The Precedent

Stephen Miran, appointed to the Federal Reserve Board and holding a vote on the Federal Open Market Committee, chose a digital asset summit as the venue to explain his dissent. The occasion was Blockworks, March 2026. The week prior, at the most recent FOMC meeting, Miran had voted for a 25 basis point rate cut — a position that put him in the minority. His audience at Blockworks was not composed of academic economists or central bank staff. It was composed of market participants with direct financial exposure to the interest rate path Miran was publicly advocating.

The juxtaposition is worth holding. A sitting Federal Reserve governor, having formally registered his disagreement with the committee's decision, traveled to a forum whose constituency stands to benefit from easier monetary policy to lay out, at length, the analytical architecture behind his vote. The setting did not appear to constrain him. Miran spoke with the fluency of someone who had worked through the argument carefully and arrived at conviction.

That argument is historical in its construction. Miran does not rest his dissent on current data alone. He reaches back, repeatedly, to specific episodes that he frames as precedents — moments when the Federal Reserve made accommodation judgments that he regards as instructive for the present situation. The Greenspan era's bet on productivity growth. The institution's long-standing doctrine of looking through supply-driven price increases. The disinflationary effects of deregulation. The global savings glut that depressed neutral rates across a prior decade. And, finally, the measurement properties of PCE inflation itself.

Each invocation carries a specific factual claim — about what the Fed actually did, what it knew at the time, and what the outcomes were. The FOMC Insight Engine contains transcripts, staff memos, speeches, and internal briefings that document the institution's own account of each of these episodes.

We went into the archives.

The Bet They Remember Differently

The first of Miran's five historical pillars is also the most load-bearing. Miran invokes the Greenspan productivity story of the 1990s as a validated precedent: when technology expands productive capacity, demand can grow faster without generating inflation, and monetary policy need not respond to apparent overheating. In his framing, "that argument makes a lot of sense in the context of a telecoms revolution" — and by extension, it should make equal sense today when AI and deregulation are doing the expanding. The proposition is stated cleanly. The archive's version of that episode is considerably less clean.

Internally, the Greenspan productivity accommodation was not an institutional consensus. It was a Chairman-driven bet that overrode the staff's quantitative models — and the Chairman said so explicitly. At the October 1999 FOMC meeting, Greenspan told the committee:

The framing matters. Greenspan was not refining staff estimates — he was dismissing the epistemic basis for them. The models that tracked labor market slack, NAIRU, and wage pressure were declared unfit for a new environment. What replaced them was the Chairman's qualitative read of productivity dynamics that conventional measurement could not yet confirm.

This created a problem the archive documents with precision. Vice Chair Alice Rivlin identified it at the September 1997 meeting, naming the epistemological trap the committee was already entering:

Rivlin was not offering a theoretical concern. She was describing the committee's actual operating logic — that the productivity narrative had become self-sealing, capable of absorbing any contradictory signal as simply the old models failing to capture the new reality. By early 1999, Michael Prell, the Director of Research and Statistics, was flagging "a certain euphoria that one senses in the markets" and warning that investors might be "exaggerating the safety of equity investments" in ways that could produce a violent reversal if any shock disturbed their complacency. Those warnings were documented, filed, and overridden.

Publicly, none of this internal fracture was visible. The New Economy thesis reached external audiences as a coherent institutional judgment — a validated framework, not a contested gamble. The gap between what the committee's own staff was recording and what the public received is one of the most significant internal-external divergences in the transcripts from this period. The institution knew more than it communicated, and what it knew was cautionary. (The pattern recurs: as The Assay documented, Judy Shelton borrowed this same Greenspan authority to authorize the same kind of supply-side accommodation — and the gap between the public narrative and the internal record was what made the borrowing possible.)

Beneath the divergence lay a problem the Chairman acknowledged in real time. At the June 1999 meeting, Greenspan conceded: "At the last meeting, I thought I saw signs that the surge of cost-cutting and hence productivity acceleration were beginning to crest. I was mistaken." The accommodation justified by a productivity thesis the Chairman could not reliably measure was already producing the equity market conditions Prell had warned about.

