We all know that what other people do influences our actions. Companies spend billions of dollars each year to surround us with pictures of happy, attractive people drinking soda, driving cars, joining fitness clubs, and all sorts of other things that they want us to do. These ads create an artificial peer group that influences us as much or more as our real friends do. What if we harnessed this energy for charitable giving?
Two researchers, Jen Shang and Rachel Croson, did just that in a series of ingenious field experiments in public radio fundraising. They varied the amount and type of “social information,” or information about other donors, that donors received and found that it had a strong effect on how much money they gave. Your nonprofit can easily do the same thing.
Shang and Croson worked with a public radio station during its pledge drive. When people called in to make a donation, the person who answered the phone sometimes told them, “We had another member; they contributed [dollar amount].” This dollar amount varied among $75, $180, and $300. The experimenters chose $75 because it was the median or typical contribution; $180 represented the 80th percentile and $300 represented the 90th percentile of gifts. Sometimes the person answering the phone said nothing about anyone else’s donation.
Giving information about high gifts led people to contribute more: an average of $120 when they heard about the $300 donation, as compared to $107 when they heard about the $75 donation or were told nothing about any other donor. Telling people about a $180 donor sometimes led people to give more, but not as much as telling them about the $300 donor. Shang and Croson concluded that giving information about other donors works, particularly when you give them information about a relatively high donation.
In a follow-up experiment, Croson and Shang set the dollar amounts higher to $600 (the 95th percentile) and $1000 (the 99th percentile). Both of these amounts were much higher than the $75 typical (median) donation. There was also a “control” condition where the callers received no information about other donations. This time, donors gave an average of $172 in the $600 condition, which was much higher than the average donation of $121 in the control condition. However, taking the dollar amount up to $1000 actually hurt donations, as the average donation in this condition was only $140 (Shang & Croson 2009).
Information about other peoples’ donations works in online giving as well. A team of researchers tracked donations to individual web pages in fundraising for runners in the London marathon (Smith 2014; Smith et al. 2013). Donors to this site could see what other people donated before them, and it influenced their giving. The average gift was 50 pounds, but when a large gift came in – one of 100 pounds or more – the next gifts went up about 10 pounds on average. When an unusually small gift came in – one of 25 pounds or less – the next gifts went down about 3 to 5 pounds.
What do we learn from this? From the public radio campaign experiments we learn that providing information about another person’s higher donation can be a good strategy but getting the amount right can be tricky. In the public radio experiment, setting the other donation at $180 was too low to have an effect and at $1000 was too high, but comparison donations of $300 or $600 worked well. In this case, $300 was at the 90th percentile of donations and $600 was at the 95th percentile. This is worth trying in your own non-profit’s case, but it may be that other nonprofits work differently from public radio stations and that other formats, such as fundraising letters, work differently than call-in pledge drives. On the positive side, there is no evidence that telling donors about another donor’s high pledge can be harmful. The best strategy is to experiment with several different higher donation levels to see which works best for your nonprofit.
From the London marathon experiment we learn that previous donations can influence subsequent ones either positively or negatively, with big donations influencing the next donations up, and small donations influencing the following ones down. If I were to design a webpage for this type of fundraiser, I wouldn’t report every single donation, for fear that low donations would have a negative influence. Instead, I’d report the average donation and also have a “hall of fame” or “star donor of the day” part of the page where I reported recent high donations.
Croson, R., & Shang, J. (2008). The impact of downward social information on contribution decisions. Experimental Economics, 11, 221-233.
Croson, R., & Shang, J. (2013). Limits of the effect of social information on the voluntary provision of public goods: Evidence from field experiments.
Shang, J., & Croson, R. (2009). A field experiment in charitable contribution: The impact of social information on the voluntary provision of public goods. The Economic Journal, 119(540), 1422-1439.