A plain-English guide to reading FOMC minutes.

The Federal Reserve publishes minutes of its policy meetings three weeks after each decision. Tomorrow's release covers the June 16-17 meeting. The paired analysis today, Yields into FOMC minutes, reads the rate market as pre-positioning for a hawkish read. Whether the read is correct depends on how the minutes are constructed and what to actually look for inside them. This piece is the manual.

What the minutes are

The FOMC minutes are the official record of the committee's policy discussion, approved by the committee and released three weeks after each meeting. They are not a verbatim transcript (that publishes with a five-year lag) and they are not a summary of decisions (that is the statement, released the day of the meeting). The minutes sit in between: a structured record of what the committee discussed and how the discussion led to the decision, including specific language used by the participants.

The minutes are collectively drafted by the Board staff under the direction of the committee secretary, then circulated for committee review and formal approval at the subsequent meeting. Any participant can request changes to language that mischaracterises their views. The final document is committee-owned, not staff-owned, which is why the minutes carry more weight than any single participant's subsequent speech.

The six sections that matter

The minutes have a standard structure. Reading them efficiently means knowing which sections carry policy signal and which are procedural.

Section one: Developments in Financial Markets and Open Market Operations

The staff market report. This section describes what financial markets have done since the prior meeting, what the desk has done in open market operations, and how the balance sheet has evolved. It is descriptive, not prescriptive. Read it for what the committee thinks the market has been telling them.

Language patterns to watch: any specific naming of market dislocations (a "stress in the repo market" mention is not boilerplate) or references to the Fed's counterparties facing difficulties. These flag operational concerns that can drive standing-facility decisions.

Section two: Staff Review of the Economic Situation

The Board staff's summary of the economic conditions the committee is looking at. Not the committee's view; the staff's. It matters because the staff sets the baseline against which committee members' views are then contrasted.

Language patterns: the staff typically hedges with words like "appear," "seem," or "consistent with." Departures from that hedge language (using "clear" or "material") are signals of staff conviction, which typically foreshadows subsequent committee language.

Section three: Staff Economic Outlook

The Board staff's projections for growth, unemployment, and inflation over the forecast horizon. This is the staff's forecast, and it feeds directly into the SEP that participants use to construct their own dot projections.

Language patterns: the staff's forecast is usually modestly more optimistic on growth and inflation than the committee median. A staff forecast that has moved more hawkish or dovish than the prior meeting's is a strong directional signal, because it typically leads committee views by two to three meetings.

Section four: Participants' Views on Current Conditions and the Economic Outlook

This is the most important section for policy signal. It summarises what the eighteen participants (twelve voting members plus six non-voting regional Fed presidents) said during the discussion. Language is aggregated: "all participants" (unanimous), "most participants" (large majority), "many participants" (majority), "some participants" (substantial minority), "a few participants" (small minority), and "one participant" (singular view).

The specific language of aggregation matters. A move from "many" to "most" is a signal of increasing committee consensus. A move from "all" to "many" is a signal of developing dissent. A view that appears as "some" or "a few" in one meeting and "many" or "most" in the next has moved from minority to majority, which is the standard pattern for how the committee shifts stance over time.

Section five: Participants' Views on Monetary Policy

The policy-specific counterpart to section four. This section describes how participants think about the current stance, the appropriate near-term policy path, and the balance of risks. It is where hike-vs-hold vs-cut language concentrates.

Specific language patterns that shift market pricing when they appear or vanish:

• "Nearly all" vs "many" participants supporting a specific decision. A shift from nearly-all to many signals a committee moving toward split votes on the specific action.
• "Data-dependent" vs "time-dependent" language. Data-dependent implies the committee is waiting for specific indicators to move policy; time-dependent implies a scheduled path. Data-dependent language typically indicates uncertainty about the next move; time-dependent implies conviction about it.
• "Balanced" vs "risks tilted" characterisations of the risk to inflation. Balanced risks support holding stance; tilted risks (upside or downside) support directional moves.
• "Meaningful" or "sustained" progress on inflation vs "some" or "further" progress. Adjectives on progress language reveal how the committee is grading the disinflation trajectory. Weaker adjectives generally preserve more optionality for the committee's next move.

