Loss Aversion: Why Your Team Fears Losing More Than It Wants to Win

Two proposals came before the same investment committee, three months apart. Read them side by side — as one of the non-executive directors eventually did — and they were, in substance, almost identical: similar capital at risk, similar expected return, similar odds. The committee approved one in forty minutes and strangled the other in a fog of requests for further analysis from which it never emerged.

The difference was not in the numbers. It was in a single framing choice. The approved proposal was presented as protecting existing revenue — a defensive investment to stop a competitor eroding a profitable line. The rejected one was presented as capturing a new opportunity — an expansion into adjacent territory. Same risk, same reward. One dressed as avoiding a loss, the other as chasing a gain.

The committee would have told you, to a member, that they had judged each case on its merits. They were wrong in the most human way possible — and the reason they were wrong is arguably the single most consequential finding behavioural science has ever handed to people who lead organisations.

The asymmetry at the bottom of the mind

In 1979, Daniel Kahneman and Amos Tversky published “Prospect Theory: An Analysis of Decision under Risk” in Econometrica — a paper that would eventually help win Kahneman the Nobel Prize in economics and become one of the most cited articles the discipline has produced. Its heart is a simple, devastating observation: people do not evaluate outcomes on absolute terms, the way rational-agent models assumed. We evaluate them as changes from a reference point — and the value function through which we feel those changes is lopsided. It is steeper on the downside. Considerably steeper.

The rough exchange rate, confirmed across decades of subsequent experiments, is that a loss weighs somewhere around twice as much as an equivalent gain. Losing R100,000 hurts roughly twice as intensely as winning R100,000 pleases. Offer people a coin flip — heads you win R1,500, tails you lose R1,000 — and most decline it, despite the clearly positive expected value. The prospective pain of the loss simply outmuscles the prospective pleasure of the larger gain.

Kahneman and Tversky called it loss aversion, and its fingerprints turned out to be everywhere. In 1990, Kahneman, Jack Knetsch, and Richard Thaler demonstrated its strange cousin, the endowment effect, with nothing more elaborate than coffee mugs: give a mug to half a lecture hall and, within minutes, the new owners demand roughly twice as much to sell it as the non-owners are willing to pay for it. Nothing about the mug has changed. What changed is the reference point — giving it up is now a loss, and losses are priced differently.

And in 1981, Tversky and Kahneman published the experiment every framing discussion returns to. An outbreak threatens 600 lives. Programme A “saves 200 people” — and most respondents choose it over a gamble. Describe the identical programme as one where “400 people die”, and preferences flip: the majority now prefer the gamble. Same facts, same mathematics, opposite decisions — because one description framed the outcome as a gain and the other as a loss, and the mind that received them is not neutral between the two.

What the asymmetry does to your organisation

Once you know the shape of the curve, you start seeing it steering rooms everywhere.

It writes the risk culture. Consider the incentives of a capable divisional manager. A bold initiative that succeeds earns a good bonus and a nod. One that fails earns a career scar that never quite fades. Under a 2:1 felt asymmetry — often amplified further by how organisations actually punish visible failure — the rational personal strategy is systematic timidity. Multiply that by every manager in the building and you get the great unmeasured cost of loss aversion: not the bad bets taken, but the good ones never proposed. The pipeline of growth ideas thins out years before it shows up in the results.

It keeps dying projects alive. Terminating a failing initiative converts a paper loss into a realised one — and realising losses is precisely what the machinery resists hardest. So projects that would never be approved today survive review after review, fed by the hope of “getting back to even”. Prospect theory’s grim addendum is that people become risk-seeking in the domain of losses: once a venture is under water, doubling down starts to feel more attractive than closing it out. Every executive has watched this happen. Most have done it.

It sits on both sides of every negotiation. A concession is a loss to whoever makes it, felt at double weight — which is why late-stage negotiations over trivial amounts turn so bitter, and why a counterpart clings to a clause worth nothing. The asymmetry also explains the power of incumbency in renewals: to the customer, switching suppliers means certain, immediate losses — familiarity, integration, relationships — set against speculative gains. The challenger’s offer must be dramatically better, not marginally better, because it is competing against the endowment effect, not just the incumbent.

