Diminishing sensitivity
The first dollar (or degree of pain) registers more than the hundredth.
What it means
Diminishing sensitivity is the principle that the marginal psychological impact of a change shrinks as outcomes move further from the reference point, so a given absolute difference feels larger near zero than far from it. It is a core building block of prospect theory: applied to gains and losses around a reference point, it gives the value function its characteristic S-shape — concave for gains and convex for losses — which in turn produces risk-aversion in the domain of gains and risk-seeking in the domain of losses. The same diminishing curve explains why people will exert effort to save $10 on a cheap item but not the identical $10 on an expensive one, and why losses are best bundled while gains are best savored separately. It is conceptually related to the older Weber–Fechner observation that perceived magnitude grows with the logarithm of stimulus intensity. In practice it shapes pricing, the framing of discounts, how bad news should be packaged, and why the same monetary amount can feel decisive or negligible depending entirely on the baseline against which it is judged.
The shape, quantified
Curvature is estimated, not assumed. Fitting a power value function to lottery choices, Tversky and Kahneman (1992) put the exponent near 0.88 for both gains and losses, and a number below one is precisely what encodes diminishing sensitivity, since an exponent of one would mean every extra dollar registered equally. At 0.88 the compression is real but gentle: doubling an outcome multiplies its felt value by roughly 1.84 rather than 2. The near-identical exponents on both sides say the psychological squeezing is roughly symmetric around the reference point, which leaves the steeper pull of losses to be carried by a separate parameter, loss aversion, rather than by the curvature itself.
What the evidence shows
The behavioral fingerprints of the curve, cautious choices among gains and bolder choices among losses, have held up unusually well for a decades-old result. Ruggeri and colleagues (2020) re-ran the original problems with 4,098 participants across 19 countries and 13 languages, adjusting only the currency; about nine in ten items replicated and twelve of thirteen theoretical contrasts held, though effect sizes were generally smaller than in 1979. So the direction of diminishing sensitivity is robust and travels across cultures, while its strength is softer and population-dependent. What replicates is the shape of preferences, not a claim about any single mechanism; attention, memory, or perceptual scaling could each produce the same bend.
Not the same as loss aversion
Diminishing sensitivity is routinely muddled with loss aversion, yet they are distinct properties of the same curve. Diminishing sensitivity is the bending: the slope grows gentler the farther you travel from the reference point, in either direction. Loss aversion is the kink at the reference point itself, where the loss arm falls away more steeply than the gain arm rises. One concerns curvature within a domain; the other, a discontinuity between domains. They predict different things, since diminishing sensitivity alone yields risk seeking over losses, while loss aversion explains why people refuse a coin flip to win $110 or lose $100. A working value function needs both, estimated separately.
Where it breaks down
The principle is a first approximation, not a law, and it frays at the edges. Concavity for gains predicts caution at every stake, yet for tiny amounts people turn risk-seeking, the 'peanuts effect.' Weber and Chapman (2005) showed that someone who rejects a scaled-up gamble will happily buy a long-shot ticket when only pennies are at stake, a reversal plain concavity cannot generate; they trace it partly to anticipated disappointment being negligible when so little rides on the outcome. Intertemporal choice supplies the mirror image: the magnitude effect, where larger sums are discounted proportionally less, implies sensitivity that rises with size rather than falling. Both anomalies sit outside a single diminishing curve.
Where it shows up: valuing many lives
The same compression that flattens the value of extra dollars can flatten the value of extra lives. Fetherstonhaugh and colleagues (1997) called it psychophysical numbing: a program saving a fixed number of people is judged more worthwhile when the endangered population is small than when it is large, so 4,500 rescued from a camp of 11,000 feels more valuable than the identical 4,500 from a camp of 250,000. The pattern underwrites the identifiable-victim effect, lopsided disaster response, and how statistical lives are priced in policy. One honest caveat: large later replications were mixed, with some designs showing the reverse, so treat scope insensitivity for lives as a real tendency here, not a settled constant.
Examples
You'll drive across town to save $10 on a $20 calculator but not on a $1,000 laptop — the same $10.
Companies announce a restructuring, a write-off and a lost contract on one call. The second and third blows land far softer than the first, so bad news is cheaper bundled.
Someone already two thousand pounds into an overdraft shrugs at another two hundred in charges. The first two hundred kept them awake; each further pound of loss stings less.
A $2,000 raise feels like a windfall to someone earning $30,000 and an afterthought to someone earning $300,000. The dollars are identical; only the baseline they land against differs.
A project already three months behind schedule absorbs one more week's slip almost unnoticed, while the same week's delay on an on-time plan would trigger alarm and escalation.
First described in Kahneman & Tversky (1979).
Key references
- Ruggeri, K., Alí, S., Berge, M. L., et al. (2020). Replicating patterns of prospect theory for decision under risk. Nature Human Behaviour, 4(6), 622-633. doi.org/10.1038/s41562-020-0886-x
- Weber, B. J., & Chapman, G. B. (2005). Playing for peanuts: Why is risk seeking more common for low-stakes gambles? Organizational Behavior and Human Decision Processes, 97(1), 31-46. doi.org/10.1016/j.obhdp.2005.03.001
- Fetherstonhaugh, D., Slovic, P., Johnson, S., & Friedrich, J. (1997). Insensitivity to the value of human life: A study of psychophysical numbing. Journal of Risk and Uncertainty, 14(3), 283-300. doi.org/10.1023/A:1007744326393
- Tversky, A., & Kahneman, D. (1992). Advances in prospect theory: Cumulative representation of uncertainty. Journal of Risk and Uncertainty, 5(4), 297-323. doi.org/10.1007/BF00122574
- Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of decision under risk. Econometrica, 47(2), 263-291. doi.org/10.2307/1914185