Key findings from just released jolt survey · Research Report

Hires and quits rates hit post-pandemic lows while layoffs hold flat — the signature of a frozen labor market

Data Source: Bureau of Labor Statistics

The Standoff

The February 2026 JOLTS release, published today by the Bureau of Labor Statistics, delivers the clearest picture yet of a labor market that has seized up on both sides. Employers are posting fewer openings and hiring at the lowest rate since the pandemic. Workers, in turn, are refusing to quit. But — and this is the critical detail — layoffs remain anchored at 1.1%, exactly where they sat through the pre-pandemic expansion. Nobody is getting fired. Nobody is moving. The labor market hasn't collapsed. It has frozen.

The numbers are stark. Hiring fell to 4.85 million in February, down from 5.35 million in January — a decline of nearly half a million in a single month. That 3.1% hires rate matches the lowest reading in the entire JOLTS history outside of the COVID-19 lockdown months of March-April 2020. Quits dropped to 2.97 million, a 1.9% rate that has now stayed at or below 2.0% for roughly eight consecutive months. During the Great Resignation of 2021-22, more than 4.5 million Americans were quitting every month. That era is definitively over.

What makes this unusual is the absence of the other shoe dropping. In a normal downturn, falling hires and quits would be accompanied by surging layoffs — employers cutting payrolls as revenue contracts. That is not happening. The layoffs rate of 1.1% in February is the same reading the economy posted in late 2019. This is not a recession pattern. It is something else: a labor market where employers have stopped expanding but haven't started contracting, and workers have stopped searching but haven't started losing their jobs. The best analogy is a game of musical chairs where the music stopped — but nobody has been told to sit down yet.

Job openings per unemployed person fell to 0.91 — below 1.0 for six of the last eight months

Data Source: Bureau of Labor Statistics

Below the Threshold

The single most-watched number in the JOLTS report is the ratio of job openings to unemployed persons. In February, it fell to 0.91 — meaning there are now fewer open positions than there are people looking for work. That is a structural threshold. For every 100 unemployed Americans, there are only 91 job openings.

The speed of this reversal is the real story. At the March 2022 peak, the ratio stood at 2.04 — more than two open jobs for every unemployed person, a level never recorded in JOLTS history dating back to 2000. In less than four years, that surplus has been entirely erased and then some. The ratio first dipped below 1.0 in early 2025 and has stayed there for six of the last eight months. In the pre-pandemic economy of 2019, the ratio averaged about 1.2 — meaning even that period, remembered as a tight labor market, had more slack than the post-pandemic boom.

The implications depend on where you sit. For the Federal Reserve, the ratio's decline is welcome: it signals that labor market rebalancing — the unwinding of the 2021-22 demand surge — is well advanced, reducing upward pressure on wages and prices. For workers, the math is less forgiving. At 0.91 openings per unemployed person, job search durations will lengthen. Workers who lose jobs now face a meaningfully harder time finding new ones than at any point since the economy exited the COVID recession. For those currently employed, the calculus is straightforward: stay put. The quits rate confirms they have received the message.

Hires collapsed to 4.85M in February — the lowest non-pandemic reading since JOLTS began tracking

Data Source: Bureau of Labor Statistics

The Mechanism: Freeze, Not Fire

The raw levels make the freeze even more visible. Job openings dropped to 6.88 million in February from 7.24 million in January — a decline of 358,000 positions in a single month. That level is down 44% from the all-time peak of 12.3 million in March 2022. But the decline in hires is the more consequential number: at 4.85 million, February's reading is the lowest non-pandemic figure in the entire JOLTS dataset going back to December 2000.

To put 4.85 million hires in context: during the pre-pandemic expansion of 2019, the economy was averaging 5.8 million hires per month — roughly a million more. Even during the sluggish recovery of 2013-14, monthly hires ran near 5.0 million. The February 2026 reading sits below that floor. It is not just low by recent standards; it is historically low by any standard.

The gap between openings and hires also tells a story. There are 6.88 million openings but only 4.85 million hires — a conversion rate of about 70%. That sounds reasonable until you consider that during the post-pandemic boom, employers were hiring at rates approaching or exceeding the number of openings. The gap now suggests either a skills mismatch (employers can't find the right workers for the jobs that remain) or employer hesitancy (positions are posted but firms are delaying final hiring decisions). Both dynamics point to a labor market stuck in limbo — not contracting, but not functioning normally either.

Construction openings halved (-51%) while retail surged 44% — the freeze is hitting sectors unevenly

Data Source: Bureau of Labor Statistics

Where the Freeze Bites Hardest

The national aggregates mask a dramatic divergence at the industry level. Construction job openings collapsed 51% over the past two years, from 413,000 in January 2024 to just 202,000 in February 2026. That is the sharpest decline of any major sector and reflects the impact of elevated mortgage rates and a pullback in commercial real estate development. The information sector — tech, media, telecom — saw openings fall 43%, from 160,000 to 91,000, extending the tech hiring drought that began with the mass layoffs of late 2022 and early 2023.

