How to Think About Incentives
To predict behaviour, look at what people are rewarded for — not what they say.
Also known as: Follow the incentives, Cui bono
intermediate Attributed to Charlie Munger (popular formulation) · Economics; behavioural analysis of incentives
Incentives — the rewards and penalties people face — shape behaviour more reliably than intentions, values, or stated goals. To understand or predict what a person or system will do, work out what it's actually being rewarded for, and expect that to happen.
What it is
Charlie Munger's advice was blunt: "Show me the incentive and I'll show you the outcome." People respond to the rewards in front of them, often without even noticing. If you want to know why a system behaves as it does, don't start with what it's supposed to do — start with what it pays off to do.
The subtle danger is the difference between the incentive you intended and the one you actually created. Reward salespeople on volume and you'll get volume — including sales that shouldn't have been made. Pay a call centre on call speed and calls get short, not resolved. Whenever you measure and reward a proxy, people optimise the proxy, sometimes at the expense of the real goal — a pattern known as Goodhart's Law.
Thinking well about incentives means two things: reading existing situations by asking cui bono ("who benefits?"), and, when you design a rule or reward, using inversion and second-order thinking to ask how a clever, self-interested person could satisfy the letter of the incentive while defeating its purpose.
Worked example
A government, wanting fewer venomous snakes, offers a bounty for each dead cobra. First-order: people hunt cobras, and numbers drop. Second-order: enterprising residents start breeding cobras to collect bounties. When the scheme is scrapped, the now-worthless snakes are released, and the population rises above where it began. The incentive was aimed at dead cobras and got exactly that — including farmed ones. The intended goal and the rewarded action had quietly come apart.
Failure mode — when it misleads
Incentive-thinking can curdle into cynicism — assuming everyone is a pure self-interested optimiser and no one ever acts on principle, which is both false and corrosive. People are also moved by identity, norms, and habit, not money alone. The framing misleads if you treat incentives as the only lever; it's a powerful first question, not the whole answer. And crude incentives you design can backfire precisely because people are cleverer than your rule.
How to apply it
- To predict behaviour, ask what the person or system is actually rewarded for.
- Read situations with "cui bono?" — who benefits from this outcome?
- When designing a reward, ask how someone could game it (invert it first).
- Watch for the gap between the metric you reward and the goal you want.
- Remember incentives aren't only financial — status, identity and norms count too.
Related entries
Sources & further reading
Poor Charlie's Almanack
by Charlie Munger (ed. Peter Kaufman) · book
Munger's talk on the psychology of misjudgment features the power of incentives prominently.
Affiliate link — we may earn a commission, at no extra cost to you.