I. INTRODUCTION

Positive recognition hits long been a trusted way of raising money on college campuses, where buildings, benches, and even the insides of library books bear the names of donors. But in an effort to spur gifts among young soon-to-be alumni, students at two Ivy League institutions are trying a different approach: publicizing the names of seniors who don't contribute to their class. (Chronicle of Higher Education) (1)

The desire for social approval is one of the reasons why individuals act more generously when their generosity is viewable by others (Hollander 1990). Studies have shown that recognizing donors by revealing their identity increases contributions to public goods and donations to charities. (2) However, very little is known about what kind of recognition is more effective in the field. For example, a charity may decide to recognize all donors by printing their names in newsletters. Alternatively, the charity may choose to recognize only the highest donors by naming buildings after them or posting their names on the walls. In our opening example, two Ivy League universities chose yet another approach by publicizing the names of seniors who did not contribute to their class, thus attempting to shame the noncontributors into giving. Although one can provide a number of arguments for or against using each of these approaches in practice, which approach is better at increasing charitable donations is an open empirical question.

We conducted a framed field experiment to investigate the impact of different recognition schemes on donations to a charity. Over 200 individuals recruited from the community participated in small groups of five to eight individuals and had an opportunity to donate part of their $10 experiment earnings to the American Red Cross. We used a between-subjects design such that each individual was randomly assigned to participate in only one treatment, making one donation decision. In the no recognition treatment, only donation amounts were displayed to the group at the end of the experiment. In the full recognition treatment, we also publicized all individuals' names next to their donations. In the positive recognition treatment, we publicized only the names of the highest donors next to their donations. In the negative recognition treatment, we publicized the names of the lowest donors (including publicizing those who gave $0).

We found that all forms of recognition have a positive impact on increasing donations relative to the baseline treatment, with selective recognition (both positive and negative) having the strongest and significant effect. In addition, we found that the underlying mechanisms for why positive and negative recognition are effective are very different. Negative recognition encourages higher donations through discouraging donations lower than $5, while positive recognition encourages maximum donations of $10.

Our findings have practical implications. For example, we show that recognizing the highest or the lowest donors by revealing their identities is the most effective recognition method. We argue that the main reason for this is that selective recognition creates tournament-like incentives, encouraging donors to seek a "positive prize" of prestige and avoid a "negative prize" of shame. Therefore, when recognizing donors or when publishing annual reports, practitioners may benefit from utilizing tournament-like incentives by recognizing their top contributors, and in some instances (as in the case with our opening example) shaming noncontributors. However, practitioners should be careful when choosing negative incentives (such as shaming noncontributors), because they also face the first-order problem of attracting and retaining donors. Given the opportunity of free entry and exit, individuals may simply avoid contributing to communal and charity groups that identify the lowest contributors.

In what follows, Section II reviews the related literature. Section III describes the experimental design. Section IV summarizes the results, and Section V provides a discussion and conclusion.

II. LITERATURE REVIEW

Related work has proposed different motivations for giving to charities. Behavioral arguments for why individuals contribute positive amounts to charities and public goods include pure altruism (Bergstrom, Blume, and Varian 1986; Harbaugh, Mayr, and Burghart 2007), "warm glow" (Andreoni 1989, 1990; Gneezy, Keenan, and Gneezy 2014), and inequality aversion (Fehr and Schmidt 1999; Rabin 1993). (3)

Empirical evidence that individuals give more when they are recognized suggests a role for recognition as an additional reason for giving (Andreoni and Bernheim 2009; Andreoni and Petrie 2004; Ariely, Bracha, and Meier 2009; Karl an and McConnell 2014; Samek and Sheremeta 2014; Soetevent 2005). Andreoni and Petrie (2004), for example, show that publicly identifying participants and their choices in a laboratory public goods experiment significantly increases contributions to the public good. Similarly, Andreoni and Bernheim (2009) provide experimental evidence that people like to be perceived as fair, which motivates them to act more pro-socially in order to enhance their social image. Soetevent (2005) further demonstrates that such concerns for social image can significantly increase contributions to charities in the field.

Theoretically, Benabou and Tirole (2006) have incorporated social image into an individual's utility function. One prediction of their model is that when recognition is introduced, an effort to preserve a high social image adds to the motivation to give to charity. In a different model, Vesterlund (2003) has shown that recognizing donors could be beneficial to the charity because publicly visible donations may send information to other potential donors. Therefore, both models make compelling theoretical arguments for recognizing donors.

