What is costly signalling?
A signal earns belief when it costs less to send for those who really have the quality it announces. Economics, biology and consumer research each reached that condition separately.
Answer to the question
What is costly signalling?
Costly signalling is the idea that a signal is believable when it costs less to send for those who have the quality it announces. Spence modelled it for hiring in 1973, Zahavi proposed it for animal displays in 1975, and its strong biological form is still disputed.
Key takeaways
- Spence (1973) showed in a worked example that schooling can sort workers when it costs a low-productivity worker twice what it costs a high-productivity one, even if it adds no productivity.
- Grafen (1990) reported that an honest signal is stable only when advertising lowers fitness and its marginal cost is lower for the better signaller.
- Penn and Számadó (2020) argued that Grafen's models support a trade-off account and that the handicap principle lacks theoretical and empirical support.
- Nelissen and Meijers (2011) ran seven experiments in which visible luxury labels brought wearers better treatment and, in some cases, financial gains.
- Simler and Hanson (2018) extended the framework to everyday life, which is their application of the biology and not a result of it.
A diploma can sort job applicants even when it teaches nothing an employer values. Michael Spence modelled this, and the result rests on one assumption about what the diploma costs different people. Employers see the credential and cannot see the person's productivity, which is the situation that makes a credential worth reading as a message. The arrangement is costly signalling, the idea that a signal earns belief when it is cheaper to send for someone who really has the quality than for someone who does not.
A Single Condition Does the Work
Spence published the model in 1973. In one of his worked examples a year of schooling costs a low-productivity worker twice what it costs a high-productivity one, and employers who see the diploma but not the productivity pay by the diploma. He stated the requirement plainly. A signal separates applicants only if its cost moves in the opposite direction to productive capability, so that the more capable find it cheaper. He called this a critical assumption, and he added that it is necessary but not sufficient, because there must also be enough signal levels inside the range where the cost difference matters.
The mechanism runs through the workers' own choices. In Spence's account education is something a person can observe and also alter, so workers invest in it in order to signal, and employers set wages according to what the credential seems to say. In his worked example the schooling adds no productivity at all and still sorts the applicants, because the price of each year differs by type. The credential works as a message only because the two kinds of applicant face different bills for it.
The condition says nothing about how large the cost is. Under costly signalling, a signal that costs every sender the same teaches the receiver nothing, since a sender without the quality could send it too. A modest signal that costs the weak far more than the strong can carry a great deal. Costly signalling therefore describes a gap in cost between types of sender, and the size of the bill is secondary.
Costly signalling in biology reached a matching condition from a different direction. Amotz Zahavi proposed in 1975 that mate choice could favour a handicap, a costly display that is believable because it is costly. Alan Grafen then built a formal model in 1990 and reported that an honest signal is stable only when advertising lowers the signaller's fitness and the marginal cost of advertising is lower for the better male. If the display were free, or cost good and poor signallers alike, it could not work as a handicap. His paper also discusses economic signalling models, so the biologist and the economist were describing the same structure.
Do Luxury Labels Pass the Test?
Rob Nelissen and Marijn Meijers took costly signalling to clothing. Across seven experiments they varied whether garments carried visible luxury brand labels. People who displayed luxury received better treatment from others, and in some cases that treatment brought financial gains. The study also examined when those benefits appear and what drives them. The authors concluded that spending on luxury goods can work as a social strategy because it signals wealth and status. The experiments tested the signal that the framework describes, a purchase whose cost is visible to the people who respond to it. What the Luxury Label Actually Buys follows that literature in more detail.
Kevin Simler and Robin Hanson carry costly signalling further in The Elephant in the Brain. They define a signal as anything used to convey information, and they call a signal honest when it reliably corresponds to an underlying trait. Because deception brings benefits without the costs of having the trait, they argue, the best signals are costly to produce and costlier still to fake.
Their examples are ordinary ones. A visit to a friend in hospital separates a loyal friend from a fair-weather one, because it costs time and comfort. A marathon says something about fitness that a claim of fitness cannot.
Their application of costly signalling to everyday human life is the authors' extension of the biology. Do Purchases Work as Self-Signalling asks a related question, about what a purchase says to its own buyer.
The Biologists Who Disagree About It
The 2020 review by Dustin Penn and Szabolcs Számadó takes a different view of what Grafen's model shows about costly signalling. In their paper they question whether Zahavi's later ideas of waste and signal selection hold together as logic. They also argue that Grafen's models support a separate idea Zahavi proposed in 1977, in which males adjust how much they invest in a signal according to their condition, which they describe as a trade-off model and not a handicap model. Their conclusion is that the handicap principle, the biological form of costly signalling, became dominant through a reading the models do not support, and that they find theoretical and empirical support for it lacking.
Grafen's paper presents its result as a vindication of Zahavi's main claims. The two accounts disagree about whether waste is what keeps a signal honest. They share the point that the signaller's cost depends on its quality.
| Account | Field | What keeps the signal honest |
|---|---|---|
| Spence, 1973 | Economics | A cost that falls as productive capability rises |
| Grafen, 1990 | Biology | A cost to fitness, with a lower marginal cost for the better signaller |
| Penn and Számadó, 2020 | Biology | Investment adjusted to condition, a trade-off and not a handicap |
Our opinion is that the cost difference between types is the part of costly signalling to rely on when the idea travels outside biology. Spence states it, Grafen's conditions contain it, and the trade-off reading keeps it. The stronger claim, that extravagance guarantees honesty in itself, is the part the dispute puts in doubt.
Common questions
What is costly signalling?
Costly signalling is the idea that a signal is believable when it costs less to send for those who have the quality it announces. Michael Spence modelled it for hiring in 1973 and Amotz Zahavi proposed it for animal displays in 1975.
Who first described costly signalling?
Spence's 1973 job market model and Zahavi's 1975 handicap hypothesis arose in economics and biology respectively. Alan Grafen's 1990 model gave the biological version a formal statement, and Penn and Számadó's 2020 review reads his models as a trade-off account instead.
Does costly signalling apply to luxury goods?
Nelissen and Meijers reported in 2011, across seven experiments, that displaying luxury brand labels brought better treatment from others and in some cases financial gains.
Sources
Simler, K., & Hanson, R. (2018). The Elephant in the Brain: Hidden Motives in Everyday Life. Oxford University Press.
Spence, M. (1973). “Job market signaling.” The Quarterly Journal of Economics, 87(3), 355-374. doi:10.2307/1882010 https://doi.org/10.2307/1882010
Zahavi, A. (1975). “Mate selection: A selection for a handicap.” Journal of Theoretical Biology, 53(1), 205-214. doi:10.1016/0022-5193(75)90111-3 https://doi.org/10.1016/0022-5193(75)90111-3
Grafen, A. (1990). “Biological signals as handicaps.” Journal of Theoretical Biology, 144(4), 517-546. doi:10.1016/S0022-5193(05)80088-8 https://doi.org/10.1016/S0022-5193(05)80088-8
Penn, D. J., & Számadó, S. (2020). “The Handicap Principle: how an erroneous hypothesis became a scientific principle.” Biological Reviews, 95(1), 267-290. doi:10.1111/brv.12563 https://doi.org/10.1111/brv.12563
Nelissen, R. M. A., & Meijers, M. H. C. (2011). “Social benefits of luxury brands as costly signals of wealth and status.” Evolution and Human Behavior, 32(5), 343-355. doi:10.1016/j.evolhumbehav.2010.12.002 https://doi.org/10.1016/j.evolhumbehav.2010.12.002