bramforgelabs.com › mental models › reward & punishment superresponse
Mental Model #11 · Charlie Munger

Reward & Punishment Superresponse

"Show me the incentive and I'll show you the outcome."

↑↑
Reward
→ more
The Principle
Responses are disproportionate, not proportional
Humans don't respond to incentives linearly. Small rewards trigger large behavioural shifts. Small punishments can eliminate behaviour entirely. The prefix super is the whole point.
↓↓
Punish
→ less

The superresponse curve

Incentive strength → Behaviour change → linear super- response gap = surprise
"
Never, ever, think about something else when you should be thinking about the power of incentives.
— Charlie Munger, Poor Charlie's Almanack
Classic illustration
FedEx overnight sorting · the experiment
01
Hourly pay
Workers paid per hour. Shifts dragged. Packages missed flights.
→
02
Change the structure
Fixed shift pay. Leave when done. One rule change.
→
03
Startling speed
Packages sorted rapidly. Same workers. Same managers. New incentive.
Aligned vs misaligned incentives
✓
Aligned: Equity compensation
Founders and employees rewarded when the company grows. Long-term thinking follows naturally.
incentive = outcome
✗
Misaligned: AUM fees
Advisors paid on assets under management, not returns. Growing the pot matters more than performance.
incentive ≠ outcome
✓
Aligned: Salary + shipped features
Engineers praised and promoted for working code, not hours logged. Output quality improves.
incentive = outcome
✗
Misaligned: Billable hours
Law and consulting firms rewarding time spent encourage complexity, delay, and churn — not resolution.
incentive ≠ outcome
⚠ Munger's warning: good people, bad systems