Toward a triadic understanding of charitable giving: How donors, beneficiaries, fundraisers, and social contexts influence donation decisions Cassandra Margot Chapman Bachelor of Arts (Hons); Bachelor of Commerce

A thesis submitted for the degree of Doctor of Philosophy at The University of Queensland in 2018 School of Psychology Abstract

This thesis applies social psychological theories about intergroup relations (e.g., Social Identity Theory, Tajfel, 1981; the Intergroup Helping Relations as Status Relations model, Nadler, 2002) to investigate the process and outcomes of charitable giving. Moving beyond the traditional focus on studying donors, this thesis uses a mixed-methods approach (surveys, experiments, thematic analysis, and literature review) to show that charitable giving is triadic, relational, and contextualized. That is, decisions about donations are informed by a triad of actors (donors, beneficiaries, and fundraisers), the relationships between them, and the wider social context. Chapter 1 provides a general overview of previous scholarly work on charitable giving and key psychological theories identified as relevant to my approach. Findings from Chapters 2 and 3 confirm that beneficiaries affect donor choices and that the relationships (e.g., shared identities) between particular donors and particular beneficiaries are especially important. Chapter 2 comprises two surveys of confirmed (N = 675) and self-reported (N = 376) donors, which show that donors are not universally generous. Instead, consistent with the social identity approach (Tajfel, 1981; Turner, Hogg, Oakes, Reicher & Wetherell, 1987), donors prefer to support charities that align with their social groups in meaningful ways. Chapter 3 goes on to present thematic analyses of responses to the question of why donors support their favorite charity (N = 1,849 from 117 countries). Results show a self/other dichotomy in donor motivation: donors are more likely to reference the self when explaining giving to religious and research charities, but are more likely to reference the other (i.e., beneficiary) when explaining giving to social service, animal, or international charities. In combination, Chapters 4 and 5 demonstrate that social contexts affect charitable responses, in part by changing the perceived relations between donors and beneficiaries. Chapter 4 presents one survey (N = 189) and one experiment (N = 290) that show how government policy on foreign aid affects private giving by changing perceptions of national giving norms. Chapter 5 presents two experiments (combined N = 440) that demonstrate how political advocacy can affect charitable giving by changing donor emotions, perceptions of efficacy, and identification with beneficiaries. Chapter 6 highlights the importance of also considering fundraisers when studying charitable giving. Mediation analyses from two samples of peer-to-peer fundraisers (combined N = 1,647) show that when fundraisers identify more with their nominated charity they take more actions to raise money and that, in turn, results in more funds raised. In particular, results show that actions which make the fundraiser more salient (e.g., adding photos, sharing personal motivations) are most effective. Chapter 7 draws together evidence from my program of research and more than 300 articles from the interdisciplinary literature to generate a new model of charitable giving: the charitable triad. This model theorizes that charitable giving is informed by

i the dyadic and triadic relationships between donors, beneficiaries, and fundraisers as well as salient aspects of the wider social context. Finally, Chapter 8 draws the entire body of work together, discusses it in relation to previous research, and presents future directions for study. The program of research presented in this thesis demonstrates the triadic nature of giving, setting a new agenda for both future research on charitable giving and fundraising practice.

ii Declaration by author

This thesis is composed of my original work, and contains no material previously published or written by another person except where due reference has been made in the text. I have clearly stated the contribution by others to jointly-authored works that I have included in my thesis.

I have clearly stated the contribution of others to my thesis as a whole, including statistical assistance, survey design, data analysis, significant technical procedures, professional editorial advice, financial support and any other original research work used or reported in my thesis. The content of my thesis is the result of work I have carried out since the commencement of my higher degree by research candidature and does not include a substantial part of work that has been submitted to qualify for the award of any other degree or diploma in any university or other tertiary institution. I have clearly stated which parts of my thesis, if any, have been submitted to qualify for another award.

I acknowledge that an electronic copy of my thesis must be lodged with the University Library and, subject to the policy and procedures of The University of Queensland, the thesis be made available for research and study in accordance with the Copyright Act 1968 unless a period of embargo has been approved by the Dean of the Graduate School.

I acknowledge that copyright of all material contained in my thesis resides with the copyright holder(s) of that material. Where appropriate I have obtained copyright permission from the copyright holder to reproduce material in this thesis and have sought permission from co-authors for any jointly authored works included in the thesis.

iii Publications included in this thesis

Chapman, C. M., Louis, W. R., & Masser, B. M. (2018). Identifying (our) donors: Toward a social psychological understanding of charity selection in Australia. Psychology & Marketing, 35(12), 980-989. doi:10.1002/mar.21150. – incorporated as Chapter 2.

