Demand, Supply, and the Meaning of Stability
The U.S. labor market’s employer side cannot be read from payroll growth alone. This chapter separates the signals of job creation, hiring intent, layoffs, worker movement, and available labor, then uses them to interpret the Federal Reserve’s mid-2026 claim that demand and supply were roughly in balance. The goal is to distinguish a labor market that is stable because it is healthy and expanding from one that is stable because both hiring demand and labor supply are moving slowly.
Each demand signal answers a different question
Payroll employment is the most familiar employer-side measure, but it is a record of jobs already on payroll. A change in payroll employment tells us how the employment stock moved over a period. It does not say how many employers wanted to hire, how many positions went unfilled, or how many workers left other jobs to take the new ones.
The Federal Reserve report describes a clear change in pace without describing a boom. Private payroll gains averaged about 30,000 per month in the second half of 2025, rose to nearly 80,000 in the first quarter of 2026, and increased to almost 100,000 in the second quarter. The report still characterizes the gains as soft by historical standards. The useful reading is therefore “hiring improved from a very weak base,” not “employers are adding jobs at an exceptional rate.”
Job openings ask a different question: how many positions employers are trying to fill. The report says JOLTS openings moved a little higher from late last year, while an alternative measure based on Indeed postings flattened during 2026 after earlier declines. The two measures point to a labor market in which hiring demand has not collapsed, but neither establishes that every opening will become a filled job. An opening can remain available because an employer is searching, because the match is difficult, or because plans change before hiring occurs.
Layoffs and claims look at separation pressure rather than new demand. The report says indicators of layoffs remained muted: initial unemployment-insurance claims moved sideways, and the JOLTS layoff rate averaged 1.1 percent in the first part of the year, close to its pre-pandemic average. Low layoffs can support a stable reading, but they do not prove that hiring is vigorous. A firm can add few workers and also dismiss few workers. That combination produces a quiet labor market, not necessarily a rapidly improving one.
Quits provide another angle. The report treats the share of workers who leave jobs voluntarily as an indicator of the availability of attractive alternatives. If quits are high, workers may feel able to move to better opportunities. If quits flatten or fall, outside options may be less plentiful, even if unemployment remains low. The measure is not a direct survey of confidence, and it does not identify why each worker left. It is a signal about mobility that becomes more useful when read with openings, hiring, and wage growth.
The “available jobs versus available workers” comparison makes the relationship more explicit. Available jobs are measured as employed workers plus job openings; available workers are measured by the labor force. The gap between them edged slightly higher in the report’s 2026 data after drifting lower the year before. That gap is not a complete measure of matching quality, but it frames the question correctly: how many potential work opportunities exist relative to the people available to fill them?
There is a hidden interface between all of these measures: matching. A vacancy is not a job in a worker’s hands, and a person in the labor force is not automatically a match for every vacancy. Industry, location, skills, schedules, pay, and timing can keep an opening unfilled even when the aggregate number of workers looks sufficient. The report does not estimate one universal matching friction, so the reader should not turn the jobs-versus-workers gap into a precise measure of bargaining power. It is a useful balance question, not a complete account of how quickly people and jobs connect.
The measures also have different failure surfaces. Payrolls can remain stable while hiring and separations offset one another. Openings can remain high while employers become slower or more selective about filling them. Claims can remain low because layoffs are low, even if voluntary mobility has weakened. Quits can fall because outside options are less attractive, but the quit rate alone cannot tell whether workers are dissatisfied, cautious, or simply staying put. A good reading names the signal’s blind spot at the same time as its information value.
Stocks, flows, and rates tell different stories
The difference among these measures is partly a difference in time. Employment is a stock observed at a point. Payroll gains are a flow over a month or quarter. The unemployment rate and layoff rate are ratios. Job openings are a stock of positions that employers report as available, while quits and claims describe movements through the labor market. Comparing them as if they were all the same kind of number can create a false contradiction.
Imagine a company that has 1,000 workers and 20 openings. If it fills 10 openings and loses 10 workers, its payroll stock does not change, but hiring and separation both occurred. If it closes 10 openings without changing its payroll, the stock still does not change, but intended demand has changed. If it keeps all 1,000 workers and stops recruiting, payroll looks stable while the path for job seekers becomes less favorable. The point is not the company example itself; it is the accounting logic. The same employment level can be produced by different flows.
The same logic applies at the economy-wide level. A monthly payroll gain can coexist with a falling number of openings if employers are filling existing vacancies and creating fewer new ones. An increase in openings can coexist with weak payroll growth if matching is slow. A low layoff rate can coexist with subdued quits if workers are reluctant to move. The signals become informative when their relationships are examined, not when one is promoted to the status of master statistic.
The time basis matters as well. The Federal Reserve’s payroll discussion uses averages and recent-quarter comparisons, while other indicators are monthly or moving averages. A payroll change for one month can be noisy; a three-month average can reveal direction but will react more slowly to a turning point. A reader should state which horizon is being used before describing acceleration or slowdown.
