The Incentives Razor
Decode behavior by following incentives before assigning blame
- Difficulty
- Moderate
- Time to result
- ~ongoing to results
- Steps
- 5
- Confidence
- 80%
The Incentives Razor is a decoding lens: before you judge a person, institution or headline, ask what the incentives of the system are. Once the incentive is visible, otherwise baffling behavior starts to make sense. The razor has two edges. As an information filter, heavily discount claims that conveniently align with the speaker's incentives, and heavily weight claims that go against their incentives or identity, because those carry far more truth. As a design tool, change the incentive rather than working harder inside the old one when you want to change an outcome. Paired heuristics make it fast to apply: never ask a barber if you need a haircut, and never attribute to conspiracy what incentives and incompetence already explain.
Origin
George Mack and Chris Williamson describe becoming obsessed with incentives, drawing on Charlie Munger's FedEx story and Naval Ravikant's razor, 'never attribute to conspiracy what is more easily explained by incentives and incompetence.' George adds his own filter: downweight anything that conveniently serves the speaker's incentives.
Core principles
- 01Once you understand the incentive, you begin to understand everything
- 02Never ask a barber if you need a haircut
- 03Discount claims that align with the speaker's incentives; weight claims that go against them
- 04Never attribute to conspiracy what is explained by incentives and incompetence
- 05Change the incentive and the outcome changes, often dramatically
How to run it
- 1
Ask what the incentives are
For any behavior, claim or institution, map the incentives of the system before you judge it.
Pro tip Once you understand the incentive you begin to understand everything.
- 2
Discount incentive-aligned claims
When someone praises the very thing that benefits them, treat the information as suspect.
Watch out Don't 'download' a recommendation just because it's confidently given — check whose interest it serves.
- 3
Weight against-incentive claims
When someone criticises their own city, job or partner, the truth content is far higher.
Pro tip People rarely swim against their own incentives, so when they do, believe it.
- 4
Check incentives before conspiracy
Before assuming coordinated malice, ask whether incentives plus incompetence already explain the outcome.
- 5
Redesign the incentive to change the outcome
To fix a system, tweak what is rewarded rather than exhorting people to try harder within it.
Pro tip One well-placed incentive tweak can compound absurdly.
Watch out Anticipate how people will game the new incentive before you deploy it.
In the wild
Britain paid ship captains per prisoner who boarded for Australia, and roughly a third died en route because the captains were already paid. Switching payment to prisoners who arrived alive pushed the survival rate from around 66% toward 99%. One incentive change, no new rules or oversight, transformed a lethal system.
→ Survival on the voyage jumped from roughly two-thirds to near-total.
Munger's story: FedEx could not get deliveries done on time, with staff working late into the next day. Someone had the idea to stop paying by the hour and start paying by the completed shift. The chronic delay problem simply vanished once pay was aligned with finishing the work.
→ A persistent delivery failure disappeared after the pay structure changed.
To cut congestion, Greece let cars with odd-numbered plates drive one day and even-numbered plates the next. Wealthy drivers simply bought a second car with the opposite plate, and congestion got worse. The designers ignored how people would game the incentive.
→ A congestion fix produced more congestion, not less.
Common mistakes
Downloading claims that serve the speaker
Believing someone's glowing take on their city, only for them to leave it three years later.
Reaching for conspiracy first
Assuming coordinated malice when incentives and incompetence already explain the outcome.
Exhorting instead of redesigning
Trying to fix a system with willpower and pressure rather than changing what it rewards.
Is it for you?
Best for
Anyone making judgment calls about people, institutions or news where motives are in play.
Not ideal for
Low-stakes situations with no hidden interests, where over-analysing motives just adds noise.
From the transcript
“once you understand the incentive you begin to understand everything”
“never as your Barber if you need a haircut cuz guess what he'll think you need a haircut”
“never attribute to conspiracy what is more easily explained by incentives and incompetence”
From the episode
13 Life-Changing Ideas You’ve Never Heard Of - George Mack - #801
George Mack