Job Openings: The New Labor Market Barometer

For most of the twentieth century, economists measured the labor market from the worker’s side. They counted the unemployed. They measured hours worked, wages earned, layoffs absorbed. The employer’s side — how many jobs were available, how desperately firms were trying to hire — remained largely invisible in the official statistics.

That changed in December 2000, when the Bureau of Labor Statistics launched the Job Openings and Labor Turnover Survey, known universally as JOLTS. Each month, the BLS surveys roughly 21,000 business establishments and government agencies, asking a deceptively simple question: how many positions are you actively trying to fill?

The answer — the job openings level — has become one of the most closely watched indicators in macroeconomics. In January 2026, there are 6.9 million unfilled jobs across the American economy. That figure is down sharply from the pandemic-era peak of 11.2 million, but it remains above the levels that prevailed through most of the 2010s. It suggests an economy where employers still want workers, even if the frenzy has passed.

But the raw level is only the beginning. JOLTS also tracks hires, separations, quits, and layoffs — the full metabolic activity of the labor market. Previous episodes in this series examined the quit rate (Episode 5) and unemployment (Episode 1). This episode turns to the demand side: how many jobs are out there, who is posting them, and what the relationship between openings and hires tells us about where the labor market is heading.

The 26-year history of the JOLTS data divides neatly into four chapters, each defined by a distinct labor market regime.

Chapter one: the early baseline (2001–2007). When the survey launched, America had roughly 5.2 million job openings. The dot-com bust cut that to 3.4 million by 2003 — a 35% decline in just two years. Recovery was slow. It took until 2005 to return to the pre-recession level, and by 2007, openings stood at 4.8 million. These were the years when economists were still learning what “normal” looked like for this new data series.

Chapter two: the Great Recession crater (2008–2013). The financial crisis destroyed employer demand. Openings plunged to 2.7 million in January 2009, the lowest reading in the survey’s history. At the trough, there were roughly six unemployed workers competing for every posted vacancy. Recovery was even slower than after the dot-com bust. It was not until 2014 that openings consistently exceeded 4 million, and it took until 2015 to surpass the pre-crisis peak of 4.8 million.

Chapter three: the steady escalation (2014–2019). Something changed in the mid-2010s. Job openings did not just recover — they accelerated. From 4.1 million in January 2014, they rose to 5.3 million in 2015, then 6.0 million in 2016, then 6.6 million in 2018, and finally 7.5 million in January 2019. Each year set a new record. In 2018, something happened that had never been recorded before in American economic data: the number of job openings exceeded the number of unemployed persons. For every jobless American, there was more than one open position waiting to be filled.

Chapter four: the pandemic supercycle (2020–present). COVID-19 briefly crashed openings to 7.1 million in January 2020 (measured before the shutdowns hit), but the subsequent recovery was unlike anything in the data’s history. As the economy reopened, employers scrambled to rehire. Openings surged to 7.2 million in January 2021, then rocketed to an astonishing 11.2 million in January 2022 — more than triple the Great Recession trough and more than double the pre-pandemic norm.

Since that peak, the story has been one of normalization. Openings fell to 10.4 million in January 2023, then 8.5 million in 2024, then 7.4 million in 2025, and now 6.9 million in January 2026. The descent has been steady — roughly 1 to 1.5 million per year — and has brought the labor market back to approximately its late-2019 posture. Still healthy, but no longer overheated.

The scale of the 2021–2022 surge deserves emphasis. In the entire pre-pandemic history of the JOLTS survey — two decades of data — job openings never exceeded 7.5 million. Then, in less than eighteen months, they gained four million to reach 11.2 million. The causes were legion: simultaneous reopening across every sector, a wave of early retirements that shrank the labor force, pandemic savings that allowed workers to be choosier, and a reshoring boom in manufacturing and logistics that created entirely new demand. No single factor explains it. The surge was the labor market’s version of a hundred-year flood.

The JOLTS history is marked by a handful of turning points that reveal the structural evolution of American labor demand. Each milestone corresponds to a shift in the balance of power between employers and workers.