The diagram’s arrows are questions, not a causal chain in which one indicator mechanically produces the next. Payrolls do not cause quits, and openings do not guarantee hires. Each signal illuminates one interface between employers and workers; the labor-market judgment is the result of interpreting the set.
Why stable does not necessarily mean strong
The denominator chapter showed why a low unemployment rate can coexist with constrained labor supply. The demand side completes the explanation. The Federal Reserve says that labor demand and supply were roughly in balance after a period of cooling, but it also says that demand had stabilized at a modest pace. Job openings were broadly flat on balance, layoffs were subdued, and payroll growth had improved while remaining soft. Supply growth was historically low because immigration slowed and participation declined with population aging.
A stable unemployment rate can therefore be produced by two slow-moving sides. If employers add jobs at a modest pace while fewer people enter the labor force, the share of active job seekers without work may change little. That is not a claim that the labor market is weak in every sense. It is a reminder that stability is a relationship between flows of workers and flows of jobs.
Productivity adds another qualification. The report says business-sector labor productivity growth since late 2019 has averaged 2.1 percent per year, faster than the average pace in the previous business cycle. It argues that strong productivity has offset historically subdued labor-force growth in the economy’s productive capacity. In other words, limited labor-force growth does not automatically imply limited capacity to produce. But productivity growth is not the same as broad hiring, and it does not tell us how evenly gains are shared. It changes the interpretation of the system; it does not settle every labor-market question.
Sector differences also matter. The report says health-care employment growth was relatively strong, while leisure and hospitality, information, and financial activities recorded declines. An aggregate payroll number can therefore be compatible with very different local experiences. A worker in a growing sector may see more openings and mobility than a worker in a declining one. A reader should not infer that every industry is “balanced” merely because the national indicators are broadly stable.
The same caution applies to the word “demand.” It can mean that employers want to add positions, that they are actively posting openings, that they are willing to raise pay to attract workers, or that they are retaining current staff by avoiding layoffs. Those are related but not identical forms of demand. The Federal Reserve’s mid-2026 description is strongest when kept at the level it actually supports: job growth picked up, layoffs stayed muted, vacancies were roughly flat, and several indicators suggested that resource utilization was broadly stable. It does not support the stronger claim that every employer was competing aggressively for labor.
This distinction helps explain why an unemployment rate can remain low while a worker’s search becomes harder. If layoffs stay low, existing workers may remain employed. If openings and quits flatten, a person trying to change jobs may face fewer alternatives. The aggregate unemployment rate can therefore describe security for people who have jobs while saying less about the prospects of people entering, re-entering, or moving through the labor market. The demand-side signals fill that gap only when their separate meanings are retained.
How to use the signals without inventing a score
A useful employer-side reading can proceed through four questions. First, are payrolls rising, falling, or merely holding steady, and over what period? Second, are employers still seeking workers through openings and postings, or are those measures flattening? Third, are layoffs and claims showing new separation pressure? Fourth, do quits and the available-jobs-versus-workers gap suggest that workers still have attractive alternatives?
The questions are deliberately plural. If payrolls accelerate, openings rise, quits remain firm, and layoffs stay low, the evidence for stronger demand is more persuasive than any one movement alone. If payrolls are soft, openings flatten, quits weaken, and layoffs remain low, the market may be stable but less dynamic. If layoffs jump while openings fall, the story changes again. These combinations are evidence patterns, not automatic diagnoses; the source does not support a universal threshold at which one pattern receives a fixed label.
A useful distinction is between confirmation and explanation. Several indicators moving together can confirm that conditions changed, but they may still leave the cause uncertain. Payroll gains and openings could rise because demand strengthened, because labor supply expanded and made hiring easier, or because a particular sector recovered. Likewise, a stable unemployment rate and low layoffs could reflect resilience, weak turnover, or constrained entry. The indicators narrow the interpretation; they do not eliminate the need for judgment about mechanisms.
The same discipline helps when looking at a particular sector or group. Start with the national picture, then ask whether the relevant population shares the same direction of payrolls, openings, separations, and wages. Do not infer a sector’s condition from the aggregate unemployment rate alone. At the same time, do not treat one industry’s decline as proof that the whole labor market is deteriorating. The level of analysis must match the claim.
The mid-2026 evidence supports a qualified conclusion. Demand had not disappeared: private payroll gains had improved, job openings were not collapsing, and layoffs were muted. But demand was modest, labor supply was growing slowly, and the overall balance was calm partly because both sides had cooled. “Stable” describes the interaction of these signals. It does not promise strong hiring, equal opportunity across sectors, or abundant choices for every worker.
That conclusion also explains why different labor-market numbers can appear to disagree. A payroll series may use one window, an openings series another, and a report may smooth a volatile measure with a moving average. Before deciding that one number contradicts another, the reader must inspect how each was made, when it was measured, and what question it was designed to answer.