Chapman, C. M., Masser, B. M., & Louis, W. R. (2018). The champion effect in peer-to-peer giving: Successful campaigns highlight fundraisers more than causes. Nonprofit and Voluntary Sector Quarterly, Online First, 1-21. doi:10.1177/0899764018805196 – incorporated as Chapter 6.

Submitted manuscripts included in this thesis

Chapman, C. M., Masser, B. M., & Louis, W. R. (under review). Self and other orientations in charitable giving: Thematic analyses from a global donor survey. Manuscript submitted for review to European Journal of Social Psychology, 6 November 2018. – incorporated as Chapter 3.

Chapman, C. M., Mirnajafi, Z., Masser, B. M., & Louis, W. R. (under review) Rage donations: Examining charitable giving as a form of collective action motivated by anger. Manuscript submitted for review to Personality and Social Psychology Bulletin, 19 December 2018. – incorporated as Chapter 5.

Other publications during candidature

Peer-reviewed journal articles

Chapman, C. M., Deane, K. L., Harré, N., Courtney, M. G. R. & Moore, J. (2017). Engagement and mentor support as drivers of social development in the Project K youth development program. Journal of Youth and Adolescence, 46(3), 644-655. doi:10.1007/s10964-017-0640- 5.

iv Chapman, C. M., Louis, W. R., & Masser, B. M. (2018). Identifying (our) donors: Toward a social psychological understanding of charity selection in Australia. Psychology & Marketing, 35(12), 980-989. doi:10.1002/mar.21150.

Chapman, C. M., Masser, B. M., & Louis, W. R. (2018). The champion effect in peer-to-peer giving: Successful campaigns highlight fundraisers more than causes. Nonprofit and Voluntary Sector Quarterly, Online First, 1-21. doi:10.1177/0899764018805196.

Masser, B.M., Davison, T.E. & Chapman, C.M. (2017). How can we encourage our voluntary non-remunerated donors to donate more frequently? ISBT Science Series, 12(1),112-118. doi:10.1111/voxs.12312

Nielsen, R. O., Chapman, C. M., Louis, W. R., Stovitz, S. D., Mansournia, M. A., Windt, J., Møller, M., Parner, E. T., Hulme, A., Bertelsen, M. L., Finch, C. F., Casals, M., & Verhagen, E. (2017). Seven sins when interpreting statistics in sports injury science. British Journal of Sports Medicine, Online First. doi:10.1136/bjsports-2017-098524

Petrovic, K., Chapman, C. M., & Schofield, T. P. (2018). Religiosity and volunteering over time: Service attendance is associated with the likelihood of volunteering, and religious importance with time spent volunteering. Psychology of Religion and Spirituality, Advance online publication., 1-11. doi:10.1037/rel0000236

Plows, S. E., Smith, F. D., Smith, J. R., Chapman, C. M., La Macchia, S. T., & Louis, W. R. (2017). Healthy eating: A beneficial role for perceived norm conflict? Journal of Applied Social Psychology, 47(6), 295-304. doi: 10.1111/jasp.12430

Book chapters

Louis, W. R., Chonu, G. K., Achia, T., Chapman, C. M., & Rhee, J. J. (2018). Building group norms and group identities into the study of transitions from democracy to dictatorship and back again. In B. Wagoner, I. Bresco, & V. Glaveanu (Eds.), The Road to Actualized Democracy: A Psychological Exploration. Charlotte, NC: IAP-Information Age Publishing.

v Peer-reviewed conference papers

Louis, W. R., Chapman, C. M., Chonu, G. K., & Achia, T. (2017). Collective action and social change. In M. C. Gastardo-Conaco, M. E. J. Macapagal, & Y. Muramoto (Eds.), Asian Psychology and Asian Societies in the Midst of Change (pp. 3-28). Quezon City, Philippines: Psychological Association of the Philippines.

Contributions by others to the thesis

Professors Winnifred Louis and Barbara Masser both supervised the research program, provided advice in all stages from conception, design, analysis, and reporting of the results, and helped to critically revise the manuscripts and thesis for clarity of argument and style.

Zahra Mirnajafi helped conceive of the idea and significantly contributed to the design for both studies reported in the Rage Donations manuscript (included as Chapter 5) and also collected data for the second study. She also helped critically revise the manuscript for clarity of argument in her capacity as co-author on the paper.