The subsequent equity market correction coexisted with the conditions staff had warned about, though attributing the bubble directly to the productivity accommodation is a causal claim the archive does not resolve.

The archive's more durable finding is not about the outcome — it is about the process. The internal record was not one-sided: senior participants including Edward Boehne and Robert McTeer publicly endorsed the productivity framework, and the staff itself adopted a somewhat more positive view of structural efficiency gains by mid-1999. The episode was a genuine institutional bet, not a Chairman acting alone against unanimous opposition — which makes the lack of resolution more, not less, relevant to Miran's appropriation of it as settled precedent. The Greenspan episode was not a validated framework available for later appropriation. It was a documented case of a Chairman proceeding despite documented internal dissent, on the basis of productivity estimates he acknowledged misjudging as they were occurring. If the precedent failed precisely because real-time measurement of supply-side capacity was unreliable, then any framework that depends on that same real-time measurement inherits the structural flaw — and the borrowed authority collapses before the case it was meant to support can stand.

The Doctrine That Required Earning

The collapse of the Greenspan productivity analogy forces Miran onto a second footing. Even if the capacity-expansion story cannot bear the weight placed on it, there remains what he calls "the classic reasoning of why a central bank should look through an oil shock as the Federal Reserve has historically done" — an appeal not to a specific episode but to a general institutional practice. The claim carries the structure of doctrinal authority: this is how central banks have always operated. The archive disagrees, and the disagreement is not marginal.

The starting point is 1973. When the Arab oil embargo struck, the Federal Reserve under Arthur Burns accommodated the shock rather than tightening to contain it. The staff's internal analysis was ambivalent from the beginning — torn between the recessionary consequences of restraint and the inflationary consequences of accommodation:

The warning was documented. It was also filtered. Public communications reframed the accommodation as prudent crisis management — calibrated policy in an era of unusual disruption. What the internal record shows is harder to defend: the institution recognized the inflationary risk of its own stance and proceeded anyway, protecting output at the cost of anchored expectations. Burns-era accommodation is widely regarded as having worsened inflation dynamics in this period, though the subsequent Great Inflation reflected multiple drivers — successive oil shocks, fiscal pressures, and wage-price spirals that monetary policy alone could neither produce nor fully contain. What the archive assigns to the accommodation is not sole causation but a consequential failure to arrest the spiral before those dynamics became self-sustaining.

Volcker arrived at the chair in 1979 as the second oil shock — the Iranian revolution — reached markets. His response was the institutional refutation of Burns's approach. Where Burns had treated the oil price increase as something to absorb, Volcker treated it as a test of credibility the institution had already squandered. In November 1979, challenging a staff forecast that appeared to minimize pass-through, he pressed: "Are you saying a 40 percent increase in imported oil prices will bring a 1-1/2 percent increase in GNP prices?" The skepticism was not rhetorical. He was not looking for permission to look through. He was establishing that the institution could no longer afford to.

Priority — not balance, not calibration. The tightening that followed was severe. It produced a deep recession and, eventually, the restoration of credibility that made the modern framework possible. The modern framework's capacity to look through supply shocks rests entirely on what Volcker purchased at that cost.

This is the history Miran's framing erases. By the Bernanke era, staff analyses had codified the specific preconditions that distinguish valid look-through from dangerous accommodation: anchored expectations, credible commitment, and the absence of second-round wage effects. This codification is reflected across transcripts, staff memos, and Monetary Policy Reports from the 2003–2019 period — an analytical index of cross-layer documentary alignment that the Engine scores at 0.85, a metric measuring the consistency of precondition language across institutional registers rather than the frequency of a single quoted formulation. The preconditions themselves appear throughout that record, though the Engine's convergence score is a derived analytical measure and not itself a directly quotable finding. The institution learned the lesson so thoroughly that it permeates every documentary register. Miran invokes the language of that converged modern doctrine and names its preconditions — anchored forward inflation expectations, a cooling labor market, no wage-price spiral. The archive's question is not whether he named them, but whether naming them is sufficient. The Bernanke-era record shows that the committee treated those preconditions not as indicators to be cited but as conditions to be earned and continuously demonstrated. As Sandra Pianalto warned in June 2007, even when expectations appeared anchored, "large and persistent price disturbances" warranted sustained attention precisely because the anchoring was a credibility bet rather than a verified structural fact — and David Wilcox's own staff acknowledged that the 2005–2008 look-through rested on a "judgmental assumption" of limited pass-through that the institution later recognized as a process failure. Miran's favorable readings on each precondition are not in dispute. What the archive establishes is that the Bernanke-era standard required a demonstrated track record of credibility maintenance, not merely a contemporaneous assertion that the relevant conditions obtain. Whether the current environment meets that standard is the open question his dissent does not resolve. If the doctrine is conditional rather than default, and the historical practice is cautionary rather than validating, the supply-side accommodation case rests on a single remaining pillar: the specific supply mechanism itself — deregulation.