Healthcare, long considered the economy's most reliable job engine, is no longer immune. Openings in healthcare and social assistance fell 31% to 1.28 million — still the single largest category of openings by volume, but meaningfully below its 1.85 million reading two years ago. Professional and business services, which includes the consulting, legal, and accounting firms that serve as leading indicators of corporate investment intentions, dropped 12%.

The lone bright spots are in consumer-facing sectors. Retail trade openings surged 44% to 701,000 — the only major sector to show a double-digit percentage gain. Trade, transportation, and utilities rose 15%. This pattern is consistent with an economy where consumer spending — fueled by still-solid wage gains among the employed — is holding up even as business investment and hiring slow. The consumer is doing the work that corporations have stopped doing.

Government openings fell 21%, a decline that likely reflects the ongoing federal workforce reduction initiatives. At 701,000, public-sector openings are down from 892,000 two years prior, with the contraction concentrated at the federal level rather than state and local governments.

Construction openings fell relentlessly since early 2024 while retail trade broke upward after mid-2025

Data Source: Bureau of Labor Statistics

What Comes Next

The time-series view across industries makes the divergence inescapable. Construction openings have declined in a near-straight line, dropping from over 400,000 at the start of 2024 to barely 200,000 by early 2026. Information-sector openings traced a similar path but from a lower starting point, falling to just 91,000 — a level that suggests the sector has absorbed its pandemic-era overhiring and is now running lean. Professional and business services, the largest white-collar employment category, drifted lower throughout the period but held above 1.2 million, suggesting the slowdown in corporate investment has translated to a hiring pause rather than a wholesale pullback.

Retail trade's trajectory is the outlier. After hovering between 480,000 and 570,000 through most of 2024, retail openings broke sharply upward in mid-2025, reaching 701,000 by February 2026. That shift coincided with several large retailers announcing store expansion plans and warehouse buildouts — a bet on the resilience of the American consumer even as the broader economy softens.

For the 7.6 million Americans currently classified as unemployed, February's JOLTS data is a concrete deterioration of their prospects. Two years ago, each of them would have had more than one open job to compete for. Now they are splitting 6.88 million openings among 7.57 million people — and the openings that remain are increasingly concentrated in retail and low-wage services rather than the higher-paying professional, construction, and technology sectors that drove wage growth during the boom. The labor market is not just frozen. It is freezing in a configuration that is less favorable to workers than the one it is replacing.

At the current trajectory — with hires declining by an average of 80,000 per month over the past year — monthly hiring could fall below 4.5 million by mid-2026, a level last sustained during the 2010 recovery from the Great Recession. Whether that threshold is reached depends on whether the freeze thaws or deepens. The JOLTS data so far offers no evidence of a thaw.

What This Analysis Cannot Tell You

The February 2026 hires reading of 4.849 million is an unusually sharp drop from January's 5.347 million. JOLTS month-to-month volatility is well-documented — the 90% confidence interval for total nonfarm hires is approximately ±300,000 at the national level, meaning the true February figure could plausibly be anywhere from 4.5 to 5.15 million. A single month's reading should not be over-interpreted, and the March 2026 release could show a partial rebound. The trend, however, has been consistently downward for over a year, which limits the chance that February is pure noise.

The JOLTS survey cannot distinguish between employer demand and employer intent. A job opening that has been posted for six months but never filled may reflect a genuine inability to find qualified workers, a 'ghost' posting maintained for compliance or pipeline purposes, or a position where the budget was quietly rescinded but the listing remains. The decline in openings could therefore overstate or understate the true contraction in labor demand. Cross-referencing with private-sector job posting data from platforms like Indeed or LinkedIn would help resolve this ambiguity, but such data was not available for this analysis.

The 'frozen market' thesis — low hires, low quits, low layoffs — is consistent with at least two distinct interpretations. The benign reading: employers are pausing to assess macroeconomic uncertainty (tariff policy, interest rate trajectory) and will resume hiring once clarity returns. The adverse reading: the economy is in a structural transition where AI-driven productivity gains and federal spending cuts are permanently reducing the equilibrium level of hiring. The JOLTS data alone cannot distinguish between these scenarios. What would change the conclusion: if the hires rate rebounds above 3.5% and the quits rate above 2.3% within two quarters, the freeze interpretation would be falsified in favor of a temporary pause.

Industry-level JOLTS estimates are based on substantially smaller samples than the national aggregate, making individual sector readings — particularly for smaller sectors like Information (91,000 openings) and Construction (202,000) — subject to wide confidence intervals. The retail trade surge of +44% could partly reflect seasonal adjustment difficulties or sample composition changes rather than a genuine demand shift. The direction of the trends across sectors is more reliable than the precise magnitudes.

Job openings per unemployed person fell to 0.91 — below 1.0 for six of the last eight months

Hires collapsed to 4.85M in February — the lowest non-pandemic reading since JOLTS began tracking

Construction openings halved (-51%) while retail surged 44% — the freeze is hitting sectors unevenly

Construction openings fell relentlessly since early 2024 while retail trade broke upward after mid-2025

Annotation