Six findings from the research, and what each one means in practice for anyone weighing a move, or trying to hire someone who is not looking for one.
The rest of this report is the evidence behind each of these, and what it adds up to for the hardest decision in a cautious market: asking a person who already holds something good to risk it for something better.
Thirteen chapters, from the science of how perceived risk forms to what it means for moving and hiring. Read straight through, or jump to what you need.
Risk is a number: the chance an event happens, multiplied by what it costs if it does. Perceived risk is a feeling. The two rarely match, and the gap between them drives more decisions than the number ever will.
People do not respond to risk as it is measured. They respond to how it feels, and feeling follows rules of its own. Half a century of research, much of it rewarded with Nobel Prizes, shows the gap between the two is not random noise. It is systematic, predictable, and large enough to reverse a decision on its own.
This matters most when the future is uncertain. In calm conditions, perception and reality drift only slightly apart. Under stress, an AI wave reshaping whole professions, conflict, a market shock, the two pull sharply apart, and the divergence itself changes behaviour. Investment pauses. Hiring freezes. People who would have moved stay where they are. None of it requires the actual risk to have risen. It only requires the felt risk to have done so.
This report maps that gap. It opens with the science of how perceived risk forms, why a loss outweighs an equal gain, why the framing of a choice reorders it, why dread crowds out arithmetic. It turns to the risks that go unseen and unpriced, then traces how uncertainty freezes both markets and careers. It closes on the one lever the evidence supports: not changing the risk, but changing what people know about it.
The distinction is not academic. It decides where capital is allocated and who accepts the next job. A firm that understands it can read a market more clearly than one that assumes people are weighing probabilities. And anyone trying to move a cautious decision-maker, an investor, a board, a senior candidate, is really in the business of closing the distance between felt and faced.
The words are used interchangeably. They should not be. The difference between a known probability and an unknown one is the difference between a problem you can price and one you cannot, and it explains why some dangers are mishandled and others ignored entirely.
The economist Frank Knight drew the line a century ago. Risk is a known event with a known or estimable probability. A coin toss, an actuarial table, a default rate. You may not like the odds, but you can compute them, price them, and insure against them. Uncertainty is a known event with an unknown probability. You can see what might happen but cannot put a reliable number on it.
A 2025 Federal Reserve working paper by Barth and Schreft sharpens this into three tiers, and the third is the one that does the damage. Beyond risk and uncertainty sits the true black swan: an event not even conceived as possible beforehand, and labelled rare only in hindsight. The three behave very differently when markets try to handle them.
The practical lesson is uncomfortable. For the known unknown, Barth and Schreft note, the probabilities are not known, "resulting in inaccurate risk pricing and imperfect risk transfer." For the black swan, "no market mechanism can exist to price the event because there is no knowledge that an event needs to be priced." Markets, in other words, systematically under-prepare for the risks that matter most, not out of carelessness but by construction.
Hold on to the distinction, because almost everything that follows is a consequence of it. People are poor at estimating known probabilities, and worse still at the uncertainty they cannot estimate at all. Into that vacuum rushes feeling, and feeling is where perceived risk is made.
The single most replicated finding in the study of risk is also the most consequential: we do not weigh gains and losses equally. A loss of a given size hurts roughly twice as much as the same gain pleases. That asymmetry quietly bends most decisions toward the status quo.
In 1979, Daniel Kahneman and Amos Tversky published Prospect Theory, now among the most cited papers in the social sciences and the basis of Kahneman's 2002 Nobel Prize. Its central claim is that people do not judge outcomes by their final wealth, as standard economics assumed. They judge them as gains and losses relative to a reference point, usually wherever they happen to be now.
The value people place on those gains and losses is not symmetric. The curve is steeper for losses than for gains. Later work put a number on it: a loss is felt around 2.25 times as intensely as an equivalent gain. The practical signature is everyday reluctance, the refusal to accept a coin-toss bet that pays slightly more than it risks, even though the maths is favourable.