The 2018 crossover deserves special attention. Until that year, there had always been more unemployed Americans than open jobs. The surplus varied — it was enormous during the Great Recession (roughly 6 unemployed per opening) and narrow during booms (about 1.2 per opening in 2007). But it was always positive. Employers, in aggregate, never had to worry about a literal shortage of available workers.

Then the math flipped. In January 2018, there were 6.6 million openings and roughly 6.5 million unemployed persons. For the first time in the data’s history, there were more jobs than jobless people. This was not a statistical anomaly — it persisted through 2019 and, after a brief pandemic interruption, became even more extreme through 2021 and 2022. At the 2022 peak, there were 11.2 million openings versus approximately 6 million unemployed — a ratio of nearly 1.9 openings per person.

By January 2026, with 6.9 million openings and approximately 7.3 million unemployed, the ratio has come back to roughly one-to-one. The historic surplus of workers over jobs has returned, but just barely. The labor market is balanced in a way it hasn’t been since 2017 — before the crossover that changed the dynamics of hiring, wages, and worker leverage for half a decade.

Not all industries post openings at the same rate. The JOLTS data includes an “openings rate” — job openings as a percentage of total employment plus openings — that reveals which sectors are most aggressively trying to hire relative to their current workforce. The December 2025 data (the latest available by industry) shows a clear hierarchy of labor demand.

Leisure and hospitality leads at 5.2%, meaning that roughly one in twenty positions in the sector is unfilled at any given time. This is structural, not cyclical. Hotels, restaurants, and entertainment venues have faced chronic staffing challenges since the pandemic, driven by lower wages relative to other sectors, demanding work conditions, and a workforce that has increasingly opted for higher-paying alternatives in warehousing, healthcare, and gig work.

Other services (5.0%) and education and health services (4.8%) round out the top three. Healthcare’s presence is no surprise — the aging of the Baby Boom generation has created insatiable demand for nurses, home health aides, medical technicians, and therapists. The Bureau of Health Workforce projects a shortage of 124,000 physicians by 2034, and the pipeline of new workers has not kept pace with retirements.

Professional and business services (4.3%) sits in the middle of the pack, reflecting a sector that is growing but no longer experiencing the frantic hiring of 2021–2022, when every consulting firm, tech company, and staffing agency seemed to have dozens of unfilled roles. The cooling of white-collar hiring has been one of the defining labor market stories of 2024 and 2025.

At the bottom of the spectrum, financial activities has the lowest openings rate at just 2.7%. Banking, insurance, and real estate are mature industries with low turnover, high compensation, and relatively stable headcounts. When a job opens in finance, it fills quickly — the sector attracts no shortage of applicants. State and local government (3.0%) is similarly low, though for different reasons: budget constraints and lengthy hiring processes keep the posted vacancy count modest even when the actual need is larger.

The industry breakdown reveals a fundamental asymmetry in the American labor market. The sectors with the highest opening rates — leisure, healthcare, other services — are predominantly in-person, physical-labor industries. They require workers to be present in a specific place at a specific time, doing work that cannot be automated or offshored. The sectors with the lowest opening rates — finance, information, government — are predominantly desk-based, knowledge-work industries where productivity per worker is high and turnover is low.

This asymmetry has profound implications for wage dynamics. The industries that cannot find enough workers are forced to raise pay to compete. Leisure and hospitality wages have risen 30% since 2020, the fastest growth of any major sector. The industries that have no trouble filling positions can hold the line on compensation. The JOLTS data, in other words, is not just a measure of labor demand — it is a map of where the bargaining power sits.

Job openings measure demand. Hires measure fulfillment. The relationship between the two is one of the most revealing indicators in labor economics — a measure of how efficiently the labor market is matching employers with workers.

In a well-functioning labor market, the hires rate should roughly track the openings rate. When employers post jobs, they fill them. When openings rise, hires rise. The two lines move together, like a call and its echo.

But the JOLTS data shows that this relationship broke down dramatically during the pandemic era — and has not fully recovered.

