Last lesson handed you a curve with a kink in it: the value function, concave on the gain side (so we tiptoe around gains, preferring a sure thing) and convex on the loss side (so we throw caution out the window to dodge a loss). It was a tidy diagram. This lesson is what happens when someone weaponizes it.
Here is the unsettling part. Which arm of the curve you stand on isn’t fixed by the situation. It’s fixed by whichever word the situation got described with — and that word is usually chosen by someone who isn’t you. Call an outcome a “gain” and you land on the cautious arm. Call the exact same outcome a “loss” and you slide to the reckless one. Nothing about the world changed. The wording did, and the wording is the steering wheel.
Before we prove it, commit to a guess — the whole point of this course is to feel the trap close on you at least once.
Before you read — take a guess
A hospital offers a surgery. Patients are told either 'this operation has a 90% survival rate' or 'this operation has a 10% mortality rate.' The two statements describe the exact same outcome. What does loss aversion predict about how patients respond?
The framing effect, defined
The analogy. Picture a window with a movable frame. Slide the frame one way and you see a sunlit garden; slide it the other and the same window shows a looming storm cloud — but it’s the same sky. The frame doesn’t change the weather; it changes which part of the weather you stare at. A description of a choice is exactly that window: it can’t change the outcomes, but it decides whether you’re looking at what you’ll gain or what you’ll lose.
The precise definition. The framing effect is the systematic finding that describing the same objective outcomes as gains versus losses reverses people’s preferences. Same numbers, same odds, same world — flip the wording from “what you keep” to “what you forfeit” and risk-averse choosers become risk-seeking ones (and vice versa). It is one of the most reliable demonstrations in all of behavioral science, and it falls straight out of last lesson’s machinery.
Here’s the mechanism, bolted directly onto the value function:
- Gain wording sets your reference point at the bottom (“you have nothing; here’s what you could win”). Outcomes are now gains, so you’re standing on the concave arm — diminishing returns, where a sure thing beats a gamble of equal average value. You play it safe.
- Loss wording sets your reference point at the top (“you have everything; here’s what you could lose”). Outcomes are now losses, so you slide to the convex arm — where the sure loss feels unbearable and a gamble that might avoid it looks tempting. You gamble.
The frame is nothing but a reference-point setter. Move the reference point with a word, and you’ve moved the chooser from one arm of the curve to the other without touching a single payoff.
Why this is loss aversion, not a separate trick
Framing isn’t a new bias to memorize — it’s loss aversion plus the movable reference point from lesson 2, working together. The asymmetry (losses ≈ 2× gains) supplies the force; the wording supplies the reference point that decides whether an outcome counts as a gain or a loss. That’s why the same person, looking at the same facts, can be cautious in the morning and reckless in the afternoon — the only thing that changed was which baseline the words handed them.
When to use it
Reach for this lens the instant a choice arrives pre-described — a pitch, a policy, a price, a medical option, a contract clause. Ask: “Is this worded as something I’d gain or something I’d lose, and who picked that wording?” If the description leans hard on one side, assume the frame was chosen to push you toward a specific arm of the value function. The defense is mechanical and we’ll formalize it below: re-describe the same facts the other way and see if your preference survives.
The Asian-disease problem
The cleanest proof on record comes from Tversky and Kahneman (1981). Imagine the United States is bracing for an unusual disease expected to kill 600 people. Two programs are proposed to fight it, and you must choose. Half the experimental subjects saw the programs described one way; the other half saw a description that was numerically identical — only the wording differed.
| Frame | ”Sure” option | ”Gamble” option | What most people chose |
|---|---|---|---|
| Gain frame (lives saved) | Program A: 200 people saved, for certain | Program B: 1/3 chance all 600 saved, 2/3 chance no one saved | ~72% chose the sure A |
| Loss frame (lives lost) | Program C: 400 people die, for certain | Program D: 1/3 chance nobody dies, 2/3 chance all 600 die | ~78% chose the gamble D |
Now walk the flip slowly, because the trap is hiding in plain arithmetic. Out of 600 people:
- “200 saved” (A) and “400 die” (C) are the same outcome — 200 live, 400 die either way.
- “1/3 chance all 600 saved” (B) and “1/3 chance nobody dies” (D) are the same gamble — a one-in-three shot at everyone surviving.
So A = C and B = D, exactly. The two frames offer the identical pair of choices. And yet a comfortable majority picks the sure thing in the gain frame and flips to the gamble in the loss frame. The same person, shown both versions on different days, will frequently contradict themselves and never notice.
Why? Pure value-function geometry. In the gain frame, “200 saved for sure” sits on the concave arm where a certain gain feels great and the risky version (which might save no one) feels scary — so people lock in the sure win. In the loss frame, “400 die for sure” sits on the convex arm where a certain loss is unbearable, and any gamble that might make the loss vanish looks worth a roll of the dice — so people reach for D. The words moved the reference point; the reference point moved the arm; the arm moved the choice.
This is your cue to go play with the toggle below. Flip between “lives saved” and “lives lost” and watch the star jump from the safe option to the gamble — on numbers that never change.
Losses loom larger
Same numbers, opposite choice
Drag the loss-aversion slider and watch the curve. Gains bend gently up; losses plunge — and the more loss-averse you are (higher λ), the deeper the drop. The kink at zero is the whole model.
At λ = 2.25, winning $100 feels like +58, but losing $100 feels like -129 — the loss hurts about 2.25× as much as the equal gain feels good. That asymmetry is loss aversion.
