"Long before it's in the papers"
June 18, 2013


Studies may have overestimated our generosity

June 17, 2013
Special to World Science  

Many past stud­ies may have over­es­ti­mated hu­man gen­eros­ity, if a new piece of re­search is any clue.

The study recre­ated a game of­ten used in lab­o­r­a­to­ry ex­pe­ri­ments to as­sess peo­ple’s will­ing­ness to give away mon­ey, or their al­tru­ism.

A downtown Las Vegas bus stop.

Par­ti­ci­pants are typ­ic­ally granted an­o­nym­ity. But the new study was de­signed to af­ford an as­sur­ance of an­o­nym­ity even more be­liev­a­ble than usu­al. It was set up so that par­ti­ci­pants would be un­aware any ex­pe­ri­ment was even hap­pen­ing—or that any de­ci­sion would even be counted, let alone watched.

Un­der this seem­ingly great­er lev­el of se­cre­cy, the lev­el of giv­ing plunged to ze­ro.

The find­ings sug­gest that the lev­els of al­tru­ism recorded in pre­vi­ous ex­pe­ri­ments may be “sub­stanti­ally in­flat­ed,” wrote the re­search­ers, Jef­frey Wink­ing and Nich­o­las Mizer of Tex­as A&M Uni­vers­ity.

The results are pub­lished in the Ju­ly is­sue of the jour­nal Evo­lution and Hu­man Be­havior.

The findings, they added, high­light the idea that “an­o­nym­ity” and “se­cre­cy” can be dif­fer­ent things, be­cause any vis­i­ble ex­pe­ri­men­tal situa­t­ion can threat­en a par­ti­ci­pant with “very sub­tle cues” that the free­dom from pry­ing eyes is not ab­so­lute.

The al­tru­ism game, known as the Dic­ta­tor Game, along with many vari­ants, is rou­tine fare in psy­chol­o­gy and eco­nom­ics ex­pe­ri­ments. Usu­ally par­ti­ci­pants are giv­en some mon­ey, along with in­struc­tions that they may share some of it with an un­known, ran­domly as­signed part­ner if they wish. An­o­nym­ity is of­ten, though not al­ways, prom­ised.

In the new ver­sion, peo­ple wait­ing for bus­es near Las Ve­gas casi­nos were ap­proached by an ap­par­ently ran­dom strang­er. This per­son would of­fer some free ca­si­no chips, con­vert­i­ble to mon­ey, say­ing he did­n’t have time to cash them in. In some cases, this strang­er would al­so sug­gest that the re­cip­i­ent could share the chips with a sec­ond strang­er, who was stand­ing some dis­tance away with his back turned, chat­ting in­to a cell phone. Both strang­ers were really ac­tors. 

The chip-giver would then leave, and the sec­ond strang­er would put down his phone and come to the bus stop. The chip re­cip­i­ent’s next move was then se­cretly not­ed. This was one situ­ation where what happened in Vegas would not stay in Vegas (though anon­ym­ity was still guard­ed). 

In the stand­ard, anonymous Dic­ta­tor Game, the over­all av­er­age dona­t­ion based on past stud­ies is 28 per­cent of the to­tal, with nearly two-thirds of peo­ple of­fering at least some­thing, Wink­ing and Mizer wrote.

But in their “real-life” re-enact­ment of the game, no one gave a thing.

One pos­si­ble rea­son might have been re­luc­tance to in­i­ti­ate a con­versa­t­ion with a strang­er, the re­search­ers spec­u­lat­ed. How­ev­er, two par­ti­ci­pants did start a con­versa­t­ion with the cell phone guy—just to re­late their stroke of good luck.

There were 60 par­ti­ci­pants in this “real-life” Dic­ta­tor Game. Thir­ty ad­di­tion­al par­ti­ci­pants were re­cruited in­to a ver­sion that com­bined as­pects of the “real-life” form and the lab­o­r­a­to­ry form. These peo­ple were ap­proached much as be­fore, but were told that it was an (an­o­nymous) ex­pe­ri­ment. This was done to as­sess wheth­er the “real-life” form and the lab­o­r­a­to­ry ver­sion were in­deed com­pa­ra­ble, apart from their in­tend­ed dif­fer­ences. 

