The clearest off-ramp may come from inflation, not slowing down —but the evidence must be broad enough to survive cross-checking.
Warsh Changed the Question
Kevin Warsh’s Jackson Hole speech was easy to read as a warning that the Federal Reserve is preparing to raise rates in September.
It was not a rate announcement.
Warsh said the Fed must be confident that “underlying inflation is moving to our objective, clearly and at sufficient speed.” But he ended by saying he was committed to “a discipline, not to a decision.”
Warsh did not mention a September hike. He made the condition for waiting much clearer.
The question is no longer simply whether inflation produced one better-than-expected report. It is whether the evidence is strong enough to persuade Warsh—and enough of the FOMC—that underlying inflation is genuinely converging toward 2%.
At first, I thought the weakening labor data might provide the clearest reason to wait.
After examining layoffs, claims, hiring and hours, that argument became less convincing.
Then I turned back to inflation. There, the evidence changed the conclusion twice.
Inflation Looks Hot—Until You Look Underneath
The hawkish case begins with numbers that are difficult to dismiss.
The PCE price index increased 3.7% over the 12 months through July. Core PCE increased 3.3%. https://www.bea.gov/news/2026/personal-income-and-outlays-july-2026
Warsh’s concern is broader than either headline.
In his Jackson Hole speech, he noted that 54% of the goods and services in the PCE basket had increased more than 3% over the previous year. That was well above the 32% share recorded across the two decades before the pandemic.
My first reading was therefore straightforward: inflation remained too high and too widespread for one favorable monthly report to change the Fed’s direction.
Then the Dallas Fed’s trimmed-mean measure complicated that conclusion.
The 12-month Trimmed Mean PCE rate was only 2.3% in July. Its six-month annualized rate was also 2.3%. https://www.dallasfed.org/research/pce
If those were the only underlying-inflation measures available, the case for another hike would look considerably weaker. Inflation would appear much closer to target than the 3.7% headline suggests.
But the cross-check could not stop there.
In August, Dallas Fed researchers examined an alternative trimmed-mean methodology. For the 12 months ending in June, the original measure was 2.2%, while the alternative was 2.6%. https://www.dallasfed.org/research/economics/2026/0813
The 2.6% figure is a June estimate, not a July update. It also represents the researchers’ analysis rather than an official Dallas Fed policy judgment.
Still, the methodological warning is there. The researchers concluded that the evidence was not clear-cut, but that the alternative measure might deserve greater weight while the distribution of price changes remains unusually skewed.
That changed my interpretation again.
Underlying inflation is probably not running at 3.7%. But the evidence is not strong enough to declare that it has already returned to 2% either.
The most reasonable working interpretation may be somewhere in the mid-2% range. That is a GMS judgment, not a single official inflation measure.
This is the Fed’s dilemma.
The headline may overstate underlying inflation.
The most dovish underlying measure may understate it.
The Labor Market Looks Weak—Until You Look Underneath
July payrolls initially appeared to offer Warsh a clear reason to wait.
Nonfarm payroll employment declined by 23,000, and employment gains in the two previous months were revised lower. The unemployment rate stood at 4.1%. https://www.bls.gov/news.release/empsit.nr0.htm
That is unmistakably weak hiring momentum.
But a weak payroll report is not the same thing as a labor market entering broad contraction.
The average private-sector workweek and aggregate weekly hours were unchanged in July. Preliminary June JOLTS data showed approximately 7.4 million job openings and 5.3 million hires, while the layoffs and discharges rate remained at 1.1%. https://www.bls.gov/jlt/
The latest initial jobless claims add another piece. The claims were 203,000 in the week ending August 22. https://tradingeconomics.com/united-states/jobless-claims
That is not what an economy-wide firing cycle normally looks like.
The labor market has clearly lost momentum.
Companies are no longer hiring aggressively. They have not started firing aggressively either.
“Low hire, low fire” remains the best short description.
This does not make the payroll decline irrelevant. It means the decline is evidence of weak labor demand—not yet proof of broad job destruction.
Warsh Does Not Need a Jobs Collapse
The Fed does not necessarily need unemployment to surge before inflation can fall.
According to the Employment Cost Index, wages and salaries for civilian workers increased 3.2% over the year through June. Private-industry wages and salaries increased 3.1%.
Those are still nominal wage increases, but they are far removed from the wage acceleration experienced earlier in the inflation cycle.
Warsh also made an important admission at Jackson Hole: wage growth has not been a reliable predictor of future inflation for a long time.
That weakens the simplistic sequence in which jobs must collapse, wages must collapse, and only then can inflation return to target.
Labor-market cooling can occur through fewer vacancies, weaker hiring and lower worker turnover. Companies do not need to dismiss existing employees simply because they require fewer additional workers.
That process could gradually reduce wage pressure without producing a recession.
Wages may no longer be the main problem. Prices still are.
The Economy Is Not Making the Decision for Him
Headline GDP also looks weaker than the economy underneath it.
Real GDP increased at a 1.5% annualized rate in the second quarter. https://www.bea.gov/news/2026/gdp-second-estimate-and-corporate-profits-2nd-quarter-2026
But real final sales to private domestic purchasers—which isolates consumer spending and private fixed investment—increased 4.2%.
The contrast is important.
Headline growth was soft. Private domestic demand was not.
The economy is therefore not weak enough to make the September decision for the Fed.
Nor did Warsh create the case for another hike at Jackson Hole.
At its July meeting, the FOMC held the federal funds target range at 3.50%–3.75% by a 9–3 vote. The three dissenters favored a 25-basis-point increase. https://www.federalreserve.gov/monetarypolicy/fomcminutes20260729.htm
The minutes also show that several participants supported an increase, while many thought additional tightening would probably be necessary if inflation failed to decline.
The majority had not decided to hike.
But the argument was already inside the room.
Warsh’s contribution was to make the burden of proof more explicit.
The Verdict: What Would Stop a September Hike?
A slowing down is not required to stop a September rate increase.
Two developments could plausibly do it.
First, convincing and broad-based disinflation.
One favorable CPI report would help, but it may not be sufficient. The stronger signal would be convergence across several measures: lower core inflation, further declines in trimmed or median measures, and a meaningful narrowing in the breadth of price increases.
Warsh does not need every inflation measure to reach 2% immediately. He needs credible evidence that the underlying trend is moving there clearly and fast enough.
Second, genuine deterioration in employment.
Another weak payroll report would matter more if it were accompanied by rising unemployment, higher initial claims, increasing layoffs and declining hours. That combination would indicate that the labor market was moving beyond weak hiring and into actual job destruction.
Current evidence does not yet meet either test decisively.
Inflation is cooler underneath than the 3.7% headline suggests.
The labor market is firmer underneath than the 23,000 payroll decline suggests.
The economy is stronger underneath than 1.5% GDP growth suggests.
The next evidence will arrive quickly.
September 1: July JOLTS is scheduled
September 4: the August Employment Situation
September 10: August PPI
September 11: August CPI
September 15–16: The FOMC meetings
August PCE will not be released until September 30. July PCE is therefore the final official PCE report available before the meeting.
That makes the August CPI and employment reports unusually important—but neither should be read in isolation.
A September hike can still be stopped without a recession.
But a merely good inflation report may no longer be enough.
Warsh has made the burden of proof clear: if the Fed is going to wait, the data need to give it a reason.
Again, we need to see what the Fed sees.