Prospect theory adds a second twist, the reflection effect. People are risk-averse when choosing between gains, preferring a sure thing to a gamble of equal expected value. But facing losses, the same people turn risk-seeking, preferring to gamble rather than accept a certain loss. The instinct to protect what we have makes us cautious in good times and reckless in bad ones.
Two honest caveats. These patterns are strongest, in aggregate, for moderate-probability choices over money, and they soften with very low probabilities, with real rather than hypothetical stakes, and in moral or social decisions. And the idea of loss aversion as a universal law has been challenged. But in the exact territory that concerns a careful decision-maker, a real choice with something to lose, the asymmetry is about as solid as behavioural science gets. For anyone weighing a move that risks a known, comfortable present against an uncertain future, the deck is emotionally stacked toward staying put.
People do not use probabilities as they are stated. They quietly convert them into decision weights, and the conversion is lawful, distorted, and the reason the same person will buy insurance and a lottery ticket in the same afternoon.
Kahneman and Tversky found that decision weights are generally lower than the probabilities they replace, with one important exception: small probabilities are overweighted. A one-in-a-thousand chance feels bigger than one in a thousand. Pair that with the certainty effect, the outsized pull of an outcome that is guaranteed rather than merely likely, and two contradictory behaviours fall out of a single rule.
The lesson for anyone reading a risk is that low-probability, high-feeling events get more weight than they deserve, and dull, certain costs get less. A vivid, frightening, unlikely outcome can dominate a decision while a quiet, near-certain one is waved through. When we turn to the risks people miss entirely, this is half the reason: the mind is busy overweighting the dramatic and discounting the mundane.
Note too what the certainty effect does to a settled situation. The known present is, by definition, certain. Anything that would replace it is merely probable. The arithmetic of decision weights, like the arithmetic of loss aversion, leans on the side of staying.
If perceived risk tracked reality, the wording of a choice could not change it. It does, completely. The most famous experiment in decision science shows identical options reversing preference depending only on whether they are described as lives saved or lives lost.
In 1981, Tversky and Kahneman put a problem to two groups. A disease threatens 600 people. One group was offered programmes described in terms of lives saved; the other, programmes described in terms of lives lost. The underlying numbers were identical. Saving 200 of 600 is the same as letting 400 die. The certain option and the gamble were mathematically the same in both versions. Only the language moved.
A fifty-point swing, from one wording to another, with the facts held constant. This is the reflection effect from the previous chapter made visible: risk-averse over gains, risk-seeking over losses. It has survived direct replication and holds even when the wording is made scrupulously precise, so it cannot be dismissed as a verbal trick.
The implication is large. Whoever frames a choice shapes the risk that is felt. A reframe is not spin. It can be the honest difference between a decision that reads as a dangerous leap and the same decision read as the protection of something worth keeping. The facts do not have to change for the answer to.
Ask people which hazards are riskiest and their rankings diverge sharply from the body counts. The gap is not ignorance of the numbers. It is a different definition of risk altogether.
Paul Slovic's 1987 paper Perception of Risk established the point with data. People can produce fatality estimates that roughly resemble the technical ones. Yet their sense of how risky something is tracks other qualities entirely, and so diverges from their own death-rate estimates. Factor analysis collapses the many qualities of a hazard into two.
One caveat of method. These are aggregate, cross-hazard patterns; at the level of a single individual, the framework explains much less of the variation. But the headline is robust: perceived risk is assembled from qualities, not quantities. New and uncontrollable beats familiar and frequent.
Carry that into any decision involving the unfamiliar. A first-time employer, an untested team, an emerging strategy, a market with no track record, all carry a dread premium that has little to do with their true odds. The novelty itself reads as danger. Whatever cannot be observed or controlled feels riskier than the numbers say it is, and the burden of proof falls on whoever is asking someone to step toward it.