Associate Professor Emma Thomas helped critically revise the Charitable Triad manuscript (included as Chapter 7) for clarity and structure of the argument in her capacity as co-author on the paper.

Fundraising consultants Martin Paul and Gavin Copey from More Strategic consultancy collected the data reported in Chapters 2 and 6 as part of a broader project. They did not have input in the design, analysis, or reporting of the results included in this thesis.

Statement of parts of the thesis submitted to qualify for the award of another degree

No works submitted towards another degree have been included in this thesis.

vi Research Involving Human or Animal Subjects

The research presented in this thesis received clearance from the University of Queensland School of Psychology Ethics Committee. Details of the various ethics approvals and the data sets they relate to are presented in the Table below. Copies of the ethics approval letters are also provided in Appendix A.

Project Name Studies Clearance Committee Survey of charitable attitudes and Chapter 3; 15-PSYCH-PHD- Psychology Ethics behaviour among UQ students Chapter 4, S4.1 29-JH

Australian donation behaviour Chapter 2; 15-PSYCH-PHD- Psychology Ethics across charity domains Chapter 6 73-JH

Prosocial conformity Chapter 4, S4.2 17-PSYCH-PHD- Psychology Ethics 4-AH Rage donations Chapter 5 17-PSYCH-PHD- Psychology Ethics 79-JS

vii Acknowledgements

Completion of this thesis would not have been possible without the tireless support, enthusiasm, and challenging critique from my two research advisors, Winnifred Louis and Barbara Masser. It’s not possible to convey just how much these two women have meant to me over the past few years but suffice it to say I cannot thank them enough. Winnifred, thank you for always championing me, believing I could master statistics, and for turning every opportunity into a teaching moment. I will be ever grateful for your energy, kindness, humor, and intellect. Barbara, thank you for giving me the space to grow but also for offering such thoughtful advice whenever I asked for it. I appreciate the passion you have for your work and your understated excellence—and my writing will be forever improved because of your gentle guidance.

Beyond my advisory team, I wish to thank all of my psychology whanau for your support and guidance throughout my candidature. Big thanks to Matthew Hornsey and Emma Thomas for your spontaneous mentoring and advice. Special thanks to Tarli Young, Mike Norwood, Emily Harris, Zahra Mirnajafi, Brendan Zietsch, Morgana Lizzio-Wilson, Tamara Butler, Shannon Stuart, Bill Bingley, and Wuri Prasetyawati for your friendship throughout the years. Thanks also to all the inspiring scholars in the Centre for Research in Social Psychology, the Social Change Lab, and the Applied Social Lab—your work is fascinating, important, and it has been so fun to be on the sidelines for your successes.

Thanks to all my friends and family who have politely held back the eye-glaze when I talked about my research and who kept my life in balance by filling it with adventure. Hugest thanks to Jayd Blunden for your constant support—I’m sure you know more about the psychology of giving than you ever hoped to and I thank you for always listening to my woes and celebrating my small wins.

Finally, I wish to thank Martin Paul, More Strategic, Pareto Fundraising, and dozens of their charity clients (too many to name here) for generously sharing their data for scholarly purposes. It has been inspiring to work with you all, and I hope that this research will be valuable for your important work.

viii Financial support

This research was supported by an Australian Government Research Training Program Scholarship. The University of Queensland’s School of Psychology also contributed student support funds for research costs and conference travel.

Keywords

Charitable giving, donations, prosocial behavior, helping, fundraising, identity, norms

Australian and New Zealand Standard Research Classifications (ANZSRC)

ANZSRC code: 170113, Social and Community Psychology, 80% ANZSRC code: 150502, Marketing Communications, 20%

Fields of Research (FoR) Classification

FoR code: 1701, Psychology, 80% FoR code: 1505, Marketing, 20%

ix Table of Contents

Table of Contents ...... x List of Tables ...... xiii List of Figures ...... xiv Chapter 1. General Introduction ...... 1 Traditional Approaches to Studying Charitable Giving ...... 2 Charitable Giving as a Group Process...... 5 Triadic Nature of Charitable Giving ...... 8 Contributions of the Thesis ...... 9 Thesis Overview ...... 10 Chapter 2. Identifying (Our) Donors: Towards a Social Psychological Understanding of Charity Selection in Australia ...... 12 Abstract ...... 13 Introduction ...... 14 Study 2.1...... 16 Material and Methods...... 16 Results and Discussion ...... 17 Study 2.2...... 21 Material and Methods...... 21 Results and Discussion ...... 22 General Discussion ...... 25 Conclusions ...... 28 Chapter 3. Self and Other Orientations in Charitable Giving: Thematic Analyses from a Global Donor Survey ...... 29 Abstract ...... 30 Introduction ...... 31 Method ...... 34 Results ...... 35 Discussion ...... 48 Conclusion ...... 51 Chapter 4. Norm Divergence Model of Prosocial Influence: Government Aid Both Crowds In and Crowds Out Citizens’ Charitable Giving Via Diverging Descriptive and Injunctive Norms ………………………………………………………………………………… 53 Abstract ...... 54

