What Cranks Availability Up

Last lesson we made the availability heuristic precise: your mind quietly swaps the hard question “how frequent is this?” for the easy one “how fast does an example come to mind?” — what Kahneman and Tversky called attribute substitution. The slogan is sharp, but it leaves one giant question dangling. If ease of recall is the thing your mind reads as frequency, then what makes recall easy? Because if the only thing that made memories easy to summon were how often they happen, the heuristic would be flawless — easy-to-recall would mean common, and you could trust it completely.

It isn’t flawless, and this lesson is the reason why. There are at least four forces that crank a memory’s availability way up without that memory being one bit more frequent. They are the engines of distortion, and once you can name them you can catch them firing in real time. They are vividness, recency, emotion, and repetition — and every famous case of a society fearing exactly the wrong thing is one or more of these engines running hot. We’ll meet each one through a real, costly misjudged risk, then watch what happens when they stack on top of each other.

Engine 1: Vividness — the concrete crowds out the abstract

The analogy. Picture a courtroom where one witness gives a flat numerical summary — “the defendant was at the scene roughly 80 percent of the relevant evenings” — and another describes, in trembling detail, the single night they watched him climb through the window. Both are “evidence.” But twelve jurors will replay the window scene in their heads on the drive home and forget the percentage by lunch. The vivid beats the statistical not because it’s more true but because it’s more retrievable.

The definition. Vividness (closely related to salience and concreteness) is the property of a memory that makes it sensory, specific, and easy to picture. A gory, concrete, named image carves a deep, easily-walked path in memory. A dry, abstract number leaves almost no trace — it slides off. Because the availability heuristic reads easy to recall as frequent, anything vivid gets a frequency bonus it never earned.

The worked example — planes versus cars. A plane crash is about as vivid as a risk gets: a spectacular fireball, wall-to-wall footage, named victims, a single moment of mass horror. A car crash is mundane — one of tens of thousands a year, rating a paragraph in the local paper if that. Now the numbers. In the United States, driving kills on the order of tens of thousands of people every year; commercial aviation, in a typical year, kills a tiny fraction of that, and per mile traveled flying is dramatically safer than driving — by most estimates, riding in a car is on the order of a hundred times more dangerous per mile than flying. Yet survey after survey finds people far more frightened of flying. The fear tracks the vividness of the crash, not the frequency of the death. You are scared of the fireball you can picture and blasé about the fender-bender you can’t be bothered to.

The identifiable-victim cousin. Vividness has a famous relative: the identifiable victim effect. Tell people that a drought threatens millions of children and you get a shrug; the number is too abstract to picture. Tell them about one named seven-year-old girl with a photograph and a story, and donations pour in — often more for the one child than for the statistical millions. As the line attributed to Stalin (and echoed by researchers) goes, one death is a tragedy; a million deaths is a statistic. A statistic doesn’t summon a face, so it doesn’t summon a memory, so it doesn’t move the needle. The single concrete victim is available; the million are not.

When it bites you

Vividness bites hardest wherever the dangerous-but-rare thing is photogenic and the dangerous-but-common thing is boring. Shark attacks (vivid, cinematic, a handful of deaths a year worldwide) terrify people far more than the mundane, far deadlier causes — the drive to the beach, the sun overhead, the heart disease building quietly for decades. The rule of thumb: the easier a risk is to make into a movie, the more your gut will overrate it.

Engine 2: Recency — what just happened masquerades as what usually happens

The analogy. Ask someone on the first sunny day after a week of rain whether “it’s been a rainy month,” and they’ll often say no — today is bright, and today is what’s loaded in the chamber. The most recently fired memory sits at the front of the queue, and the front of the queue gets mistaken for the whole queue.

The definition. Recency is the simple fact that events from the last few days or weeks dominate recall: they’re freshest, so they come fastest, so the heuristic reads them as most typical. “What just happened” gets promoted, without permission, to “what usually happens.” Recent events haven’t gotten any more frequent — they’ve just gotten more available, and availability is the currency your gut spends.

The worked example — disaster insurance. This one is beautifully documented. Purchases of flood and earthquake insurance spike sharply right after a flood or earthquake — and then decay month by month as the memory fades, often dropping back to baseline within a few years even in places where the underlying risk hasn’t changed at all. Think about how backwards that is. The actual probability that your house floods next year is roughly the same the week before a flood as the week after one — arguably the ground is wetter and the immediate risk higher before the next dry spell, not after a disaster has passed. But people buy when the memory is freshest, not when the risk is highest. The flood didn’t change the odds; it changed the availability of the image of a flooded living room, and the buying tracks the image.

