Researchers often find that after several weeks of compassion meditation practice, people perform more acts of kindness and caring outside the lab, such as visiting a retirement home or telling a coworker what they appreciate about her. They also report more feelings of compassion toward suffering people. They act more kindly toward strangers. They become less subject to the “bystander effect,” whereby everyone assumes that someone else will step up and come to the aid of a stranger in need. They more readily offer an exhausted woman a chair, as occurred in a 2015 study at Stanford University’s Center for Compassion and Altruism Research and Education. Good news, right?

I’m skeptical. Scientists, like humans generally, fall into the trap of looking for explanations that suit them best. Meditation may indeed do all that is claimed. But has research really demonstrated that fact?

The way these studies are usually conducted, they cannot rule out two alternative explanations for the effects attributed to meditation. One is a placebo response: People who practice compassion meditation might believe it makes them kinder, better people—and expectation makes it so. The other possible explanation is a desire to please the researchers: If volunteers guess what effect the scientists are looking for, they may consciously or unconsciously produce it. In either case, meditation itself wouldn’t be doing what researchers think it is.

These questions were running through my mind when I encountered a 2016 study on compassion meditation and generosity by researchers at the University of Colorado, Boulder. In general, in compassion meditation, you focus on suffering individuals, then groups of suffering people, then all of suffering humanity. In each case, you express the wish that they be free from suffering. It has been a mystery what compassion meditation actually does to produce the compassionate behavior and thoughts. So when Colorado’s Yoni Ashar and his colleagues set out to “show how compassion meditation changes how we think and feel about suffering people”— which would presumably lead people to be more generous—their study design was unusually exacting.

Generosity wanes if people perceive the world as full of threats and looming scarcity rather than abundance and security.

They showed 200 participants fictional biographies and photographs of people in need. The volunteers rated the following: their feelings of warmth toward each suffering individual; how much distress they felt at the person’s situation; how much blame they felt the individual deserved for his plight; and how much the person would benefit from a donation.

The volunteers could then donate up to $1 to each of the sufferers. Contrary to the belief that distress makes people less generous by making them turn away from suffering to relieve their own discomfort, the researchers found that greater distress predicted greater generosity. But so did these factors: thinking the person was blameless; believing a donation would actually help; and feeling warmth toward the sufferer. Curiously, having values, interests, and demographic features in common didn’t make people more generous, the researchers reported in the journal Emotion.

That challenged a widespread idea that we’re more generous to People Like Us (PLUs). We might be. But if so, it’s because we feel greater warmth toward PLUs and are more likely to think they’re not to blame for their suffering, rather than because similarity directly triggers generosity.

For the compassion meditation part of the generosity study, 58 new participants listened to biographies of suffering people—an orphaned child, a cancer victim, a homeless veteran— and were asked how much of the $100 the researchers gave them they’d like to donate to the person whose story they just heard. Then the group was split in three: some listened again to one biography; others engaged in a guided compassion-meditation session every day for four weeks; the rest were given a placebo, a nasal spray they were told contained an empathy-increasing hormone (a.k.a. H2O).

The participants were asked again if they would like to donate up to $100 to the individual they most recently learned about. The meditators reported significantly more compassion (positive feelings toward those in need) than the “compassion placebo” group did, suggesting that this effect is a real, direct consequence of meditating. But the effect on generosity wasn’t straightforward. The meditating group did not make more generous donations after four weeks of daily compassion meditation than they had before—despite the increase in generosity-associated feelings. Ashar and his colleagues aren’t sure how to explain why feelings didn’t translate into behavior, but they find one ray of hope in the data: The meditators’ giving did not drop off as sharply as the other groups’ did (the familiar phenomenon of “donor fatigue”).

It’s surprising that fundamental questions about generosity—what thoughts and feelings trigger it?—are still unanswered. In a world of vast unmet needs, where figuring out how to bring out the best in people could go far to alleviate suffering, that’s a troubling knowledge gap. But researchers are making some headway. They are learning, for instance, that generosity does not seem to be an instinctive, default behavior: When experimenters gave one child three marbles and another just one, only one-third of the kids spontaneously evened things out, according to a 2015 report from Germany’s Max Planck Institute for Evolutionary Anthropology. Perhaps generosity is an emotional or cognitive skill that must be learned. If so, it might explain this: Although three-year-olds in societies as different as urban California and hunter-gathering Aka shared treats that scientists gave them only one-third of the time (and became less likely to be generous through age seven), as they got older their generosity matched the norms of their culture—more evidence that generosity is a skill that one learns…or doesn’t.

It is certainly possible to absorb lessons for or against generosity, said Notre Dame sociologist Christian Smith. But generosity wanes if people perceive the world as full of threats and looming scarcity rather than of abundance and security—one of the individual traits most predictive of individual generosity.

Smith’s research, along with similar research elsewhere, throws cold water on two commonly cited findings about generosity. One is that generous people are happier and healthier because they donate. Smith’s research suggests another explanation: Because generous people view the world as safe, secure, and abundant, their happiness might derive not from being generous but from a generally sunny outlook. A second myth: When people act generously, their brain’s reward circuitry becomes more active, therefore the very act of giving must make us feel good. Of course, it can. But a 2015 study by scientists at Caltech found a different explanation for reward circuits humming like crazy: They are furiously calculating whether to give and how good or bad they’ll feel about that decision.

Perhaps the strongest message from the science of generosity is that the more adversity someone has experienced, the more compassion she feels and the more generous she’s likely to be. I’m reminded of this every time I see someone who looks destitute drop a few coins into a panhandler’s cup while expensively dressed commuters rush past. That fits with Ashar’s conclusion that belief in the sufferer’s blamelessness and expecting a donation to make a difference predict generosity. Someone who knows what it is to suffer also knows how outside forces can land one in deep poverty through little fault of one’s own, and how wonderful it can be to have a dollar for a McDonald’s coffee.

This article also appeared in the August 2016 issue of Mindful magazine.