x Introduction ...... 55 Study 4.1...... 59 Results ...... 60 Discussion ...... 62 Study 4.2...... 63 Method ...... 64 Results ...... 66 Discussion ...... 71 General Discussion ...... 71 Chapter 5. Rage Donations: Examining Charitable Giving as a Form of Collective Action Motivated by Anger ...... 76 Abstract ...... 77 Introduction ...... 78 Study 5.1...... 80 Method ...... 80 Results ...... 83 Discussion ...... 88 Study 5.2...... 88 Method ...... 89 Results ...... 90 Discussion ...... 94 General Discussion ...... 95 Conclusion ...... 97 Chapter 6. The Champion Effect in Peer-to-Peer Giving: Successful Campaigns Highlight Fundraisers More Than Causes ...... 98 Abstract ...... 99 Introduction ...... 100 Method ...... 104 Results ...... 108 Discussion ...... 112 Conclusion ...... 117 Chapter 7. The Charitable Triad: How Donors, Beneficiaries, and Fundraisers Influence Charitable Giving ...... 118 Abstract ...... 119 Introduction ...... 120

xi The Charitable Triad: Donors, Beneficiaries, and Fundraisers ...... 121 Donors as Agents of Charity ...... 123 Beneficiary-Centered Approaches to Understanding Charitable Giving ...... 128 Donor-Beneficiary Dyad: Who Gives Depends on Who Will Receive ...... 131 Fundraisers Promote Charitable Giving ...... 134 Donor-Fundraiser Dyad: Who Gives Depends on Who Asks...... 138 Fundraiser-Beneficiary Dyad: Who Asks Determines If They Receive ...... 139 Putting Charitable Giving in Context ...... 140 The Charitable Triad Model: Charitable Giving as Triadic, Relational, and Contextual ...... 144 Conclusion ...... 147 Chapter 8. General Discussion ...... 149 Donors: Understanding Who Gives and Why ...... 149 Beneficiaries: Understanding Who Is Most Likely to Receive and Why ...... 150 Social Contexts Influence Giving ...... 151 Fundraisers: Understanding Who Is Most Effective at Asking for Donations and Why ...... 152 The Charitable Triad ...... 153 A Word on Measurement ...... 153 Implications for Practice ...... 154 Strengths, Limitations, and Future Directions ...... 155 Conclusion ...... 157 References ………………………………………………………………………………………. 158 Appendix A: Evidence of Ethics Clearance ...... 194 Ethics Clearance: 15-PSYCH-PHD-29-JH ...... 194 Ethics Clearance: 15-PSYCH-PHD-73-JH ...... 195 Ethics Clearance: 17-PSYCH-PHD-4-AH ...... 196 Ethics Clearance: 17-PSYCH-PHD-79-JS ...... 197 Appendix B: Supplementary Materials for Chapter 2 ...... 198 Appendix C: Supplementary Materials for Chapter 6 ...... 199

xii List of Tables

Table 1...... 19 Table 2...... 20 Table 3...... 23 Table 4...... 24 Table 5...... 36 Table 6...... 38 Table 7...... 43 Table 8...... 46 Table 9...... 47 Table 10...... 61 Table 11...... 68 Table 12...... 69 Table 13...... 84 Table 14...... 86 Table 15...... 91 Table 16...... 92 Table 17...... 107 Table 18...... 109 Table 19...... 110 Table 20...... 198 Table 21...... 200 Table 22...... 201

xiii List of Figures

Figure 1. Results of the moderated mediation analysis of the effect of perceived aid levels on international donation likelihood via norms, moderated by identification ...... 61 Figure 2. Summary of results of the moderated mediation analysis of the effect of manipulated aid levels on international donation willingness and behavior via perceived aid levels and norms ...... 70 Figure 3. Mediation analysis for Study 5.1 (NFL players kneeling) ...... 87 Figure 4. Mediation analysis for Study 5.2 (Abortion) ...... 93 Figure 5. Mediation model showing how fundraiser’s identification with the cause predicts fund’s raised indirectly via their best practice actions of solicitation, signaled investment, high target, and signaled efficacy (2013 sample) ...... 111 Figure 6. The charitable triad model: charitable decisions are influenced by characteristics of the three actors, but also the interactions between particular donors, beneficiaries, and fundraisers within a wider social context ...... 122

