Measuring What Exhaustion Conceals
Section II · THE CARNEGIE SYSTEM · Andrew Carnegie · Volume I
The Mechanism
Worker exhaustion is invisible in standard metrics until you change the denominator. Productivity per hour, defect rates, injury rates, and turnover costs reveal what total output conceals.
The Story
Captain Jones at Edgar Thomson proved eight-hour shifts outperformed twelve-hour shifts: 7-12% increase in tonnage per man-hour, 25% reduction in defect rates, proportional reduction in injuries, net profitability improvement when full costs were allocated. His successors reverted to twelve-hour shifts by measuring different metrics that concealed the costs workers bore. Fifty years later, Kellogg's six-hour experiment documented the same pattern: 4% increase in hourly productivity, one-third reduction in reject rates, near-zero turnover. The proof persisted only as long as someone with power maintained the measurement system that revealed it. When Jones died, the measurement died. When Kellogg's reformers retired, the priorities changed.
Application Scenarios
The remote work debate.
Organizations measuring "time in seat" rather than "output per hour" repeat Carnegie's successors' error. Run the Captain Jones experiment yourself: for one quarter, measure your team on deliverables per hour of focused work rather than hours logged. Track quality metrics (error rates, revision cycles, customer complaints) alongside quantity. If per-hour output rises while total hours fall, you have found the Jones proof in your own data. The harder step: publish the results internally. If leadership ignores them in favor of return-to-office mandates, you have also found the Jones mortality: the proof lives only as long as someone with authority protects the measurement system.
Any organization evaluating burnout or attrition.
Your exit interviews say "better opportunity" or "compensation." These are polite fictions. The real cause often lives in the denominator nobody checks. Pull three months of timesheets for the department with the highest attrition. Calculate output per hour, not total output. If output per hour has been declining while total hours have been rising, your people are not leaving for better pay. They are leaving because the organization has been extracting more hours for less productivity per hour, and the people smart enough to notice are the ones with the best outside options. This is the same dynamic Captain Jones documented: the twelve-hour shift looks productive until you measure what it costs in quality, safety, and the systematic departure of your best operators.
Critical Warning
The eight-hour proof is fragile. It persists only as long as someone with power maintains the measurement system that reveals it. The proof requires institutional commitment, not just analytical demonstration.