It is the engine of resistance to change. Every reorganisation asks people to surrender certainties they currently hold — role, status, mastery of the old system — in exchange for benefits that exist, for now, only in a slide deck. Held certainties are endowed; promised benefits are not. Run that through the loss-averse value function and rational-looking resistance falls out automatically. Your people are not failing to understand the strategy. They are pricing it with the only value function human beings have.

Leading with the grain of the curve

Awareness alone is worth a great deal here, but there is a discipline to apply. I call it keeping a Loss Ledger — a deliberate, four-part accounting that goes wherever consequential decisions are being made.

First, audit the frame — including your own. Before any major decision, ask the committee’s question out loud: which way is this proposal framed, and how would it read if we reversed it? The strongest debiasing move available to a leadership team is simply to state every major option both ways — as the gain it might capture and the loss it might prevent or incur — and let the room notice how differently the two versions feel. The feeling is the bias, caught in the act.

Second, name the losses honestly. In any change or negotiation, someone is losing something real, and pretending otherwise hands the loss all of its power. Unnamed losses go underground and reappear as resistance. Named ones — “this closes a plant some of you built; that is a genuine loss and we will treat it as one” — can be grieved, offset, and traded transparently. People will accept remarkable losses that are acknowledged; they will fight trivial ones that are denied.

Third, price the loss of inaction. The status quo wins by default because its losses are invisible — the eroding share, the departing talent, the compounding technical debt. Make them visible and concrete: standing still costs us this, by this date. This is the legitimate use of loss framing — not manufacturing fears that do not exist, but ensuring real ones carry their true weight against the imagined safety of doing nothing. The line between the two is the line between influence and manipulation, and it sits exactly where the evidence sits.

Fourth, build the new endowment early. If people defend what they own, give them ownership of the future state before you ask them to release the old one. Pilots people can touch, roles named early, teams co-designing their own transition — all of it converts the destination from a speculative gain into something already partly theirs, so the endowment effect begins working for the change instead of against it. The mug experiment runs in both directions.

The negotiator’s corollary: package with the grain

For anyone who negotiates, prospect theory carries one more gift, and it comes from Richard Thaler’s work on mental accounting in the 1980s. Because the value curve is steep near the reference point and flattens further out, how outcomes are bundled changes how they feel — even when the totals are identical. Thaler’s rules of hedonic framing follow directly: separate the gains, combine the losses.

Two concessions offered together register as one pleasant event; offered a week apart, they register as two, and the second buys you goodwill the bundle never would. The reverse holds for pain. A price increase delivered alongside the service downgrade and the new payment terms is one bad day; delivered in three instalments over a quarter, it is three, each freshly weighed on the steep part of the curve. Every experienced dealmaker knows the shape of this truth instinctively — it is why bad news is released on one Friday and product announcements are drip-fed — but knowing the mechanism lets you apply it deliberately: unbundle what you give, consolidate what you take, and never let a negotiation’s final memory be a fresh small loss.

The same logic governs how concessions should be framed at the table. A concession described from your reference point — “we’re giving up the exclusivity clause” — invites the other side to bank it and move on. The identical concession framed from theirs — “this means your team can sign the two regional partners you mentioned” — lands on their gain curve, where it is actually felt. You worked hard for that concession. Make sure it is experienced in the currency that counts.

None of this manufactures value that does not exist; the totals are the totals. What it does is stop real value from being lost in translation between the arithmetic and the asymmetric machine that will feel it — which, in the end, is the machine that signs.

The curve is not the enemy

A final word in defence of the asymmetry, because it is easy to leave this subject treating loss aversion as a design flaw. It is not. It is a survival architecture — built for a world where a single unguarded loss could be final, and still useful in ours: the same weighting that makes your committee too timid also makes it appropriately hard to bet the company.

The failure mode is not having the curve. It is not knowing it is there — mistaking its pull for analysis, its flinch for prudence, letting a framing choice made by whoever wrote the proposal decide what a room full of experienced people believes it independently concluded.

The non-executive director who spotted the twin proposals did exactly one thing about it. At the next meeting, she asked for every major recommendation to be presented under both frames, gain and loss, side by side on one page. It slowed nothing down. It changed almost every conversation that followed.

The curve stayed lopsided, of course. It always will. But the room could finally see it — and a bias you can see is a bias that has lost most of its vote.