Availability heuristic

type: effect · origin: study · evidence: supported · themes: judgements and probabilities

Claim assessed: People judge how frequent something is by how easily examples come to mind.

Published 1 October 2026, updated 7 October 2026 · how we work

In short

Tversky and Kahneman (1973) called the availability heuristic the habit of judging how frequent or likely something is by how easily examples come to mind. Frequent things are usually also easier to recall, so the clue is often good. But ease also depends on other things: how well known, how recent or how dramatic an example is. That is where the errors come from. Later studies showed that it is not only what you recall that matters, but also how hard it is to recall.

What it says

The first demonstrations. In their 1974 overview, Tversky and Kahneman describe an experiment in which participants heard a list of well-known names, men and women, and were then asked whether the list contained more men or more women. In some lists the men were more famous, in others the women. Participants wrongly judged that the group with the more famous names was more numerous. The authors also note that recency and salience matter: seeing a house burning probably affects you more than reading about a fire in the local paper.

Causes of death. Lichtenstein, Slovic, Fischhoff, Layman and Combs (1978) asked people to estimate how often people die from various causes. The estimates were consistent but systematically distorted: rare causes were overestimated, frequent ones underestimated, and some specific causes were exaggerated relative to others that were just as frequent. The authors linked the distortions to how much people are exposed to those causes and how easy they are to remember or imagine. Participants could not correct them even when explicitly asked to.

Ease, not just content. Schwarz and colleagues (1991) separated the two. In their first experiment, 40 female students described either 6 or 12 situations in which they had behaved very assertively, or situations in which they had behaved unassertively. A pretest had shown that 6 examples were easy to find and 12 were hard. Then they rated their own assertiveness.

If only the content had mattered, those who found 12 assertive examples would have rated themselves the most assertive. The opposite happened: after 12 assertive examples they rated themselves less assertive than after 6, and after 12 unassertive examples they rated themselves more assertive. The authors’ explanation: if examples are hard to find, you conclude that the thing can’t be frequent for you. In another experiment, the effect disappeared when participants could attribute the difficulty to an external cause.

The meta-analysis. Weingarten and Hutchinson (2018) pooled 263 studies on this effect. On average, the standard “few versus many examples” method produces a moderate effect. A third to half of it runs through the feeling of ease, so the standard explanation is supported, but it doesn’t explain everything. They also found signs of publication bias: after correction, the standard effect drops from r ≈ 0.25 to 0.16–0.19, that is by a quarter to a third, but it doesn’t disappear. Until then there were no replications with public results.

Example

Tversky and Kahneman (1974) give an example from everyday experience: when you see a car overturned by the side of the road, accidents seem more likely for a while. The risk hasn’t changed; only the example has become more available. In the same way, after news of a plane crash, flying seems more dangerous than it did a week before.

How to apply it

The steps below are a practical way to apply the idea, proposed by us based on the studies cited.

1. When you estimate how often something happens, look for numbers. The examples that come to mind are a sample distorted by what is memorable, recent or covered in the media.
2. Ask why the example comes easily. Is it frequent, or just dramatic, recent or much discussed? Lichtenstein et al. (1978) linked the errors precisely to exposure and memorability.
3. Don’t rely on knowing about the effect. In 1978, participants could not correct the distortions even when asked to.
4. Watch judgements about yourself. How easily you recall examples of courage, laziness or success is not a good measure of how often they happen. Notes keep track better than memory.

Limits and nuances

• Availability is often a good clue. Tversky and Kahneman present it as a useful clue, because frequent things are usually easier to recall. Errors arise when ease is influenced by something other than frequency.
• Not all classic demonstrations held up. A famous example, also described in the 1974 overview, asked whether an English word more often starts with the letter “r” or has it in third position. Sedlmeier, Hertwig and Gigerenzer (1998) redid the question in three studies and found that people’s judgements generally followed the real letter frequencies.
• Schwarz’s first experiment was small. The effect in Experiment 1 was only marginally significant, with 9–10 people per group. The authors note that the pattern repeated in their later experiments, and the 2018 meta-analysis confirms a moderate effect, though probably inflated by publication bias.
• Not every distortion comes from availability. In Lichtenstein et al. (1978), overestimating rare causes and underestimating frequent ones is a general pattern of the estimates, separate from the exaggeration of specific causes.
• The meta-analysis includes a few studies by Lawrence Sanna. Of the 142 papers pooled by Weingarten and Hutchinson, six are co-authored by Sanna, who has had papers retracted for fabricated data. We don’t rely on his studies, but we can’t remove their weight from the meta-analysis’s overall result.
• What we read. We read Tversky and Kahneman (1974) and Schwarz et al. (1991) in full. Of Weingarten and Hutchinson (2018) we read the accepted manuscript, from the University of Pennsylvania repository: the abstract, the method, the main results, the publication-bias analysis and the general discussion, but not all appendices. The 1973 article was not accessible to us; the demonstrations above come from the 1974 overview. For Lichtenstein et al. (1978) and Sedlmeier et al. (1998) we read the abstracts.
• Revised on 7 October 2026. We read the 2018 meta-analysis, known until now only from its abstract, and added the effect sizes before and after correcting for publication bias, plus the note on Sanna’s studies. The evidence level has not changed.

Sources

• Amos Tversky, Daniel Kahneman (1973). Availability: A heuristic for judging frequency and probability
• Amos Tversky, Daniel Kahneman (1974). Judgment under uncertainty: Heuristics and biases
• Sarah Lichtenstein, Paul Slovic, Baruch Fischhoff, Mark Layman, Barbara Combs (1978). Judged frequency of lethal events
• Norbert Schwarz, Herbert Bless, Fritz Strack, Gisela Klumpp, Helga Rittenauer-Schatka, Annette Simons (1991). Ease of retrieval as information: Another look at the availability heuristic
• Peter Sedlmeier, Ralph Hertwig, Gerd Gigerenzer (1998). Are judgments of the positional frequencies of letters systematically biased due to availability?
• Evan Weingarten, J. Wesley Hutchinson (2018). Does ease mediate the ease-of-retrieval effect? A meta-analysis

See also: Anchoring, Bystander effect, Confirmation bias, Conjunction fallacy, Negativity bias, Serial position effect