Section six: Committee Policy Action

The formal record of the decision, the vote, and any dissents. Procedural for the vote, but the dissenting statements (published at the end of the section) are substantive: a dissenter's stated reasons are the most explicit signal available of an alternative committee view.

Language patterns: dissents that name specific data releases or specific policy variables (rather than general stance) are typically more consequential; they signal that the dissenter will hold their position at future meetings unless the data they cite shifts.

Reading the June 2026 minutes: what to actually watch for

Tomorrow's release covers a meeting with unusual features: Warsh withheld his own dot on the SEP (the first chair to do so since 2012), and nine of eighteen officials projected at least one rate hike by end-2026. Both are without recent precedent. Three specific reads carry outsized weight.

The count of "some" versus "many" participants favouring a hike. Nine of eighteen dot projections in the SEP does not tell us the language weight. If the minutes say "many participants" (implying ten or more), the hike scenario is more consequential than the SEP alone suggests. If the minutes say "some participants" (implying fewer than nine, with some of the SEP hikes coming from non-voting members whose views the aggregation may weight less), the hike scenario is less consequential.

The specific inflation trajectory language. Warsh's July 1 statement said "prices are too high." If the committee language in the minutes uses similar absoluteness (words like "unacceptable," "elevated," "concerning" rather than "gradual," "mixed," "progressing") the committee is signalling less patience with the current trajectory than the May minutes suggested.

The description of the labour market's role. The June meeting predated Thursday's July NFP print of +57,000. The committee will not have seen that print. Their characterisation of labour market conditions therefore predates the softening; what matters is whether they described the labour market as an active constraint on policy or as a data input they were monitoring. Active constraint language suggests the print will constrain future action; monitoring language suggests it feeds into the data window but does not automatically reshape the stance.

Common misreadings

Treating the entire minutes as directional signal. The minutes have thousands of words. Most are procedural, descriptive, or the staff's baseline. The directional signal is in sections four and five, concentrated in aggregation-language shifts. Reading linearly and reacting to individual sentences produces whipsaw. Reading structurally and scoring against the aggregation patterns produces signal.

Confusing dissent for stance change. A single dissenter on a hold does not shift the committee's stance. It signals that the committee has a member with a published alternative view. That view can become the consensus over multiple meetings if data supports it, or remain a permanent minority position. The stance itself shifts when the language of the majority moves.

Reading the minutes as prediction. The minutes describe a meeting three weeks in the past. Data that has released since the meeting will have shifted the committee's actual views. Tomorrow's minutes describe a committee that had not seen the June PCE print, the June NFP print, or the July Brent collapse. What the minutes reveal is the committee's baseline three weeks ago, which is a starting point for the current stance, not the current stance itself.

The Warsh factor

A specific reading complication with the current committee: Chair Warsh's communications style has moved actively away from the Powell-era practice of using minutes and speeches to guide market expectations toward pre-committed paths. Warsh's speech content has been shorter, more assertive, and less path-committing. The June statement was cut from 341 words to 130. The June SEP had a withheld chair projection. Every element of Warsh's communication has reduced signal-per-word compared to the prior chair.

That reduction in signal density means the minutes carry more weight per word than they did under Powell. There is less other Fed communication to triangulate from, so the minutes' specific language matters more relative to the baseline. Read them close.

Where this fits

The FOMC minutes framework sits alongside the other Fed communications references we have published:

• A plain-English guide to the SEP and dot plot. The quantitative counterpart to the qualitative minutes.
• A plain-English guide to forward guidance. The four eras of Fed guidance and where Warsh's style fits.
• A plain-English guide to monetary policy frameworks. The frameworks the minutes' language reveals which the committee is actually applying.

Read together, they cover the full communications stack from formal framework to individual speech, with the minutes sitting at the highest-signal-density point of that stack. Tomorrow's release will test the framework on a committee with a genuinely novel chair and a genuinely split committee. The signal will be in the language density; the reader's job is to be ready to read for it.