FOMC Minutes, July 2026 — A 9–3 Hold and a Calendar Fight

The 9–3 vote was known on decision day; the minutes add that the hawkish case circulated well beyond the dissenters, and that the chair himself opened a structural debate about how often the committee should meet. Both threads point to the same place: Jackson Hole on August 28 is now the market's next real signal.^[1][2][3]

Three dissenting votes sounds striking, and it is — but not for the reason usually given. Three-vote dissents have happened before. What makes July 2026 uncommon is that all three dissents pointed the same way: Hammack, Kashkari and Logan each wanted to *raise* the target range by a quarter point, not to cut or to wait.^[1][7]

The closest parallel is September 2016, when Esther George, Loretta Mester and Eric Rosengren dissented together, each preferring to raise the target range at that meeting. The structure is nearly identical a decade apart: a decision to hold, with three officials arguing the committee was already behind.^[4]

The distinction matters when reading commentary. A split committee where dissents cancel out — one wanting tighter policy, another looser — signals genuine uncertainty about direction. Three dissents pushing in a single direction signals something else: a bloc that agrees with each other and disagrees with the majority.^[1][7]

The September 15-16, 2026 meeting has now taken place. On the Federal Reserve's FOMC calendar, that meeting is listed with a Statement, an Implementation Note, a Press Conference and Projection Materials — so the decision and the dot plot have been published. This page does not report what the September decision was: it has not opened that statement.^[6]

Not every meeting does. The Federal Reserve collects projections «four times a year, in connection with the Federal Open Market Committee's (FOMC's) meetings in March, June, September, and December». So the October 27-28 meeting will produce a statement but no updated dot plot, while December 8-9 will have one again. Expecting projections at every meeting is a common way to misread the calendar.^[5][6]

The minutes are a separate release. On the same calendar, the July 28-29 meeting carries a line reading "Minutes: PDF | HTML (Released August 19, 2026)", while the September 15-16 entry carries no such line at the time this page was checked on 18 September 2026. That absence is what the calendar shows; it is not a statement about when the minutes will appear.^[6]

The reaction function has flipped relative to the pattern most readers absorbed in 2024 and 2025. Back then, a weak employment report was read as raising the odds of a cut. With three officials already pushing to tighten and the debate centred on whether to *hike*, the mapping runs the other way: a strong print strengthens the hawkish case, and a soft one reduces the pressure to move.^[1][7]

This is why the same headline number can be described as 'good news' or 'bad news' depending on which year's framework the writer is using. When comparing coverage across time, it is worth checking whether the piece assumes the committee is debating cuts or hikes — the interpretation of identical data reverses between the two.^[1]