The Supply They Couldn't Count

With the historical precedent contested and the look-through doctrine conditional on earned credibility, the specific mechanism Miran advances must now carry the supply-side argument alone. That mechanism is deregulation — a "deregulatory wave" that would, in his construction, "drag on inflation by about half a percent a year for the next few years." The analogy he offers is precise: "if you've got a smokestack that you can run eight hours a day with strict carbon regulations and then the carbon regulations ease and you can run it 16 hours a day you don't need to do additional investment." Existing capacity unlocked. No offsetting demand. Persistent disinflation without new stimulus. The archive treats this claim with a skepticism that runs to the institution's own research division.

Michael Prell, Director of the Division of Research and Statistics, addressed deregulation's price effects at the height of the New Economy expansion that Miran's framework treats as the template:

One-time shocks. The phrase is precise in its damage. A 30-50 basis point persistent annual effect requires a continuous disinflationary channel — regulatory easing that generates lasting downward pressure on price levels, year after year. What Prell's division observed were discrete, non-repeating adjustments: prices fall when a regulatory constraint lifts, then the effect exhausts itself. The smokestack runs longer hours, but that transition happens once. The model cannot sustain the persistence Miran requires.

The archive is not, however, one-sided on this question. Greenspan argued publicly and internally that deregulation may have done more to restrain inflation than the institution's models could readily document — precisely because the efficiency gains were diffuse and accrued gradually across the economy rather than appearing as discrete, measurable inputs. Laurence Meyer, a Federal Reserve Board governor during the late New Economy expansion, estimated that the productivity acceleration associated with structural improvements including deregulatory reforms was lowering inflation by approximately 1.5 percentage points at its peak. These are not peripheral observations. They reflect genuine analytical disagreement within the institution about whether the disinflationary channel was a level shift or a persistent flow — and that disagreement was never resolved to consensus. The archive shows a mixed record, not settled skepticism.

This ambiguity matters more because Miran cites contemporary research, not only the 1990s precedents. A 2026 Federal Reserve staff paper — Garcia and Yakov — estimates a persistent 0.3 percent annual inflation drag attributable to current deregulatory initiatives.

The archival evidence assembled here predates that research and cannot evaluate its identification strategy or the regulatory domain it models. What the archive can establish is the distinction that older evidence does and does not address: Prell's one-time-shock characterization describes the price-level channel of product-market deregulation — the effect of lifting a specific regulatory constraint on a specific class of goods or services. That framing does not automatically govern the question Garcia and Yakov appear to examine, which may involve a different scope and sequencing of regulatory change. Miran's 0.5 percent estimate may or may not survive contact with the archive's skepticism about persistence; on the basis of the 1999 internal record alone, the claim is contested rather than settled.

Publicly, Greenspan framed deregulation as fostering "competitive market efficiencies" that restrained inflation — macro language that built the New Economy narrative and traveled widely. Internally, staff models could not validate that framing as a quantifiable, ongoing channel. The institution's public storytelling and its analytical record diverged at precisely the point that mattered most for policy. Miran inherits Greenspan's public framing; the internal record belongs to Prell — and the honest summary of the archive on this pillar is not refutation but a mixed record that does not cleanly support Miran's specific quantitative claim.