While there is agreement among researchers that recognizing donors by revealing their identities is important for increasing contributions to public goods and donations to charities, very little is known about which kind of recognition is more effective in the field. For example, one could recognize all donors (Andreoni and Petrie 2004), or selectively recognize only the highest or the lowest donors (Samek and Sheremeta 2014). On the one hand, full recognition incorporates both positive and negative recognition, since recognizing all donors implies also recognizing the highest and the lowest donors. On the other hand, recognizing only the highest or only the lowest donors singles them out and makes their social identities more salient. This singling out amplifies the desire to seek prestige by the highest donors (Harbaugh 1998; Hollander 1990) and the desire to avoid shame by the lowest donors (Frank 1988; Gilbert 1998; Ketelaar 2004). In a sense, singling out creates tournament-like incentives, encouraging donors to seek a "positive prize" of prestige and avoid a "negative prize" of shame. (4) Such incentives should strengthen the effects of prestige under positive recognition and shame under negative recognition, thus generating higher donations than full recognition. Ultimately, this is an empirical question that we address with our experiment.

III. EXPERIMENTAL DESIGN

In the Harrison and List (2004) taxonomy, our experiment is a "framed field experiment." The experiment involved 205 individuals who were recruited from families participating in an early childhood study in a suburb of Chicago. (5) Upon arriving to the experiment, participants in groups of five to eight individuals had the opportunity to donate part of their endowment to the American Red Cross. Individuals were assigned to a session date and time in advance without knowing which treatment would be carried out during that session; as such, we expect balance across treatment groups. Group size varied based on how many participants arrived to each session. Each group participated in a separate session and room.

At the beginning of the session, all participants identified themselves by writing their first and last name on a nametag that was displayed to the others at all times. Participants sat in a circle facing each other and were given a walled lapboard (that we call a "privacy screen") so that all decisions could be made in private. Participants were instructed not to communicate with one another. Participants were given the instructions, which included a description of the American Red Cross (see Appendix S1, Supporting Information). The researcher also handed out a short quiz on understanding and went over the responses with the participants.

Each participant earned $10 for coming to the experiment, which was announced in advance and distributed at the end of the experiment. Participants received an additional endowment of $10 (ten $1 bills) to use in the experiment. Participants placed the amount that they wanted to give to the Red Cross in one envelope (labeled "give") and the amount that they wanted to keep in the other envelope (labeled "keep"). This decision was made quietly and in private. The researcher collected the "give" envelopes and tallied the totals at the back of the room. The donation decision was only made once.

At the end of the decision, all individual donations (including $0 donations) were displayed on the board at the front of the room sorted from the largest to the smallest. We conducted four treatments using a between-subjects design in which we experimentally manipulated the recognition information available to participants at the end of the experiment (see Table 1). In the no recognition treatment, no additional information was available about the amounts that each person gave. In the full recognition treatment, we told participants prior to their decision that the first and the last name of each group member would be written next to each donation amount at the end. In the positive recognition treatment, participants learned that the first and the last name of the two highest donors would be displayed, but that the remaining participants would not be recognized. Finally, in the negative recognition treatment, participants learned that the first and the last name of the two lowest donors (including $0) would be displayed, and the other participants would not be recognized. In case of ties, the order of names was randomly determined.

At the end of the experiment, participants responded to a short questionnaire, which included basic demographic questions as well as questions about charitable giving habits, tendency to care about what others think, and one's perceived social status. The experiment lasted approximately 30 minutes. The researchers sent all the donated money to the Red Cross via web payment within a week following the experiment, and participants were instructed that they could receive a confirmation email if they wished.

IV. RESULTS

A. Average Donations

Figure 1 summarizes the average donations by treatment. In the positive recognition treatment, the average donation is $6.93, followed by $6.29 in the negative recognition treatment, then by $6.02 in the full recognition treatment, and finally by $5.26 in the no recognition treatment. Even in the baseline no recognition treatment, individuals on average give more than 50% of their available earnings to the Red Cross. Compared to the real donation experiment with university students of Eckel and Grossman (1996), individuals in our experiment gave more. A major difference in our experiment is that even in the baseline treatment, all donation amounts are immediately revealed, while in Eckel and Grossman's experiment, no donation amounts are revealed and the experiment was double blind. In addition. Carpenter. Connolly, and Myers (2008) find that community members (as in our experiment) tend to give more than students to charity.

