This practice has grown in popularity, with companies such as Bomba’s, Tom’s, and Warby Parker making charitable donations, whether cash or in kind, on behalf of consumers. Giving-by-proxy is even expanding outside of the retail context, with some companies beginning to use the promise of a philanthropic gift as an incentive for strong employee performance.
What’s not known, however, is how this practice influences consumer behavior down the line. Does an experience with giving-by-proxy make people more or less likely to be charitable themselves? Any change could have a big impact on the net social impact of these initiatives.
When she began investigating this question, Maryam Kouchaki, a professor of management and organizations at the Kellogg School, could have seen it going either way. Her past research has found that when people see themselves as moral, they give themselves permission to behave in self-indulgent and self-interested ways, a phenomenon called “moral licensing.” At the same time, she thought giving-by-proxy could also serve to prime the pump for future beneficence.
Happily, it’s the latter. Giving-by-proxy experiences make people more willing to be charitable in the future. Kouchaki, along with Kellogg doctoral student Samantha Kassirer and Jillian J. Jordan of Harvard Business School, found that people are more prosocial following a giving-by-proxy experience—even when they have no control over the beneficiary of the initial donation.
“Giving-by-proxy has an indirect impact, meaning that people who are part of these efforts are more likely to donate later or do more good work,” Kouchaki says. “There’s a spillover effect.”
“Buy one, donate one” and future do-gooding
The researchers designed several experiments to investigate the relationship between giving-by-proxy and subsequent charitable behavior. In the first of these, 756 participants completed an online experiment for which they received $1.50, plus (in some cases) a $0.50 bonus in the form of a charitable donation made on the participant’s behalf.
The participants were randomly assigned to one of three groups: a group in which they selected the charity that would receive their bonus, a group in which the charity was chosen for them, and a control group.
Next, they completed ten simple word-search games. For participants in the two giving-by-proxy groups, each completed word search earned a five-cent charitable bonus; in the control group, no bonuses were awarded.
Then, the researchers asked participants if they would like to be notified about future studies. Those who said yes were asked whether they preferred studies where they could earn bonuses for charity or ones where they could earn bonuses for themselves. This decision, between a charitable option and a more selfish one, was the main variable the researchers focused on in their analysis.
Both giving-by-proxy groups were much more likely to make the charitable choice than the control group, the results showed: 27 percent of controlled and 29.8 percent of autonomous giving-by-proxy participants expressed interest in research with bonuses for charity, as compared with 17.4 percent of control-group participants.
The negligible difference between the two giving-by-proxy groups intrigued Kouchaki. “Intuitively, you might think there would be an extra bump in charitable behavior if you have choice,” she says. “There’s a lot of research on the importance of autonomy that would suggest it should matter.” Across their studies, the choice did not matter.
Giving-by-proxy and charitable credit
Next the researchers set out to understanding why giving-by-proxy prompts prosocial behavior. They suspected that giving-by-proxy—whether they chose the recipient or not—might prompt people to take charitable “credit” for that donation, motivating future prosocial behavior.
For this experiment, 685 new participants were once again divided into three groups: a giving-by-proxy group where they could choose the charity, one where they could not, and a control group. Then, they completed a word search where they earned $1, plus either a $0.25 charitable bonus (in the two giving-by-proxy groups) or no bonus (in the control group). Participants once again made a choice between participating in future studies with bonuses for either themselves or for charity.
Finally, participants answered a series of questions designed to assess why giving-by-proxy might prompt greater willingness to give. In particular, the researchers asked participants to rate how much moral or charitable “credit” others would assign their behavior during the study.
As in the first experiment, participants in both giving-by-proxy groups made the charitable choice at roughly equal rates—and at significantly higher rates than the control group. While there was no difference between the control and giving-by-proxy groups in how much moral credit they took, there was a difference in charitable credit-taking.
Notably, the two giving-by-proxy groups felt equally able to take credit for their charitable actions. To Kouchaki, this helps explain why choosing (or not) the recipient of the bonus ultimately didn’t affect participants’ willingness to make the charitable choice. “It doesn’t matter whether you benefitted charity A that you prefer or charity B that some else picked—others (including yourself) are going to see you as charitable,” she says.
Giving-by-proxy in the wild
For their final experiment, the researchers tested the downstream effects of giving-by-proxy in a real-world setting, using Amazon’s recently shuttered AmazonSmile program. Under this program, customers shopped at an identical-looking website, AmazonSmile.com, where purchases benefited a charity of their choice.
The researchers recruited 603 new participants, gave them a $3.50 Amazon gift card, and asked them to choose from one of six items to purchase. Half of the participants (the giving-by-proxy group) were told they would be making the purchase through AmazonSmile and given a brief description of the program, while the rest were simply directed to Amazon.
Then, participants were told they would be entered into a raffle to win $50. If they won the raffle, they could either keep the winnings or donate half to a charity of their choice. They also answered the same questions about charitable credit used in the previous study.
Once again, participants were significantly more likely to donate the potential raffle winnings to charity in the AmazonSmile group (39.5 percent) than in the control group (29.2 percent). Charitable credit-taking was also higher in the AmazonSmile group, which, statistical analysis revealed, drove the higher rates of charitable behavior.
Giving at scale
To Kouchaki, the research offers an optimistic lesson. Giving-by-proxy programs not only have a short-run positive effect—they also have the potential for broader impact. “Giving encourages future giving, and creating that type of culture is important,” she says. “You have to be mindful of how you do it, obviously, but you can use these programs to create momentum.”
She hopes more companies will experiment with these programs; she’s particularly intrigued by the growing practice of prosocial performance bonuses. “There’s already research showing these programs are effective at increasing employees’ motivation—it motivates them and creates more loyalty, as well as these indirect spillover effects. These are really positive programs.”