Incentive Design Entry #0379 Classified Declassified

The reason a target hit exactly is suspicious

Genuine performance is spread out. When results pile up immediately above a threshold and thin out below it, the distribution is reporting the incentive.

No visual record attached The written record below is complete.
Plate 103 — a histogram with one wall and no doorway.

Intuition test — answer before you read on

Resolution times cluster just inside a target window and almost none fall just outside. What does the shape indicate?

A department reports ninety-five per cent of cases resolved within the target window, against a target of ninety-five. The following quarter it reports ninety-five point one. The distribution of individual resolution times, plotted, shows an unusual mass of cases finishing minutes before the deadline and almost none finishing minutes after it.

What everyone sees

Hitting a target exactly is read as competence: the team understood the goal and delivered it. Managers treat the tight clustering as evidence of process control, which is a reasonable interpretation of a well-run operation and an equally good description of a well-managed number.

What is actually happening

Discontinuities at incentive thresholds are a standard diagnostic. Bennett and colleagues documented ambulance response times bunching just inside an eight-minute target, and Burgstahler and Dichev found earnings distributions with a marked deficit just below zero and a surplus just above. Where a threshold carries a reward, the underlying distribution is reshaped near the cut point while remaining plausible everywhere else.

Why it stays hidden

The pattern hides because the summary statistic is the only figure reported. A mean or a headline percentage cannot show bunching; only the full distribution can, and full distributions are rarely published. Each individual case near the line also has a legitimate explanation, so the effect is only visible in aggregate — the one view the reporting format omits.

Look at the shape, not the summary. A cliff at the threshold is evidence about the incentive, not about the work.

Look at the shape, not the summary. A cliff at the threshold is evidence about the incentive, not about the work.

The hidden part — entry #0379

Collect this card

Look at the shape, not the summary. A cliff at the threshold is evidence about the incentive, not about the work.

0 / 10,000 collected

Sources & further reading 3
  1. Bennett, Bevan & Hood, "Have Targets Improved Performance in the English NHS?", BMJ, 2006
  2. Burgstahler & Dichev, "Earnings Management to Avoid Earnings Decreases and Losses", Journal of Accounting and Economics, 1997
  3. Bevan & Hood, "What’s Measured Is What Matters: Targets and Gaming in the English Public Health Care System", Public Administration, 2006

Circulate this file

Annotations are reserved for archive members.

Sign in to annotate