A finding from a 2008 paper in Science has had an unusual second life in the popular conversation about happiness. The paper, by Elizabeth Dunn, Lara Aknin, and Michael Norton at the University of British Columbia, reported that people who were given a small windfall and asked to spend it on someone else were happier at the end of the day than people asked to spend the same amount on themselves. The headline result has been repeated in TED talks, self-help books, and corporate wellness materials for fifteen years, and the popular version is now treated as a settled fact about how happiness works.
The research is real and the finding is interesting. The popular version also tends to leave out several things that anyone trying to act on the finding would want to know. This piece is about what the research actually shows, and what is worth keeping and discarding from the way the result has been told.
What the original experiment did
The headline experiment, Study 3 in Dunn, Aknin, and Norton’s 2008 paper “Spending Money on Others Promotes Happiness” in Science, worked like this. The researchers approached 46 University of British Columbia undergraduates on campus in the morning. Each was given an envelope containing either $5 or $20, in cash, with instructions to spend it by 5 p.m. that day. Half were told to spend the money on a bill, an expense, or a gift for themselves. The other half were told to spend it on a gift for someone else or a donation to charity. Researchers called participants in the evening and asked them to rate their happiness. The group that had spent the money on others reported greater happiness, on average, than the group that had spent it on themselves. The amount of money did not matter much; the direction of spending did.
This is the experiment the popular version is almost always referring to. It is worth pausing on the size of the sample. Forty-six undergraduates is not a large study by current standards in social psychology, and a great deal has happened to the field’s evidentiary standards since 2008, much of it driven by the discovery that many widely cited findings from that era do not replicate cleanly when tested with larger samples and pre-registered methods.
What the cross-cultural paper added
The claim that the effect “held across countries, income levels, and cultures” comes from a separate paper. Aknin, Barrington-Leigh, Dunn, Helliwell, and colleagues’ 2013 paper “Prosocial Spending and Well-Being: Cross-Cultural Evidence for a Psychological Universal,” in the Journal of Personality and Social Psychology, brought a much larger evidence base to bear on the question. Study 1 in that paper used Gallup World Poll survey data from 136 countries and found that the association between prosocial spending and self-reported happiness held in poor and rich countries alike. Two further studies ran small experimental tests in Canada and Uganda, and reported that recalling a past instance of prosocial spending caused a measurable lift in mood in both settings.
This is a stronger evidence base than the original 2008 study, but it is worth being precise about what each piece of it shows. The 136-country survey data is correlational. It demonstrates that around the world, people who report giving more also report being happier, but it cannot rule out that happier people give more rather than the other way around. The causal experimental evidence in the paper covers two countries, not all 136. The broader claim of a psychological universal is the authors’ interpretation of these two pieces of evidence read together, not a finding any single study established on its own.
What the replication record looks like
The original 2008 paradigm has been retested several times in the years since. The results are mixed. A 2022 close replication by Garam Kim and colleagues, published in PLOS One, ran 133 participants through the same procedure as Dunn, Aknin, and Norton’s Study 3 and did not find a statistically significant effect using the original analysis. They reported a weaker effect using a different happiness measure, in the same direction as the original finding. This is less than a full failure to replicate, but it is also less than a confirmation.
A larger registered replication by Lara Aknin and colleagues in 2020, with a sample of 730, found stronger support for the original effect (p < .001), though using a modified paradigm that removed several differences between the prosocial and personal spending conditions to isolate the effect more cleanly. The current state of the evidence is that the prosocial spending effect is probably real, but smaller and more dependent on specific conditions than the popular version of the finding suggests. The most defensible summary is that giving money to someone else can produce a modest mood lift under some circumstances, more reliably than the framing “spending on yourself makes you happy” would predict, but less reliably than the headline framing of the original study implied.
What the popular version leaves out
The version of the finding that circulates in popular discussion tends to make three moves that the underlying research does not support.
The first is to treat the effect as large. The actual effect sizes in the original paper and the replications are modest. Prosocial spending produces a measurable but small lift in mood, not a transformation of someone’s day. A reader who tries the experiment on themselves and notices that they feel only slightly better afterward is having an experience consistent with the research, not a failed one.
The second is to treat the effect as universal across kinds of giving. The research is specifically about spending money. It does not, on its own evidence, generalize cleanly to time, attention, effort, or emotional labor. Some related research suggests these forms of giving also lift mood, but the strong version of “prosocial behavior makes you happy” is a synthesis of multiple lines of work, not a single finding.
The third is to treat the effect as advice. “Spend money on others to be happier” reads as a recommendation. The research is descriptive. It tells us something about what tends to happen when people spend money in a particular direction. It does not establish that strategically organizing one’s spending around mood maximization is a good idea, and the authors of the original studies have been careful, in their own writing, not to make that leap.
What is worth keeping from the finding
The useful version of the result, stripped of the overclaiming, is something like this. People appear to derive emotional benefit from spending money on others in a way that is at odds with the assumption built into a lot of consumer culture, which is that the route to happiness runs through accumulating things and experiences for oneself. The research does not show that prosocial spending is a route to lasting wellbeing. It shows that, in the moment, the small act of directing money outward seems to produce something the equivalent act of directing money inward does not, and that this pattern shows up in modest but reasonably consistent form across several studies and several cultures.
This is genuinely interesting. It is also less dramatic than the popular framing of the finding. The popular framing tends to oversell the evidence, to imply that science has revealed a hidden lever for happiness, and to obscure the fact that the lever, on close examination, is small and conditional. The research is most useful when held at its actual size: a real, modest, replicable-with-caveats effect that complicates the standard consumer story without quite replacing it with the inspirational alternative the talks and books tend to draw from it.
What this is not an argument against
None of this is an argument against giving. Generosity has its own justifications, none of which depend on whether it produces a measurable mood lift in the giver. People have given to others throughout recorded history for reasons including love, duty, religion, community membership, and the ordinary recognition that other people’s lives matter. The case for any of these reasons is not strengthened or weakened by the precise size of the effect in a 2008 study at the University of British Columbia.
The research is a small data point in a larger picture about how human beings work. It is worth knowing about. It is also worth knowing in the form the evidence actually supports, rather than in the inflated form it tends to take by the time it reaches the talks and the books. The honest version is more modest than the popular one, and more durable, because it does not depend on any single small study holding up cleanly. It points at something real about people, even if the precise size and shape of that something is still being worked out by the field that first noticed it.
