Unlike the kind of helping where you have to do something – help an old lady cross the street, volunteer to give meals to the dispossessed, go door-to-door to fundraise for a cause – virtue signalling often consists of completely costless actions, such as changing your profile picture or saying you don’t like a politician’s stance on immigration. Bartholomew complains that “saying the right things violently on Twitter is much easier than real kindness”.

In everyday discourse, the people who accuse others of virtue signaling are often not interested in doing real moral analysis – mostly, they want to discredit their political opponents (“My allies are heroically rallying for a just cause, people on the other side are virtue signaling”). It might be more illuminating to look at what science says on the subject.

Over the past few decades, scientists in a variety of fields have developed sophisticated analyses of signaling as a general phenomenon – how humans (and other animals) send signals designed to convey information to other individuals. The insights of signaling theory have had a huge impact on biology and the social sciences. They also tell us that virtue signaling is more nuanced and more interesting than the picture painted by conventional wisdom and political rhetoric. As it turns out, there are bad and good things about virtue signalling – but probably not for the reasons we think.

Why do we scold virtue signalers for having it easy? The urge to dismiss someone’s actions because they took no effort is powerful. But does it not make more sense to focus on what that action actually achieves? Why do we often focus on the costs people pay rather than how effective they are at making the world better?

Why do we have the strong emotions we have about virtue signaling, and is it actually good or bad?

A few decades ago, biologists and economists struggled with similar questions.

Why are peahens so attracted by the peacocks with the most extravagant tails – which are very costly to maintain but otherwise seemingly useless?

Why do employers care that you put yourself hundreds of thousands of dollars in debt to get an Ivy League degree in sociology with no obvious relevance to the job?

In the 1970s, the zoologist Amotz Zahavi and the economist Michael Spence offered a provocative answer. They argued that the cost paid by the peacock (or the college graduate) is the whole point.

Their argument is a bit subtle, so it is worth carefully looking at how it works.

Communication is difficult because individuals have incentives to lie. Employers are looking for certain qualities (intelligence, conscientiousness, ambition) in their employees. They could ask the people they interview if they are intelligent and conscientious, but why wouldn’t the job candidates simply lie?

Instead, employers select their employees on the basis of signals that are difficult to fake, such as university degrees. In general, having the qualities that employers value makes it easier to get a degree. People who do not have the right mix of intelligence, conscientiousness and ambition will find college more difficult, and either drop out or spend much more time completing their studies. People who anticipate that getting a degree would be too costly for them will opt out.

So, in principle, even if nothing you had learnt was relevant to the job you want, completing the degree still sends a valuable signal to potential employers: you are the kind of person for whom this high-effort achievement is easy enough. Because it sends a valuable signal, it is in your interest to get a degree, and in the employer’s interest to hire you on its basis.

A similar argument applies in the biological domain, but with natural selection in the driver’s seat. Growing an extravagant tail is moderately costly for a healthy peacock – but a diseased bird would put its life at risk if he spent that much energy growing the ornament. Therefore, only the peacocks in good enough condition can afford to grow an elaborate tail. As such, natural selection favours peahens who prefer peacocks with a long tail, because these peahens mate with healthy males, and get healthy offspring as a result.

Costly signals – signals that are honest because of the fact that they are costly – are ubiquitous. Why do people give flowers to their romantic interests, or take them to overpriced restaurants? Probably because these acts are costly: were the suitor not interested in a long-term relationship, he would have little incentive to invest such effort. His gifts function not because roses are particularly useful items, but because they are a costly signal of his commitment.

Here is why this matters for virtue signalling. Dishonesty is a major problem in the moral domain. People want to appear good, because it wins them friends and social status. Our moral sense evolved because people who convince others of their moral qualities reap such social benefits. But what prevents someone from pretending to be a good person, reaping all the social benefits, and not following through?

Throughout human evolution, being able to discriminate true allies (who stick with you no matter what) from fair-weather friends (who abandon you when you fall ill) could make the difference between life and death. As such, humans are obsessed with moral hypocrisy. We carefully scrutinise potential romantic partners, friends or team members for signs that they’re not only in it for the money. And since – per the logic of costly signalling – the costs that people are willing to pay are a reliable signal of their commitment, we pay extra attention to these costs when we evaluate other people.

Social psychologists have found that, when we see someone perform an altruistic act, we’re suspicious that they’re really being altruistic if they derive some benefit from the act. Clever cognitive psychology experiments even show that we categorise other people on the basis of the costs they are willing to pay to benefit their group – but not on the basis of the amount of benefits they actually provide.

This is probably why we find virtue signalers irritating. They are doing things that might gain them social status – the approval of society, a place on the right side of history. But are they actually committed to the causes they support? Or are they just interested in the social benefits?

When they are not paying any meaningful costs, virtue signallers activate the alarm bells that millions of years of evolution put in our heads to protect us from fair-weather friends and other moral hypocrites.

So let’s concede that some virtue signalling is fake; but does that mean it is bad? Here it is useful to take a step back from our default mode of thinking. Evolution designed our brain to make us good at small-scale interaction, but we not very good (or especially concerned) at evaluating the large-scale social effects of things. As such, it is easy for a polemist to throw discredit on someone who virtue-signals by pointing out that there is no guarantee that the virtue-signaler actually shares your moral values. But is this the right yardstick by which to evaluate these signals?

