the reasoning in terms that make the departure intelligible and prevent it from being read as a compromise.
5. Designed Communication, Not Improvised Communication
Minimalism at the centre is defensible, but it should be accompanied by a disciplined framework for the communication of conditional intentions. The Committee should consider articulating, in a few sentences, the principal conditions under which it would accelerate or slow the pace of adjustment. This would reduce the burden placed on individual speeches and would allow the institution to retain flexibility without leaving the market to infer its reaction function from the remarks of single members.
6. Periodic Reconsideration of the Policy Framework
The strategy meetings should be used for exactly the purpose for which they have been proposed: to examine whether the assumptions underlying the policy framework remain valid. The principal questions of the moment are evident. They include the level of the neutral rate in an economy experiencing a technology-driven investment boom, the reliability of inflation measures in the presence of data gaps, the effect of tariffs and energy shocks on the inflation process, and the appropriate response to shocks that originate on the supply side but become embedded in demand. A short public summary of the issues examined would preserve the independence of the discussion without leaving the market to speculate.
7. Accountability and Human Judgment
Whatever analytical tools are employed, the authority to take consequential decisions must remain with accountable policymakers. They should be able to explain not only what was decided but why alternative responses were judged less appropriate. In an environment in which public confidence in institutions is fragile and the central bank is under political pressure, the capacity to give a coherent account of its reasoning is a component of its independence.
X. Objections Reconsidered
A balanced assessment must acknowledge that greater flexibility carries risks, and the events of November 2026 sharpen several of the objections raised in the original paper.
The first objection is that a reduced number of scheduled decision meetings weakens transparency and increases uncertainty about timing. This is partly correct. Financial markets benefit from knowing when decisions will be made, and the proposal to designate certain meetings as strategy meetings makes the calendar less informative about when policy may change. The answer is not to retain every meeting as a decision meeting but to ensure that the rules governing intermeeting action are transparent.
The second objection is that more frequent deliberation would encourage short-termism and excessive intervention. The risk is real, particularly in an environment in which the data are noisy. The appropriate response remains the separation between the frequency of deliberation and the frequency of intervention. A committee can review the evidence weekly and still hold its stance for months if the information does not warrant a change.
The third objection is that non-standard increments complicate communication. The December debate shows why this objection has force: in a period when credibility is under scrutiny, an unusual number may be interpreted as indecision. The objection is an argument for clear explanation, not for removing the Committee’s discretion, but it is also a reason for exercising that discretion sparingly.
The fourth objection concerns strategic surprise and the risk of overreaction. A 50-basis-point move, if it is subsequently seen to have been unnecessary, would impose avoidable costs on borrowers and on financial stability and could damage the reputation for competence. The scale of the real policy rate, the breadth of the inflation pattern and the strength of the real economy reduce that risk but do not remove it, which is why the evaluation of the final data before the meeting is essential.
The fifth objection concerns artificial intelligence, and here the events of the autumn add a consideration. The technology that offers better analytical tools is also contributing to the economic disturbance that the tools must interpret, and models trained on earlier regimes may be least reliable when they are most needed. The solution is neither to reject the technology nor to defer to it, but to subject it to independent validation, human oversight and continuing scepticism about whether the relationships it has learned still hold.
A sixth objection, specific to the present, concerns political perception. A decisive December increase after a hold on the eve of an election could be read in two ways: as a sign that the Committee was constrained by the electoral calendar before and is free of it now, or as a sign of the ordinary operation of a data-dependent institution. The only reliable defence against the first reading is consistency, which requires that the Committee’s actions follow from its stated objectives whether or not they are politically convenient.
XI. Conclusion: From Calendar-Based Policy to Adaptive Monetary Governance
The central issue in monetary-policy design is not whether central banks should meet six, eight, twelve or fifty-two times a year, nor whether interest rates should invariably change by 25 basis points. These are secondary questions, to be evaluated by their contribution to the quality of decisions. The primary question is whether an institution can interpret changing conditions, distinguish information from noise, anticipate the strategic behaviour of other agents and act at the appropriate time and magnitude.