The chart below plots the openings rate and hires rate for total nonfarm employment from 2001 to 2026. Through 2017, the two lines moved in broad parallel. The openings rate ranged from about 2% to 4%, and the hires rate ranged from about 3% to 4%. There was always a gap — not every opening leads to a hire in the same period — but it was stable and predictable.

Starting in 2018, a divergence appeared. The openings rate climbed sharply — from 3.7% to 4.3% to 4.8% — while the hires rate barely moved, hovering between 3.7% and 3.9%. Employers were posting more and more jobs, but the rate at which they were actually filling those positions was flat. The gap represented a matching problem: there were jobs and there were workers, but they were not connecting.

The pandemic blew this divergence wide open. By January 2022, the openings rate had reached 7.0% while the hires rate sat at just 4.3%. The gap of 2.7 percentage points was the widest in the survey’s history. Employers were frantically posting positions — offering signing bonuses, raising wages, relaxing credential requirements — and still could not fill them fast enough. The labor market had become a seller’s market for workers in a way never previously recorded.

The normalization since 2022 has been striking. The openings rate has fallen from 7.0% to 4.2% in January 2026, a decline of 2.8 percentage points. But the hires rate has also fallen — from 4.3% to 3.3%, a decline of 1.0 percentage point. The gap has narrowed from 2.7 points to 0.9 points, but it has not closed.

This persistent gap — openings materially above hires even as the market cools — suggests that some of the matching inefficiency from the pandemic era is structural. Several explanations have been proposed:

Geographic mismatch. The jobs are in different places than the workers. Remote work opened up some roles to a national talent pool, but many of the hardest-to-fill positions — healthcare, hospitality, construction — require physical presence in specific locations where housing costs have made it difficult for lower-wage workers to live.

Skills mismatch. The economy’s rapid transformation during the pandemic — toward digital services, logistics, and healthcare — created demand for skills that the existing workforce does not always possess. Retraining takes time, and the American system of workforce development is fragmented.

Wage expectations mismatch. Workers who experienced the leverage of 2021–2022 — when they could demand higher wages and better conditions — may be reluctant to accept positions at the lower compensation levels that employers are now offering as margins tighten.

Whatever the cause, the data is clear: the American labor market fills a smaller share of its posted openings than it did before the pandemic. In January 2016, for every 4.0% openings rate, there was a 3.6% hires rate — a conversion of 90%. In January 2026, for every 4.2% openings rate, there is a 3.3% hires rate — a conversion of just 79%. Something has changed in the plumbing of the labor market, and it has not changed back.

Economists increasingly focus on a single metric derived from the JOLTS data: the number of unemployed persons per job opening. This ratio captures the balance of power in the labor market more precisely than any other single indicator.

When the ratio is high — many unemployed per opening — it is an employer’s market. Firms can be selective, wages stagnate, and workers accept whatever they can find. When the ratio is low — few unemployed per opening — it is a worker’s market. Employees can be choosy, wages rise, and employers must compete to attract talent.

The trajectory tells a remarkable story. During the Great Recession, there were 4.4 unemployed Americans for every job opening. If you were out of work in January 2009, the arithmetic was bleak — you were competing against more than four other people for every position that existed. The recovery from that ratio was achingly slow. It took five years to get below 2.0, seven years to get below 1.5, and ten years to reach the historic crossover of 1.0.

The pandemic created the mirror image. By 2022, the ratio had fallen to 0.5 — meaning there were two open jobs for every unemployed person. Workers had never had so much leverage in the history of the JOLTS survey. This was the environment that produced the “Great Resignation,” record wage growth for low-income workers, and a cultural shift in worker expectations that is still playing out.

At 1.1 in January 2026, the ratio has returned to near-balance. There is approximately one unemployed person for every open job — a state that economists would characterize as a labor market in equilibrium. It is neither the desperation of 2009 nor the frenzy of 2022. It is something like normalcy — if normalcy is something the American labor market can sustain.

