Why rewarding attendance lowers performance

A factory introduces an attendance bonus. Absenteeism falls and output per hour also falls. More people are present and less work is done per person. The two numbers moved in opposite directions in response to the same incentive.

What everyone sees

The bonus looks like a productivity tool: if people show up, work gets done. The distinction between presence and productivity is not visible in the policy, because the policy measures the first and assumes the second.

What is actually happening

Kerr described the folly of rewarding A while hoping for B: incentive systems routinely reward the measurable proxy while hoping for the unmeasured outcome. Prendergast reviewed the principal-agent literature and noted that when a measurable proxy replaces the actual objective, agents optimise the proxy. An attendance bonus is a textbook case: the measurable input (showing up) is rewarded, and the unmeasured output (productive work) is hoped for. Workers who would have been absent are now present but not necessarily engaged, and the bonus gives no reason to be.

Why it stays hidden

The divergence hides because attendance feels like a necessary condition for performance and is therefore treated as nearly equivalent to it. Managers monitor the metric that moved in the right direction — attendance — and attribute the output decline to other factors. The connection between the reward and the decline is obscured by the time lag and by the fact that each individual is now technically complying with the policy.