Rubaltelli Asymmetric cost and benefit perceptions in willingness‐to‐donate decisions 2019 Accepted

1 Asymmetric cost and benefit perceptions in willingness-to-donate decisions Enrico Rubaltelli1,* , Dorina Hysenbelli1, Stephan Dickert2,3, Marcus Mayorga4, Paul Slovic4,5 University of Padova Department of Developmental and Socialization Psychology via Venezia, – 35131 Padova, Italy University of Klagenfurt Department of Psychology Universitätsstr 65-67 – 9020 Klagenfurt, Austria Queen Mary University of London School of Business and Management Mile End Campus – London E1 4NS, United Kingdom Decision Research 1201 Oak St Suite 200 – Eugene, OR, 97491, USA University of Oregon Department of Psychology 1227 University of Oregon – Eugene, OR 97403, USA * Corresponding author: Email: enrico.rubaltelli@unipd.it Phone: 0039 8276541 Fax: 0039 8276511 RUNNING HEAD: Cost and benefits in donation decisions Word count: 10638 Tables: Figures: 12 Acknowledgement This material is based upon work supported by the National Science Foundation under Grant Nos 1227729 and 1427414 Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and not necessarily reflect the views of the National Science Foundation Asymmetric cost and benefit perceptions in willingness-to-donate decisions Abstract Charitable giving entails the act of foregoing personal resources in order to improve the conditions of other people In the present paper, we systematically examine two dimensions integral to donation decisions that have thus far received relatively little attention but can explain charitable behavior rather well: the perceptions of cost for the donor and benefit for the recipients In line with current theories in judgment and decision making, we hypothesize that people weigh these dimensions subjectively and perceive them asymmetrically, consistent with prospect theory Costs for the donor are typically perceived as losses, whereas benefits for recipients are perceived as gains In four studies, we presented several scenarios to participants in which both donation amounts (costs) and number of lives helped (benefits) were manipulated while keeping the ratio of costs and benefits constant Results from Studies and showed that willingness to help decreased as donation amounts and number of lives helped increased Additionally, Studies and provide evidence for a solution to reduce the asymmetry and increase donation amounts as the number of lives at risk increase (WORD COUNT: 180) Keywords: Prosocial behavior; donation decisions; cost; benefit; prospect theory Although people may help others for many different reasons (e.g., Bekkers & Wiepking, 2011), they usually forego personal resources (i.e., time or money) to improve someone else’s wellbeing It seems reasonable to assume that the benefactor will engage in a subjective appraisal of the degree of benefit that can be achieved with a specific amount of resources The psychological underpinnings of such an appraisal likely involve both emotional aspects (e.g., sympathy for the victims) as well as more deliberative ones (e.g., impact of the donation), and the comparison between benefits and costs is considered instrumental in establishing whether the act of helping is worth it or not This may especially be the case when personal resources are limited, and the donor needs a compelling reason to give them away In other words, the decision whether to help or not is likely influenced by the perceived tradeoff between the resources a donor must give up and her perception of the beneficial effect of the helping intervention The goal of the present paper is to investigate whether the subjective weights attached to the personal cost and benefit of the helping action can lead donors to evaluate lives at risk in a non-normative way (i.e., placing different value on lives depending on the number of individuals at risk; Dickert, Västfjäll, Kleber, & Slovic, 2012; Slovic, 2007) In line with prospect theory (Kahneman & Tversky, 1979), we present evidence that donors put more emphasis on the cost of giving up an increasing amount of money compared to the benefit of helping increasing numbers of lives, leading to a decrease in their willingness to help Perceived cost for the donor and perceived benefit for the recipients We investigate whether a person’s decision to donate to a particular fund-raising campaign depends on their subjective judgments about the cost they incur, and the benefit provided to the recipients The tradeoff between these two constructs has thus far not received much attention in research on donation decisions but should be taken into consideration to understand how a donor assesses the utility of a donation In the current paper, we focus on the monetary cost people face when contributing to a charity organization1, and benefits are conceptualized as the positive