xiv Chapter 1. General Introduction

Charities and other nonprofits are tasked with diverse and important missions: caring for people in need, educating children, promoting religion, encouraging sport and the arts, researching the causes and cures of diseases, protecting animals and the environment, and much more. Financial donations are the lifeblood of these organizations. Many people make charitable donations each year—55% of American, 61% of British, and over 80% of Australian adults (Charities Aid Foundation, 2017a; Giving Australia, 2016; Osili & Zarins, 2018)—and without these contributions, many essential services would go undelivered. Understanding why donors give (and why they do no not) is therefore important. This thesis is specifically concerned with charitable giving, or donating money to organizations that help others outside the family circle (Bekkers & Wiepking, 2011b). Psychology as a discipline has spent decades examining prosocial behavior, which is action intended to benefit others (see, for example, Dovidio, Piliavin, Schroeder, & Penner, 2006; Eagly & Crowley, 1986; Fischer et al., 2011; Ma, Tunney, & Ferguson, 2017; Penner, Dovidio, Piliavin, & Schroeder, 2005; Saucier, Miller, & Doucet, 2005; Shariff, Willard, Andersen, & Norenzayan, 2015). However, the various forms of prosocial action may be qualitatively different. Helping research, for example, investigates dyadic contexts when bystanders respond to individuals in need (e.g., Eagly & Crowley, 1986; Stürmer, Snyder, Kropp, & Siem, 2006) or the conditions under which an individual from one group will help people from another group (e.g., M. Levine & Manning, 2013; Nadler, 2002; van Leeuwen, 2017; Zagefka & James, 2015). I argue that giving decisions are often different from traditional helping decisions because of the triadic nature of the phenomenon— giving is most frequently a response to a third party asking on behalf of the beneficiary. There is also evidence that different forms of giving (e.g., money, time, blood, or body parts) have different underlying motives (Chell, 2016; Clary & Snyder, 1999; Clary et al., 1998; Konrath & Handy, 2017; L. Lee & Piliavin, 1999). For these reasons, I focus my attention specifically on financial charitable giving. My aim is to generate new psychological insights by integrating disparate interdisciplinary approaches and applying psychological theories to the study of charitable giving. Traditionally, research on charitable giving has asked who gives to charity and why (e.g., Bekkers & Wiepking, 2011b; Bekkers & Wiepking, 2011c). Yet these literatures are filled with contradictory findings that suggest untested moderators. I argue in this thesis that any attempt to understand charitable giving that focuses only on the donor will be incomplete. Donors do not give in isolation but instead give to particular beneficiaries. As such, characteristics of beneficiaries will determine whether or not a gift will be made (see, for example, Zagefka & James, 2015). The relationship between donors and beneficiaries can also influence giving (such as shared identities;

