Remote Work Tripled in U.S. Job Ads—Employer Choice Still Decides

A working paper released on August 19 finds that the share of U.S. job advertisements offering at least one remote day a week more than tripled between 2019 and 2026. The Dallas Fed release covers nearly 600 million vacancy postings in the United States, United Kingdom, Canada, Australia and New Zealand and uses a classifier assessed at 98% accuracy.

The immediate answer for applicants is that occupation and location matter, but neither determines whether a vacancy will offer remote work. Employer decisions remain a major dividing line: companies in the same industry seeking talent in the same occupations advertise substantially different arrangements. Separate current data point in the same direction—the Indeed Hiring Lab’s August UK analysis put remote or hybrid language in 16.9% of postings at the end of June, close to the series peak despite pressure for more office attendance.

Five countries share the rise, not one uniform market

The new study shows a common post-2019 expansion but not equal availability across countries. Its national comparison accounts for differences in occupational mix, reducing the risk that a country appears more remote simply because its vacancies are concentrated in occupations that can more readily be performed away from a workplace.

• United States: advertised remote availability moved into a growth tier of more than three times its 2019 share.
• Australia, Canada, New Zealand and the United Kingdom: each recorded growth of at least five times its 2019 share.

Those multiples describe change from different starting points; they are not a ranking of where applicants now have the greatest chance of finding remote work. The study also identifies substantial variation within countries, including differences among cities, industries and occupations.

Current hiring conditions can move differently from the longer trend. A Bullhorn labor-market update dated July 28 described remote openings as stabilizing after their pandemic-era peak, with a seasonal rebound in the second quarter of 2026 and flexibility still differentiating employers competing for specialized workers. That does not contradict the working paper: a market can remain far above 2019 levels while growing slowly or retreating from a later high.

Employer choice survives controls for job type

The study’s most consequential result is the variation that remains after broad explanations are narrowed. Employers within the same industry, recruiting in the same occupational categories, still differ markedly in how often their advertisements explicitly offer remote work.

• Occupation sets the feasibility boundary. Tasks, equipment, security requirements and in-person service needs affect whether remote work is possible.
• Location shapes the available market. National rules, local labor supply and an employer’s operating footprint can limit where a remote employee may be based.
• The employer sets the advertised offer. Companies can organize comparable work differently even when they compete for similar candidates.

This is why an occupation-level label such as “remote-capable” is incomplete. It describes whether work could plausibly be performed remotely, not whether a particular organization has designed the role that way or is willing to make the arrangement explicit during recruitment.

The company comparison also should not be read as proof of a single policy covering every vacancy. Business units, teams and managers may advertise different requirements inside one organization. The evidence establishes employer-level variation in postings, not perfect consistency within each employer.

Job advertisements measure offers, not where employees work

The classifier identifies vacancies whose text says a new employee may work remotely at least one day per week. The measure therefore includes both hybrid and fully remote roles, which can impose very different commuting and location requirements.

It also measures stated hiring terms rather than completed employment arrangements. A company may permit flexibility without mentioning it in an advertisement, negotiate different terms during recruitment or alter an arrangement after a person starts. The dataset does not establish whether each vacancy was filled or where the eventual employee performed the work.

That boundary matters when comparing job-posting figures with workforce surveys. Postings describe the flow of advertised opportunities for prospective hires; employee surveys describe the stock of people already working under arrangements that may have been agreed earlier. A remote-work share from one population cannot be substituted directly for the other.

Company-level searches expose the variation occupation filters hide

For candidates, the findings make the employer a necessary unit of comparison. Occupation filters can identify relevant work and location filters can define the legal or practical hiring area, but company-level results reveal which organizations are currently putting remote or hybrid terms into writing.

A useful comparison holds the role as constant as possible: review similar vacancies across competing employers, then examine whether flexible language recurs across several openings at each company or appears only in an isolated advertisement. The wording still needs to distinguish fully remote work from a hybrid schedule and to identify any country, state, province or time-zone restrictions.

This approach does not predict what a company will offer in future negotiations. It separates observable vacancy terms from assumptions based on an occupation’s reputation, a city’s technology profile or an employer’s general public position on office attendance.

The expansion is established; its next direction is not

The August release is a working paper, not a final peer-reviewed publication. Its scale and human-checked classifier strengthen the measurement of advertised remote work, while the independent 2026 hiring indicators show that flexible postings remain a meaningful part of current recruitment in both the U.S. and U.K.

What is established as of August 24 is the large increase since 2019 and the persistence of differences among comparable employers. What remains unresolved is whether those company-level gaps will narrow as remote hiring stabilizes or widen as organizations make different choices about offices, recruitment areas and job design.