Public Charity and Who Is Watching
Public charity comes with an audience. A widely read book and the experiments it relies on suggest the audience changes what people do, though less certainly than either would like.
Key takeaways
- Kevin Simler and Robin Hanson argue in The Elephant in the Brain that giving is partly a display, and they lean on a 2007 study to show it.
- In that study, led by Vladas Griskevicius, a romantic frame of mind raised women's helping in public and left it unchanged in private.
- Men in the same frame of mind became more willing to spend on conspicuous luxuries rather than necessities.
- A 2017 study by Joël Berger, replicating earlier work on status labels rather than the 2007 experiments, found no social reward for visible luxury or green labels, so producing a signal is not the same as being paid for it.
- For fundraisers, public recognition is a bet worth testing alongside a quiet route to give, not a lever to assume.
At a charity dinner the auctioneer asks for pledges, and hands go up around the room. Each one is seen by every table. The same people, alone at home with a donation page, might give more, less or nothing at all. Public charity comes with an audience, and a widely read book and a psychology lab have each argued that the audience is doing more of the work than anyone at the dinner would say.
What does The Elephant in the Brain claim about giving?
Kevin Simler and Robin Hanson's The Elephant in the Brain argues that people act on motives they don't report, and often can't see in themselves. Their chapter on charity applies that to giving. Generosity, they suggest, is partly a display: a way of showing others what kind of person you are, which would explain why so much giving happens where it can be seen.
The argument is uncomfortable rather than cynical. The authors do not dismiss the giving itself, and they do not say causes fail to benefit. They say the reasons people give themselves for giving are incomplete, and that the missing reasons tend to be social.
They know this is a hard claim to test. Motives are private, and almost any behaviour can be described as secretly strategic after the fact. That is a real risk with arguments about hidden motives. If giving in public shows a desire to be seen, and giving anonymously shows a subtler desire to be seen as modest, the theory can absorb any behaviour and never be wrong.
So the book reaches for experiments that make a prediction which could fail, and the one it leans on in the charity chapter is a 2007 study of romantic motives. Its own notes flag that experiments priming romantic motives have a mixed record. That caution deserves to travel with the study wherever it's cited, this article included.
What did the 2007 experiments actually find?
Vladas Griskevicius led the work, with Joshua Tybur, Jill Sundie, Robert Cialdini, Geoffrey Miller and Douglas Kenrick, across four experiments in the Journal of Personality and Social Psychology. The theory under test was costly signalling: a display that costs something is believable because it costs something, so generosity could work as a signal of qualities a partner might value. If so, a romantic motive should raise generosity, and specifically where someone can see it.
People were put into a romantic frame of mind and asked what they would spend and how willing they were to help. As Simler and Hanson describe one version of the set-up, participants imagined an ideal first date before answering. The outcomes were stated willingness, the kind of answer people give about what they would do, not money handed over.
In men, the romantic motive raised willingness to spend on conspicuous luxuries and not on basic necessities. In women, it raised helping in public and left helping in private unchanged.
The women's result carries the weight. A simple mood account predicts that a warmer frame of mind makes people more helpful everywhere. That isn't what happened. Helping rose where it would be seen and stayed flat where it would not, which is the pattern a signal predicts and a mood does not.
Helping rose where it would be seen and stayed flat where it would not, which is the pattern a signal predicts and a mood does not.
The authors also report that the romantic motive made men more helpful where helping allowed a display of heroism or dominance, and led women to spend more publicly on helpful causes. The usual summary, that men buy and women help, is too neat, and it tends to drop the crossovers that make the paper more interesting than its headline. Both groups moved towards some form of visible display. They reached it by different routes.
Is it romance, or just being watched?
The experiments cannot fully separate two readings. In one, a romantic motive switches on signalling in particular, as the evolutionary account predicts. In the other, a romantic frame of mind simply makes people more aware of how they look to others, and anything that did the same would produce the same public tilt.
The public and private contrast cannot choose between them, because both predict it: a person signalling to a partner and a person simply conscious of being seen would each help more in public than in private. What the contrast does rule out is a plain warm mood, which would raise helping everywhere. We lean towards the signalling reading because it is the one the experiments were designed to test, but the design does not let us bet much on it.
A second line of evidence points the same way. In 2010 Griskevicius, Tybur and Bram Van den Bergh reported that status motives raised people's preference for green products over more luxurious ones when shopping in public, not in private, and when the green option cost more rather than less. We have written about that study separately, and the parallel is close: the same restriction to public settings, found with products rather than with helping. It shares authors with the 2007 paper, so it isn't independent confirmation, but it makes the same prediction in a different domain and finds the same restriction to public settings.
Does a visible status signal actually pay?
Producing a display and being rewarded for it are different things, and the second has been tested for a different kind of display: status labels on clothing. Joël Berger's 2017 study in PLOS ONE was an extended replication of earlier research on whether visible luxury and green labels earn their wearers better treatment. It did not replicate the 2007 charity experiments, and it tells us about labels rather than about giving. Across five experiments, run double-blind in both average and poorer neighbourhoods, it found no positive effect of a luxury label anywhere, no effect of a green label, and a tendency for the luxury label to do worse in poorer areas.
