The "9-to-5 model" never described how most U.S. workers actually worked. It described an industrial-era expectation that became a white-collar default through the post-war period and that has been eroding for forty years — first through professional-services overwork, then through manufacturing-shift atomization, then through the rise of independent and contingent work, and finally through the pandemic-era remote-work shift that broke whatever was left of the model. The genuinely interesting question is not whether the 9-to-5 model is dead. It is. The interesting question is what comes after it, what the empirical record says about which alternatives actually work, and who gets access to them.
The argument here is that "flexibility" is not one thing. It is a stack of distinct policy and product designs — schedule control, location flexibility, hours reduction, time-off generosity, and predictability — each with its own evidence base and its own distributional consequences. Nick Bloom's WFH Research, the 4 Day Week Global trial data, the Iceland four-day-week trials, the OECD's working-time research, and Susan Lambert's research on schedule predictability for hourly workers all converge on a clear picture: flexibility done well produces measurable productivity and well-being gains; flexibility done badly produces income volatility, work intensification, and worse outcomes for the workers it claims to serve.
The five dimensions of flexibility
Schedule control
The ability to choose when one works, within constraints. The empirical evidence on schedule control for knowledge workers is strongly positive. The longitudinal data from Microsoft's Work Trend Index, the WFH Research project, and a long tradition of organizational-psychology research (including Ellen Galinsky's body of work at the Families and Work Institute) finds that schedule control is among the most-valued and most-predictive flexibility dimensions for retention. The catch is that schedule control is rare for hourly and shift workers, who experience the opposite — schedule unpredictability — as a primary source of household financial volatility.
Location flexibility
The hybrid and remote question. Bloom, Barrero, and Davis's continuing WFH Research project has produced the most-cited body of data on remote work since 2020. The findings: hybrid arrangements (typically 2–3 days remote per week) produce higher self-reported well-being and equivalent or modestly higher productivity than fully in-office; fully remote produces mixed outcomes that depend heavily on individual fit and role; the in-office productivity premium that some firms have invoked to justify return-to-office mandates is, on average, smaller than the workers value at $10,000-plus per year in non-cash compensation. For the full treatment, see our flagship Remote Work, Year Six →.
Hours reduction
The four-day workweek experiments. The 4 Day Week Global trials, evaluated by researchers at Cambridge, Boston College, and Autonomy across more than 60 UK companies in 2022, produced reductions in burnout, turnover, and stress with no measurable revenue decline. 92% of participating firms chose to continue. Subsequent replications in Belgium, Iceland, Spain, Portugal, and Japan have produced consistent results. Microsoft Japan's 2019 trial showed a 40% productivity-per-worker increase. The four-day week is one of the most-studied employment innovations of the past decade and the empirical record is unusually strong.
Time-off generosity
The U.S. is one of two OECD countries without statutory paid annual leave (the other is the Marshall Islands), per the OECD Family Database. The European mandatory minimum is 20 days of paid leave plus public holidays. The U.S. average for full-time private-sector workers is about 11 days according to BLS National Compensation Survey data, with very large variance and roughly a quarter of private-sector workers having no paid vacation at all. The empirical evidence on paid leave's effects on productivity, retention, and health is well-developed, including work by Maya Rossin-Slater and Jenna Stearns on parental leave specifically.
Schedule predictability
The often-overlooked dimension, and the most consequential one for the bottom half of the wage distribution. Susan Lambert's research at the University of Chicago, the Aspen Institute Job Quality Center's body of work, and Joan Williams's work at UC Hastings document that unpredictable scheduling — last-minute shift changes, clopening shifts, on-call expectations — is the most documented driver of mental-health and household-financial degradation among hourly workers. The "fair scheduling" ordinances in Oregon, Seattle, San Francisco, and New York City all target this dimension, and the empirical evaluations show measurable reductions in worker stress and household financial volatility.
The distributional question
The honest framing is that flexibility is currently distributed in approximately the opposite shape of where it would do the most good. Knowledge workers in white-collar professional roles have the most schedule control, the most location flexibility, the most generous time off, and the least scheduling unpredictability. Hourly workers in retail, food service, healthcare support, and warehouse occupations have the least of all five. The "flexibility" conversation in HR and management literature has been disproportionately focused on the workers who need it least.
The OECD's Negotiating Our Way Up: Collective Bargaining in a Changing World of Work (2019) and the Aspen Institute's Job Quality Center work argue that the most consequential flexibility reforms for the bottom of the wage distribution are predictable scheduling, advance notice, and minimum-shift guarantees — not the four-day-week or schedule-control conversations that dominate the white-collar discourse.
The employer policy levers that work
Three policy and operational changes have the strongest empirical record.
Coordinated hybrid for knowledge workers. The "core hours" or "coordinated hybrid" model — defined in-office days for collaboration and protected remote days for focus — outperforms both fully-in-office and fully-remote on retention and productivity in the WFH Research data.
Four-day-week pilots with full productivity backing. The structured trials, with rigorous evaluation, have outperformed expectations consistently. Firms that have committed permanently (Buffer, Bolt for a period, several UK firms post-trial) have maintained the outcomes.
Predictable scheduling for hourly workers with two weeks' advance notice and minimum-shift guarantees. Where state and city ordinances have required this (Oregon, Seattle, San Francisco, NYC), the empirical evaluations show worker-side gains that exceed the small implementation costs to employers. The case for federal predictable-scheduling legislation, often discussed and never passed, is strong.
The four-day-week question, taken seriously
The four-day-week is the most discussed and the most studied of the alternative-schedule models. The accumulating evidence base, including the 2023–2024 follow-up evaluations of the UK trials by Autonomy and Cambridge researchers, supports the conclusion that the model works at a structural level for many but not all employer types. The conditions under which it works: knowledge work where output is what is measured rather than hours, organizations with sufficient management discipline to redesign workflows rather than just shorten them, and roles where the customer-coverage demands can be met across a five- or seven-day week with four-day-per-worker scheduling.
The model is harder, but not impossible, for retail, healthcare, and customer-facing roles. The Iceland trials, the Spanish pilots, and several U.S. health-system pilots have produced workable variants for these settings.
The harder honest reading is that the four-day-week is a workload redesign disguised as a scheduling change. Firms that adopt it without redesigning workflows produce work intensification rather than work reduction, and the well-being gains disappear. The firms that succeed are the ones that took the four-day-week as an excuse to do the workload-rationalization they should have done anyway.
The 9-to-5 model is dead. The replacement is not "flexibility" in the singular. It is a stack of five distinct policy choices, currently distributed in roughly the opposite shape of where they would do the most good.
The "redefine work-life balance" frame has produced a generation of articles that are right about the direction and vague about the destination. The destination, at the level of the evidence, is the five-dimension stack: schedule control, location flexibility, hours reduction, time-off generosity, and schedule predictability. The high-leverage policy and employer choices in each dimension are now well-documented. The remaining work is mostly distributional — getting these dimensions to the workers who do not currently have access to them. That work runs through federal paid-leave legislation, state predictable-scheduling laws, sectoral collective bargaining where applicable, and employer-level redesign for the firms that have not yet caught up. None of it requires more rhetoric about "flexibility." It requires execution on policy choices the evidence already supports.
Updated May 21, 2026. This piece was substantively rewritten as part of NWLB's 2026 editorial refresh.
