Job Search Statistics 2026: Openings, Hires, Unemployment, and Movement

July labor-market data show a mixed picture: employers reported 7.271 million job openings, while the household survey found 27.0% of unemployed people had been out of work for at least twenty-seven weeks. That equals roughly one in 3.7 unemployed people. Strong demand and long individual job searches can exist at the same time.

What the latest numbers say

The July JOLTS release, published on September 1, 2026, measures openings, hires, and separations at employers. The August Employment Situation measures unemployment and employment from a household survey. They describe related but different parts of the labor market.

The main July measures were:

These rates are monthly JOLTS rates tied to an employment-based denominator. They are not the percentage of all adults who found work, nor the percentage of unemployed people who received an offer.

The June figures were revised after the preliminary report. Openings were revised down by 177,000 to 7.2 million, hires by 16,000 to 5.3 million, and total separations by 14,000 to 5.3 million. A preliminary figure is an initial estimate; a revised figure incorporates later information. Revisions do not necessarily change the broad story, but they can change comparisons between months.

Openings and hiring are different flows

Openings are a stock: they describe positions available at a point in time. Hires are a flow: they count people brought onto payroll during the month. Because the measures use different concepts, subtracting hires from openings does not produce a count of vacancies left unfilled.

Where openings are concentrated

The industry mix matters because job seekers do not compete in one national market. The July openings were spread across private industries and government, with the largest reported totals in education and health services, trade and transportation, and professional and business services.

The share column below is a PlainReads calculation using each industry’s reported openings divided by total July openings of 7.271 million. Industry levels are reported in thousands.

Unemployment and search duration

The August unemployment rate was 4.1%, representing 7.0 million unemployed people. The unemployment rate uses the labor force as its denominator: people who are working plus people actively looking for work.

Among the unemployed, 1.9 million people had been jobless for twenty-seven weeks or longer. They represented 27.0% of all unemployed people, not 27.0% of the entire adult population. This is why the long-term unemployment figure can remain high even when openings exceed hires.

The labor force participation rate was 61.6%, and the employment-population ratio was 59.1%. Both use the civilian noninstitutional population as their broad denominator, but one measures participation in the labor force while the other measures employment.

Other August measures add context. 4.4 million people were working part time for economic reasons, meaning they wanted full-time work or more hours but faced economic conditions affecting their schedules. 5.7 million people were outside the labor force but wanted a job. Of those, 1.7 million were marginally attached, including 441,000 discouraged workers. These groups should not simply be added together: the marginally attached and discouraged categories describe subsets of people outside the labor force.

Movement, wages, and job-seeking behavior

The Federal Reserve’s 2025 SHED Appendix B reports that 23% of adults applied for a new job, 13% started a new job, 8% voluntarily left a job, and 7% were laid off during the prior twelve months.

Across all adults, the applied-to-started comparison is 1.77 — PlainReads calculation, based on 23 divided by 13. It is not an application success rate. The questions are separate, people can submit multiple applications, and not every application belongs to a person who ultimately started a job during the period. The maximum unweighted respondent count was 12,934, and the number of respondents differed by item.

What job seekers can control

The data cannot tell one person which employer will respond. They can, however, support a more focused search.

Start with the sectors where openings are concentrated, then narrow by location, role, pay, and required skills. Treat the national openings total as context rather than as a promise. A targeted application that matches the employer’s requirements is more informative than a large application count by itself.

Track applications, follow-ups, interview stages, and outcomes so the search can improve over time. The job application tracker can help organize that process. For practical guidance on using tools during the search, see Land the Offer with AI. If the current field has weak fit or limited opportunity, the career change guide can help frame a transition around transferable skills.

The statistics also support patience without passivity. Long searches do not prove that a candidate lacks value. They may reflect timing, geography, occupation, screening systems, or competition. At the same time, the national data cannot identify which explanation applies to a particular search.

Methodology and limitations

The JOLTS figures refer to July and come from an employer-side survey. Openings and hires are not interchangeable measures, and separations include quits, layoffs and discharges, and other separations. June values were revised, so comparisons with earlier releases should use the latest published estimates.

The Employment Situation figures refer to August and come from the household survey, with payroll information providing an establishment-based view of employment. Unemployment is a labor-force concept, while people outside the labor force are not counted as unemployed unless they meet the relevant search and availability conditions.

Seasonality also matters. Hiring, layoffs, school-year employment, holidays, and summer work can produce recurring patterns. A single monthly observation should be read alongside the broader movement in the published series.

Finally, these statistics show association, not causation. The coexistence of 1.44 openings per hire and 27.0% long-term unemployment does not prove that openings cause longer searches or that long searches cause employers to retain openings. The measures do not identify the effects of skills, pay, location, employer screening, transportation, health, or application quality.

Sources

• BLS JOLTS July release
• BLS Employment Situation August release
• Federal Reserve SHED Appendix B

Frequently asked questions

Does more than one opening per hire mean every job seeker has a good chance?

No. The 1.44 openings-to-hires result is an aggregate comparison between a stock and a monthly flow. It does not account for occupation, location, qualifications, pay, timing, or employer screening.

Why can long-term unemployment remain high when openings exceed hires?

Openings and unemployment describe different populations and processes. A reported opening may not match the skills or location of a particular unemployed person. Hiring can also take time, and the 27.0% long-term unemployment share reflects the duration of individual searches rather than the total supply of jobs.

Is the applied-to-started comparison an application success rate?

No. The 1.77 comparison divides separate shares of adults who applied and started a job. People can submit multiple applications, and the survey questions do not link a particular application to a particular start.

Should job seekers focus on the industries with the most openings?

Those industries are useful starting points, not guaranteed paths. The largest reported concentration was in private education and health services, trade, transportation, and utilities, and professional and business services. Fit, requirements, geography, and application quality still determine an individual outcome.