Conversion Science · established evidence

What a Government Actually Measures: Inside the BLS American Time Use Survey

Last reviewed 2026-07-20. Written by Chandranshu Kumar, Founder, Raveneye Global. · 8 min read

Nearly every widely quoted "hours per day" media statistic, Nielsen's Total Audience Report, DataReportal's global overview, GWI's panel data, comes from a vendor with a commercial stake in the number, usually an advertising, media-measurement, or social-analytics company. The American Time Use Survey (ATUS), run annually by the US Bureau of Labor Statistics, is different on three counts: it is government-run rather than advertiser-funded, it measures a whole day rather than a single medium in isolation, and its full methodology is disclosed rather than proprietary. In 2025, ATUS found television remained the single largest leisure activity at 2.6 hours a day of 5.2 hours total daily leisure, and that time spent on computers and games for leisure rose from 25 minutes a day in 2015 to 37 minutes in 2025, while socializing time fell from 41 minutes to 35 minutes over the same period. ATUS is not a perfect instrument. It relies on self-reported recall rather than passive tracking, and it was not built to measure marketing-relevant categories like "AI-answer engine use." But as the closest thing to neutral ground truth on how a population actually spends its day, it is the anchor figure set that any honest media-diet claim should be checked against, not the vendor panel that happens to support the argument being made.

Why "who funded the number" matters before "what the number says"

A large share of the media-time statistics circulating in marketing content trace back to firms whose core business is selling advertising, audience measurement, or social-media analytics. Nielsen's Total Audience Report, DataReportal's Digital Global Overview (with We Are Social and Meltwater), and GWI's panel research are all genuinely useful, methodologically disclosed to varying degrees, and worth citing with attribution. They are also produced by companies with a commercial interest in showing large, growing numbers for digital and social media time, because those numbers justify the ad inventory and subscription products those same companies sell.

This is not an accusation of fabrication. It is a structural observation about incentives, and it is the reason a government-run, methodologically transparent alternative matters as a cross-check, not a replacement.

What ATUS actually is

The American Time Use Survey is conducted annually by the US Bureau of Labor Statistics. Unlike a media-specific panel that asks "how much time did you spend on social media," ATUS asks respondents to account for an entire day, all twenty-four hours, across every activity category from sleep and work to leisure, care work, and media use. That whole-day design is a meaningfully different instrument from a single-medium panel: it forces every activity to compete for the same fixed twenty-four hours, which is a closer structural match to how attention actually behaves than a survey that only asks about the media categories a vendor wants to sell against.

The survey is disclosed in full: methodology, sample design, and questionnaire are all public, which is not the case for several of the vendor panels commonly cited in marketing content.

What the most recent results show

The 2025 results, released in 2026, found television watching remained the single largest leisure activity, at 2.6 hours a day out of 5.2 hours of total daily leisure. Two directional shifts stand out over the prior decade: time spent "playing games or using a computer" for leisure rose from 25 minutes a day in 2015 to 37 minutes a day in 2025, while time spent socializing and communicating fell from 41 minutes a day to 35 minutes, with the share of people socializing on a given day dropping from 38 percent to 30 percent. These are the kind of decade-scale directional shifts that a government instrument, run consistently on the same methodology year after year, is well suited to capture.

The limits worth naming honestly

ATUS is not without real constraints, and an honest treatment of the data says so plainly. It relies on self-reported time diaries rather than passive, device-level tracking, which introduces the same recall and rounding biases present in any self-report instrument. It was designed to capture broad activity categories, not the granular, marketing-relevant distinctions a business actually needs, such as which specific platforms carry attention within the broad "computer or leisure" category, or how much time goes to newly emerging categories like AI-chatbot use, which the survey's categories were not built to isolate.

It is also a population-wide, US-national instrument. It says nothing about how time use differs by the specific demographic, geographic, or interest-based audience a given business actually serves, which is exactly the domain-specific gap a population-level government survey, by design, cannot fill.

