The belief that time and productivity are the same thing is remarkably persistent. Clock in, clock out, measure the hours — and assume the hours measure the work. It is a model built for a different era, when most work was physical. Knowledge work broke this relationship a long time ago.

The diminishing returns of long hours

Analysis of output data found that the relationship between hours worked and productivity is not linear — it curves sharply downward. Employees who work more than 50 hours a week see significant drops in output per hour, and beyond 55 hours the decline becomes steep enough that additional hours produce almost nothing of value (Pencavel, 2015). Sustained cognitive work depletes resources that require rest to restore.

The problem with activity-based measurement

Microsoft’s research on hybrid work identified “productivity paranoia” — managers who feel employees are not doing enough, even when output metrics suggest otherwise. The response is typically more visibility: more check-ins, more status updates. This actively harms the productivity it is meant to protect. Frequent interruptions fragment deep work, and employees optimize for looking productive rather than being productive (Microsoft, 2022).

Evidence from shorter workweeks

A large-scale UK pilot of four-day workweeks found that participating companies saw a 35% improvement in productivity, with 92% electing to maintain the model after the trial ended (Autonomy, 2023). When employees have less time, they use it more deliberately. Meetings get shorter. Work is structured around outputs rather than schedules.

Autonomy as a productivity lever

Research consistently finds that employees who have more control over how they organize their time are more engaged, report lower burnout, and produce better work (Garton & Mankins, 2020). Autonomy allows people to match the type of work to the conditions under which they do it best — and employees who can make those choices deliberately will consistently outperform those who cannot.

References