The evidence of 10 November 2026 supports that framing. A third consecutive month of core inflation at 0.3 percent, a strong employment report, solid growth, a real policy rate close to zero and a surge in long-term yields collectively imply that the Committee’s next move is likely to be an increase and that its size is a matter of judgment. The Federal Reserve’s decision to reduce the number of decision meetings reflects a valid insight about the distinction between deliberation and intervention, but it has not yet been matched by an intermeeting protocol or a communications framework adequate to the new calendar. Governor Waller’s speech has demonstrated how conditional commitment can make a signal credible, and how a minimalist centre can leave individual members to perform the function of guidance. And the gap in the official inflation statistics has shown why artificial intelligence, however valuable, cannot replace the integrity of the underlying data or the exercise of institutional judgment.
Several developments could change this assessment. A decline in wholesale-price inflation and in the Committee’s preferred measure of consumption deflation, together with weaker employment data in early December, would strengthen the case for the conventional increment and for patience. A further acceleration, or evidence that inflation expectations are becoming unanchored, would strengthen the case for a larger move and for a more explicit commitment to further tightening. A financial dislocation, particularly one tied to the technology sector or to the bond market, would introduce the nonlinear scenario that no calendar and no convention can adequately anticipate.
The ultimate objective should be to replace mechanical procedural rigidity with disciplined institutional flexibility. This does not mean abandoning rules, transparency or accountability. It means ensuring that conventions remain subordinate to the purposes they were created to serve. The future of monetary policy should not be defined by how often central bankers meet or how many basis points they add or subtract. It should be defined by how effectively they interpret uncertainty, anticipate change and exercise judgment, and by whether they can explain that judgment to those who bear its consequences.
Selected Research and Institutional References
Academic references are those of the original draft that can be verified as published scholarship; the 2026 sources are the primary statistical and policy releases and contemporaneous reporting relied upon for the evidence of November 2026. Market-implied probabilities vary across sources by time of observation.
1. Brainard, William C. (1967). Uncertainty and the Effectiveness of Policy. American Economic Review, 57(2), 411–425.
2. Faust, Jon, and Lars E. O. Svensson (2001). Transparency and Credibility: Monetary Policy with Unobservable Goals. International Economic Review, 42(2), 369–397.
3. Blinder, Alan S., Michael Ehrmann, Jakob de Haan and David-Jan Jansen (2024). Central Bank Communication with the General Public: Promise or False Hope? Journal of Economic Literature, 62(2), 425–457.
4. Hubert, Paul, and Fabien Labondance (2021). The Signaling Effects of Central Bank Tone. European Economic Review.
5. Bank for International Settlements (2023). Statistics Shape the Setting of Monetary Policy.
6. Lubik, Thomas A., and Jacob Titcomb (2026). Data Revisions and Uncertainty: How They Affect Policy. Federal Reserve Bank of Richmond, Economic Brief 26-01. [Carried from the original draft; to be re-verified before circulation.]
7. Assessing the Credibility of Central Bank Signals: The Case of Transitory Inflation (2022). Economics Letters, 220, 110875. [Carried from the original draft; to be re-verified before circulation.]
8. Central Bank Communication of Uncertainty (2025). Journal of International Money and Finance, 157, 103385. [Carried from the original draft; to be re-verified before circulation.]
9. Board of Governors of the Federal Reserve System (28 October 2026). Federal Reserve Issues FOMC Statement. Press release.
10. Federal Reserve Board. Federal Open Market Committee: Meeting Calendars, Statements and Minutes.
11. Bank of Canada. Monetary Policy Decision-Making Process.
12. U.S. Bureau of Labor Statistics (10 November 2026). Consumer Price Index: October 2026.
13. U.S. Bureau of Labor Statistics (6 November 2026). The Employment Situation: October 2026.
14. Waller, Christopher J. (10 November 2026). Remarks at the Bank of England Watchers Conference, London, as reported by Axios, The Hill and CNBC.
15. CNBC (10 November 2026). Consumer prices rose 0.4% in October, more than expected; 10-year Treasury yield races back to 5%.
16. Axios (10 November 2026). Price pressures persisted in October; Get ready for rate hike(s).
17. CNN (10 November 2026). US inflation rose more than expected last month.
18. Reuters (28 October 2026). Federal Reserve Chairman Kevin Warsh announces changes in how future FOMC meetings will go.
19. CNBC (28 October 2026). Fed interest rate decision, October 2026: Fed holds rates steady.
20. Associated Press and CBS News (6 November 2026). Employers added 165,000 jobs in October; unemployment rate dipped to 4.1%.
21. J.P. Morgan Asset Management. Fewer Federal Reserve meetings under Kevin Warsh: what it could mean for investors.