Life is rife with coordination problems. Consider passing someone on the street going the other way. You both have a shared incentive to coordinate about which side of the street to walk on, so that you don’t bump into each other. Even though the other person is a complete stranger, there is no particular reason she would try to deceive you. In such circumstances, people send signals (eg. move suddenly and flamboyantly toward one side) to successfully coordinate. Mathematical models show that these costless signals can be crucial in helping people solve otherwise thorny coordination problems.

Coordination is crucial in the moral domain too. Imagine you live in a society that practises slavery, and you think you are the only one morally revulsed by it. Should you speak out about your concerns? If you think that everyone else is indifferent, you might be afraid that others will think you are weird, that the people benefiting from the system will punish you, and that you stand no chance to make a difference anyway.

The paradox is that, even if many people are in this situation – everyone is concerned but convinced that no one else is – they might fail to act, despite having the majority opinion.

But speaking up can start a chain reaction. The more individuals raise their voice to denounce what they see as a moral problem, the more the initially silent people realise they are not alone and speak up in turn.

Loud and public signals are especially effective as establishing common knowledge of a moral norm ­making sure that everyone knows about the moral norm, that everyone else knows that everyone knows about the moral norm, that everyone knows that everyone knows that everyone knows (and so on).

Psychology experiments have demonstrated that common knowledge is a powerful determinant of social behaviour: people are much more likely to coordinate on a joint action when everyone knows that everyone knows that working together will generate good outcomes.

In addition to fostering coordination, common knowledge prevents people from hiding behind the veil of plausible deniability. To get away with selfish behaviour, we often pretend to ignore its consequences. If you can plausibly say that you didn’t know about the poor working conditions of people in sweatshops, people will judge you less harshly for buying cheap clothes. But if many people virtue signal by campaigning for better workers’ rights, the issue rises to common awareness – and, when everyone can expect everyone to know, it is harder for you to claim ignorance as a defence.

Viewing morality as a coordination game suggests that public opinion can undergo rapid shifts, as society coordinates on new moral norms. And this is indeed what we observe: public opinion on a variety of subjects – such as racism and gay rights – has shifted dramatically in a progressive direction over the past few decades (sometimes within a few weeks).

In sum, virtue signalling can be a powerful force for social change, by creating common knowledge around a moral issue that people would otherwise ignore (out of complacency or selfishness).

Importantly, this works even when there is no guarantee the people who are sending the signals are particularly virtuous or committed to the cause.

Mark and Bob both made a $1,000 donation to a charity. Before making his donation, Mark read extensively about different charities, seeking the one with the best ROI – the charity that runs programmes with the most impact on people’s lives per dollar invested. In the process of his search, Mark would often read about an organisation that does great things, but then drop it from consideration because there was no evidence that it would use his money in the best possible way. He finally settled on a charity that provides a treatment against parasitic worms that plague many people in Africa.

Bob was watching TV one evening when he saw an ad for a charity that gives teddy bears to convalescent children in hospitals. Bob was overwhelmed by emotion at the plight of the children, and immediately made a transfer to the charity.

Mark’s donation will make a much greater impact on people’s lives. But he comes off as cold and calculating, and we can’t repress a nagging doubt about the kind of person he is. Would he still have donated if he had found that none of the organisations he read about was effective enough? Bob by contrast leaves no doubt about his good nature: he spontaneously gave his money without the slightest care about crude materialistic concerns such as economic efficiency.

The tale of Mark and Bob illustrates one of the bleak insights of signalling theory: the concrete impact of an altruistic act (how much it actually helps other people) is often dissociated from the signal it sends about what kind of person we are. This means that, to convince people that you are good, the most persuasive signal you can send is often not what will actually produce the most good.

Why might people be motivated to send extreme signals of their commitment to a cause? The answer lies in the fact that, when we send virtue signals, many of the things we seek – such as friends and social status – are rival goods. The people you want to be friends with can have only so many friends, so you want to convince them that they should pick you. The social groups you belong to can have only so many leaders and, in order to gain influence, it helps to show that you are more committed to the group’s ideology than is the average member.

When individuals send signals to try to convince others that they are better than average, the result is often what signalling theorists call a ‘runaway’: an arms race toward more and more extreme signals. If peahens want to mate with the peacocks whose tails are more extravagant than the average peacock, natural selection favours peacocks with increasingly more extravagant tails over successive generations. If everyone has a high-school diploma, students need to start getting bachelor’s degrees in order to become distinctive in the eyes of employers.

In the moral domain, runaway signaling happens when people try to elevate their moral status by doing and believing things that not everyone else does. For example, everyone is opposed to killing humans: saying that you think people should not kill each other does not set you apart from others. But not everyone is opposed to eating animals, so being vegetarian or vegan can effectively increase your moral standing.

When we see someone virtue signalling, we often have strong reactions – sometimes admiration, sometimes annoyance or contempt. But these intuitions are the product of psychological mechanisms that are designed to help us evaluate if that person could be a good friend or a good ally, not to help us evaluate whether the person’s action will have a positive impact on the world.

The emerging science of signaling shows that these things can often come apart. Keeping this insight in mind is essential as we navigate an increasingly noisy world.