1 Prendergast & Maggie, 2013). Salient factors in the social context (e.g., the identities or norms that are made salient; M. Levine, Prosser, Evans, & Reicher, 2005) also affect the nature of these relationships. Further, charitable gifts are rarely offered spontaneously: most giving occurs in response to being asked (Bryant, Jeon-Slaughter, Kang, & Tax, 2003). The requester is usually a third party—a fundraiser—and characteristics of fundraisers can influence whether or not gifts are made (e.g., Landry, Lange, List, Price, & Rupp, 2006). The relationships between fundraisers and both donors and beneficiaries should also influence giving, although these relationships have not yet been extensively studied. This thesis generates a new conceptual model for understanding charitable giving: the charitable triad. In this view, the decision of a donor to give is informed by characteristics of donors, beneficiaries, and fundraisers, the dyadic and triadic relationships between them, and the wider social context. This thesis comprises six manuscripts presenting a research program that demonstrates how charitable giving is triadic, relational, and contextualized. Chapters 2 and 3 examine how particular beneficiaries or charities influence who gives and why. These studies also highlight identities as important drivers of donation decisions. Chapters 4 and 5 show how social contexts—specifically social norms, government policy, and advocacy—can affect donor behavior, by influencing the donor’s self-conceptions (identities or perceived norms), feelings, and identification with beneficiaries. Chapter 6 demonstrates the importance of the fundraiser in eliciting donations, showing the relationships between donors and fundraisers also inform donor choices. Based on this program of research and a review of over 300 articles from the interdisciplinary literature, Chapter 7 puts forward a novel conceptual model for the study of charitable giving: the charitable triad. Finally, Chapter 8 presents a general discussion of the findings, meaning, and contribution of the thesis as well as limitations and suggestions for future research. Traditional Approaches to Studying Charitable Giving Charitable giving is of interest to various disciplines, especially psychology, economics, sociology, nonprofit studies, and marketing. The dominant focus of past research on charitable giving across these disciplines has been on understanding who gives, and why (e.g., Bekkers & Wiepking, 2011b; Bekkers & Wiepking, 2011c; Konrath & Handy, 2017; Wiepking & Bekkers, 2012). Who Gives to Charity? A primary focus of research within sociology, non-profit studies, and marketing has been to identify who donors are. Research has examined a range of demographic characteristics. Broadly speaking, donors are more likely to be female, older, married, religious, educated, and wealthy (Bègue, 2014; Bekkers & Wiepking, 2011c; Brooks, 2003b; Choi & DiNitto, 2012; Forbes & Zampelli, 2013; Helms & Thornton, 2012; Herzog & Yang, 2018; Hill & Vaidyanathan, 2011; 2 Hughes & Luksetich, 2008; Korndörfer, Egloff, & Schmukle, 2015; Leslie, Snyder, & Glomb, 2013; Manesi, Van Lange, Van Doesum, & Pollet, 2018; McGregor-Lowndes & Crittall, 2014; Mesch, Brown, Moore, & Hayat, 2011; Mesch, Osili, Ackerman, & Dale, 2015a; Mesch, Rooney, Steinberg, & Denton, 2006; Midlarsky & Hannah, 1989; Neumayr & Handy, 2017; Rajan, Pink, & Dow, 2009; Rooney, Mesch, Chin, & Steinberg, 2005; Sibley & Bulbulia, 2015; Stavrova & Siegers, 2014; L. Steinberg & Monahan, 2007; Tremblay-Boire & Prakash, 2017; Vaidyanathan, Hill, & Smith, 2011; Wiepking & Bekkers, 2012; Wiepking, Bekkers, & Osili, 2014; Wiepking & James, 2013; Wiepking & Maas, 2009). A range of individual psychological differences have also been examined with reference to charity donors, many of which are correlated with demongraphics. For example, people higher in empathy—the ability to understand and respond appropriately to other people’s thoughts and emotions (Baron-Cohen, 2011)—are more likely to give to charity (Batson, Chang, Orr, & Rowland, 2002; Bekkers & Ottoni-Wilhelm, 2016; Davis, 1983; Kim & Kou, 2014; Luengo Kanacri et al., 2016; Mesch et al., 2011; Neumayr & Handy, 2017; Wiepking & Maas, 2009; Zhou, Wildschut, Sedikides, Shi, & Feng, 2012). On the other hand, people higher in social dominance orientation—a preference for group-based hierarchy in society—are far less willing to help others and give to charity (M. A. Brown, 2011; Freeman, Aquino, & McFerran, 2009; Halabi, Dovidio, & Nadler, 2008; Pratto et al., 2013). People who report a strong moral identity—who feel it is important to be seen as a moral person—are also more generous (Aquino & Reed, 2002; Saerom Lee, Winterich, & Ross, 2014; Winterich, Mittal, & Ross Jr, 2009; Winterich, Zhang, & Mittal, 2012). Finally, people with a universalist value orientation—believing that all people are worthy of care and concern—also give more to charity (Joireman & Duell, 2007; Schwartz, 1994). The associations reported above, however, are not universally observed (e.g., Andreoni, Brown, & Rischall, 2003; Bennett, 2012; Chapman, Louis, & Masser, 2018; Forbes & Zampelli, 1997; Helms & Thornton, 2012; Mesch, Osili, Ackerman, & Dale, 2015b; Piff, Kraus, Côté, Cheng, & Keltner, 2010; Piper & Schnepf, 2008). Inconsistencies in the associations between donor characteristics and giving suggest untested moderators, which I propose include characteristics of the beneficiary. A recent review article on disaster giving by Zagefka and James (2015) identified psychological aspects of beneficiaries that influence whether or not they will receive help. These include: the identifiable victim effect, where beneficiaries who are identified by name or photograph receive more support than beneficiaries that are anonymous (see also Dickert, Kleber, Västfjäll, & Slovic, 2016; Genevsky, Västfjäll, Slovic, & Knutson, 2013; Jenni & Loewenstein, 1997; Kogut & Ritov, 2005); victim blame, where beneficiaries who are blamed for their circumstances receive less support (see also Fong & Luttmer, 2011; Leeuwen & Wiepking, 2012;

