Tainted Altruism form a single scale (a =87).On a separate page,partici- less moral than individuals who are selfish and do no pants also good at all:in addition,providing relevant counterfactual information appea s to revers this effect Results that the charitable en efit was previous studie as the by-product of some other selfish intention.Therefore,it A 2x 2 analysis of variance revealed a significant interac- could be that this effect arises only in cases in which 1335 an en s the agent b naves cha 001.A To address this possibility we asked participants in information was provided,participants evaluated the tar Experiment 4 to evaluate a real-world prosocial cause get more negatively whe he donated money to the Gap (RED)c ed 50%c 634eeo2esed4e40 the fight the =052.In contrast.when the counterfactual information malaria).In this experiment.the key manipulation was was provided.the pattern of results reversed,such that whether the profit to the company(the 50%of profits no participants rated the target more po ely when he donated to charity)was highlighted e 0. inclu ions:In the 0.23),K(141)=3.22.p=.002.Moreover,within-domain the Gap.but no information about any charitable pro comparisons revealed that the presence of counterfactual grams.In the altruism condition,participants read about information increased ratings of the targ et in the charity 14 was sta sm con profit to the compa Hayes,2008)to determine whether agreement with the other 50%)were stated.Finally.in the counterfactual counterfactual item explained the effect on ratings of the information condition,after reading that the Gap (RED) The wo-w mnteraction was use cha ption that the tar acted as altruistically money to charity as he could have"was the mediator,and the two main Our hypothesis was based on the previous results effects were included as covariates.This analysis indi cated with the counte the tion was highlighted.However,following the logic of Discussion Experiment 3,we predicted that when the relevant coun terfactua informatio participants would that h effect.In the case of charity,presenting the counterfac Metbod tual information (i.e.. that the target could have no ted money)significantly increased ratings of mora Participants wer nterfactual infom that u sed in th have donated money to charity)had the opp osite effect. were randomly ass subjects Moreover,agreement with the counterfactual item fully conditions.An additional 8 adults participated,but sub mediated morality ratings of the target onses in less than 20 s (the me on pre Experiment 4:Ambiguous Intentions In the control condition,participants read basic infor Experiments 1 through 3 established that people rate mation about the Gap company.In the altruism condi- individuals who do good for self-interested reasons as tion,participants read the same information as in theTainted Altruism 5 form a single scale (α = .87). On a separate page, partici￾pants also rated whether the target “acted as altruistically as he could have” (1 = absolutely not, 9 = absolutely). Results A 2 × 2 analysis of variance revealed a significant interac￾tion between domain (charity, advertising) and the pres￾ence/absence of counterfactual information, F(1, 141) = 13.35, p < .001. A series of planned contrast analyses indicated that, as predicted, when no counterfactual information was provided, participants evaluated the tar￾get more negatively when he donated money to charity (M = 5.70, SE = 0.26) than when he invested the same money in advertising (M = 6.34, SE = 0.21), t(141) = 1.96, p = .052. In contrast, when the counterfactual information was provided, the pattern of results reversed, such that participants rated the target more positively when he gave to charity (M = 6.53, SE = 0.21) than when he invested the same money in advertising (M = 5.49, SE = 0.23), t(141) = 3.22, p = .002. Moreover, within-domain comparisons revealed that the presence of counterfactual information increased ratings of the target in the charity domain, t(141) = 2.50, p = .013, but significantly reduced them in the advertising domain, t(141) = 2.67, p = .009. We then conducted a bootstrap analysis (Preacher & Hayes, 2008) to determine whether agreement with the counterfactual item explained the effect on ratings of the target. The two-way interaction was used as a predictor variable, ratings of the target were used as the dependent variable, perception that the target “acted as altruistically as he could have” was the mediator, and the two main effects were included as covariates. This analysis indi￾cated that agreement with the counterfactual item signifi￾cantly mediated the two-way interaction, b = 0.55, SE = 0.22, 95% confidence interval = [0.21, 1.09]. Discussion Results from this experiment provide support for the proposition that counterfactual thinking explains the effect. In the case of charity, presenting the counterfac￾tual information (i.e., that the target could have not donated money) significantly increased ratings of moral￾ity. In the comparison case, however, presenting the counterfactual information (i.e., that the target could have donated money to charity) had the opposite effect. Moreover, agreement with the counterfactual item fully mediated morality ratings of the target. Experiment 4: Ambiguous Intentions Experiments 1 through 3 established that people rate individuals who do good for self-interested reasons as less moral than individuals who are selfish and do no good at all; in addition, providing relevant counterfactual information appears to reverse this effect. However, one limitation of the previous studies was that the charitable benefit was always presented as the by-product of some other selfish intention. Therefore, it could be that this effect arises only in cases in which people or organizations explicitly use charity as a means to an end—that is, the agent behaves charitably only to achieve some other goal. To address this possibility, we asked participants in Experiment 4 to evaluate a real-world prosocial cause, the Gap (RED) campaign. In 2006, Gap pledged 50% of the profits earned from the sale of certain (RED) products to fight the spread of infectious disease (HIV/AIDS and malaria). In this experiment, the key manipulation was whether the profit to the company (the 50% of profits not donated to charity) was highlighted. This experiment included four conditions: In the con￾trol condition, participants were given information about the Gap, but no information about any charitable pro￾grams. In the altruism condition, participants read about the Gap and the (RED) campaign, but only the donation was stated (50% to charity). In the tainted-altruism condi￾tion, participants read about the Gap (RED) campaign, and both the donation and the profit to the company (the other 50%) were stated. Finally, in the counterfactual￾information condition, after reading that the Gap (RED) raised money for charity and earned a profit, participants were reminded that the Gap did not have to donate any money to charity. Our hypothesis was based on the previous results: Participants would view the charitable effort as tainted (by the potential for self-interest) and would therefore judge the company more negatively when information about profits was highlighted than when only the dona￾tion was highlighted. However, following the logic of Experiment 3, we predicted that when the relevant coun￾terfactual information was provided, participants would rate the company much more favorably. Method Participants were 206 adults (mean age = 39.4 years; 62% female, 38% male) who were recruited from the same online pool as that used in the previous experiments and were randomly assigned to one of four between-subjects conditions. An additional 8 adults participated, but sub￾mitted responses in less than 20 s (the same criteria as used in all of the previous studies). The information pre￾sented to participants was taken from Wikipedia. In the control condition, participants read basic infor￾mation about the Gap company. In the altruism condi￾tion, participants read the same information as in the Downloaded from pss.sagepub.com by Cai Xing on January 21, 2014