Beneath framing and dread sits something more basic. People read risk and benefit off a single emotional impression of a thing. Improve the impression and the perceived risk falls, even though the real risk has not moved.
Finucane, Alhakami, Slovic and Johnson named this the affect heuristic in 2000. The mind takes a shortcut: rather than appraise risk and benefit separately, it reads both off one overall good-or-bad reaction to the item. This is why, inside people's heads, risk and benefit run in opposite directions, the things we like feel low-risk and high-benefit, even though in the real world risk and benefit tend to rise together.
Two findings make this more than a curiosity. First, the coupling tightens under time pressure, when there is no room to deliberate and feeling has the floor. Second, and decisively, it is causal: tell people about an item's benefit and their sense of its risk falls; tell them about its risk and their sense of its benefit falls. The emotional impression is the lever, and moving it moves both judgments.
This is the hinge of the whole report. If perceived risk is anchored to an overall impression, then perceived risk can be lowered without touching the actual risk, by improving the impression and by supplying information that changes the feeling. That is not manipulation in any dark sense. It is the legitimate work of communication, and the rest of this report turns on it.
Three honest limits. The effect is asymmetric: perceived risk responds more readily to new information than perceived benefit does, so the strongest route is to address the source of fear directly rather than simply pile on upside. It is probabilistic, not guaranteed. And it is weaker in people who think more reflectively. A reliable aggregate lever, then, not a switch.
Uncertainty is not only a private feeling. Measured across an economy, it has a consistent and well-documented effect. When it rises, firms stop. Investment is shelved and hiring is paused, even when the underlying business has not changed.
Nicholas Bloom's 2009 paper The Impact of Uncertainty Shocks set out the mechanism. When uncertainty spikes, the option value of waiting rises, so firms adopt a "wait and see" posture, pausing investment and hiring until the fog clears. Activity drops sharply, then rebounds once uncertainty subsides. Uncertainty jumps after major shocks: Bloom lists the Cuban Missile Crisis, the JFK assassination, the 1973 oil shock and 9/11. A fair caveat: the clean rebound is less robust in data after 1983, and the real-options story is one mechanism among several.
Baker, Bloom and Davis then made uncertainty measurable. Their Economic Policy Uncertainty index, built from newspaper coverage and validated against twelve thousand human-read articles, shows that rises in policy uncertainty precede declines in investment, output and employment across twelve major economies.
The authors are careful: the relationship is predictive, not proven causal, and the macro result is sample-dependent. But two features matter for this audience. The effect concentrates in policy-sensitive sectors, and finance is named among them. And the damage compounds when money is tight.
For an investment firm or a finance function, the chain is direct. When uncertainty rises, the firms that hire pause, and the freeze lands hardest exactly where this audience sits. In the boardroom it is a real-options calculation. But it is mirrored by a parallel freeze in individuals, and that one, as the next chapters show, is driven as much by feeling as by arithmetic.
Everything that freezes a firm freezes a worker too. When uncertainty rises, voluntary moves dry up, and the people most worth hiring are the least likely to jump. The reason is the same loss aversion and dread that govern every other risk.
Job mobility is procyclical and driven by quits. When confidence falls, voluntary moves fall with it. US job-to-job flows hit a post-war low after the 2008–09 recession, and the rate roughly halved between 2000 and 2012, a structural decline in mobility that predates any single downturn. The recent climate has earned its own label in the data, the "great stay": hiring and quitting subside together, and people sit tight.
The mechanism is simply this report applied to a career. A move risks a known, comfortable present, and loss aversion weights that prospective loss twice over. The new employer is unfamiliar and less controllable, so it carries a dread premium. The downside of a bad move is vivid and overweighted; the upside is merely probable. In an anxious market, the option value of waiting rises for the individual exactly as it does for the firm. Every bias points the same way: stay.
This is the cruel asymmetry for anyone trying to hire in an uncertain market. A downturn shrinks the available pool precisely at the senior, discerning end, the people who already hold something good and will not put it at risk without a compelling, de-risked reason. The market does not just go quiet. It goes quiet where it matters most.