3 Loseke & Fawcett, 1995; Zagefka, Noor, Brown, de Moura, & Hopthrow, 2011); stereotypes, or fixed and simplified ideas of what a beneficiary should look like (see also Cuddy, Fiske, & Glick, 2007); and perceived need, where beneficiaries who are seen to be needy receive more help (see also Bennett & Kottasz, 2000; Leeuwen & Wiepking, 2012; Loseke & Fawcett, 1995; Polonsky, Shelley, & Voola, 2002; Zagefka, Noor, Brown, Hopthrow, & Moura, 2012). These literatures demonstrate that qualities of the beneficiary also affect giving decisions. This thesis also seeks to examine the importance of beneficiaries in understanding who gives—not just identifying psychological characteristics of beneficiaries but also examining how attributes of the beneficiary are differently perceived by different donors. Chapter 2 includes two studies that explore whether the associations between some of the donor demographic characteristics outlined above and charitable giving depend on the category of charity being considered, as different categories highlight different beneficiaries. Why Do People Give to Charity? A focus on characteristics alone helps us understand who gives, and to whom, but doesn’t tell us much about why people give. Various reviews have summarized evidence about what motivates donors to give (e.g., Bekkers & Wiepking, 2011b; Konrath & Handy, 2017). Diverse motives have been put forward, but so far there has been little theoretical discussion of when and why different motives take precedence over others. Scholars’ understanding of donor motivations is primarily derived from research in economics, marketing, and psychology. Sargeant and Woodliffe (2007b) identified various motives for charitable giving, including altruism, empathy, emotions (e.g., sympathy, guilt, fear), social justice needs, prestige (i.e., seeking public recognition), and tax benefits. Similarly, Bekkers and Wiepking (2011b) reviewed around 500 articles to identify eight key mechanisms that they propose drive charitable giving: awareness of need (knowing donations are needed), solicitation (being asked), costs and benefits (pragmatic considerations such as tax benefits and thank you gifts), altruism (caring about the charity’s output or consequences for beneficiaries), reputation (considering the social consequences of making a gift), psychological benefits (feeling good about oneself), values (acting out one’s priorities in the world), and efficacy (whether or not the donation will make a difference). Finally, Konrath and Handy (2017) recently argued that donors are motivated by six key concerns: trust (confidence in charities to deliver services), altruism (concern about the wellbeing of the less fortunate), social (surrendering to social influence), tax benefits (reducing the price of giving), egoism (gaining public praise), and constraints (not having money to spare). These reviews highlight that different research disciplines emphasize different motives, without necessarily integrating across approaches or attempting to theorize when certain motives 4 would take precedence over other. Considering various identified motives, I suggest motivations may fall within a dichotomy of motives that are more self-oriented (e.g., reputation, social motives, egoism, prestige, benefits, constraints) or other-oriented (e.g. altruism, empathy, efficacy, need). I propose that the other—or beneficiary—may also influence which motives are strongest. In Chapter 3, I present a thematic analysis of donors’ explanations of why they support their favorite charity to test if and how donors are motivated to by self- and other-oriented concerns. Further, this chapter considers how the type of charity or beneficiary may influence the motive for giving. So far I have presented evidence of who gives and why, problematized a traditional focus on donors, and shown that considering beneficiaries as well as donors helps to clarify inconsistencies in previous evidence. A new focus on beneficiaries provides a framework for categorizing donor motives as self- or other-oriented, raising new theoretical questions as well as illuminating new applied problems of fundraising in each case. However, donors and beneficiaries do not interact merely as individuals; sometimes groups are salient. Indeed, I argue throughout this thesis that charitable giving is often an intergroup process and that identities strongly influence charitable decisions. Below, I review existing research that has considered helping as an intergroup process and argue that these approaches inform charitable giving (a unique type of helping behavior) but have yet to be systematically applied to the context of giving. Charitable Giving as a Group Process Although charitable giving is usually the action of an individual donor, beneficiaries are rarely individuals. Instead, beneficiaries are groups that warrant support (e.g., cancer patients, hungry children, disaster victims). I argue that giving decisions therefore make group identities salient. Helping research has often considered how group dynamics influence the voluntary provision of help to others (e.g., M. Levine & Manning, 2013; Nadler, 2002; van Leeuwen, 2007; Zagefka & James, 2015). Much of this work has used charitable giving as a measure of helping