2 Newman,Cain proposal,previous research has found that people are Method tha 2009.H 2004 In additio ema nal reminders of money can lead people to engage in a studies we included only the participants who took at number of antisocial behaviors,such as reduced least 20 s to read the vignette(s)and respond to the items ne other peopl .Good distin sum. Ito be the minimum amount o are many reasons to think that people may criticize chari table efforts that also provide personal gains.However,it failed to meet this requirement Participants were randomly assigned to one of three is unclear why such may be perceived as worse ing nothat this conditions.In two (between-subjects)conditions,the elated to the wh ity of different counterfactuals:When people co nsider a r per week voluntee ring at her place of work.One charitable effort that also realizes a personal benefit, ast betw (charitable-benefit condition),whereas anothe as conside d at nization)did not alth man's intentio self in ed he behave as altruistically as he or she could have. As a ervers are likely to one type less shelter or the coffee shop).Participants in the third t In thi ）condit The firs arget (o more altruistic;as a result,the y do not view the agent We predicted,therefo as ssed the target's morality.Participar nts rated hov re,that action hat pro and pe ethical he was (1 he was (1 e efit.We refer to this as the tainted-altruism effect of his tested this predicti on acros range of cont an ini highly correlated (a 92)and were averaged to create the the effec secon me of a hiring decision.In Expe riment 3 we te ted the sp ved orality. mhe targ cific proposal regarding counterfactual reasoning.In beneficial"they were (uot atall 9=tery hen Experiment 4, we examined whether this effect results extent to which they "made the world from eplic chan to an eficiab),and the er plac not at al ity These onducted ove ndms were als highly c period from April 2012 to March 2013,and 10.3%of the sure of benefit. participants were involved in at least two experiments Experiment 1:Between-versus within. Results Subjects Comparison We first compared ratings across the two between-sub In Experiment 1,we tested the tainted-altruism effect as jects conditions.Participants judged the target to be sig volunte ed at t e home he h0=562E-026100-227. that the self-interested behavior resulted in a charitable 025.In addition,they rated the agent's actions as equally benefit,whereas another group read that the behavior beneficial when he volunteered at the homeless shelte resulted in a neutral benefit.A third group of participants =6.33.SE=0.24)and when he volunteered at the evaluated both scenarios hop(M=5.83,SE=0.28),p=.18 muy21.214 2 Newman, Cain proposal, previous research has found that people are reluctant to accept monetary compensation for tasks that are perceived to be social favors (Ariely, Bracha, & Meier, 2009; Heyman & Ariely, 2004). In addition, even sublimi￾nal reminders of money can lead people to engage in a number of antisocial behaviors, such as reduced helpful￾ness toward other people (Vohs, Mead, & Goode, 2006). In sum, there seems to be an important psychological distinction between charity and profitability; hence, there are many reasons to think that people may criticize chari￾table efforts that also provide personal gains. However, it is unclear why such efforts may be perceived as worse than doing no good at all. We propose that this effect is related to the accessibil￾ity of different counterfactuals: When people consider a charitable effort that also realizes a personal benefit, there is an inherent contrast between charitable behavior and selfish behavior. People consider the same behavior as it might occur in the absence of self-interest and ulti￾mately conclude that the person (or organization) did not behave as altruistically as he or she could have. As a result, observers are likely to rate the target negatively. However, when someone is only selfish, only one type of behavior is present. In this case, people do not spontane￾ously consider whether the person could have been more altruistic; as a result, they do not view the agent negatively. We predicted, therefore, that actions that pro￾duce both charitable and personal benefits will be evalu￾ated as worse than equivalent self-interested behaviors that produce no charitable benefit. We refer to this as the tainted-altruism effect. We tested this prediction across a range of contexts. Experiment 1 was an initial demonstration of the effect. In Experiment 2, we replicated the effect in the context of a hiring decision. In Experiment 3, we tested the spe￾cific proposal regarding counterfactual reasoning. In Experiment 4, we examined whether this effect results from explicit use of charity as a means to an end or merely from the presence of profitability alongside char￾ity. These experiments were conducted over a 1-year period from April 2012 to March 2013, and 10.3% of the participants were involved in at least two experiments. Experiment 1: Between- Versus Within￾Subjects Comparison In Experiment 1, we tested the tainted-altruism effect as well as the extent to which people explicitly endorsed it. Participants read about a target whose behavior was motivated by self-interest. One group of participants read that the self-interested behavior resulted in a charitable benefit, whereas another group read that the behavior resulted in a neutral benefit. A third group of participants evaluated both scenarios. Method Participants were 162 adults (mean age = 35.7 years; 53% female, 47% male) who were recruited from an online subject pool maintained by a private university. In all studies, we included only the participants who took at least 20 s to read the vignette(s) and respond to the items. This was predetermined to be the minimum amount of time necessary to read through the materials. In this study, an additional 11 adults completed the survey but failed to meet this requirement. Participants were randomly assigned to one of three conditions. In two (between-subjects) conditions, they read a brief vignette. The vignette described a man who, in order to gain a woman’s affection, spent several hours per week volunteering at her place of work. One group of participants read that the woman worked at a homeless shelter (charitable-benefit condition), whereas another group of participants read that she worked at a coffee shop (neutral-benefit condition). Both vignettes stated that although the man’s intentions were self-interested, he nev￾ertheless did a good job at helping out (at either the home￾less shelter or the coffee shop). Participants in the third (within-subjects) condition read both vignettes. In all conditions, participants then rated the target (or targets) along two separate measures. The first measure assessed the target’s morality. Participants rated how ethical he was (1 = completely unethical, 9 = completely ethical), how moral he was (1 = completely immoral, 9 = completely moral), and the extent to which they approved or disapproved of his actions (1 = definitely not, 9 = definitely so). Responses to these items were highly correlated (α = .92) and were averaged to create a single measure of morality. The second measure assessed the perceived benefit of the target’s actions. Specifically, participants rated the actions in terms of “how beneficial” they were (1 = not at all, 9 = very ben￾eficial), and the extent to which they “made the world a better place” (1 = not at all, 9 = very much so). Responses to these two items were also highly correlated (r = .66, p < .001) and were averaged to produce a single mea￾sure of benefit. Results We first compared ratings across the two between-sub￾jects conditions. Participants judged the target to be sig￾nificantly less moral when he volunteered at the homeless shelter (M = 4.75, SE = 0.28) than when he volunteered at the coffee shop (M = 5.62, SE = 0.26), t(100) = 2.27, p = .025. In addition, they rated the agent’s actions as equally beneficial when he volunteered at the homeless shelter (M = 6.33, SE = 0.24) and when he volunteered at the coffee shop (M = 5.83, SE = 0.28), p = .18. Downloaded from pss.sagepub.com by Cai Xing on January 21, 2014