No force has widened the gap between felt and faced risk faster than artificial intelligence. Worry about it far outruns exposure to it, and the worry, not the exposure, is what shapes how people behave.
The clearest illustration of this report's whole thesis is sitting in recent survey data. A Pew Research survey of 5,273 US workers, fielded in October 2024, found 52% worried about the future of AI in the workplace against just 36% hopeful, and a third expecting it to mean fewer opportunities for them personally. Yet only 16% actually use AI much in their own work, and 63% barely use it at all.
PwC's 2025 global workforce survey adds the decisive twist: optimism tracks familiarity, not exposure. Daily AI users were far likelier than non-users to report benefits to their productivity, security and pay, and 69% of daily users were optimistic about the future of their role, against 44% of non-users. Those who use the tool fear it least. Recall that dread is the perception of the unfamiliar and the uncontrollable. To the people who have not yet touched it, AI scores high on both.
None of this denies that parts of finance and accounting work are genuinely exposed to automation. They are. But "exposed" is not "eliminated," and measured task exposure sits well below the perceived threat. The fear is running ahead of the facts, which is exactly what the science of perceived risk predicts it would.
For senior finance careers, the consequence is concrete. Two layers of perceived risk now sit on every move: the ordinary fear of changing jobs, and a fresh, AI-shaped dread about the future of the work itself. Both are perceptions. Both run ahead of the measurable reality. And both, crucially, respond to the same remedy: information that turns the unknown into the known.
Here is the constructive turn. Because perceived risk is built from feeling, framing and the unknown, it can be lowered without changing the actual risk. The evidence points to a small set of levers that work.
The science of risk communication, built by Slovic, Fischhoff and Sandman, and the economics of information from Michael Spence, converge on a practical toolkit. None of these levers touches the underlying risk. Each one changes what is known and felt about it.
Read the list again and notice what it is not. It is not a louder pitch, and it is not a bigger number. It is the disciplined replacement of the unknown with the known, and of free-floating fear with specific, credible detail. That is the entire craft of moving a cautious decision-maker, and it is where this report turns from evidence to application.
Put the evidence together and one decision comes into focus: a senior professional weighing whether to move. It is the perfect storm of perceived risk, and reading it correctly is the whole job of attracting the people who are not looking.
Every mechanism in this report stacks on one side of that decision. Loss aversion: the current role is the reference point, and leaving it registers as a loss, weighted roughly twice over. Dread and the unknown: a new employer is unfamiliar and less controllable, and carries a premium of fear that has nothing to do with its real odds. Probability weighting: the vivid risk of a bad move is overweighted while the upside stays merely probable. Uncertainty: in an anxious market the option value of waiting rises for the individual exactly as it does for the firm. And now an AI-shaped layer of dread sits on top of all of it. The strongest, most discerning people feel every one of these most acutely, because they have the most to lose.
There is a risk on the other side of the ledger that almost no one prices: the risk of staying. Career stagnation, a platform quietly deteriorating, a function that has stopped growing. These are the known unknowns of chapter eight, mispriced precisely because they are gradual and invisible. The move is over-feared. The stay is under-feared. Both errors come from the same place.
This is why an inflated offer so often fails to move the right person, and can even raise the perceived risk by reading as a sign that something is wrong. Loss aversion and dread are not bought off with money. What answers them is the toolkit of the previous chapter, applied with care to a single human decision.
None of this fights human nature. It works with it. The discerning candidate's caution is not an obstacle to be overcome with pressure; it is a rational response to genuine uncertainty, and it deserves a genuine answer. The firms and the advisers who understand perceived risk do not argue with the fear. They dissolve it, by turning an unknown opportunity into a known one and letting a clear-eyed person decide. The same move, seen clearly, is a different decision. That, in the end, is what it means to attract the people who were never going to answer an advert.