Subtraction Is Reading As Progress
The Decision Signal is a curated executive intelligence brief built to sharpen how decision makers see, think, and act. In this brief, three signals show how three headline numbers improved for reasons that had little to do with the performance behind them, and what that costs a plan already in motion.
Friday's employment report did not start anything. It removed the last reading that made the year look intact. Payrolls fell 23,000 in July, and the Bureau of Labor Statistics took another 103,000 out of May and June. Private payrolls rose 30,000; the headline decline came from a 49,600 drop in local government education that may prove to be a seasonal adjustment artifact. Underneath it, participation has slipped since January to 61.4% and job growth has averaged 34,000 a month across the past year.
The unemployment rate sat at 4.3% or above from January through May. It fell in June, to 4.2%, and again in July, to 4.1%, while household employment dropped 87,000 and the labor force contracted 264,000, taking 177,000 out of the unemployed count. Two consecutive drops, both on falling participation, and both read as recovery.
Unemployment continues to fall for the wrong reasons.
Daniel Zhao, Chief Economist, Glassdoor, August 2026
The question is how long the numbers have been saying this.
- Labor: Participation fell to 61.4% and took the rate down with it.
- Demand: Wendy's revenue rose 1.7% while U.S. same-store sales fell 7%.
- Earnings: S&P 500 earnings growth falls from 50.4% to 32% without two companies.
The unemployment rate fell again in July, to 4.1%, in a month when payrolls dropped 23,000 and 264,000 people left the labor force. It was the second straight month the rate fell because participation did.
Payrolls fell 23,000, and the Bureau of Labor Statistics cut May by 66,000 and June by 37,000. Private payrolls rose 30,000, and the headline loss came from a 49,600 decline in local government education that may reflect seasonal adjustment. Household employment fell 87,000 while the labor force shrank 264,000, which took 177,000 out of the unemployed count and pulled the rate down. Participation fell to 61.4% from 61.8%, and is down 0.7 points since January. Wage growth slowed to 3.2%. Every one of those measures has been public each month. The one that leads the coverage is the only one that improved.
↗ Bureau of Labor Statistics — Employment Situation Summary, July 2026Wendy's grew revenue 1.7% in the second quarter while U.S. same-store sales fell 7%. It cut the dividend and withdrew 2026 guidance the same morning.
Revenue reached $570.6 million, up 1.7%, while U.S. same-store sales fell 7% and U.S. systemwide sales fell 8.2%. Adjusted EBITDA dropped 15.4% to $124.1 million, the quarterly dividend was cut to $0.07, an annualized $0.28, and the company pulled its full year outlook. President and Chief Executive Bob Wright said in August 2026 that the company is clearly not performing at its potential, and named traffic, the value proposition and franchisee economics as the reasons. The gain did not come from selling more. Wendy's attributes it mainly to advertising fund revenue and to restaurants bought back from franchisees. Both changed what the revenue line counts. Same-store sales compare the same restaurants in both periods, and that measure fell 7%.
↗ The Wendy's Company — The Wendy's Company Reports Second Quarter 2026 ResultsSecond quarter earnings growth for the S&P 500 stands at 50.4%, against 23.1% when the quarter closed. Remove Alphabet and Amazon and it is 32.0%.
86% of reporting companies beat earnings estimates, above the five year average of 78%, and the average beat was 29.2%, the highest FactSet has recorded since it began tracking the measure in 2008. The gap between 50.4% and 32.0% is two companies, and much of what they contributed is investment related gains inside their GAAP results. No customer paid for any of it. Ten of eleven sectors grew and the forward multiple sits at 20.0, above both the five and ten year averages. The index is reporting an average, and an average carries whatever is heaviest in it.
↗ FactSet — S&P 500 Earnings Season Update: August 7, 2026A labor market, a restaurant chain and the earnings of the index are not the same story. Each headline improved for its own reason, and none of those reasons is what a reader would assume from the number.
The unemployment rate fell because 177,000 people stopped being counted as unemployed, most of them by leaving the labor force. Wendy's revenue rose because the line now counts advertising funds and restaurants it did not count a year ago. Index earnings rose 50.4% because two companies sit inside the average; without them the figure is 32%, which is still real growth.
Count, mix, and concentration: three ways a headline number moves for reasons that have little to do with the performance a reader infers from it. This is composition passing as condition, and every one of these figures was public when it published, in the measures operating plans are built on. Nothing was concealed and nothing needed to be. The number was never wrong. It was answering a narrower question than the one the plan asked of it.
The Decision Signal’s posture for the week of August 10, 2026 is Deploy.
All five postures are live. Click on any zone to read this week’s signals from there, and what standing in it would cost.
◇ DEPLOY: Run the operating review on units, customers and headcount this month. Show every rate and average beside the absolute count underneath it. A rate that improves while its count falls is the thing to catch, and the fourth quarter plan is built on measures that can do exactly that. Make the change before the preliminary benchmark estimate lands in August.
The executive is running a fourth quarter plan on headline measures that have moved for reasons unrelated to the performance underneath them.
The risk is the reading, not Friday's print. A plan fixed today is already late, and one fixed after the benchmark estimate lands is fixed in public.
Which parts of the fourth quarter plan rest on rates and averages, and which of those stops this month?
A number that improves and a condition that improves are two different things. Reporting rarely separates them, and planning calendars rarely ask.
The unemployment rate fell because people stopped looking for work. A restaurant company grew revenue while its comparable restaurants sold less. An earnings quarter that beat estimates by the widest margin since 2008 rests substantially on two companies and on gains that were never sold to a customer. Each figure is accurate. Not one describes what a reader would take from the headline, and the labor figures have done it two months running.
None of it was hidden. It was published monthly, in the measures operating plans are built on, and read as improvement each time. The work now is deciding what this company stops doing on the strength of information it already had.
Respectfully,
PJ Bickett
After reading this briefing, what is your immediate posture?