difference a donation will make for the people in need People are less willing to donate when they perceive their donation as a “drop-in-the-bucket,” that is, when a contribution is not seen as making much of a difference toward solving the problem (Small, Loewenstein, & Slovic, 2007) Recent studies have suggested that people’s tendency to help the highest proportion of individuals in need, rather than the most in absolute numbers, can be interpreted as a way to support the fundraising programs whose perceived effectiveness is bigger (Bartels, 2006; Erlandsson, Bjorklund, & Backstrom, 2014, 2015; Fetherstonhaugh, Slovic, Johnson, & Friedrich, 1997) In addition, helping can induce people to experience positive emotions because they are responsible for having a positive impact on others’ lives (Andreoni, 1990; Dickert, Sagara, & Slovic, 2011) However, people may also experience negative affective reactions when spending money in general (Prelec & Loewenstein, 1998; Raghubir, 2006; Knutson, Rick, Wimmer, Prelec, & Loewenstein, 2007; Chatterjee & Rose, 2012), and also in decisions involving helping other people (Rubaltelli & Agnoli, 2012; Genevsky, Västfjäll, Slovic, & Knutson, 2013) Consistent with these findings, changing the frame used to present a donation appeal can influence people’s emotional reactions and, consequently, their willingness to help For instance, Breman (2011) demonstrated in a field study that people were more willing to donate when they were asked to pledge to donate but their credit card was charged at a later stage rather than immediately The explanation for this result was that people tend to discount a cost when it is postponed in time As a result, they are more willing to help when the cost of the donation is not charged immediately Sussman, Sharma and Alter (2015) found that framing a donation as exceptional (i.e., the only chance to help Charity X) increases willingness to help compared to framing it as an ordinary expense (i.e., this year’s chance to help Charity X) We hypothesize that costs and benefits are represented differently by donors, such that changes in costs may not be matched by changes in benefits when the number of lives at risk increases According to egalitarian moral perspectives and related forms of utilitarianism all lives should be valued equally (e.g., Baron & Szymanska, 2011; Sinnot-Armstrong, 2011) and additional lives at risk could be valued even higher if their loss threatens the survivability of an entire group (Dickert, Västfjäll, Kleber, & Slovic, 2015; Slovic, Fischhoff, & Lichtenstein, 1982).2 However, despite agreeing that every life should be valued equally, people often fail to anticipate the decrease in willingness to give when the cost for the donor increases Consistent with this notion, research on donation decisions and the underlying psychological processes demonstrated that people are more willing to give when just one (or only a few) lives are at stake or when donation amounts are not too high (Kogut & Ritov, 2005a, 2005b; Small, Loewenstein, & Slovic, 2007; Cameron & Payne, 2011; Rubaltelli & Agnoli, 2012) While this research has focused on psychological mechanisms underlying donation decisions, the subjective perception of cost and benefit has not been measured with regard to increasing lives at risk Based on the existing literature on charitable giving, the proportion of participants who decide to make a donation is expected to decrease as the amount of money required (i.e., cost) and the number of lives helped (i.e., benefit) increase at the same rate This prediction is motivated by the properties of the value function of prospect theory, which is steeper in the loss domain than in the gain domain (i.e., showing greater sensitivity to losses than gains; Kahneman & Tversky, 1979; Tversky & Kahneman, 1992) In the domain of charitable giving, the cost for the donor falls in the loss domain, since the donor is giving away part of her resources and experiencing negative affect because of this (Rubaltelli & Agnoli, 2012; Genevsky et al., 2013) Conversely, the number of lives helped falls in the gain domain, since the donation allows improving the life of needy individuals As a result, contrary to most literature on prospect theory, in our case the loss is experienced directly by the donors, who donate their own money, whereas the gain is experienced by other people who are at the receiving end of the helping action Nonetheless, we still expect to find results that are consistent with prospect theory prediction that perceived cost for the donor should follow a steeper increase than perceived benefit for the recipients as the number of lives in need increases As a consequence, the disutility of donating more money should be felt more than the utility of saving more lives Therefore, we hypothesize that: H1: The slope