The economy is sending two signals at once. The Fed is listening to the louder one.

There is a version of the two weeks to August 14, 2026 that reads as good news. Consumer prices rose just 0.1% in July. Core inflation — the measure that strips out food and energy — slipped to 2.5%, which is within touching distance of the Federal Reserve's 2% objective. Every headline figure landed exactly where forecasters expected.

There is another version that reads as a warning. American employers cut 23,000 jobs in July. The two months before that were revised down by a combined 103,000, which means the spring hiring that economists thought they saw largely did not happen. Retail sales fell 0.6% — the sharpest monthly drop in more than a year.

Both versions are true. That is the problem.

What the numbers actually say

Take the inflation report first. Headline CPI rose 3.4% over twelve months, one tenth lower than June's 3.5%. On the month it rose 0.1%. Core inflation rose 0.2% on the month and 2.5% over the year.

That distribution is the entire story. If you remove energy from the calculation, US inflation is running close to the Fed's target. If you leave it in, inflation is running a full 1.4 percentage points above it. Households do not get to remove energy from their budgets, which is why headline inflation is what people feel and core inflation is what central banks tend to act on.

The labour market went the other way

On August 7, the Bureau of Labor Statistics reported that nonfarm payrolls fell by 23,000 in July. The unemployment rate held at 4.1% and average hourly earnings rose 3.2% over the year — neither of which looks like a recession.

The revisions did more damage than the headline. May's job gain was cut from 129,000 to 63,000. June's was cut from 57,000 to 20,000. Together, 103,000 jobs that the market believed existed in early summer no longer do.

Underneath the total, the composition is unusual. Local government education shed 50,000 positions. Retail trade lost 19,000 and financial activities lost 14,000. Health care, which has carried US job growth for years, added 22,000 — and was almost the only sector doing so.

One number in that report deserves more attention than it gets: the labour force participation rate is 61.4%, down 0.7 percentage points since January. A falling participation rate can hold the unemployment rate down without a single person finding work, because people who stop looking stop being counted as unemployed. A 4.1% unemployment rate alongside falling payrolls and falling participation is a weaker picture than 4.1% alone suggests.

Then the consumer stopped

On August 14, the Census Bureau reported that retail and food services sales fell 0.6% in July to $763.6bn, the largest monthly decline in more than a year. Excluding autos, sales fell 0.3%. The control group, which strips out autos and gasoline and feeds directly into GDP calculations, fell 0.2%.

And the Fed voted to do nothing — over three objections

On July 29, the Federal Open Market Committee left its target range for the federal funds rate at 3.50% to 3.75%, with interest on reserve balances at 3.65%. The vote was 9–3.

The direction of the dissent is what makes this meeting unusual. Beth Hammack of the Cleveland Fed, Neel Kashkari of Minneapolis and Lorie Logan of Dallas all preferred a 25 basis point increase. Not a cut. In an economy shedding jobs, three regional Fed presidents wanted money to be more expensive.

Especially prudent at these uncertain times.Fed Chair Kevin Warsh, describing the decision to hold, July 29, 2026

Warsh, who took the chair earlier in 2026, characterised the three dissents as a healthy "family fight," said the Fed would not rely on a "magic wand" to bring inflation down quickly, and pointed out that the committee's eight weeks under his leadership were being measured against five years of elevated prices.

Why the two signals do not cancel out

The textbook version of monetary policy assumes the two halves of the Fed's mandate move together. When the economy weakens, inflation falls, and the central bank can cut rates to support employment without endangering price stability. Everything is a single dial.

That is not the configuration in front of the committee. The weakness is in employment and consumption. The inflation is in energy — a supply-side pressure that higher interest rates address only indirectly and slowly. Raising rates to bring down the price of oil means suppressing demand across the entire economy to affect one input. Cutting rates to support hiring risks validating the idea that 3.4% is the new normal.

What the market thinks happens next

After the CPI release, futures traders put the probability of the Fed holding rates in September at 64%, up from 52% the previous day, according to CME FedWatch. Note the framing: for most of this cycle, the question priced into that market has not been the size of the next cut. It has been whether the next move is a hike.

That expectation now has three scheduled tests before the September 16 decision:

What would change this story

• Energy rolls over. Gasoline is up 24.6% year over year. If crude — $82.40 on Friday, August 14, up 33% in twelve months — retreats, headline inflation falls mechanically over the following months and the hawkish case loses its main exhibit.
• Core reaccelerates. Core at 2.5% is what makes a hold defensible. If August core turns back up, the three dissents become a bloc rather than a protest.
• The labour market cracks properly. A negative print alongside a rising unemployment rate — rather than a flat one — changes the balance of risks immediately.
• The consumer keeps spending. Home Depot, Target, Lowe's and Walmart all report in the week of August 17. Four strong quarters would reframe July's retail sales drop as noise.