The second worked example — chasing markets. Investors do the identical thing with money. After a vivid crash, the memory of losing is fresh and available, so people pull out and stay in cash — right at the bottom, when expected returns are highest. After a vivid boom, the memory of easy gains is fresh, so money pours in — right at the top, when it’s most overpriced. “Buy high, sell low,” the most reliable way to lose money, is recency-driven availability wearing a suit: the recent direction feels like the permanent direction.

When it bites you

Recency bites whenever you generalize from your latest data point. One bad meal makes a restaurant “terrible”; one great quarter makes a stock “a sure thing”; one mugged friend makes the neighborhood “dangerous now.” The tell is a sudden, sharp swing in your sense of how common something is that lines up suspiciously well with something that just happened to you or near you. When your estimate of frequency jumps right after a single recent event, suspect recency, not new truth.

Engine 3: Emotion — fear and dread burn memories in deep

The analogy. Think of memory like ink on paper. A neutral fact is written in pencil — faint, easily erased. An emotionally searing experience is carved in with a knife. Years later you may not remember what you ate last Tuesday, but you remember exactly where you were standing during the worst scare of your life. Emotion doesn’t just record memories; it engraves them, and engraved memories come back instantly.

The definition. Emotion (especially fear, dread, and outrage) deepens how memories are encoded and speeds how they’re retrieved. This connects to a sibling shortcut called the affect heuristic: when something carries a strong emotional charge, that feeling itself becomes the shortcut you judge by — a dread-soaked risk feels enormous regardless of its real numbers. The scarier something feels, the more frequent it seems, and those two quantities are very nearly unrelated.

The worked example — driving after 9/11. This is the most haunting case in the whole subject. After the September 11, 2001 attacks, flying felt unbearably dangerous — it was vivid, emotional, recent, and replayed everywhere. So in the months that followed, large numbers of Americans chose to drive instead of fly for trips they’d normally have flown. But driving, mile for mile, is far deadlier than flying. Researchers — most notably the psychologist Gerd Gigerenzer — estimated that the extra driving caused on the order of about 1,500 additional road deaths in the year after the attacks: more people, by this estimate, died on the roads fleeing the fear of flying than would have died had they simply flown. Sit with the shape of that. The dread of the vivid risk (planes) drove people straight into the arms of the mundane risk (cars), and the mundane risk quietly collected its toll. The fear was real; the danger it pointed at was inverted.

Why dread distorts so reliably. Risk researchers find that certain qualities make a hazard feel far worse than its body count warrants: being catastrophic (many die at once, as in a crash), uncontrollable (it happens to you), dreaded, and unfamiliar. Terrorism, nuclear accidents, shark attacks, and stranger kidnapping all light up every one of these dials — so they feel huge. Meanwhile the things that quietly kill the most people — heart disease, car crashes, the slow accumulation of ordinary risk — push none of those dread buttons, so they feel small. Dread and frequency aren’t just different; they’re often pointed in opposite directions.

When it bites you

Emotion bites every time you’re frightened or outraged. The strength of the feeling is not evidence about the odds — it’s evidence about how the event was encoded. Terrorism, plane crashes, sharks, kidnapping, and rare violent crime all feel common because they feel terrible; the quiet killers feel rare because they feel boring. The tell: when a risk fills you with dread, deliberately ask for its actual base rate before you let the feeling set your behavior, because dread and frequency are nearly independent.

Engine 4: Repetition — exposure volume impersonates event frequency

The analogy. Hear a song once and it’s a song. Hear it forty times on the radio and it’s everywhere — you start to feel like the whole world is playing it, when really the same single track is just being replayed. Repetition doesn’t multiply the events; it multiplies your exposures to one event, and your gut can’t tell those apart. Forty replays of one story feel like forty stories.

The definition. Repetition is the inflation of availability through sheer frequency of exposure — media replay, social feeds, word of mouth. Every time you encounter a story, its memory trace gets reinforced and easier to summon. Crucially, coverage volume is not event frequency. A single rare event, replayed ten thousand times across every channel, can become more available than ten thousand common events that each happened once and were never reported.