When comparing the four treatments, we find that all forms of recognition increase donations relative to the baseline treatment, and recognizing only the highest or only the lowest donors has the strongest and significant effect. Compared to no recognition, individuals on average donate $1.67 more when positive recognition is introduced (Wilcoxon Mann-Whitney test, p value = .01) and donate $1.03 more when negative recognition is introduced (Wilcoxon Mann-Whitney test, p value = .05). Individuals also donate $0.76 more in full recognition relative to no recognition; however, this is not significant at the conventional level (Wilcoxon Mann-Whitney test, p value = .12). We do not find any significant differences between the positive recognition, negative recognition and full recognition treatments (all p values > .20).

When separating our data by small group sizes (i.e.. five or six individuals within a group) and large group sizes (i.e., seven or eight individuals within a group), we find exactly the same directional effects of the different treatments, with all forms of recognition increasing donations relative to the baseline treatment. It appears that the differences between treatments are more pronounced in smaller groups, but we do not have enough data to find statistically significant differences in treatment effect by group size.

B. Distribution of Donations

Figure 2 displays the cumulative distribution of donations by treatment. It is apparent that the cumulative distribution of donations in the positive recognition treatment first-order stochastically dominates the no recognition treatment (two-sample Kolmogorov-Smirnov test, p value = .06). However, there are no statistically significant differences between the full recognition and no recognition treatment (two-sample Kolmogorov-Smirnov test, p value =. 14), and between the negative recognition and no recognition treatment (two-sample Kolmogorov-Smirnov test, p value = .12). Similarly, there are no significant differences between the positive recognition, negative recognition, and full recognition treatments (all p values > .20).

To look more closely at the distribution of donations, Figure 3 displays the histogram of donations across four treatments. Several interesting patterns emerge. First, we see anchor points at a gift of $5 (50% of the endowment, and 25% of total earnings) and $10 (100% of the endowment, and 50% of total earnings). Second, these anchor points become stronger when recognition is introduced. Under no recognition, 40% of participants donate less than $5, 28% donate exactly $5, and 25% donate exactly $10. Full recognition significantly reduces donations less than $5 (from 40% to 20%, Fisher's exact test, p value = .02) and increases (although not significantly) donations of $5 (from 28% to 41%, Fisher's exact test, p value =. 16). Similarly, negative recognition significantly reduces donations less than $5 (from 40% to 19%, Fisher's exact test, p value = .02) and increases (although not significantly) donations of $5 (from 28% to 38%, Fisher's exact test, p value = .30). However, positive recognition does not significantly reduce donations less than $5 (from 40% to 24%, Fisher's exact test, p value = .14), but it significantly increases maximum donations of $10 (from 25% to 49%, Fisher's exact test, p value = .01).

In summary, we find that the underlying mechanisms of why different types of recognitions are effective are different. Negative recognition and full recognition encourage higher donations through discouraging donations lower than $5, while positive recognition encourages maximum donations of $10.

V. DISCUSSION AND CONCLUSION

We conducted a framed field experiment to investigate how recognition impacts charity donations. We varied recognition in four separate treatments and found that all forms of recognition have a positive impact on increasing donations relative to the baseline treatment, with positive recognition and negative recognition having the strongest and significant effect.

A potential mechanism for why positive and negative recognitions are effective at increasing charitable donations is that selective recognition (both positive and negative) creates tournament-like incentives. Recognizing the highest donors activates the desire to seek prestige (a "positive prize") from being a high donor, thus increasing the proportion of donors who contribute the full amount of their income. Recognizing the lowest donors activates the desire to avoid shame (a "negative prize") from being a low donor, thus decreasing the proportion of donors who do not contribute and increasing the proportion of donors who contribute half of their income. Compared to full recognition, recognizing only the highest or only the lowest donors singles these donors out and makes their social identities more salient, amplifying prestige or shame (since all attention is focused only on the two donors). Therefore, while all forms of recognition positively impact charitable donations, positive and negative recognition have the strongest effect. (6)