The re­sults from this hy­brid Dic­ta­tor Game were si­m­i­lar to the lab­o­r­a­to­ry ver­sion, the re­search­ers said, which has been re­peat­ed in at least 129 pub­lished stud­ies since 1986.

The new find­ings are im­por­tant be­cause coop­era­t­ion and al­tru­ism is an ac­tive ar­ea of stu­dy, with one great rid­dle be­ing their ev­o­lu­tion­ary ori­gins, the re­search­ers not­ed. The new find­ings, they added, may “de­mand a reevalua­t­ion” of the “true na­ture” of these hu­man qual­i­ties as ex­hib­ited in lab­o­r­a­to­ry ex­pe­ri­ments.

* * *

Send us a comment on this story, or send it to a friend

Sign up for

On Home Page         


  • Stu­dies may have over­estimated our gener­osity

  • Re­search pai­nts new pic­ture of “dino­bird” fea­thers


  • Was black­mail essen­tial for marr­iage to evolve?

  • Plu­to has even cold­er “twin” of sim­ilar size, studies find

  • Could simple an­ger have taught people to coop­erate?

  • Diff­erent cul­tures’ mu­sic matches their spe­ech styles, study finds


  • F­rog said to de­scribe its home through song

  • Even r­ats will lend a help­ing paw: study

  • D­rug may undo aging-assoc­iated brain changes in ani­mals

Many past studies may have overestimated human generosity, if a new piece of research is any clue. The study recreated a game often used in laboratory experiments to assess people’s willingness to give away money, or their altruism. Participants are typically granted anonymity. But the new study was designed to afford an assurance of anonymity even more believable than usual. It was set up so that participants would be unaware any experiment was even happening—or that any decision would even be counted, let alone watched. Under this seemingly greater level of secrecy, the level of giving plunged to zero. The findings suggest that the levels of altruism recorded in previous experiments may be “substantially inflated,” wrote the researchers, Jeffrey Winking and Nicholas Mizer of Texas A&M University. The results, they added, highlight the idea that “anonymity” and “secrecy” can be different things, because any visible experimental situation can threaten a participant with “very subtle cues” that the freedom from prying eyes is not absolute. The altruism game, known as the Dictator Game, along with many variants, is routine fare in psychology and economics experiments. Usually participants are given some money, along with instructions that they may share some of it with an unknown, randomly assigned partner if they so choose. Anonymity is often, though not always, promised. In the new version, people waiting for buses near Las Vegas casinos were approached by an apparently random stranger. This person would offer some free casino chips, convertible to money, saying he didn’t have time to cash them in. In some cases, this stranger would also suggest that the recipient could share the chips with a second stranger, who was standing some distance away with his back turned, chatting into a cell phone. Both strangers were really actors. The chip-giver would then leave, and the second stranger would put down his phone and come to the bus stop. The chip recipient’s next move was then secretly noted. In the standard Dictator Game, the overall average donation based on past studies is 28 percent of the total, with nearly two-thirds of people offering at least something, Winking and Mizer wrote. But in their “real-life” re-enactment of the game, no one gave a thing. One possible reason might have been reluctance to initiate a conversation with a stranger, the researchers speculated. However, two participants did start a conversation with the cell phone guy—just to relate their stroke of good luck. There were 60 participants in this “real-life” Dictator Game. Thirty additional participants were recruited into a version that combined aspects of the “real-life” form and the laboratory form. These people were approached much as before, but were told that it was an (anonymous) experiment. This was done to assess whether the “real-life” form and the laboratory version were indeed comparable, apart from their intended differences. The results from this hybrid Dictator Game were similar to the laboratory version, the researchers said, which has been repeated in at least 129 published studies since 1986. The new findings are important because cooperation and altruism is an active area of study, with one great riddle being their evolutionary origins, the researchers noted. The new findings, they added, may “demand a reevaluation” of the “true nature” of these human qualities as exhibited in laboratory experiments.