The unemployment rate in the United States in July was 4.1 percent.
This is one of those numbers we have been taught to understand without being taught very much about what it means. We absorb the shorthand: four percent is good. Six percent is worrying. Eight percent is bad. Ten percent is a crisis. The number arrives on the first Friday of every month and moves markets and interest-rate expectations and presidential approval ratings, and economists appear on television to explain whether it has moved by one-tenth of one percentage point in one direction or the other, and somewhere in Washington people at the Federal Reserve study it as they decide what money should cost for more than three hundred million Americans.
In July of 2026, the number was 4.1 percent.
There is another number.
It is 24.9 percent.
That is the percentage of the American labor force that the Ludwig Institute for Shared Economic Prosperity considers functionally unemployed: people who have no job, people who want full-time work but can find only part-time work, and people whose earnings fall below what the institute defines as a basic living-wage threshold. Nearly one in four Americans in the labor force meets one of those three conditions. The Bureau of Labor Statistics counts the majority of them as employed.
And then there is another number, the one I have been unable to stop thinking about since I first encountered it. For American women in the labor force, that same measure of functional unemployment stood at 31 percent in July. Nearly one in three. Not one in three women sitting at home without a job, which would be inaccurate, and accuracy matters enormously here. One in three women participating in the American labor force is either unemployed, unable to find the full-time work she wants, or working for wages so low that she remains poor.
The official unemployment statistic calls most of these women employed.
And technically, they are.
That may be the problem.
Her name is Nicole. She is 31 years old and lives in Jackson, Mississippi, with her ten-year-old daughter, Kylie, who reads and draws and is learning to play the piano. Nicole works two jobs. One is dispatching for truck drivers, which she can do from home, about 36 hours a week. The other is nursing shifts at a hospital, twelve hours at a time, though those shifts are not guaranteed. Nicole has lived in federally subsidized housing since around 2015, the year Kylie was born. She has told her story publicly, in a first-person essay published by “Ms. Magazine” in November of 2025, as part of a series called Front and Center produced in partnership with a Jackson-based nonprofit called Springboard to Opportunities.
“I work both jobs”, she wrote, “not by choice but out of necessity.”
“Budgeting is a constant struggle because you can’t budget money you don’t have, right?”
“Some weeks, it feels like I’m just working to survive.”
She is, on any given first Friday of the month, counted by the Bureau of Labor Statistics as employed. Her hours at the dispatch job, plus her hours at the hospital when they come, plus the household of one adult and one child she is holding together on those hours and those wages, do not appear in the 4.1 percent figure that will be printed in the newspapers the following morning.
Nicole is one of the people the 4.1 percent number tells us almost nothing about. She works. That answers the government’s question. It does not answer hers.
“How can I be making less and also getting less help?”
She was talking, when she said that, about a specific episode. Her orientation pay at a job had been counted as regular income, and it had lowered her food stamps at exactly the moment she needed them more, and she had tried to speak to a supervisor about it, and they kept telling her to talk to someone else, and she called and called and called between different offices, and the whole process took time and energy she did not have to spare because she was already stretched thin, and in the end the answer she was given was: “well even if it’s just a one-time payment, we count it.”
The country counts her orientation pay. The country does not, in any statistic the country publishes on the first Friday of the month, count the specific arithmetic of what her working life is actually producing for her and her daughter.
________________________________________________________________________________________________________
The Bureau of Labor Statistics does not hide how it calculates unemployment. The methodology is published for anyone who wants to read it. If you performed any work at all for pay or profit during the survey reference week, the government generally considers you employed. Part-time work counts. Temporary work counts. A person who wants forty hours but receives fifteen is employed. A person working a job that does not remotely provide enough money to live on is employed.
The government even publishes broader measures. Its U-6 measure, which includes people working part-time for economic reasons and some people who are marginally attached to the labor force, was 7.9 percent in July, roughly twice the headline number.
But the number Americans hear is 4.1. The number printed in headlines is 4.1. The number described as “the unemployment rate”, in ordinary American speech, is 4.1. Perhaps there was a time when this definition captured something closer to the economic reality Americans experienced, when the distinction between having a job and not having one told us considerably more about whether a person could support herself. But employment has changed while our cultural understanding of the word has remained remarkably intact. In ordinary American life, having a job is supposed to mean something beyond appearing in the employed column. We expect sustained work to provide some reasonable possibility of supporting a life. That proposition becomes much harder to defend when a person can work, and work regularly, and still be poor.
