Chapter 30: Rare Events

Type: Chapter 30

Category: Psychology

Difficulty: Intermediate

Reading time: 6 min

Core idea

Rare events receive disproportionate attention and weight in decisions. There are two related but distinct mechanisms:

1. Probability overestimation: people overestimate the frequency of vivid, memorable rare events — shark attacks, plane crashes, terrorist attacks. This is availability heuristic: ease of recall inflates estimated probability.

2. Decision weight overestimation: even when probabilities are assessed correctly, people weight rare events disproportionately in decisions. A 1% chance of disaster receives more than 1% of the decision weight. This is the probability weighting function from prospect theory.

Both mechanisms produce over-attention to rare events and under-attention to statistically more significant but less vivid risks. Together, they explain why people pay heavily for unlikely protections (insurance against exotic risks) while underestimating common, mundane hazards.

Why it matters

Two routes to overweighting rare events

The overweighting of rare events has two distinct sources that are often conflated:

The availability route: vividness, recency, and media coverage inflate the estimated probability of rare events. People who overestimate the probability of plane crashes are not necessarily distorting the probability weighting function — they may simply have wrong probability beliefs driven by availability.

The decision-weight route: even with accurate probability estimates, the probability weighting function overweights low probabilities in decision calculations. A person who correctly estimates that a particular lottery has 1-in-10 million odds may still buy a ticket — because 1-in-10 million is weighted at more than 1-in-10 million in the subjective decision process. The decision weight is not the probability itself.

These two routes interact: availability inflates estimated probability, and the probability weighting function inflates the decision weight of whatever probability is estimated. The combination can produce very large overweighting of rare events.

When vivid rare events receive explicit attention

A key modifier: explicit attention to a rare event amplifies its decision weight. If you are asked to imagine a catastrophic failure before making a decision, the failure’s probability and weight both increase in the decision process. This has implications for:

• Risk communication: alerting people to rare but catastrophic risks can improve their protective behavior — but can also produce fear and distorted policy preferences if the risk is rarer than the framing suggests.
• Terrorism: terrorism succeeds partly because the reaction it provokes is disproportionate to the statistical risk. Loss aversion applied to loss-domain rare events (dread risks) produces outsized resource allocation.

Dread risk vs. statistical risk

Kahneman distinguishes dread risks — events that are catastrophic, uncontrollable, affect many people simultaneously (nuclear accidents, terrorist attacks, plane crashes) — from statistical risks — the mundane, distributed risks that kill more people but generate less fear (car accidents, diabetes, heart disease).

Dread risks produce disproportionate policy responses because they are available, emotionally activated, and involve controllability violations. Statistical risks receive proportionate or inadequate policy responses because they are statistically diffuse and emotionally inert.

Key takeaways

Key takeaways

• Two routes to rare event overweighting: availability (inflated probability estimates from vividness) and probability weighting (decision weights exceed true probabilities even when probabilities are correctly assessed).
• The probability weighting function overweights low probabilities: a 1% chance receives more than 1% of the decision weight — producing lottery-buying and insurance-buying from the same weighting mechanism.
• Explicit attention amplifies both overestimation and overweighting: asking someone to vividly imagine a failure increases both their estimated probability and their decision weight for that outcome.
• Dread risks vs. statistical risks: catastrophic, uncontrollable rare events trigger disproportionate fear and policy response; mundane, distributed risks are systematically under-attended.
• Terrorism and dread risk: the disproportionate attention and resource allocation triggered by terrorism is partly a rational response to uncertainty and partly a systematic probability weighting distortion.
• Policy implication: effective risk communication requires addressing both the availability-driven overestimation and the weighting-driven distortion — statistics alone do not correct either.

Mental model

click to zoom

Two routes to overweighting rare events — availability-driven probability overestimation and decision-weight overestimation.

Read it as: Rare events are overweighted through two separate routes that compound. Availability inflates the estimated probability — vivid events feel common. Probability weighting inflates the decision weight given to whatever probability is estimated — even correctly-estimated low probabilities receive more than proportional weight in decisions. Together, these routes produce behavioral responses to rare events that are systematically disproportionate to actuarial risk.

Practical application

Risk communication and availability cascades

Communicating risk statistics to people who have just experienced a vivid rare event is often ineffective — their availability-inflated probability estimate is resistant to statistical correction. Effective risk communication in the aftermath of dramatic events must address the emotional availability directly (acknowledge the event’s impact) before statistical correction can register.

Design implications:

• Insurance product design: the overweighting of low-probability losses makes exotic insurance products profitable even at negative expected value for customers. The probability weighting function is the psychological subsidy for many insurance add-ons.
• Security theater: highly visible security measures (airport shoe removal, visible police patrols) may provide less actual risk reduction than their cost implies — but they address the decision weight of dread risk rather than the actuarial risk, which is what drives policy preference.
• Pandemic and crisis communication: early, vivid, emotionally-loaded communication about disease risks can trigger the dread risk response — producing protective behavior but potentially also panic and policy overreaction. Calibrating the emotional load of risk communication is a practical challenge.

Example

A food company’s product was involved in one recall three years ago, affecting 40 customers with minor illness. Current safety standards are well above industry average. But surveys show that 35% of former customers avoid the brand. The recall was a rare event, but the availability-plus-weighting combination has produced persistent brand damage far exceeding the actual risk the event represented.

The company’s standard response — statistical reassurance (“less than 0.001% risk”) — is ineffective because it tries to correct an availability-inflated probability estimate with a number, while the decision weight remains inflated even if the probability belief is updated. Effective brand recovery requires reducing the emotional activation of the event (lowering availability) alongside providing statistical information.

Related lessons

Related

Chapter 29: The Fourfold Pattern The theoretical framework — how probability level and domain interact to produce four distinct risk attitudes. Chapter 13: Availability, Emotion, and Risk The affect heuristic and availability cascades — one of the two routes to rare event overweighting. Prospect Theory The probability weighting function within prospect theory — the second route to rare event overweighting.