Same numbers, opposite choice: the framing flip
A disease threatens 600 people. The two programs below are numerically identical across the frames — only the wording changes. Toggle the frame and watch which option the curve above prefers.
Program A: save 200 people for sure
★ people lean hereProgram B: ⅓ chance to save all 600, ⅔ chance to save no one
In the gain frame the curve is concave, so a sure 200 saved beats the risky gamble: most people play it safe (risk-averse).
Match each way of framing a choice to the behavior loss aversion predicts it will produce:
Pick a term, then click its definition.
Everyday framing
The Asian-disease problem is a lab tidy enough to make the effect undeniable, but the wild version is everywhere — every label, ad, menu, and contract you meet has already been framed by someone with a motive. Same data, chosen wording.
| The fact | Gain frame (feels better) | Loss frame (feels worse) | Why it works |
|---|---|---|---|
| A yogurt that is 10% fat by weight | ”90% fat-free" | "10% fat” | Spotlights what you keep (health), not what you lose |
| Ground beef labeling | ”75% lean" | "25% fat” | Same meat — the lean framing reliably sells better |
| A surgery with 1-in-10 death rate | ”90% survival rate" | "10% mortality rate” | Surgeons recommend the survival-framed option more often |
| Paying with cash vs. card | ”cash discount" | "credit-card surcharge” | A “surcharge” is a loss you pay; a “discount” is a gain you earn |
| A product warranty | ”protect your purchase" | "avoid losing your $800 if it breaks” | Selling against a loss (the convex arm) drives the sale harder |
Notice the pattern in the last two rows: marketers don’t always pick the gain frame. When they want you to act — buy the warranty, upgrade the plan — they deliberately reach for the loss frame, because the convex arm makes you scramble to avoid a loss. “Don’t miss out,” “only 2 left,” “your protection expires Friday” are all loss frames engineered to push you off the cautious arm. The frame is chosen to suit the seller’s goal, not yours.
Framing is the cheapest incentive there is
This is the moment the prerequisite incentives course pays off. Recall the definition from that course: an incentive is anything that changes the payoff of a behavior — a carrot or a stick, money or status. Every incentive there costs something: a bonus costs cash, a fine costs goodwill, a metric costs measurement.
A frame costs nothing. It changes behavior without changing a single payoff or a single probability. The disease programs offer the identical lives and odds in both frames; the yogurt has the same fat either way; the surgery’s real risk is fixed. Reword the description and you’ve redirected the choice — for free, invisibly, leaving no fingerprints on the actual incentives.
That makes framing the most efficient lever in the entire incentives toolkit, and the most dangerous, precisely because it’s invisible. There’s no line item on a spreadsheet that reads “we framed this as a loss.” It’s just words — which is exactly why it slips past your guard.
The two ways framing fools you
(1) “It’s just words.” This is the fatal underestimate. The words are the lever — they set the reference point that decides which arm of the value function you stand on, and that decides your choice. Treating a frame as cosmetic is like treating the steering wheel as decoration because it’s “just plastic.” (2) “I’m reading a neutral description.” There is no neutral description. Every choice that arrives pre-worded was framed by someone, and that someone — the marketer, the politician, the negotiator across the table — already picked the arm they’d like you on. The question is never whether you’ve been framed. It’s which way, and by whom.
Spot the trap. A gym pitches its annual plan two ways to two groups: Group 1 hears 'save $120 a year versus paying monthly'; Group 2 hears 'lose $120 a year by paying monthly instead of annual.' Sign-ups for the annual plan jump in Group 2. What's the cleanest read?
Recap
You arrived with a static curve and you’re leaving with a steering wheel. Pin down these five:
- The framing effect: describing the same objective outcomes as gains vs. losses systematically reverses preferences — no change to payoffs or odds, just wording.
- It’s loss aversion plus a movable reference point: gain wording → concave arm → play safe; loss wording → convex arm → gamble. The word sets the reference point, the reference point sets the arm, the arm sets the choice.
- The Asian-disease problem (Tversky & Kahneman, 1981): ~72% pick the sure thing framed as “200 saved,” but ~78% pick the gamble framed as “400 die” — even though the options are numerically identical (A = C, B = D).
- Everyday framing is everywhere: “90% fat-free” vs. “10% fat,” “90% survival” vs. “10% mortality,” “cash discount” vs. “credit surcharge.” Sellers pick the loss frame when they want you to act.
- A frame is the cheapest incentive there is — it bends behavior for free, with no payoff or probability changed, which makes it both ubiquitous and invisible. There is no neutral description; whoever wrote the choice already picked your frame.
Check yourself: framing effects
In the Asian-disease problem, why do people flip from the sure option to the gamble when the frame changes from "lives saved" to "lives lost"?
Check your answer to continue.
Where this goes next
You can now catch a frame in the act — name the gain or loss wording, predict which arm of the value function it parks you on, and re-describe the choice both ways before deciding. One reflex defuses most of it: re-frame every important choice as both a gain and a loss, then pick your own reference point deliberately rather than inheriting the one a marketer handed you. If the choice survives both framings, it’s a real preference; if it flips, you’ve just caught loss aversion steering you.
But framing is only one place the movable reference point misbehaves. Lesson 4, The Endowment Effect & Probability Weighting, follows the same asymmetry into three more habits: why you suddenly overvalue anything the moment it becomes yours (the endowment effect), why you’d rather do nothing than risk an active mistake (status-quo bias), and why the same rational person buys both a lottery ticket and an insurance policy on the same afternoon (probability weighting). The reference point is about to start playing even stranger tricks.