JOLTS reports a stock of job openings and monthly flows of hires and separations. Equal vacancy totals can conceal very different turnover beneath them. Learn to distinguish those measures, check their reference periods and read them alongside payroll jobs and unemployment.

What JOLTS is and what it actually measures

JOLTS stands for the Job Openings and Labor Turnover Survey. It is produced by the US Bureau of Labor Statistics, the same agency that publishes the payrolls report. The data comes out once a month, at 10:00 AM US Eastern Time; convert the official release date and time to your own zone.

Job openings are a stock measured on the last business day of the month, with specific BLS availability and recruiting criteria. Hires cover additions to payrolls during the month. Separations include quits, layoffs and discharges, and other separations such as retirements, transfers and deaths. Stocks and flows have different definitions and reference periods.

NFP estimates payroll jobs and their net change, rather than unique people. JOLTS measures a vacancy stock and monthly hires and separations. These complement one another, but all hires are not necessarily fills of previously reported vacancies, and differences in method and coverage prevent treating every headline as one simple accounting identity.

Ten vacancies at each end can hide substantial activity

Track specific open positions in our own register. New vacancies appear during the month, some are filled and others are withdrawn without a hire.

Scroll sideways to see the full diagram

Reconcile our vacancy registerStart with 10 open positions. Adding 12 gives 22. Fill 8 specific registered positions, leaving 14. Withdraw 4, leaving 10. The unchanged final stock hides flows during the month.VACANCIES: STOCK AND FLOWSOur invented register of open positions10221410Start+12 new−8 filled−4 removed10 + 12 − 8 − 4 = 10 open at the end8 filled: from this specific register4 withdrawn: removed without a hire

The invented register has fixed coverage and unique position identifiers. There are no duplicates or other changes. All 8 fills concern the tracked vacancies; all 4 withdrawals occur without filling them. This is our reconciliation, not an operational formula published by BLS. BLS: JOLTS Technical Note confirms that openings are a stock on the last business day, while hires and separations cover events throughout the month and have distinct coverage.

Bars show successive stocks after the listed changes, not official JOLTS readings.

Do not automatically substitute all JOLTS hires for the 8 fills in this register. Hiring need not concern a previously reported opening. The example separates stocks from events and checks whether a flow actually refers to the same records.

Three ratios with different denominators

Use a hypothetical matched-month model with 8 million openings, 152 million employed, 3.04 million quits and 6.4 million unemployed. The job-openings rate is 8/(152 + 8) × 100 = 5%; the quits rate is 3.04/152 × 100 = 2%. Openings per unemployed person is 8/6.4 = 1.25, while unemployed per opening is 6.4/8 = 0.8. The final two are ratios of counts, not percentages. Using 8/152 would give about 5.26%, a different ratio from the stated openings rate.

Match month and seasonal-adjustment status before combining series. A ratio above one does not match particular people to the skills or locations of particular vacancies, nor guarantee wage increases. State which way round the ratio is quoted.

The quits rate measures voluntary separations as defined by BLS. Quits can occur for reasons other than a better offer, so the aggregate rate does not identify each worker’s motivation. Read it alongside openings, hires, other separations and wage information rather than treating it as direct proof of confidence or wage pressure.

Why the Fed reads JOLTS so closely

The Fed’s mandate covers maximum employment and price stability. Labour-market data may inform both objectives, but no single vacancy ratio determines policy. When reporting a historical ratio or a policymaker’s interpretation, identify the observation period and original document rather than presenting it as the current committee’s view.

Openings, quits and wages can inform a conditional assessment of labour-market pressure. They do not create a mechanical chain from vacancies to wages, inflation and a required policy decision. Compare the report with other data and the Fed’s dated assessment before stating a policy hypothesis.

Check the current Fed statement and speeches to establish which measures policymakers discuss. A historical Powell reference does not fix the current committee’s priorities or make smaller JOLTS components irrelevant.

The lag, the report's Achilles heel

Check the report’s reference month and actual release date. Publication lags and ordering relative to payrolls can vary; do not assume every release describes exactly one month earlier or arrives six weeks after the same event. Openings refer to the last business day, while hires and separations cover the month. Later monthly and annual revisions can change earlier estimates.

Check JOLTS’s reference month against the latest payrolls report and weekly initial jobless claims. Differences in publication lag do not establish a fixed ranking of market reactions. Record the actual surprise, revisions and observed currency movement for each release.

How JOLTS fits the labour-market picture

Claims, payroll jobs, unemployment and JOLTS measure different aspects of the labour market. In a hypothetical JOLTS flow calculation, 5.6 million hires minus 5.4 million separations gives +0.2 million, or 200,000. That is not automatically the published NFP change: BLS describes definitional and measurement differences and an alignment procedure. Likewise, subtracting hires during a month from the month-end vacancy stock does not calculate unfilled vacancies.

At one fictional establishment, assume fixed coverage and 100 employees initially. A month with 10 hires and 10 separations ends at 100; another with 20 and 20 also ends at 100. The sums of counted events are 20 and 40, not necessarily 20 or 40 unique people. A person hired and separated in the month can enter both flows. Keep gross flows alongside net change.

Agreement between indicators can inform a hypothesis, but does not establish a necessary rate or dollar response. Quits do not identify each person’s motive, and each series has its own coverage and uncertainty. Record the data first, then the policy interpretation and concurrent information. The fundamental analysis section provides the wider framework.

Your next step

Save the openings consensus and its timestamp before release, alongside the originally published previous value. In a hypothetical report, new 8.0 million, consensus 7.8 million, revised previous 8.1 million and original previous 8.3 million give surprise +0.2 million, monthly change −0.1 million and revision −0.2 million. The −0.3 difference versus the original value mixes the last two. Preserve both vintages and check estimation uncertainty; a numerical change is not automatically statistically significant.

Read openings, quits, hires and the other separation categories using their own definitions and reference periods. No one component is a direct measurement of the wage pressure policymakers will act on. Compare successive releases and other labour data without treating a few observations as a forecast. The article on the Fed decision and the dollar provides additional context.