Remote work statistics

Choose the measure before choosing a headline. BLS’s reference-week question captures any paid work from home, including mixed-location weeks. Census’s commuting measure instead identifies where people usually worked. Both can be correct because the questions are different. Their percentages should not be described as competing estimates of precisely the same behavior.

The dates differ too. An annual observation, a particular month, and a one-time employee survey cover different windows. This page keeps each window visible rather than merging them into a single current rate. For a time series, use the same source and definition throughout, and check whether the data are seasonally adjusted before interpreting small monthly changes.

BLS distinguishes teleworking some hours from teleworking all hours in the reference week. A person can fall into the all-hours category during one week without having a contract that requires remote work indefinitely. Conversely, someone with an ongoing hybrid schedule can have a week with no remote hours.

Hours and people also need separate labels. The average telework-hours figure is calculated among teleworkers; it does not mean everyone in the workforce worked that many hours from home. The average number of teleworkers describes an average reference-week population, rather than everyone who ever used the arrangement during the year.

Pew and Gallup focus some questions on people whose jobs can be performed remotely. Those groups exclude many jobs that require an in-person location. A high share within the eligible group can therefore coexist with a lower share across the workforce. Removing “remote-capable” from a citation would materially change its meaning.

BLS’s age comparison also needs context. It shows a difference between groups, but does not isolate the role of occupation, education, industry, or job tenure. Use it to describe who reported teleworking in that period. It cannot establish that changing a worker’s age would itself change access to remote work.

Gallup’s schedule-control question asks who decides a hybrid arrangement. That is a separate issue from where employees work. A team can share the same broad location category while its members have very different choices over which days they attend or who approves a change.

For practical planning, separate location, availability, and coordination. A survey about working from home does not establish when colleagues are expected to respond, which meetings require attendance, or how a team should handle time-zone overlap. Those questions need an explicit team agreement. Treat the location figures as context for that conversation rather than a complete description of a working arrangement.

Pew’s return-to-office question describes a hypothetical change for people already working from home at least sometimes. Respondents reported how likely they would be to stay. This is useful evidence about the importance of flexibility to that group, but it is not a measurement of subsequent resignations.

Actual behavior can depend on job alternatives, pay, family responsibilities, and how a policy is implemented. The survey does not assign employers a universal expected turnover rate. When citing the result, keep the condition and eligible group in the sentence rather than replacing them with a claim that a share of all workers will quit.

The BLS tables warn that the 2025 annual averages exclude October, when data were not collected during the federal government shutdown. They therefore are not strictly comparable with full-year averages. A chart should show that gap in its note rather than imply that the missing month was measured or estimated in the usual way.

Census also explains that its standard annual commuting chart omits the experimental pandemic-year estimates for 2020. Missing or differently constructed observations are reasons to qualify a trend, not invitations to fill it with an unrelated survey. For a defensible comparison, keep the question, eligible population, period, and source consistent before discussing a change.