Job openings are a leading indicator — they signal where the labor market is going before it gets there. When openings fall, hiring slows. When hiring slows, unemployment rises. The lag is typically six to twelve months. This makes the JOLTS data a critical early warning system for recessions and recoveries alike.

The current trajectory offers a cautiously optimistic reading. Openings have declined from 11.2 million to 6.9 million — a 38% drop — without triggering a significant rise in unemployment. The January 2026 unemployment rate of 4.3% (discussed in Episode 1) is elevated from the 2023 lows but remains historically moderate. This suggests that the economy has achieved what Federal Reserve officials call a “soft landing” in the labor market — cooling demand without crashing into contraction.

However, the hires rate deserves close attention. At 3.3%, it is at the lowest level since 2014. While lower openings can simply mean less overheated demand, lower hires mean less actual job creation. If the hires rate continues to decline, it will translate into slower employment growth and eventually higher unemployment — regardless of how many openings are posted.

The relationship between openings and hires is, in many ways, the Rosetta Stone of the current labor market. Openings tell you what employers want. Hires tell you what they are actually doing. And when the gap between wanting and doing narrows not because matching improves but because both decline together, it is a signal that the cycle may be turning.

For now, at 6.9 million openings, the American economy still has more unfilled jobs than it did at any point before 2018. The demand for workers has not disappeared. It has simply returned to a level that the supply of workers can approximately satisfy. Whether that equilibrium holds — or whether openings continue their decline toward the 5–6 million range that characterized the mid-2010s — will be one of the defining economic questions of the next twelve months.

Economists have one more tool for interpreting the JOLTS data: the Beveridge Curve, which plots the job openings rate against the unemployment rate. In a stable labor market, these two variables have an inverse relationship — when openings are high, unemployment is low, and vice versa. The curve traces out a downward-sloping arc.

But the position of the curve matters as much as the slope. When the curve shifts outward — moving to the northeast, with both higher openings and higher unemployment at every point — it signals a structural deterioration in matching efficiency. The labor market is generating plenty of vacancies but cannot fill them, even though workers are available.

This is exactly what happened during and after the pandemic. The Beveridge Curve shifted dramatically outward between 2020 and 2022, suggesting that the matching function of the labor market had been damaged. Geographic immobility, skills gaps, and changed worker preferences all contributed. Since 2023, the curve has been rotating back toward its pre-pandemic position, but it has not fully returned. The labor market in 2026 operates with a modest structural inefficiency that was not present in 2019 — a friction that manifests as slightly higher openings for any given level of unemployment.

This matters for monetary policy. If the Beveridge Curve has permanently shifted, then achieving the same unemployment rate requires tolerating a higher level of unfilled vacancies — and the wage pressure that goes with them. The Federal Reserve, which watches the JOLTS data as closely as any economic release, must calibrate its interest rate decisions with this structural shift in mind.

The JOLTS survey is barely 26 years old — a newcomer by the standards of federal statistics. The unemployment rate has been published since 1948. The Consumer Price Index dates to 1913. JOLTS, launched in December 2000, has existed through only two complete business cycles: the 2001 recession and recovery, and the 2008–2009 crisis and its long aftermath. The pandemic cycle is still unfolding.

Yet in those 26 years, the data has revolutionized how economists think about the labor market. Before JOLTS, the narrative was one-dimensional: unemployment goes up, the economy is bad; unemployment goes down, the economy is good. JOLTS added the employer’s perspective, revealing that the labor market has two sides and they do not always move in sync.

The 2022 experience was the ultimate proof of concept. Unemployment was low — just 4.0% — but the labor market was anything but calm. With 11.2 million unfilled positions, employers were in crisis. Wages were soaring. Workers were quitting in record numbers (as Episode 5 explored). The unemployment rate said “everything is fine.” The openings data said “the labor market is on fire.” The JOLTS data was right.

As the series continues, the next episode examines the darkest corner of the JOLTS data: layoffs and discharges — the involuntary separations that spike during recessions and signal the onset of economic contraction. If job openings are the barometer of demand, layoffs are the barometer of distress.