Costly signals are expensive actions that reveal private information in Game Theory. Because the signal is hard to fake, it helps the receiver judge quality, intent, or type more accurately.
Costly signals are actions in a signaling game that give other players information, but only because sending the signal is expensive enough to separate different types. In Game Theory, that cost is what makes the signal believable. If anyone could do it easily, it would not tell the receiver much.
The basic idea comes up when one player knows something private and another player has to decide how to respond. The sender might know their quality, commitment, skill, or reliability, while the receiver can only observe the signal. A costly signal works because high-quality senders are more willing or able to pay the cost than low-quality senders.
That cost does not have to mean money every time. It can be time, effort, risk, lost opportunities, or a price that would hurt a weak type more than a strong type. For example, in Spence's Job Market Model, education can work as a signal not just because it teaches skills, but because completing it is harder for lower-productivity workers to justify.
The key question is whether the signal separates types. If the action is too cheap, everyone sends it and the signal becomes noise. If it is too expensive, even the good types may avoid it. The useful range is where the cost creates a gap between types, so the receiver can update beliefs and choose a better response.
Costly signals are part of signaling games and information revelation. They contrast with cheap talk, where messages can be sent without cost and are much easier to fake. In a signaling problem, you usually track who knows what, what signal each type chooses, what beliefs the receiver forms, and whether the outcome is separating, pooling, or somewhere in between.
Costly signals explain how Game Theory handles situations where one side knows more than the other side. That is a big deal in models of jobs, markets, politics, ads, warranties, and any setting where trust is not automatic. Without a credible signal, the receiver has to guess, and that guess can lead to bad choices.
This term also helps you tell the difference between simple communication and strategic communication. A message only changes beliefs if the receiver thinks the sender had a reason to send it. Costly signals create that reason, so they are a core tool for solving signaling games and predicting whether information gets revealed.
When you work problems in this topic, you often ask whether a signal separates high and low types, whether it supports a Bayesian Nash Equilibrium, and whether the receiver should trust the action. That makes the term useful far beyond one example. It shows up anytime the game is about credibility, not just choice.
Keep studying Game Theory Unit 8
Visual cheatsheet
view galleryAsymmetric Information
Costly signals matter only when one player knows something the other player does not. Asymmetric information creates the gap that signaling tries to close. If both sides already know the same thing, the signal is unnecessary and does not change the game in a meaningful way.
Signaling Theory
Signaling theory is the broader framework that explains why and when actions can reveal hidden information. Costly signals are one of its main mechanisms. The theory asks which signals are believable, which types send them, and how receivers should update their beliefs after seeing them.
Bayesian Nash Equilibrium
In a signaling game, a costly signal only matters if it fits into an equilibrium where beliefs and actions make sense together. You often test whether each type wants to send the signal and whether the receiver's response is rational given those beliefs. That is where Bayesian Nash Equilibrium comes in.
cheap talk games
Cheap talk games are the contrast case. Messages there are free, so they are easier to bluff and less likely to separate types on their own. Costly signals carry more weight because paying a cost can make the message credible in a way cheap talk cannot.
A problem set or quiz will usually ask you to decide whether a sender's action is a real signal or just noise. You might be given two types, two possible actions, and a receiver response, then asked whether the costly action creates a separating equilibrium or a pooling one. Sometimes the task is to explain why a signal works, such as why education, warranties, or advertising can reveal quality.
When you answer, name the private information, identify the cost, and say what belief the receiver forms after observing the signal. If the cost does not distinguish types, point out that the signal loses credibility. If the game includes payoffs, check whether each type would actually choose the signal given the receiver's likely response.
Cheap talk is communication with no direct cost, so it can be easy to fake and may not separate types. Costly signals, by contrast, require a real sacrifice, which can make the action credible and change the receiver's belief.
Costly signals are expensive actions that reveal private information in a signaling game.
The cost is what makes the signal credible, because not every type can or wants to pay it.
A strong costly signal usually separates types, while a weak or cheap one may be ignored.
The receiver uses the signal to update beliefs and choose a response based on what the action suggests.
If the signal is too easy to copy, it stops working as information.
Costly signals are actions that a sender takes at a real cost to prove something about their private information. In Game Theory, the cost makes the signal more believable because low-value or low-quality types are less likely to copy it.
The sender chooses whether to take an action that is observable to the receiver. If the cost is structured right, high types are willing to signal and low types are not, so the receiver can infer something useful from the choice.
No. Cheap talk is costless communication, so it can be persuasive but is easier to fake. A costly signal requires effort, money, risk, or another sacrifice, which can make it much more credible.
Education credentials are a classic example in Spence's Job Market Model. The degree may signal ability not only because of what you learned, but because completing the credential is harder for lower-productivity workers to justify.