The financial deregulation cases raise a separate analytical question. Governor Martha Seger raised the alarm from inside the building as the thrift crisis deepened:

Robert Forrestal, President of the Federal Reserve Bank of Atlanta, had noted two years earlier that rising bank failures were already prompting concern about systemic soundness and its potential repercussions across the broader economy. Seger's warning and Forrestal's concern operate on a distinct channel from the supply-side price mechanism Miran invokes. The S&L crisis and the fragility it exposed were consequences of financial deregulation — the removal of constraints on depository institutions' asset allocation and risk-taking — rather than of product-market or regulatory-burden deregulation of the kind the smokestack analogy describes. Treating one as a direct rebuttal to the other requires an analytical bridge the archive does not provide: financial stability risk and supply-side price-level effects are separable welfare questions, and the institutional record addresses them in separate registers. What the S&L episode does establish is that deregulatory episodes can simultaneously produce some of the efficiency gains Greenspan and Meyer identified while also unlocking risk that required emergency intervention and constrained subsequent policy space. The net welfare assessment of a specific deregulatory wave depends on which channels dominate — a question the archive flags as unresolved rather than settled in either direction. If the supply-side architecture is contested — precedent uncertain, doctrine conditional, mechanism disputed — Miran requires an alternative basis for arguing policy is too tight. He has two remaining pillars: the claim that neutral rates are structurally depressed, and the claim that inflation measurement itself overstates the true price level.

The Inflow They Came to Regret

With the supply-side architecture dismantled, Miran's case for easier policy pivots to a separate foundation: the neutral rate itself may be structurally depressed, making current policy tighter than it appears regardless of supply dynamics. Miran advances several channels for this structural depression. Two of them — declining working-age population growth and the long-run path of fiscal consolidation — engage a substantial academic and internal Federal Reserve literature. Internal deliberations during the 2010s, including staff reviews of the Laubach-Williams and Holston-Laubach-Williams models, attributed between one and two percentage points of the r-star decline to demographic factors alone; Richard Clarida stated publicly in 2019 that "most papers estimate that demographics can explain between 1 and 2 percentage points of the decline in r*," and the archival record contains detailed quantification of fiscal channels as well. Those arguments are not assessed here. The archive predates the current demographic and fiscal configuration, and testing them would require a separate analytical framework.

The third channel Miran advances is stablecoin-driven capital inflow, and it is this channel — not the demographic or fiscal arguments — that the archive can speak to directly. "if you have huge inflows from the rest of the world into US dollar denominated savings, that's going to weigh on the neutral rate" — and optimistic projections for stablecoin adoption could produce flows "maybe not quite as big as the global savings glut, but let's say half as big." The archive contains a detailed record of what the savings glut actually was, how it operated, and what the institution concluded about it afterward. What follows tests the savings-glut analogy specifically. The archive's finding on this channel does not support the analogy.

The savings glut, as Bernanke's own staff documented, was not diffuse retail capital seeking dollar assets. It was concentrated, institutional, and sovereign — $715.3 billion in foreign official purchases representing East Asian central bank reserve accumulation, petrodollar recycling through OPEC sovereign vehicles, and current account flows intermediated through state balance sheets. The mechanism required sovereign-scale institutional infrastructure operating across years of sustained current account imbalances. Stablecoin demand operates through crypto-native infrastructure and retail channels. The analogy fails before the question of scale is even reached.

The archive's more consequential finding concerns what the savings glut did to the economy that received it. Staff identified the risk years before the crisis. As early as September 2004, the Bluebook flagged it internally:

Yield compression from foreign official flows was feeding the housing bubble three years before the crisis made the connection undeniable. In May 2007, Bernanke acknowledged publicly that "financial markets are priced for perfection, which implies some risks on that score" — while concluding there was "yet no indication of significant spillover from housing." The internal warning had been legible for years; the public posture remained sanguine. By March 2007, the spillover was already in motion. Janet Yellen, then President of the San Francisco Fed, named what the institution's benign framing of the glut had permitted:

The archive shows that the Fed recognized — at least internally — that depressed real rates could fuel housing misallocation, and multiple participants later linked these dynamics to financial stability risk; however, an unambiguous institutional consensus attributing the crisis to the savings glut is not established in the documentary record. What is established is that the savings glut compressed yields, and was identified by multiple participants as a contributing factor in the dynamics that preceded the financial crisis. Miran invokes it as a model for what stablecoin inflows might replicate. The archive treats it as a cautionary demonstration of what unchecked yield compression produces. This is an analytical distinction, not an archival finding: the savings glut operated through sovereign balance sheets and institutional flows — concentrated, sustained, and intermediated through state infrastructure — whereas stablecoin demand, by contrast, appears to flow through different infrastructure and counterparty channels, though contemporary data on stablecoin transmission to Treasury term premia remains limited. The analogy fails before the question of scale is even reached, but the failure rests on structural inference rather than archival evidence about stablecoin flows specifically. With every historical analogy now either mechanistically inapplicable or retrospectively cautionary, his dissent arrives at its final claim: that the inflation data themselves overstate the true price level.

The Number They Already Fought Over

Where the prior four claims rested on historical analogy, this one rests on arithmetic. Miran argues that "a lot of the inflation excess over target has been a result of some quirks of measurement of things like portfolio management services which is basically just the stock market going up biasing the way we measure inflation up by 30 to 40 basis points." The mechanism is precise: portfolio management fees in PCE are imputed from equity returns, so when stock prices rise, measured services inflation rises with them, independent of any change in what households actually pay. No historical parallel is required here. The bias Miran identifies is real.

The archive confirms that imputed components in PCE — including healthcare services and financial services — can shift measured inflation by several tens of basis points, consistent with the order of magnitude Miran identifies. Staff analysis from a November 2017 memo to the committee documented that the healthcare services category alone was contributing about 30 fewer basis points to core PCE than during the 2002–2007 period — a gap large enough to shift the committee's assessment of whether policy was tight or accommodative. Miran's specific mechanism concerns portfolio management services tied to equity returns; the archival examples most directly documented here involve healthcare services and broader imputation methodologies. The measurement concern is real; the specific channel Miran isolates requires additional documentation. These decompositions were not academic exercises. They were policy inputs. When David Wilcox, then Director of the Division of Research and Statistics, summarized the implications for the committee, he translated the arithmetic directly into a rate-setting conclusion:

The archive documents a broader class of measurement anomalies — healthcare-services imputation, imputed banking services, CPI-to-PCE revisions — that establish the institutional precedent for taking bias claims seriously. Miran's specific channel, portfolio-management services tied to equity-market gains, operates through a different mechanism than those the archive directly documents: where the archival cases involve BEA methodological revisions to non-market service prices, Miran's channel links asset-market returns directly to imputed financial-services costs in a way the transcripts address only obliquely. The existence question and the policy-use question are therefore separable. On the existence question, James Bullard identified in June 2017 that rising equity prices were having a "major effect on this index" through precisely the imputed financial-services channel Miran describes, and David Wilcox estimated a specific 9-basis-point shortfall in core PCE attributable to non-market price components that month — evidence that the mechanism was understood internally even if not fully codified. On the policy-use question, Wilcox's own translation of the arithmetic to the committee carries the same inferential structure Miran employs: a known measurement bias implies that the policy rate consistent with the true inflation target is lower than the headline reading suggests. The staff said it. The committee heard it. (As The Measure documented, the institution's relationship to its own inflation statistics has been more fraught than either Miran's confidence or the committee's public certainty would suggest — the choice of gauge has at times served institutional convenience as much as analytical rigor.)

What is in dispute is the direction of inference. Working with the same measurement framework, Janet Yellen reached the opposite conclusion about what revisions revealed:

The same data, the same methodology, the same revision process — and two entirely different policy implications drawn from identical analytical material. One reading supports accommodation; the other demands restraint. The archive does not resolve which is correct, because the committee never did.

Governor Elizabeth Duke identified the institutional consequence of that unresolved disagreement:

Duke was describing a committee that had access to exactly the measurement research Miran cites and still could not reach consensus on what the adjusted numbers meant for policy. The 30 to 40 basis points Miran identifies fall within the range the institution's own staff quantified. Miran's technical sophistication is not the question. What the archive establishes is that the step from "the number has a known bias" to "therefore inflation is not overly problematic" is not a technical inference but a contested policy judgment — one the institution has already fought over without resolution — which means presenting it as settled analytical ground is precisely the error the committee, in its own deliberations, identified as the practice that fractures credibility.