Double-blind means that the people carrying out and taking part in a study do not know which condition is being tested, which removes the chance that expectations shaped the result. That design feature is why the 2017 study carries weight on this question.
Set beside the 2007 findings, that is less of a contradiction than it looks. People can reliably reach for a display whose payoff is doubtful. The fact that a signal is produced is not evidence that anyone reads it, let alone rewards it. For public charity that is almost a comfort, because the cause receives the money whether or not the display impresses anyone.
What should a fundraiser do with this?
Treat visible recognition as a reasonable bet to test, not a law. Named donor lists, pins and pledges read out at a dinner all attach an audience to a gift, and an audience is the condition these experiments varied. But the experiments measured willingness in a laboratory under a romantic motive. They did not measure donations, and many fundraising appeals rest on reputation among colleagues, neighbours or friends, which is a motive the research did not touch.
There is a second reason for care, and it is our concern rather than a finding. Public charity changes the gift as well as the giver, and it is plausible that some people would rather not give in front of others. The 2007 experiments only measured what visibility added under a romantic motive. They could not see whether it put anyone off, and none of the sources here measures that, so a fundraiser leaning on public charity should watch for it rather than assume it away.
The practical design follows from that gap. Offer public charity as a choice rather than a default, keep a quiet route open for donors who want one, and compare what each brings in. The 2007 paper gives a reason to expect visibility to help with some givers. It gives no reason to assume it helps with all of them, and a policy that only rewards the visible gift may lose the anonymous one.
For public charity appeals the evidence is a starting hypothesis rather than a playbook. Measuring that properly means watching more than the total raised. Worth tracking, at a minimum:
- the average gift, with and without recognition
- the share of donors who choose to be named
- how many give again the following year
- how many quietly stop
Public charity can raise the first of those numbers and lower the others, which is why a single total is a poor test.
It is also worth being clear about what none of this shows. Generosity that responds to an audience is not fake generosity. Something can be sincere and still be switched on by being seen, and most human behaviour has both parts. Thorstein Veblen made the related point about consumption in 1899: a signal needs an audience able to read it. Public charity provides one. Whether that audience changes how much good gets done is an experiment we have not found anyone running with real money.
Common questions
Do people give more when others can see?
In these experiments, women in a romantic frame of mind helped more in public and no more in private. The study measured willingness in a laboratory rather than real donations, so it shows a direction, not an amount.
Did men and women behave differently?
Yes. The romantic motive raised men's spending on conspicuous luxuries and women's public helping. Men also helped more where helping showed heroism, and women spent more publicly on good causes.
Is the 2007 study reliable?
It is a widely cited four-experiment paper, and we have found no direct replication either way. Research priming romantic motives has a mixed record, which Simler and Hanson's own notes acknowledge, so the public and private pattern deserves more weight than the size of any effect.
Does this mean generosity is not genuine?
No. It shows that some generosity responds to being observed, which is about what prompts it, not about whether the motive is sincere.
Sources
Simler, K., & Hanson, R. (2018). The Elephant in the Brain. Oxford University Press. The charity chapter draws on the 2007 study, and the book's notes flag the mixed record of romantic-priming experiments. Read in full by this publication, 2026-09-14.
Griskevicius, V., Tybur, J. M., Sundie, J. M., Cialdini, R. B., Miller, G. F., & Kenrick, D. T. (2007). 'Blatant benevolence and conspicuous consumption: when romantic motives elicit strategic costly signals.' Journal of Personality and Social Psychology, 93(1), 85-102. Four experiments; inducing mating goals in men increased willingness to spend on conspicuous luxuries but not basic necessities; in women mating goals boosted public but not private helping; mating motivation induced more helpfulness in men in contexts allowing displays of heroism or dominance, and led women to spend more publicly on helpful causes. Verified at the published record 2026-10-06. doi:10.1037/0022-3514.93.1.85
Veblen, T. (1899). The Theory of the Leisure Class. Read from the primary text by this publication, 2026-09-20. Cited for the condition that a signal requires an audience able to read it.
Griskevicius, V., Tybur, J. M., & Van den Bergh, B. (2010). 'Going green to be seen: status, reputation, and conspicuous conservation.' Journal of Personality and Social Psychology, 98(3), 392-404. doi:10.1037/a0017346. Status motives raised preference for green products in public but not in private, and when the green option cost more. Verified at the published record 2026-10-05.
Berger, J. (2017). 'Are luxury brand labels and “green” labels costly signals of social status? An extended replication.' PLOS ONE, 12(2), e0170216. States that some experiments in the previous study were not conducted in a perfectly double-blind manner, and resolves this by replicating variations of the original design double-blind, adding a green-label condition, and running in both average and below-average socioeconomic neighbourhoods. Found no positive effects of the luxury brand label in any of five experiments, no green-signalling effect, and a negative tendency of the luxury label in poor neighbourhoods; concludes a signalling account must take the characteristics of observers into account. Verified at the published record 2026-10-01. doi:10.1371/journal.pone.0170216