Clawback Incentive
Give the reward first, then make performance the condition for keeping it
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 99%
The clawback incentive reverses a conventional bonus. Instead of promising a reward after performance, give it at the beginning and explain that it will be taken back if a defined target is missed. The mechanism uses loss aversion: once people possess something, surrendering it hurts more than merely failing to receive it later. John List also describes the reward as a commitment device because it brings a future benefit into the present while preserving a consequence for inadequate effort. Match the metric to the setting: use controllable inputs for children when possible, but observable outputs when workplace inputs cannot be measured. Repeated use requires checking whether the effect depreciates and whether stress, resentment, or quitting offsets the gain.
Origin
List describes using clawbacks with Chinese manufacturing workers, preschool teachers, students, and his own children.
Core principles
- 01People feel the loss of something they own more strongly than a delayed gain
- 02Moving a future benefit into the present can create commitment
- 03The target should be observable and understood in advance
- 04Repeated incentives must be checked for depreciation and side effects
How to run it
- 1
Choose the target
Define the behavior or outcome that matters and can be verified. Prefer a controllable input when the recipient cannot fairly control the final output.
Pro tip For a child, reward study time or books read rather than a curved exam grade.
Watch out An uncontrollable target makes the incentive unfair.
- 2
Set the reward period
Use a bounded period with a clear beginning and end. List's manufacturing experiment granted bonuses on Monday and cleared them on Friday.
Pro tip Start with a short period that makes the contingency easy to understand.
Watch out A distant review weakens the connection between effort and consequence.
- 3
Grant the reward first
Put the valued reward in the recipient's possession at the start. Explain that meeting the target lets them keep it.
Pro tip Use a reward that is meaningful without being coercively large.
Watch out Do not disguise the clawback condition after giving the reward.
- 4
Apply the stated rule
Measure the agreed target at the end of the period and either leave the reward in place or take back the specified amount. Keep the rule consistent.
Pro tip Record the target before the period begins.
Watch out Moving the goalposts destroys trust.
- 5
Audit repeated effects
Compare later rounds with the first and inspect side effects as well as performance. Stop or redesign the incentive if depreciation, stress, or departures outweigh the benefit.
Pro tip Ask recipients whether the structure helps them commit.
Watch out A one-round improvement does not prove the incentive remains healthy over time.
In the wild
Workers in a Chinese manufacturing plant received a bonus on Monday. Teams kept it on Friday if they met a goal based on the number of non-defective items produced. List compared this clawback condition with a standard end-of-week bonus during a six-month experiment.
→ The clawback produced stronger effort than the standard incentive while remaining effective across repeated rounds.
List gives a child baseball cards or a movie pass on Monday and makes keeping it conditional on an hour of intensive mathematics study each night through Friday. The target is study effort rather than an exam score the child may not fully control.
→ List says his children nearly always meet the weekly goal and keep the reward.
Common mistakes
Rewarding an uncontrollable output
A recipient should not lose a reward because of factors outside their control, such as other students' performance on a curved exam.
Ignoring side effects
Performance gains can be outweighed if the incentive causes excessive stress, resentment, or attrition.
Assuming the first effect lasts
Behavioral incentives can depreciate, so repeated rounds need measurement rather than assumption.
Is it for you?
Best for
Measurable behaviors or outcomes where a valued reward can be granted safely before a short performance period.
Not ideal for
Ambiguous targets, people who cannot control the metric, or settings where taking back the reward would be harmful or impractical.
From the transcript
“at the beginning of the year i give them the incentive but then i tell them if they don't perform i will take away the…”
“the idea here is give somebody some money that will induce them to work harder”
“many workers actually like the clawback because they view it as a commitment device”
From the episode
#566: John List — A Master Economist on Strategic Quitting, How to Practice Theory of Mind, Learnings from Uber, Optimizations to Boost Donations, the Primitives of Decision-Making, and How Field Experiments Reveal Hidden Realities
John List