Availability, Emotion, and Risk

Core idea

After a major disaster — a flood, an industrial accident, a nuclear plant incident — people's perceived risk of that event spikes dramatically. Insurance sales surge, safety equipment sells out, policy changes are enacted. Over time, memory fades, the emotional salience decreases, and perceived risk declines back toward — sometimes below — pre-disaster levels. Then the cycle repeats.

This cycle is not irrational mood fluctuation. It is a predictable consequence of two heuristics operating in tandem: availability (ease of recall inflates probability estimates) and affect (emotional valence shapes judgments of risk and benefit). Paul Slovic, who pioneered the psychometric study of risk perception, documented the relationship: people who like a technology judge it as low-risk and high-benefit; people who dislike it judge it as high-risk and low-benefit. The risk-benefit correlation is driven by the emotional response, not by independent evidence about both dimensions.

Why it matters

The affect heuristic and risk-benefit correlations

In objective reality, risks and benefits are often uncorrelated — some things are risky and beneficial, some are safe and beneficial. But in public perception, risks and benefits correlate negatively: high-benefit activities feel low-risk, and high-risk activities feel low-benefit. Slovic traced this to the affect heuristic: the overall emotional tag on a technology — nuclear power, pesticides, handguns — determines responses to questions about both its risks and its benefits. The emotional tag is the common cause.

Author's argument: People do not reason from evidence about risks to conclusions about benefits, or vice versa. They have an emotional reaction to a technology as a whole, and that reaction determines their estimates of both risk and benefit simultaneously.

This produces a structure where changing someone's mind about one dimension automatically shifts the other. Show people evidence that nuclear power has lower benefit than they thought — and their perceived risk rises. Show them evidence of higher benefit — and their perceived risk drops. The information does not travel through reasoned inference; it travels through affect.

Availability cascades

Kuran and Sunstein described availability cascades: self-reinforcing cycles in which a perceived risk becomes increasingly salient through media coverage, which increases public concern, which generates more coverage, which increases salience further — until the risk is dramatically over-estimated relative to actuarial reality. Meanwhile, risks that receive little coverage stay underestimated.

The cascade dynamics are not driven by the underlying reality of the risk. They are driven by what makes a good story — vividness, identifiable victims, novelty, agency (someone did this to us). These narrative features drive coverage, coverage drives availability, availability drives perceived probability and severity.

Expert vs. public disagreement on risk

Risk experts — people who use actuarial data to estimate risk — consistently disagree with the public about which risks are most serious. The public dramatically overestimates risks that generate vivid images (terrorism, plane crashes, shark attacks) and underestimates mundane statistical killers (heart disease, car accidents, falls). This disagreement is not a failure of public intelligence — it is a predictable output of the availability-affect system operating on a media environment optimized for vividness.

Key takeaways

Mental model

Copy sourceZoom

Practical application

Practical implications for risk communication and decision-making:

• Do not present only statistics: actuarial data without emotional resonance does not update affect tags. Risk communicators need to find ways to make statistically significant risks feel salient — not just accurate.
• Watch for availability cascades in your organization: when one type of incident receives intensive attention (internal investigation, all-hands, policy changes), check whether the response is proportional to its actual frequency or driven by recency and vividness.
• Separate "how do I feel about this?" from "what is the data?": the affect heuristic means your emotional reaction to a technology or policy is not neutral evidence about its risk-benefit profile. Both components of your judgment originate from the same emotional tag.

Example

A city is deciding whether to invest in upgrading pedestrian crosswalks across the city, or in a high-visibility surveillance and patrol program near a park where a violent crime occurred six months ago. The crime received two weeks of front-page coverage; the pedestrian crossing deaths — four per year, consistent for a decade — have never appeared in the local news.

The city council votes overwhelmingly for the patrol program. From an actuarial standpoint, the pedestrian upgrades would prevent more deaths per dollar spent. But the park crime is available and emotionally activated; the pedestrian deaths are neither. The availability-affect system governs the budget even though a purely statistical analysis would allocate differently.

Related lessons