It has been a commonly held belief, grounded in theoretical research (Benabou and Tirole 2006; Vesterlund 2003) and laboratory experimentation (Andreoni and Petrie 2004; Rege and Telle 2004), that positive recognition matters. Our study provides further evidence that positive recognition is effective in increasing charitable donations in the field. (7) However, our results also point out the effectiveness of negative recognition. In our opening example, we quoted a story from the Chronicle of Higher Education in which two Ivy League universities attempted to increase donation amounts from their soon-to-be alumni by "publicizing the names of seniors who don't contribute to their class." This may appear like a controversial strategy (and indeed it generated many debates after the article was published); however, the bottom line is that negative recognition also works. Because shame appears to be a powerful motivator to encourage giving, one may ask the question: why don't all social groups, charity organizations, and online communities practice displaying the lowest contributors? While these institutions face the problem of increasing contributions, they also face the first-order problem of attracting and retaining contributors. Given the opportunity of free entry and exit, individuals may simply avoid contributing to communal and charity groups that identify the lowest contributors. Therefore, in the long run a negative recognition policy may backfire. (8) We see the long-term effects of recognition as a fruitful avenue for future research.

doi: 10.1111/ecin.12448

ANYA SAMEK and ROMAN M. SHEREMETA *

* We thank three anonymous referees and the Editor of this journal for their valuable suggestions. We also want to thank Tristin Ganter, Phuong Ta, and Amanda Chuan for valuable research assistance. We have benefited from helpful comments of Sera Linardi, John List, Lise Vesterlund, and participants at the Science of Philanthropy Initiative Conference. Finally, we wish to thank the John Templeton Foundation through the Science of Philanthropy Initiative for research funding. Any remaining errors are ours.

Samek: Assistant Professor, Center for Economic and Social Research & Department of Economics, University of Southern California, Los Angeles, CA 90089. Phone 773-609-0425, Fax 213-821-2716, E-mail anyasamek@gmail.com

Sheremeta: Assistant Professor, Weatherhead School of Management, Case Western Reserve University. Cleveland. OH 44106; Economic Science Institute, Chapman University, Orange, CA 92866. Phone 216-368-4271, Fax 216-368-5039, E-mail rshereme@gmail.com

REFERENCES

Andreoni, J. "Giving with Impure Altruism: Applications to Charity and Ricardian Equivalence." Journal of Political Economy, 97, 1989, 1447-58.

--. "Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving." Economic Journal, 100, 1990, 464-77.

Andreoni, J., and B. D. Bernheim. "Social Image and the 50-50 Norm: A Theoretical and Experimental Analysis of Audience Effects." Econometrica, 77, 2009, 1607-36.

Andreoni, J., and R. Petrie. "Public Goods Experiments without Confidentiality: A Glimpse into Fund-Raising." Journal of Public Economics, 88, 2004, 1605-23.

Ariely, D., A. Bracha, and S. Meier. "Doing Good or Doing Well? Image Motivation and Monetary Incentives in Behaving Prosocially." American Economic Review, 99, 2009, 544-55.

Benabou, R., and J. Tirole. "Incentives and Prosocial Behavior." American Economic Review, 99, 2006, 544-55.

Bergstrom. T., L. Blume, and H. Varian. "On the Private Provision of Public Goods." Journal of Public Economics, 29, 1986, 25-49.

Carpenter, J., C. Connolly, and C. K. Myers. "Altruistic Behavior in a Representative Dictator Experiment." Experimental Economics, 11, 2008, 282-98.

Cason, T. N., W. A. Masters, and R. M. Sheremeta. "Entry into Winner-Take-All and Proportional-Prize Contests: An Experimental Study." Journal of Public Economics, 94, 2010, 604-11.

Clingingsmith. D., and R. M. Sheremeta. "Status and the Demand for Visible Goods: Experimental Evidence on Conspicuous Consumption." Working Paper, 2016. Available at SSRN: https://ssrn.com/abstract=2699325

Croson, R., and J. Y. Shang. "The Impact of Downward Social Information on Contribution Decisions." Experimental Economics, 11, 2008, 221-33.

Dechenaux, E., D. Kovenock, and R. M. Sheremeta. "A Survey of Experimental Research on Contests, All-Pay Auctions and Tournaments." Experimental Economics, 18, 2015, 609-69.

Eckel, C. C., and P. J. Grossman. "Altruism in Anonymous Dictator Games." Games and Economic Behavior, 16, 1996, 181-91.

