Conviction-Humility Decision Rule
Match the speed of action to forecast confidence and the cost of waiting
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 6
- Confidence
- 91%
The Conviction-Humility Decision Rule separates knowledge of current conditions from confidence about the future. Leaders first assess the present state, the objectives they must balance, and whether the directional response is obvious. They then evaluate conviction in the forecast rather than pretending uncertainty does not exist. Strong evidence can justify decisive movement, especially when delay threatens one side of the mandate. Weak conviction favors preserving optionality and learning with time, provided current conditions are stable enough to wait. Humility is not indecision: it means recognizing forecast limitations, hearing competing views, and avoiding false precision. Conviction is not certainty: it is sufficient evidence to act proportionately while remaining prepared to adjust.
Origin
Tom Barkin explained how he balances conviction with humility when making monetary-policy decisions amid tariff and economic uncertainty on Masters in Business.
Core principles
- 01Act decisively when the direction and forecast are clear
- 02Wait for information when forecast conviction is weak
- 03Remain aware of personal forecasting limits
- 04Balance risks on both sides of the decision
- 05Move faster when evidence makes delay more costly
How to run it
- 1
Define the mandate
State the objectives that the decision must balance and the acceptable range for each. Make trade-offs explicit before considering a forecast.
Pro tip Use observable outcomes rather than vague goals.
Watch out An undefined mandate lets confidence substitute for relevance.
- 2
Assess the present
Separate what is known about current conditions from assumptions about the future. Determine whether the current state already demands action.
Pro tip Ask whether the direction would be obvious if no forecast were available.
Watch out Do not let a dramatic scenario overwrite stable current evidence.
- 3
Rate forecast conviction
Evaluate the clarity, consistency, and evidentiary support of the forecast. Listen to informed disagreement and identify the assumptions driving each view.
Pro tip Express conviction as a range and list what evidence would change it.
Watch out Confidence and accuracy are not the same thing.
- 4
Compare action with waiting
Estimate the consequences of acting now, acting later, and being wrong in either direction. Include the informational value of waiting.
Pro tip Favor reversible steps when risks are balanced.
Watch out Waiting is itself a decision with costs.
- 5
Choose a proportionate move
Act decisively when conviction and urgency are high; preserve optionality when conviction is weak and current conditions allow it. Match the size of the move to the strength of the evidence.
Pro tip Predefine escalation conditions before new information arrives.
Watch out Humility should not become a permanent excuse for inaction.
- 6
Learn and adjust
Review new evidence and compare outcomes with the assumptions behind the decision. Change course without treating revision as failure.
Pro tip Record which assumptions proved wrong to calibrate future conviction.
Watch out Do not defend a prior decision merely to appear consistent.
In the wild
Inflation remained above target, unemployment was low, and policy was modestly restrictive, but the destination of tariff and fiscal policy was unclear. Rather than act on a low-conviction forecast, Barkin favored waiting to learn more while current conditions left room to do so.
→ The decision preserved flexibility until the policy environment became clearer.
A founder sees stable current revenue but receives conflicting signals about demand six months ahead. Instead of committing to a large permanent team, the founder runs a limited hiring round, defines demand thresholds for further expansion, and reviews the decision after two sales cycles.
→ The company learns without accepting the full cost of an overconfident forecast.
Common mistakes
Confusing humility with passivity
Waiting is appropriate only when the value of information exceeds the cost of delay. Some conditions require action despite an uncertain forecast.
Pounding the table too early
Forceful presentation does not increase forecast quality. Strong claims should follow strong evidence rather than personality or status.
Demanding false precision
Complex systems rarely permit exact forecasts of peaks, troughs, or timing. Decisions should remain robust across a reasonable range of outcomes.
Is it for you?
Best for
It is best for leaders making reversible or staged decisions when present conditions are tolerable but future outcomes are unclear.
Not ideal for
It is not ideal when delay creates irreversible harm or when an emergency requires immediate action despite uncertainty.
From the transcript
“if, like me, you don't have that much conviction in the forecast, then you say, let's wait and see where we go.”
“I'm humble enough about my own forecasting ability. It just makes me think that, you know, I'll learn more. I'll learn more with time.”
“I think we're always going to be wrong.”
From the episode
Richmond Federal Reserve President and CEO Tom Barkin
Tom Barkin