of the donor’s perception of costs will be steeper than the slope of the perception of benefits for recipients H2: Participants will become less willing to help as the number of lives at risk and the cost of helping increase at a constant ratio Finally, we assume that people’s willingness to help is a function of perceived benefit for the recipients and cost for the donor, and we expect that people’s asymmetric valuation of cost and benefit leads to a point where benefit no longer outweighs the cost At this point, people’s willingness to help should decline or abate altogether Within a range of values up to which someone is willing to donate money, one would expect willingness to help to remain constant when both cost and benefit increase at the same rate However, this should not be the case if losses are perceived asymmetrically compared to gains Based on the above reasoning, we expect that: H3: The asymmetric slopes of cost for the donor and benefit for the recipients will predict participants’ willingness to help Across four studies, we test these hypotheses and show that the differential perception of cost for the donor and benefit for recipients can predict the decrease in willingness to help as the number of lives at risk increases The first study provides initial evidence for this effect The second study extends the results measuring the cost/benefit tradeoff with a single bipolar question rather than two separate questions The third study demonstrates that presenting donation amounts as part of a bundle of different aid programs rather than a lump sum can be an effective way to increase willingness to help by changing people’s perceptions of the cost relative to the benefit Specifically, each program in the bundle is associated with a smaller donation amount compared to the lump sum condition Similarly, each program in the bundle condition helps fewer lives If people have a differential perception of cost and benefit, then breaking down these two dimensions should reduce the perception of cost with less impact on the perception of benefit Therefore, the goal of Study is to further understand how the cost/benefit tradeoff impacts donation decisions, while at the same time showing an effective way to influence people’s weighting of the cost dimension Finally, the fourth study examines whether this effect is asymmetrically influenced by splitting the lump sum into several smaller donation amounts or the overall number of lives helped into several smaller groups Study In Study 1, we presented participants with seven scenarios in which they were asked to donate money to Kenyan children who live in conditions of extreme poverty The amount of money and the number of children varied for each scenario, but the amount donated to a single child (i.e., the ratio between cost and benefit) was kept constant across scenarios Method Participants One hundred and fifty-one participants (43% female; mean age 32 years, range from 18 to 66) took part in the study They were recruited on Amazon Mechanical Turk, completed the questionnaire online, and received a fee of $0.40 All participants were located in the United States Amazon Mechanical Turk is commonly used to recruit adult participants for online studies and has been validated by Buhrmester, Kwang and Gosling (2011) and Paolacci, Chandler and Ipeirotis (2010) Since we have conducted donation studies on Amazon Mechanical Turk in the past, we used the TurkGate system to filter out participants who had already taken part in previous studies (Goldin & Darlow, 2013) Materials and procedure Participants completed an online questionnaire that was about 6minute long to complete In the questionnaire, they were told to imagine they had been contacted by a humanitarian aid organization that was raising funds to help extremely poor people living in the Kakuma refugee camp in Kenya Participants were informed that their donation would be used to buy food, clothes and medicine for children living in the refugee camp We presented seven different scenarios within-subjects These scenarios were presented in a different random order for each participant and for each scenario we repeated the same cover story outlined above Participants were instructed to consider each donation request separately and independently from other scenarios presented In each scenario, participants were asked whether or not they wanted to make a donation and we varied the amount of the donation and the number of children helped (see Table below) After each scenario, participants were asked to answer two questions, which measured how much the donation was perceived as a cost for the donor and as a benefit for the recipients Both were answered using 7-point scales ranging from (not at all) to (very) After completing all scenarios, participants provided demographic information