Edwards, J. T., and J. A. List. "Toward an Understanding of Why Suggestions Work in Charitable Fundraising: Theory and Evidence from a Natural Field Experiment." Journal of Public Economics, 114, 2014, 1-13.

Fehr, E., and K. M. Schmidt. "A Theory of Fairness, Competition, and Cooperation." Quarterly Journal of Economics, 114, 1999, 817-68.

Feltovich, N., R. Harbaugh, and T. To. "Too Cool for School? Signaling and Countersignalling." RAND Journal of Economics, 33, 2002, 630-49.

Frank, R. H. Passions within Reason: The Strategic Role of the Emotions. New York: W. W. Norton & Co., 1988. '

Gerber, A. S., D. P. Green, and C. W. Larimer. "Social Pressure and Voter Turnout: Evidence from a Large-Scale Field Experiment." American Political Science Review, 102, 2008, 33-48.

Gilbert, P. "What Is Shame? Some Core Issues and Controversies," in Shame: Interpersonal Behaviour, Psychopathology and Culture, edited by P. Gilbert and B. Andrews. New York: Oxford University Press, 1998, 3-38.

Gneezy, U., E. A. Keenan. and A. Gneezy. "Avoiding Overhead Aversion in Charity." Science, 346, 2014, 632-35.

Harbaugh. W. T. "The Prestige Motive for Making Charitable Transfers." American Economic Review, 88, 1998. 277-82.

Harbaugh, W. T., U. Mayr, and D. R. Burghart. "Neural Responses to Taxation and Voluntary Giving Reveal Motives for Charitable Donations." Science, 316, 2007, 1622-25.

Harrison, G. W., and J. A. List. "Field Experiments." Journal of Economic Literature. 42. 2004, 1009-55.

Hollander, H. "A Social Exchange Approach to Voluntary Cooperation." American Economic Review, 80, 1990, 1157-67.

Jones, D., and S. Linardi. "Wallflowers: Experimental Evidence of an Aversion to Standing Out." Management Science, 60. 2014, 1757-71.

Karlan, D., and M. A. McConnell. "Hey Look at Me: The Effect of Giving Circles on Giving." Journal of Economic Behavior and Organization. 106, 2014. 402-12.

Ketelaar, T. "Ancestral Emotions, Current Decisions: Using Evolutionary Game Theory to Explore the Role of Emotions in Decision-Making," in Darwinism, Public Policy and Private Decisions, edited by C. Crawford and C. Salmon. Mahwah, NJ: Erlbaum, 2004, 145-68.

Lacetera, N., and M. Macis. "Social Image Concerns and Prosocial Behavior: Field Evidence from a Nonlinear Incentive Scheme." Journal of Economic Behavior and Organization, 76, 2010, 225-37.

Ledyard, J. "Public Goods: A Survey of Experimental Research," in Handbook of Experimental Economics, edited by J. Kagel and A. E. Roth. Princeton, NJ: Princeton University Press, 1995.

Mago. S., A. Samek, and R. M. Sheremeta. "Facing Your Opponents: Social Identification and Information Feedback in Contests." Journal of Conflict Resolution, 60. 2016, 459-81.

Niederle. M., and L. Vesterlund. "Do Women Shy Away from Competition? Do Men Compete Too Much?" Quarterly Journal of Economics, 122, 2007, 1067-101.

Rabin, M. "Incorporating Fairness into Game Theory and Economics." American Economic Review, 83, 1993, 1281-302.

Rege, M., and K. Telle. "The Impact of Social Approval and Framing on Cooperation in Public Good Situations." Journal of Public Economics, 88, 2004, 1625-44.

Samek, A. S., and R. M. Sheremeta. "Recognizing Contributors: An Experiment on Public Goods." Experimental Economics, 17, 2014, 673-90.

Shang, J., and R. Croson. "A Field Experiment in Charitable Contribution: The Impact of Social Information on the Voluntary Provision of Public Goods." Economic Journal, 119, 2009, 1422-39.

Sheremeta, R. M. "Overbidding and Heterogeneous Behavior in Contest Experiments." Journal of Economic Surveys, 27. 2013, 491-514.

--. "Behavioral Dimensions of Contests," in Companion to Political Economy of Rent Seeking, edited by R. D. Congleton and A. L. Hillman. London: Edward Elgar. 2015.

