Tell people that 85% of cabs in a city are Green and 15% are Blue, then describe a witness who identified the cab as Blue (80% accuracy). Most people say "Blue." The correct answer, incorporating the base rate, is actually Green (the math gives ~41% probability of Blue). But now tell them the Blue company is involved in a disproportionate number of accidents. Suddenly, they use the base rate perfectly. What changed?

The Framework

Kahneman's Chapter 16 distinguishes two types of base rates that produce dramatically different cognitive responses. Statistical base rates (85% of cabs are Green) are abstract, population-level numbers with no causal mechanism — System 1 ignores them because they don't fit into a story. Causal base rates (Blue company causes more accidents) provide a causal mechanism — a reason that's easy to integrate into a narrative — and System 1 processes them automatically.

The distinction explains a persistent puzzle in the base-rate literature: sometimes people respect base rates and sometimes they don't. The answer is that System 1 processes causal information automatically but treats statistical information as irrelevant background. If you want someone to incorporate a base rate into their judgment, you must transform the statistical base rate into a causal story.

Where It Comes From

Kahneman presents this distinction in Chapter 16 of Thinking, Fast and Slow as the resolution of conflicting findings in the base-rate neglect literature. Tversky and Kahneman's original work showed dramatic base-rate neglect in the Tom W and Linda problems. But subsequent research found cases where base rates were used. The causal/statistical distinction explains both findings: System 1 uses base rates that have causal force and ignores base rates that are merely statistical.

> "People who are told that there is a 10.3% chance of getting disease X from unprotected sex don't know what to think, but they will certainly worry if they learn that 1 in 10 participants got the disease." — Thinking, Fast and Slow, Ch 16

Cross-Library Connections

Hormozi's case study approach in $100M Offers intuitively applies this distinction: abstract conversion statistics ("87% success rate") are statistical base rates that prospects ignore. Individual transformation stories ("John went from $50K to $500K") are causal narratives that prospects process automatically.

Berger's storytelling principle in Contagious explains why stories spread and statistics don't: stories provide causal structure (beginning → challenge → resolution) that System 1 processes naturally. Statistics are abstract numbers that System 1 ignores.

The Implementation Playbook

Sales and Marketing: Transform statistics into stories. "95% customer satisfaction" is a statistical base rate that prospects ignore. "Here's how Sarah, who was skeptical just like you, transformed her business" is a causal narrative that activates System 1.

Risk Communication: "1 in 10 people who don't wear seatbelts die in crashes" (causal — not wearing it causes death) is more effective than "10% crash fatality rate for unbelted passengers" (statistical). Convert statistical risks into causal mechanisms.

Medical Communication: "Smoking doubles your risk of heart disease" (causal) is more effective than "Heart disease prevalence is 12% among smokers vs. 6% among non-smokers" (statistical). The causal framing integrates into the patient's story; the statistical framing is ignored.

Decision-Making: When you need to consider base rates in your own decisions, convert the statistical information into a causal mechanism. "90% of restaurants fail within 5 years" is easy to ignore. "Restaurants fail because they run out of cash when initial excitement fades and fixed costs persist" provides a causal mechanism you can evaluate against your own situation.

Key Takeaway

Statistical base rates are ignored because System 1 doesn't process abstract numbers. Causal base rates are used because they fit into narratives. If you need someone (including yourself) to incorporate base-rate information, convert it from a statistic into a cause.

Continue Exploring

[[Representativeness Heuristic]] — The mechanism that causes base-rate neglect: judging by resemblance rather than statistics

[[Narrative Fallacy]] — Why causal stories are processed and statistical information is ignored

[[Conjunction Fallacy]] — Another case where vivid causal detail overrides statistical logic

📚 From Thinking, Fast and Slow by Daniel Kahneman — Get the book