Gene Ludwig served as Comptroller of the Currency under President Bill Clinton from 1993 until 1998 and spent much of his career in banking and financial regulation. He now runs an institute in Washington whose research argues that some of the economic statistics Americans rely upon most heavily describe the economy more favorably than many Americans actually experience it. The title of his 2025 book is The Mismeasurement of America. There is an entire political philosophy contained in that title.
Because measurement sounds bloodless until you consider what measurement does. We measure unemployment because we want to know whether Americans can find work. We measure inflation because we want to know whether Americans can afford things. We measure wages because we want to know whether their lives are becoming economically better or worse. These measurements inform monetary policy, fiscal policy, elections, legislation, business decisions, and the stories the country tells about itself. If the measurement is incomplete, the story becomes incomplete too. And if the story becomes incomplete for long enough, the people the story fails to describe eventually stop appearing in the country’s picture of itself at all.
The Ludwig Institute begins with the same federal labor data the BLS uses. It asks a somewhat different question. Not simply: do you have a job? But: do you have enough work, and does that work pay enough for you to live?
The difference between those two questions is the difference between 4.1 percent and 24.9 percent.
The difference between those two questions is Nicole.
The disparities inside the larger number are what break your heart when you sit with them.
For white workers, LISEP calculated functional unemployment in July at 23.8 percent. For Hispanic workers, 26.7 percent. For Black workers, 27.3 percent. For men, 19.5 percent. For women, 31 percent. The gender gap widened in July to 11.5 percentage points, up from nine the month before. Women’s functional unemployment rose 1.6 points in a single month while men’s declined. It is now at its highest level since March of 2021, when the country was still climbing out of the pandemic collapse.
If 19.5 percent of American men in the labor force are functionally unemployed, something is wrong. If 31 percent of women are, something more is wrong. If Black and Hispanic workers experience functional unemployment at substantially higher rates than white workers, then a national statistic that does not illuminate those differences can produce the comforting impression of a labor market whose prosperity is much more evenly distributed than it actually is.
It would be reasonable to ask how much of these disparities is produced by LISEP’s alternative definition. So set LISEP aside for a moment and look at the federal government’s own data. In the first quarter of 2026, the median full-time American woman earned $1,098 a week. The median full-time American man earned $1,362. Hispanic full-time workers had median weekly earnings of $984. Black full-time workers, $985. white full-time workers, $1,263. Black men, working full time, earned 72.6 percent of what white men earned working full time. These are the government’s own figures for the workers the government most confidently considers employed. Different measures reveal the same broad pattern of unequal economic outcomes. LISEP’s methodology does not create the underlying racial and gender disparities; they remain visible even among the workers whom the federal government most securely classifies as employed.
I employ about thirty people at our winery in Oregon, and nearly half of our staff is Hispanic, many of Mexican descent, which makes it difficult for me to encounter 26.7 percent as an abstraction.
Women are disproportionately represented in part-time and low-wage work, and several of the reasons are well documented. Caregiving still falls disproportionately upon women, while women remain heavily represented in service occupations where wages are lower and schedules less predictable. A 2024 Oxfam analysis found that 42 percent of American working single parents earn less than $17 an hour, compared with 12 percent of partnered parents; among Latina women the share earning low wages was 40 percent, and among Black women 35 percent. The National Women’s Law Center reports that more than seven in ten American part-time workers raising children are women.
________________________________________________________________________________________________________
There is another way to see the same distortion. In the first quarter of 2026, the Bureau of Labor Statistics reported median weekly earnings for full-time wage and salary workers of $1,235. The Ludwig Institute’s True Weekly Earnings measure, which includes part-time workers and unemployed people actively seeking work and is adjusted for inflation, calculated the same quarter’s figure at $1,018. The difference is not because one organization discovered hundreds of dollars the other somehow misplaced. The difference is whom they count.
The headline BLS earnings measure counts only workers with full-time hours. LISEP counts everyone in the workforce, including the people whose hours are too few and the people whose hours have run out. Both numbers are honest arithmetic. They answer different questions. But one question produces an America earning $1,235 a week and rising. The other produces an America earning $1,018 a week and falling. Two hundred and seventeen dollars a week, annualized, is more than eleven thousand dollars, a striking difference produced by the radically different populations each measure chooses to include.
What matters, then, is which America policymakers believe they are governing. Which America the Federal Reserve believes it is calibrating monetary policy for. Which America a president describes when announcing that the economy is strong. Which America members of Congress imagine when deciding what workers require from the country and what they should be expected to provide for themselves.