--. "Impulsive Behavior in Competition: Testing Theories of Overbidding in Rent-Seeking Contests." Working Paper, 2016. Available at SSRN: https://ssrn.com/abstract=2676419

Soetevent, A. "Anonymity in Giving in a Natural Context--A Field Experiment in 30 Churches." Journal of Public Economics, 89, 2005. 2301-23.

Spence, M. "Job Market Signaling." Quarterly Journal of Economics, 87. 1973, 355-74.

Vesterlund. L. "The Informational Value of Sequential Fundraising." Journal of Public Economics, 87. 2003, 627-57.

SUPPORTING INFORMATION

Additional Supporting Information may be found in the online version of this article:

Appendix S1. Instructions

(1.) From the Chronicle of Higher Education: http://chronicle.com/article/Students-at-2-Ivy-League/125056/.

(2.) Studies showing the positive effect of recognition include Andreoni and Petrie (2004). Rege and Telle (2004), Soetevent (2005), Karlan and McConnell (2014), Jones and Linardi (2014), and Samek and Sheremeta (2014). Social recognition has also been found effective in voter turnout (Gerber. Green, and Larimer 2008). blood donation (Lacetera and Macis 2010), consumption of private goods (Clingingsmith and Sheremeta 2016). and tournaments (Mago, Samek. and Sheremeta 2016).

(3.) It is also possible that individuals incur a utility loss by contributing below the socially acceptable amount, whereby the acceptable amount can be signaled by the fundraiser through recommended amounts (Edwards and List 2014) or through giving information about a past amount given by a previous donor (Croson and Shang 2008; Shang and Croson 2009).

(4.) It is well documented that tournament-like incentives are powerful at motivating individuals (Dechenaux, Kovenock, and Sheremeta 2015). Moreover, such incentives are even more powerful than predicted by the standard game theory (Sheremeta 2013, 2015. 2016).

(5.) Recruitment for the study was done through the "Living Lab" at the Chicago Heights Early Childhood Center in Chicago Heights, IL. Over 2,000 subjects make up this pool of households from Chicago Heights and the surrounding area, and are invited to participate in field experiments on a semiregular basis. Participants in this study had not participated in a similar study in the past, and we asked participants to indicate whether they knew anyone else in their group, separating them into separate groups if they said they knew anyone. In a survey following the experiment, we also asked participants to state whether anyone in their group was a friend, and about 7% of participants said they were friends with someone in their group. Our results do not change when we control for this in the analysis.

(6.) One could formally model a signaling game to explain some nonmonotonicities observed in our data (Feltovich, Harbaugh. and To 2002; Spence 1973). Although such an exercise is outside the scope of our paper, potentially it could explain why participants in the negative recognition treatment move from giving some amount below $5 to giving $5 and why participants in the positive recognition treatment shift from $5 to $10. This is an interesting exercise for future research.

(7.) Interestingly. Samek and Sheremeta (2014) find that in a laboratory experiment negative recognition increases contributions to the public good while positive recognition does not. There are a number of important differences between our two studies which may explain different findings. First, unlike the public goods game, individual donations do not directly impact the payoffs of the group members who view the recognition information. In the public goods game, participants who are identified as top contributors may be viewed as "suckers." preventing them from contributing to the public goods, while it is not the case with charitable donations to the Red Cross. Moreover, individuals receive no monetary benefit from the amount of total gifts given to the Red Cross, abstracting away from concerns about fairness, inequality-aversion, and free-riding, which are important social motivators in public goods games (Ledyard 1995).

(8.) It is also possible that positive recognition may have a similar effect of reducing entry in the long run. First, by creating tournament-like incentives, positive recognition may crowd out intrinsic motivation to donate. Second, it is well-documented that many individuals do not like tournaments (Cason, Masters, and Sheremeta 2010: Niederle and Vesterlund 2007) and so positive recognition may reduce entry of such individuals.

Caption: FIGURE 1 Average Donations by Treatment

Caption: FIGURE 2 Cumulative Distribution of Donations by Treatment

Caption: FIGURE 3 Histogram of Donations by Treatment

TABLE 1
Summary of Treatments

Treatment       Display of     Linking       Number of    Number of
                 Donation     Identities    Individuals    Groups
                 Amounts     to Donations

No                 Yes           None           57            9
recognition

Full               Yes           All            51            8
recognition

Negative           Yes        Lowest two        52            8
recognition

Positive           Yes       Highest two        45            7
recognition