How to use ATUS honestly alongside vendor data

The correct use of ATUS is as an anchor, not a replacement. Vendor panels like Nielsen's Total Audience Report or GWI often provide finer-grained, more current, and more platform-specific detail than ATUS captures, and are genuinely useful when attributed properly and read with their commercial origin in mind. ATUS is the instrument to check those vendor claims against when a directional trend is in dispute, precisely because it has no commercial stake in the answer.

A media-diet claim that cites only a single advertiser-funded panel, with no cross-check against a disclosed, non-commercial source, is a weaker piece of evidence than one that names both and shows where they agree and where they diverge.

The evidence

Key findings, with their sources

In 2025, television watching remained the single largest leisure activity among Americans at 2.6 hours per day, out of 5.2 hours of total daily leisure time.

established U.S. Bureau of Labor Statistics, American Time Use Survey, 2025 results (released 2026, bls.gov/news.release/atus.htm).
Time spent "playing games or using a computer" for leisure rose from 25 minutes per day in 2015 to 37 minutes per day in 2025.

established U.S. Bureau of Labor Statistics, American Time Use Survey, 2025 results.
Time spent socializing and communicating fell from 41 minutes per day in 2015 to 35 minutes per day in 2025, with the share of people socializing on a given day dropping from 38% to 30%.

established U.S. Bureau of Labor Statistics, American Time Use Survey, 2025 results.
Nielsen's Total Audience Report figures (a vendor-sourced comparator) moved from roughly 11 hours 27 minutes per day (Q1 2019) to over 12 hours 20 minutes per day (April 2020), with an explicit methodological caveat that simultaneous multi-device use inflates the additive total.

established but vendor-sourced Nielsen, Total Audience Report releases, 2018-2020 (nielsen.com).

Reference

Glossary

American Time Use Survey (ATUS): An annual survey run by the US Bureau of Labor Statistics that measures how Americans spend their entire day across all activity categories, using disclosed government methodology.
Time diary methodology: A survey method where respondents account for their activities across a full day, forcing every activity to compete for the same fixed twenty-four hours, rather than asking about a single medium in isolation.
Vendor-sourced data: Statistics produced and published by a commercial firm, such as an advertising, media-measurement, or analytics company, which may have a business interest in the direction of the finding.
Self-report bias: Systematic inaccuracy in survey data that arises because respondents recall and report their own behavior imperfectly, a limitation present in ATUS and most time-use instruments.

Straight answers

Frequently asked questions

What is the American Time Use Survey?

It is an annual survey run by the US Bureau of Labor Statistics that measures how Americans spend an entire day across all activity categories, using publicly disclosed government methodology rather than a proprietary vendor panel.

Why trust ATUS over a vendor report like Nielsen or DataReportal?

ATUS is government-run with no commercial stake in the outcome and measures a whole day rather than a single medium in isolation. Vendor panels are genuinely useful and methodologically disclosed to varying degrees, but they are produced by firms with a business interest in showing large digital or social media numbers, which is a reason to cross-check, not a reason to dismiss them.

What are the real limits of ATUS data?

It relies on self-reported time diaries rather than passive tracking, its categories were not designed to capture granular, marketing-relevant distinctions like specific platforms or emerging categories such as AI-chatbot use, and it reports national population averages that do not reflect any one business's specific audience.

Did Americans spend more time on computers and games for leisure in 2025 than in 2015?

Yes. ATUS found leisure time on computers and games rose from 25 minutes a day in 2015 to 37 minutes a day in 2025, while time spent socializing fell from 41 minutes to 35 minutes over the same period.

Provenance

Sources

U.S. Bureau of Labor Statistics, American Time Use Survey, 2025 results, released 2026, bls.gov/news.release/atus.htm (established, primary)bls.gov
Nielsen, Total Audience Report, recurring releases, nielsen.com (established, vendor-sourced)
DataReportal (with We Are Social/Meltwater), Digital: Global Overview Report, annual, datareportal.com (established, vendor/compiled)
Reuters Institute for the Study of Journalism, University of Oxford, Digital News Report, annual (established, primary, disclosed methodology)
GWI (formerly GlobalWebIndex), The Global Media Landscape, annual, gwi.com (vendor panel data)

Every figure above is attributed to a real, dated source and tagged with its evidence tier. Where a claim could not be verified to a primary source, it is not stated as fact.