and were presented with a debriefing page explaining method, hypotheses, and expected results of the study (full materials are available in the Supplementary Online Materials) Table -Results Donation decisions Most participants (88%) were willing to donate when presented with a scenario asking them to help two children with a $5 donation, whereas only a few (14%) were willing to donate when the scenario asked them to help ninety children with a $225 donation A substantial proportion of participants (60%) refused to make a donation when the cost reached $50 for helping 20 children (see Table for the proportion of donations in each scenario) We ran a multilevel logistic model using R software (version 3.4.3; R Development Core Team, 2012) and the lme4 package (Bates et al., 2015) to estimate participants’ random effects Scenario was the within-subject predictor and donation decisions in each of the seven scenarios served as the dependent variable Results revealed a significant effect of scenario, χ2 (6) = 243.59, p < 00001 As expected, the percentage of participants who made a donation decreased as the size of the donation (and the number of children helped) increased (see Figure 1) Table 10 Contrast effects showed that, for each scenario, participants’ willingness to help was lower than for the previous one (e.g., less people made a donation in the second scenario than in the first, and so on; always p < 00001) This shows that the effect we found is already present when cost is relatively low and not just when it is so high to make the donation amount almost unrealistic or too high for many people to consider it Figure -Cost - benefit perceptions Examination of the ratings indicated that the average perceived benefit for recipients was higher than the average perceived cost for the donor in the first five scenarios, and the scenario factor had a significant effect on both ratings such that they increased as donation amounts and number of lives at risk increased Consistent with donation decisions, the percentage of people who rated benefit higher than cost decreased as cost and benefit of helping increased (see Table 2) Finally, perceived cost and benefit were significantly correlated across the seven scenarios, but the correlation was higher for participants who did not donate (r = 55, p < 001) than for participants who did (r = 17, p < 001) The difference between the two correlations was statistically significant (z = 3.84, p = 0001) In line with these findings, we ran a multilevel linear model with scenario, dimension (cost vs benefit), and the interaction between scenario and dimension as predictors and participants’ ratings in each scenario as the dependent variable Results revealed a significant effect of scenario, χ2 (6) = 1173.06, p < 00001, and a significant effect of the dimensions, χ2 (1) = 238.57, p < 0001 Contrast effects showed that ratings were higher in all scenarios compared to the first one (t values = 6.44 or higher, ps < 00001) On average, ratings were higher for perceived benefit than perceived cost (t = 15.45, p = 04) In addition, we found a significant interaction effect, χ2 (6) = 272.13, p < 00001 (see Figure 2) 38 References Anderson, N H (1965) Averaging versus adding as stimulus-combination rule in impression formation Journal of Experimental Psychology, 70, 394-400 http://dx.doi.org/10.1037/h0022280 Anderson, N H (1981) Foundations of Information Integration Theory New York: Academic Press Andreoni, J (1990) Impure altruism and donations to public goods: A theory of warm-glow giving The Economic Journal, 100, 464-477 http://doi.org/10.2307/2234133 Baron, J., & Szymanska, E (2011) Heuristics and biases in charity In D M Oppenheimer & C Y Olivola (Eds.), The Science of Giving: Experimental Approaches to the Study of Charity (pp 215-235) New York, NY: Psychology Press Bartels, D M (2006) Proportion dominance: The generality and variability of favoring relative savings over absolute savings Organizational Behavior and Human Decision Processes, 100, 76-95 https://doi.org/10.1016/j.obhdp.2005.10.004 Bates, D., Maechler, M., Bolker, B., Walker, S., et al (2015) lme4 Retrieved from https:// github.com/lme4/lme4 Bekkers, R., & Wiepking, P (2011) A literature review of empirical studies of philanthropy: Eight mechanisms that drive charitable giving Nonprofit and Voluntary Sector Quarterly, 40, 924-973 https://doi.org/10.1177/0899764010380927 Buhrmester, M D., Kwang, T., & Gosling, S D (2011) Amazon’s Mechanical Turk: A new source of inexpensive, yet high-quality, data? Perspectives in Psychological Science, 6, 3-5 http://doi.org/10.1177/1745691610393980 Breman, A (2011) Give more tomorrow: Two field experiments on altruism and intertemporal 39 choice Journal of Public Economics, 95, 1349-1357 http://doi.org/10.1016/j.jpubeco.2011.05.004 Cameron, C D., & Payne, B