A new breakdown every week — read the newsletter →
PeopleBusinessesTools
Strategies Mental ModelsDecision Tools Business ModelsFrameworksMoats
Learn Book SummariesReading Lists GuidesQuote CollectionsLearning Paths
Artificial IntelligenceNewsletter AboutContact
Human nature

Incentives

Show me the incentive and I'll show you the outcome — people respond to rewards.

Overview

The incentives model holds that to understand or predict behaviour, you look first at what the system rewards and punishes. People, often unconsciously, move toward what benefits them — so a system's incentives, more than its stated goals or good intentions, shape what actually happens. Munger called misjudging incentives one of the most common and costly errors.

The corollary is powerful: if you want to change behaviour, change the incentives, not the exhortations. And beware incentives that quietly reward the wrong thing.

When to use it

Predicting or shaping how people will behave in any system.

How to apply it

Step 1

Ask what's rewarded

In any system, identify what actually earns people money, status, or ease.

Step 2

Follow the incentive, not the intention

Predict behaviour from the rewards, not the stated goals.

Step 3

Look for perverse incentives

Check whether the system accidentally rewards the wrong behaviour.

Step 4

Change behaviour by changing incentives

To shift what people do, shift what pays — not just what you ask.

Common pitfalls

  • Assuming stated goals or good intentions will override the actual incentives.
  • Creating perverse incentives that reward the wrong behaviour.
  • Ignoring non-monetary incentives — status, ease, belonging — which often dominate.

Frequently asked questions

Why are incentives so powerful?

Because people reliably move toward what benefits them, often unconsciously — so the rewards in a system shape behaviour more than its stated aims.

What is a perverse incentive?

One that unintentionally rewards the wrong behaviour — like paying for speed and getting unsolved problems closed quickly.

Related models