Diminishing Sensitivity of Value — a Deep Dive into Prospect Theory

Remember how we left off with reference points and the whole drama between gains vs losses? Good. Because now we zoom in on the shape of the feeling: how value changes as outcomes move away from that reference point. This is where the magic (and most of your bad financial decisions) happens.

What is diminishing sensitivity? (Quick, slightly dramatic definition)

Diminishing sensitivity means: as you move farther from your reference point, each additional unit of gain or loss matters less emotionally. The first $50 feels a lot; the next $50 feels like less of a party.

Micro explanation: If going from $0 to $50 makes you very happy, going from $50 to $100 still helps, but not by as much. Similarly, losing your first $50 hurts a lot; losing the next $50 hurts, but a bit less.

The value function — how psychologists drew feelings

Kahneman and Tversky modeled this with a value function v(x) that is:

• Concave for gains (diminishing sensitivity to increases above the reference),
• Convex for losses (diminishing sensitivity to increases in loss below the reference),
• Steeper for losses than for gains (loss aversion).

In formula-speak (don’t panic):

v(x) = x^α for x ≥ 0
v(x) = −λ(−x)^β for x < 0

where α, β are between 0 and 1, and λ > 1 captures loss aversion. Empirical studies often find α ≈ β ≈ 0.88 and λ ≈ 2.25 — so losses sting about twice as much as equivalent gains.

Why diminishing sensitivity matters (real-world fireworks)

1. Risk attitudes flip depending on domain

   • For gains: Because the curve is concave, people prefer a sure thing over a gamble with the same expected value — classic risk aversion.
   • For losses: Because the curve is convex (mirror image), people prefer gambles over sure losses — classic risk seeking.

   This is called the reflection effect. Flip a problem from gains to losses and watch people switch from hedging to gambling like a soap opera plot twist.

2. Explains common behaviors

   • Salary raises: That 3rd promotion feels less ecstatic than the 1st because of diminishing sensitivity.
   • Insurance buying vs gambling: People buy insurance (pay a certain premium) to avoid a big loss, but also buy lottery tickets (small chance of big gain) — both behaviors fit the curvature and loss aversion.
   • Disposition effect in investing: People sell winners too early (diminishing sensitivity makes extra gain feel smaller) and hold losers too long (risk-seeking hoping to escape small losses).

3. Bridges to biases you already met

   • You learned about loss aversion in biases. Diminishing sensitivity explains how loss aversion combines with curvature to produce risk preferences.
   • It also feeds into status quo bias: because small changes are felt less, sticking with what you have seems emotionally cheaper than risking a small chance of a big improvement.

Simple, concrete example (because numbers tame the beast)

Imagine two choices:

A: A sure gain of $100.
B: 50% chance to gain $200, 50% chance to gain $0.

Expected value: both are $100 on average.

If v(x) = x^0.5 (a concave function), then v(100) = 10, while the gamble's expected value is 0.5v(200) + 0.5v(0) = 0.5*14.14 = 7.07.

So v(A) > v(B) — the sure $100 feels better, so you avoid the gamble. That's diminishing sensitivity in action.

Flip to losses:

A: A sure loss of $100.
B: 50% chance to lose $200, 50% chance to lose $0.

Using the convexity for losses, the gamble may feel less bad in expectation, so people often choose it — they gamble to avoid the certain sting. Reflection effect.

Intuition through analogies (because your brain loves stories)

• Think of emotional impact like color saturation on a photo: the first adjustments make the image pop; later adjustments barely change it.
• Or imagine pain: a first hot water burn hurts a lot; adding another degree to an already severe burn still hurts, but the extra sting feels relatively smaller.

These analogies explain why a small scandal ruins a politician early on but later scandals often barely move public opinion: diminishing sensitivity.

How this differs from familiar economics ideas

• Classical expected utility also uses concave utility to explain risk aversion, but prospect theory anchors to reference points and treats gains/losses asymmetrically.
• Diminishing sensitivity is not about objective money amounts only; it’s about changes relative to a reference. Two people with different reference points will react differently to the same absolute change.

Practical consequences (useful and dangerously applicable)

• Marketing: Framing discounts relative to a reference price exploits diminishing sensitivity — "$20 off" feels better if your reference is $100 than if it's $1,000.
• Negotiations: Getting someone to re-anchor their reference point can change whether they accept risk.
• Policy: Safety regulations and insurance design are more effective when they account for how people overweight small probabilities and the diminishing sensitivity to repeated gains/losses.

Quick limitations and nuance

• Diminishing sensitivity is an empirical regularity, not a law. The α parameter can vary across people and contexts.
• When stakes are huge, different cognitive processes (e.g., deliberative thinking) can change risk attitudes.
• Prospect theory also includes probability weighting — people don’t perceive probabilities linearly — which interacts with diminishing sensitivity.

Key takeaways (tl;dr you can tattoo these)

• Diminishing sensitivity = decreasing emotional impact for additional gains or losses as you move away from the reference point.
• It makes the value curve concave for gains and convex for losses, producing risk aversion for gains and risk seeking for losses.
• It pairs with loss aversion to explain many puzzling choices: insurance vs lottery, disposition effect in investing, and framing effects in marketing.

This is the moment where the concept finally clicks: our feelings about money, pain, or praise are not linear. The world is measured in emotional inches, not dollars.

Quick practice question (try it mentally)

You can either take a sure $1,000 or a 10% chance to win $10,000 (0 otherwise). Which would you pick, and why? Think in terms of diminishing sensitivity and the shape of the value function — not just EV.

If you liked this, next we can graph real value functions, simulate choices with different α and λ, or unpack how probability weighting mixes into the whole mess. Want to gamble on which one you pick? (I bet you pick the sure thing.)