K (2011) Escaping affect: How motivated emotion regulation creates insensitivity to mass suffering Journal of Personality and Social Psychology, 100, 115 http://doi.org/10.1037/a0021643 Caserotti, M., Rubaltelli, E., & Slovic, P (2019) How decision context changes the balance between cost and benefit increasing charitable donations Judgment and Decision Making, 14, 187-198 Chatterjee, P., & Rose, R L (2012) Do payment mechanisms change the way consumers perceive products? Journal of Consumer Research, 38, 1129-1139 http://doi.org/10.1086/661730 Dickert, S., Sagara, N., & Slovic, P (2011) Affective motivations to help others: A two-stage model of donation decisions Journal of Behavioral Decision Making, 24, 361-376 http://doi.org/10.1002/bdm.697 Dickert S., Kleber J., Västfjäll D., & Slovic P (2016), Mental Imagery, Impact, and Affect: A Mediation Model for Charitable Giving PLoS ONE 11 (2): e0148274 http://doi.org/10.1371/journal.pone.0148274 Dickert, S., Västfjäll, D., Kleber, J., & Slovic, P (2012) Valuations of human lives: Normative expectations and psychological mechanisms of (ir) rationality Synthese, 189, 95-105 http://10.1007/s11229-012-0137-4 Dickert, S., Västfjäll, D., Kleber, J., & Slovic, P (2015) Scope insensitivity: The limits of intuitive valuation of human lives in public policy Journal of Applied Research in Memory and Cognition, 4, 248-255 http://doi.org/10.1016/j.jarmac.2014.09.002 40 Erlandsson, A., Bjorklund, F., & Backstrom, M (2014) Perceived utility (not sympathy) mediates the proportion dominance effect in helping decisions Journal of Behavioral Decision Making, 27, 37-47 http://doi.org/10.1002/bdm.1789 Erlandsson, A., Björklund, F., & Bäckström, M (2015) Emotional reactions, perceived impact and perceived responsibility mediate the identifiable victim effect, proportion dominance effect and in-group effect respectively Organizational Behavior and Human Decision Processes, 127, 1-14 http://doi.org/10.1016/j.obhdp.2014.11.003 Fetherstonhaugh, D., Slovic, P., Johnson, S M., & Friedrich, J (1997) Insensitivity to the value of human life: A study of psychophysical numbing Journal of Risk and Uncertainty, 14, 283-300 http://doi.org/10.1023/A:1007744326393 Goldin, G, & Darlow, A (2013) TurkGate (Version 0.4.0) [Software] Available from http://gideongoldin.github.com/TurkGate/ Genevsky, A., Västfjäll, D., Slovic, P., & Knutson, B (2013) Neural underpinnings of the identifiable victim effect: Affect shifts preferences for giving The Journal of Neuroscience, 23, 17188-17196 http://doi.org/10.1523/JNEUROSCI.2348-13.2013 Jaeger, T F (2008) Categorical data analysis: Away from ANOVAs (transformation or not) and towards logit mixed models Journal of Memory and Language, 59, 434-446 http://doi.org/10.1016/j.jml.2007.11.007 Kahneman, D., & Tversky, A (1979) Prospect theory: An analysis of decision under risk Econometrica, 47, 263-291 http://doi.org/10.2307/1914185 Knutson, B., Rick, S., Wimmer, G E., Prelec, D., & Loewenstein, G (2007) Neural predictors of purchases Neuron, 53, 147-156 http://doi.org/10.1016/j.neuron.2006.11.010 41 Kogut, T., & Ritov, I (2005a) The “identified victim” effect: An identified group, or just a single individual? Journal of Behavioral Decision Making, 18, 157-167 http://doi.org/10.1002/bdm.492 Paolacci, G., Chandler, J., & Iperiotis, P G (2010) Running experiments using Amazon Mechanical Turk Judgment and Decision Making, 5, 411-419 Pittarello, A., Caserotti, M., & Rubaltelli, E (In press) “Three is better than two”: Increasing donations with the attraction effect British Journal of Psychology Prelec, D., & Loewenstein, G (1998) The red and the black: Mental accounting of savings and debt Marketing Science, 17, 4-28 http://doi.org/10.1287/mksc.17.1.4 R Development Core Team (2015) R: A language and environment for statistical computing Vienna, Austria: R Foundation for Statistical Computing Retrieved from: http://www.Rproject.org/ Raghubir, P (2006) An information processing review of the subjective value of money and prices Journal of Business Research, 59, 1053-1062 http://doi.org/10.1016/j.jbusres.2006.09.013 Rubaltelli, E., & Agnoli, S (2012) The emotional cost of charitable donations Cognition & Emotion, 26, 769-785 http://doi.org/10.1080/02699931.2011.613921 Rosseel, Y (2012) lavaan: An R package for structural equation modeling Journal of Statistical Software, 48, 1-36 Sinnot-Armstrong, W (2011) Consequentialism In E N Zalta (Ed.), Stanford Encyclopedia of Philosophy (http://plato.stanford.edu/archives/win2011/entries/consequentialism); Stanford, CA: Stanford University 42 Slovic, P (2007) “If I look at the mass I will never act”: Psych numbing and genocide Judgment and Decision Making, 2, 79-95 Slovic, P., & Lichtenstein, S (1971) Comparison of Bayesian and regression approaches to the study of information processing in judgment