A country governs the economy it can see. Which means there are consequences when the instruments through which it sees that economy fail to capture how people are actually living.
There is a temptation in writing about statistics like these to announce that the official unemployment rate is a lie. I do not think that is right. The official unemployment rate measures precisely what the government says it measures. It counts people without jobs who are actively looking for work as a percentage of the labor force. Economists understand its limitations. The Bureau of Labor Statistics publishes alternative measures alongside it, in tables anyone can find if she is willing to look. The more interesting problem is not that the number is false. The problem is that we have allowed one narrow measurement to become synonymous with something much larger.
We say “unemployment rate” when what many Americans hear is “how difficult it is for Americans to find adequate work”. Those are not the same thing, and somewhere in the distance between those two meanings are millions of Americans who have jobs and are still poor. They disappear statistically because they showed up for work.
Nicole shows up for work. She dispatches trucks from home for thirty-six hours a week. She works twelve-hour shifts at a hospital when the shifts are given to her. She is showing up. And the country’s most-quoted statistic about the labor market she is inside does not describe her.
In July, LISEP’s functional unemployment rate rose for the fourth consecutive month, from 24.7 to 24.9 percent. It had been 23.6 percent in March. At the same time, labor-force participation was falling, meaning fewer Americans were participating in the labor market at all. LISEP warned that the combination, rising functional unemployment alongside declining participation, could indicate a labor market weakening beneath the headline numbers. The official unemployment rate moved the opposite direction. It fell.
One headline measure improved slightly. Another measure of labor-market adequacy deteriorated for the fourth consecutive month. Both statements can be mathematically true. That should bother us more than it does.
But I keep returning to the women. Thirty-one percent. And to Nicole, whose working life illustrates precisely what the headline unemployment number cannot tell us, and to the daughter she is raising in the subsidized apartment they have called home since the year Kylie was born, while her mother works two jobs and still struggles to make the arithmetic come out.
“My ideal future is one where we aren’t living paycheck to paycheck”, Nicole wrote. “Where I can pay all our bills, provide stability, and even take a trip on the weekends for fun, just to enjoy life together. I want more for Kylie and me.”
She works. She works both jobs, at all hours, taking whatever twelve-hour hospital shifts are offered to her, dispatching trucks from home for the roughly full-time thirty-six hours in between. She looks at abandoned hotels and imagines turning them into housing for people experiencing homelessness. Someday she wants to build something like that herself. The woman struggling to make the arithmetic of her own life work is already imagining how the country might make the arithmetic work for somebody else.
The unemployment statistic does not ask her any of this. On the first Friday of every month, it asks a much narrower question: did she work?
There is something deeply American about our faith in numbers. GDP rises. Unemployment falls. Wages increase. Inflation moderates. The market reaches another record. We collect the numbers and arrange them into a story about ourselves, and because numbers possess the appearance of objectivity, eventually the story begins to feel like fact.
But every statistic contains a human decision about what deserves to be counted. Someone has to decide what “employed” means. Someone has to decide what “unemployed” means. Someone has to decide whether a woman working twenty hours because no one will give her forty belongs among the economic successes or the economic failures. Someone has to decide whether a man earning wages that cannot keep him above poverty is evidence of a functioning labor market because he possesses a job. Someone has to decide whether a woman working two jobs and still saying she is “just working to survive” belongs among the economic successes or the economic failures.
And after those decisions have been made often enough and for long enough, we forget they were decisions at all. They become numbers. Then the numbers become headlines. Then the headlines become the country we believe we live in.
The unemployment rate is 4.1 percent. That number is true. So is 24.9 percent. So is 27.3 percent for Black workers. So is 26.7 percent for Hispanic workers. So is 19.5 percent for men. And so is 31 percent for women.
The question is not which number we would prefer to believe. The question is what happens to a country when the number it chooses as its shorthand for the health of the labor market no longer describes the working lives of millions of people inside it.
Because perhaps the most consequential people in the American labor market are not only the people we count as unemployed. Perhaps they are the people we do not know how to count at all.
The mother who receives twenty-five hours because twenty-five are all she can get.
The warehouse worker who finishes one shift and opens an app to begin another.
The person who wakes before sunrise, works all day, receives a paycheck every other Friday, and remains poor.
And Nicole in Jackson, dispatching trucks from home, waiting to learn whether there will be another twelve-hour shift at the hospital, raising a ten-year-old girl who draws and plays the piano and has already learned three or four songs.
The country has a word for Nicole.
Employed.
The word is accurate.
It just doesn’t tell us very much about her life.