The worked example — the crime “epidemic” that wasn’t. Imagine a city where a particularly grim crime gets wall-to-wall coverage for a month: the same case, the same footage, the same anxious panel discussions, over and over. Ask residents afterward and they’ll swear crime is exploding — even in years where the actual crime statistics are flat or falling. They’re not lying or stupid; their availability is genuinely flooded with crime examples. But every example traces back to the same handful of cases replayed endlessly. Coverage volume went up; event frequency didn’t. The “epidemic” lives in the broadcast schedule, not in the streets.

The social-circle version. Repetition doesn’t need a newsroom. If three friends in a row mention a layoff, you start to feel like everyone is getting laid off — your personal feed got repetitive, so the experience feels common. The same handful of anecdotes, circulating and recirculating through your social circle, distorts your sense of how widespread something is. You can be flooded with examples of a thing that is, in the wider population, genuinely rare.

The illusory-truth sibling. Repetition has a darker cousin worth naming: the illusory truth effect — a claim repeated enough starts to feel true, regardless of whether it is, simply because familiarity is comfortable and your brain misreads “I’ve heard this before” as “this is correct.” It’s the same fluency trick aimed at belief rather than frequency: the more available a statement, the truer it feels. Propaganda, advertising, and viral misinformation all run on this engine — repeat, repeat, repeat, until availability does the persuading.

When it bites you

Repetition bites whenever your sense of “how common is this?” is being fed by a channel that repeats — news, social media, a chatty friend group, an algorithm optimizing for engagement. The tell: notice when many of your “examples” of something trace back to the same source or the same few cases seen multiple times. If you can’t summon many independent instances, you may be counting replays, not events.

When the engines stack: the quadruple-availability monster

Here’s the part that makes availability so dangerous in the real world: the engines don’t fire one at a time. They stack, they reinforce each other, and the worst distortions are the ones where all four run at once. The clearest example is a major terrorist attack.

Walk it through. A terrorist attack is vivid — concrete, cinematic, horrifying images burned into every screen. It is emotional — pure dread, the catastrophic-uncontrollable-unfamiliar trifecta maxed out. It is recent — it dominates the news the moment it happens. And it is repeated — replayed endlessly across every channel for days or weeks. That’s all four engines firing simultaneously: quadruple availability. No wonder terrorism dominates risk perception so far out of proportion to its frequency — across most years and most countries, the number of people killed by terrorism is minuscule next to the number killed by car crashes, heart disease, or ordinary falls. But terrorism wins every availability race by a mile, so the gut hands it a frequency, and a danger, that the numbers never support. The fear isn’t crazy; it’s the predictable output of four amplifiers stacked on one event.

This is why the most distorted risks tend to share a profile: dramatic, emotional, freshly broadcast, and endlessly replayed. The more engines an event recruits, the more your sense of its frequency detaches from reality. Here is the whole quartet in one view:

When the engines actually help

It would be a mistake to conclude that availability is always wrong — it usually isn’t, which is exactly why your mind trusts it. Ease of recall is a fair proxy for frequency under one condition: when your sample of memories is itself a fair, well-mixed sample of reality. Drawing on stable, well-sampled personal experience, availability works beautifully. How often does your bus run late? How common is rain in your town in November? How often does the office printer jam? You’ve personally sampled these many times, in rough proportion to how often they actually happen, and none of the four engines is systematically skewing the sample — so “how easily can I recall a late bus?” really does track “how often is the bus late?” Trust the heuristic here.

The proxy gets poisoned the instant your memories stop being a fair sample — and the four engines are precisely the things that poison the sample. Anything filtered through media, emotion, repetition, or a single recent shock is not a fair sample. The news shows you the rare and dramatic; emotion over-weights the dreadful; repetition multiplies a few cases into a felt flood; recency over-weights the latest event. So the practical rule is a clean split: trust availability for things you’ve personally and repeatedly sampled in ordinary life; distrust it — hard — for anything your sense of which comes mainly from screens, feelings, or a single recent event. The skill isn’t to abandon the heuristic. It’s to know which side of that line you’re standing on.

Recap

Four engines crank availability up independent of frequency, and naming the one that’s firing is most of the defense:

You now know what makes a memory leap to mind whether or not it’s frequent — and you noticed that one institution recruits all four engines at industrial scale. Next, in How the News Rewrites Your Risk Map, we put the headline machine under the microscope: why coverage runs inverse to frequency, the shark-attack-versus-heart-disease gap drawn to scale, and an interactive chart where you watch a population’s fears drift clean away from what actually kills it.