Organizational Behavior and Human Processes, 6, 649-744 http://dx.doi.org/10.1016/0030-5073(71)90033-X Slovic, P., Fischhoff, B., & Lichtenstein, S (1982) Response mode, framing, and information processing effects in risk assessment In R Hogarth (Ed.), New Directions for Methodology of Social and Behavioral Science: Question Framing and Response Consistency, no 11 San Francisco, CA: Jersey-Bass Slovic, P., Finucane, M L., Peters, E., & McGregor, D G (2002) The affect heuristic In T Gilovich, D Griffin, & D Kahneman (Eds.), Heuristics and Biases: The Psychology of Intuitive Judgments (pp 397-429) New York: Cambridge University Press Small, D A., Loewenstein, G., & Slovic, P (2007) Sympathy and callousness: The impact of deliberative thought on donations to identifiable and statistical victims Organizational Behavior and Human Decision Processes, 102, 143-153 http://doi.org/10.1016/j.obhdp.2006.01.005 Sussman, A B., Sharma, E., & Alter, A L (2015) Framing charitable donations as exceptional expenses increase giving Journal of Experimental Psychology: Applied, 21, 130-139 http://dx.doi.org/10.1037/xap0000047 Tversky, A., & Kahneman, D (1992) Advances in prospect theory: Cumulative representation of uncertainty Journal of Risk and Uncertainty, 5, 297-323 http://doi.org/10.1007/BF00122574 43 Västfjäll, D., Mayorga, M., & Slovic, P (2015) Pseudoinefficacy: negative feelings from children who cannot be helped reduce warm glow from children who can be helped Frontiers in Psychology, 6, 616 https://doi.org/10.3389/fpsyg.2015.00616 44 Footnotes We focus on monetary cost, however the same reasoning holds for other limited resource (e.g time) From a moral standpoint, we should strive to achieve the greater good for the greatest number of people, rather than helping more when there are only one or few individuals at risk (see, Slovic, 2007) This paper is based on this moral normative definition of how people should value lives at risk, however other normative standards can be used From an economical standpoint, it may be rational to help less when costs increase (von Neumann & Morgenstern, 1947; Bernoulli, 1954) 45 Table Donation amounts and number of lives helped in the seven scenarios used in Study Amount of the donation Number of children helped Scenario Scenario Scenario Scenario Scenario Scenario Scenario $5 $15 $25 $50 $75 $150 $225 children children 10 children 20 children 30 children 60 children 90 children NOTE: Throughout all scenarios we kept constant the ratio of children helped with each amount of money, which was equal to two children for every five dollars Scenarios are here reported in order from the lowest donation to the highest one, but the order of presentation was randomized for each participant 46 Table Donation decisions and perceived cost and benefit by scenario in Study Donation request Willingness to Perceived cost for Perceived benefit Participants donate the donor for the children who perceive (% of yes) Benefit > Cost % M SD M SD % Scenario 88% 2.13a 1.56 4.51 1.86 76% Scenario 75% 3.12a 1.80 5.06 1.76 73% Scenario 68% 3.72a 1.81 5.15 1.68 64% Scenario 40% 4.79a 1.81 5.44 1.63 48% Scenario 32% 5.26 1.71 5.53 1.56 41% Scenario 20% 5.96 1.56 5.80 1.51 23% Scenario 14% 6.24b 1.46 5.99 1.51 18% NOTE: Ratings were provided on 7-point Likert scales ranging from (not at all) to (very) Scenarios are here reported in order from the lowest donation to the highest one, but the order of presentation was randomized for each participant The percentage of participants who perceived the benefit higher than the cost was computed as benefit minus cost (score ranging from -6 to 6) and corresponds to all values above The difference between the rating of the cost for the donor and the rating of benefit for the recipients is significant at: a p < 01; b p < 10 47 Table Amount of the donation and number of children helped in each scenario in Study 2, with willingness to donate Amount of the donation Number of children helped Willingness to donate Scenario $5 children 73% Scenario $15 children 66% Scenario $25 10 children 58% Scenario $50 20 children 42% Scenario $75 30 children 38% Scenario $150 60 children 34% Scenario $225 90 children 30% NOTE: Throughout all scenarios we kept constant the ratio of children helped with each amount of money, which was equal to two children every five dollars Scenarios are reported in order from the lowest donation to the highest one, but the order of presentation was randomized for each participant 48 Table Mean ratings and standard deviations of the cost/benefit tradeoff measure in each scenario of Study Donation request Cost/benefit tradeoff 95% C.I Participants who perceived Benefit > Cost M SD Low/High bounds % $5/2 children 6.59 2.32 6.35/6.83 64% $15/6 children 6.10 2.50 5.84/6.36 59% $25/10 children 5.72 2.56 5.46/5.98 51% $50/20 children 5.15 2.68 4.87/5.43 45% $75/30 children 4.94 2.71 4.66/5.22 42% $150/60 children 4.60 2.77 4.31/4.89 38% $225/90 children 4.48 2.75 4.19/4.77 39% NOTE: Ratings were provided on a 9-point Likert scale ranging from =“much more costly than beneficial” to =“much more beneficial than costly”; numbers were not shown on the scale and the mid-point was labeled “equally costly and beneficial” The percentage of participants who perceived the benefit higher than the cost corresponds to all answers that were above the mid-point on the cost/benefit bipolar scale Scenarios are here reported in order from the lowest donation to the highest one, but the order of presentation was randomized for each participant 49 Table Donation decisions and perceived cost and benefits by condition and donation amount (Study 3) Condition Willingness Perceived cost for the Perceived benefit for Benefit > to donate donor recipients Cost $150 $165 Lumpsum Bundle $150 M SD $165 % of yes 46% % of yes 42% M SD 5.07 1.84 5.42 1.55 71% 68% 4.17 1.74 4.25 1.62 $150 M SD $165 $150 $165 M SD % % 5.87 1.47 5.66 1.49 49% 42% 5.80 1.44 5.76 1.46 66% 70% NOTE: Ratings of cost and benefit were provided on 7-point Likert scales ranging from (not at all) to (very) The percentage of participants who perceived the benefit higher than the cost was computed as benefit minus cost (score ranging from -6 to 6) and corresponds to values above 50 Table Donation decisions and perceived cost and benefits by condition (Study 4) Condition Willingness Perceived Perceived Benefit > to donate cost for the benefit for Valence Cost donor recipients % of yes M SD M SD % M SD C1 Lump-sum 66% 4.95 1.61 5.65 1.58 54% 3.66 1.13 C2 Bundle 83% 4.03 1.86 5.76 1.39 71% 4.11 78 C3 Children bundle 71% 4.69 1.76 5.78 1.52 63% 4.00 78 C4 Money bundle 75% 4.60 1.73 5.67 1.30 62% 3.86 85 NOTE: Ratings of cost and benefit were provided on 7-point Likert scales ranging from (not at all) to (very) For the bundle, children bundle, and money bundle conditions, the table reports average ratings across information that was split in three different aid programs The percentage of participants who perceived the benefit higher than the cost was computed as benefit minus cost (score ranging from -6 to 6) and corresponds to all values above 51 Figure captions Figure Percentage of participants who decided to help in each of the seven hypothetical scenarios of Study Figure Ratings of cost and benefit across the seven hypothetical scenarios of Study Shades indicate 95% confidence intervals Figure Donation decisions in the seven hypothetical scenarios of Study as predicted by the difference between cost and benefit For ease of representation, we grouped the scenarios into three different levels, but the analyses were done considering all seven scenarios independently from each other High values on the x-axis indicate that people rated the donation request much more costly than beneficial Figure Percentage of participants who decided to help in each of the seven hypothetical scenarios of Study Figure Ratings of cost and benefit across the seven hypothetical scenarios of Study Error bars indicate standard errors of the mean Figure Donation decisions in the seven hypothetical scenarios of Study as predicted by the tradeoff between cost and benefit For ease of representation, we grouped the scenarios into three different levels, but the analyses considered all seven scenarios independently from each other High values on the x-axis indicate that people rated the donation request much more costly than beneficial Figure Percentage of participants who decided to make a donation in each condition of Study Figure Ratings of cost and benefit in the four conditions of Study Shades indicate 95% confidence intervals 52 Figure Model assessing the role of cost and benefit as mediators of the effect of condition (lump sum vs bundle) on donation decisions Figure 10 Percentage of participants who decided to make a donation in each condition of Study Figure 11 Ratings of cost and benefit for each condition of Study Shading indicates 95% confidence intervals Figure 12 Model assessing the role of cost and benefit as mediators of the effect of condition (lump sum, C1 vs bundle, C2) on donation decisions ... diminishing perceived benefit and thus increasing willingness to help Furthermore, in Study 4, we replicated these findings using lower donation amounts (and number of children helped) Again,... recipients and cost for the donor, and we expect that people’s asymmetric valuation of cost and benefit leads to a point where benefit no longer outweighs the cost At this point, people’s willingness... participants’ decisions in each of the seven scenarios, we computed the difference between cost and benefit ratings, thus obtaining a score ranging from -6 (benefit much higher than cost) to (cost much