but has theorized more broadly in relation to helping or other forms of prosocial behavior. Although I argue throughout this thesis that charitable giving is unique in its triadic nature, theories and evidence from the wider helping literature can shed light on the mechanisms at work in charitable giving decisions. In particular, three theories—Social Identity Theory (Tajfel, 1981), the Intergroup Helping as Status Relations model (Nadler, 2002), and the Strategic Outgroup Helping model (van Leeuwen, 2017)—are relevant to the work presented throughout this thesis. Social Identity Theory and Preference for Helping Ingroup Beneficiaries Social Identity Theory (and its offshoot, Self-Categorization Theory; Tajfel, 1981; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987) proposes that people hold both individual (personal) and group-based (social) identities that may be relevant in or made salient by certain contexts. The social groups that a person belongs to shape their self-definition and their attitudes and behavior as 5 group members (Turner et al., 1987). When a given identity is salient, perceptions of what other people in one’s group approve of or do (i.e., group norms) influence a range of socially desirable behaviors (see, for example, Cialdini, Reno, & Kallgren, 1990; Hysenbelli, Rubaltelli, & Rumiati, 2013). Identities inform helping in emergencies (see review by M. Levine & Manning, 2013) and have been theorized to be relevant to charitable giving (Aaker & Akutsu, 2009). Indeed, empirical research has demonstrated that identities and their associated norms affect donor choices such as the decision to donate or the value of gifts (Charnysh, Lucas, & Singh, 2015; Croson, Handy, & Shang, 2009; Hysenbelli et al., 2013; Kessler & Milkman, 2018; Zagefka, Noor, & Brown, 2013). Identities also inform who receives help. Donors are more willing to give when they perceive the beneficiary to be similar to them, particularly when a shared identity is salient (Charnysh et al., 2015; T. K. James & Zagefka, 2017; M. Levine et al., 2005). T. K. James and Zagefka (2017) demonstrate this preference for helping ingroup victims in a series of vignette studies that piped in the participant’s own country (vs. a fictional country) and invited donations to disaster victims. In their third study, for example, participants reported they would hypothetically make a £5.59 donation when they thought the victims were from their own country, compared to only £3.86 when victims were from another country. In a series of studies on intergroup helping, Levine and colleagues also provided evidence that people prefer to help ingroup members and demonstrated that whether or not help is offered depends in part on the particular identity made salient (M. Levine, Cassidy, Brazier, & Reicher, 2002; M. Levine et al., 2005; M. Levine & Thompson, 2004; S. Reicher, Cassidy, Wolpert, Hopkins, & Levine, 2006). For example, British participants were more willing to donate to help victims of a natural disaster in Europe (vs. South America) when their shared European identity was highlighted, despite showing no preferences between helping European and South American victims when their British identity was highlighted (M. Levine & Thompson, 2004). A similar study in India found that Hindu people showed a greater willingness to help Muslim disaster victims when they were encouraged to think of themselves as sharing a superordinate identity (i.e., their Indian nationality) with the victims (Charnysh et al., 2015). These studies highlight how contextual cues can influence giving decisions by making particular identities salient. The social identity approach outlined above proposes identities will inform behavior and that people will show a preference for helping others within their own important groups. This preference is demonstrated in the three studies reported in Chapters 2 and 3, which show that donors prefer to give to causes that align with the priorities of their social groups and that donors are actively motivated to help beneficiaries with whom they share important identities.

6 Nonetheless, there will be some contexts in which people may favor helping people outside of the bounds of their own group, especially when there are benefits to the group for doing so. Models of Strategic Helping of Outgroup Beneficiaries Despite strong evidence for a preference to help people within one’s own groups, a significant portion of charitable giving occurs across group borders. This includes obvious outgroup helping, such as donations to international appeals and support of animal welfare. But it also includes donations to dissimilar others within the borders of one’s nation or species. Indeed, the study reported in Chapter 3 shows that particular kinds of charitable giving—specifically to support animals and international beneficiaries—are most likely to be motivated by aspects of the beneficiary (vs. the self). Two models have been proposed to explain when and why people may help outgroups for strategic reasons. The Intergroup Helping as Status Relations model, based in part on Social Identity Theory, was first put forward in 2002 by Nadler and has since been elaborated (Halabi & Nadler, 2017; Nadler, 2002, 2016). According to this model, groups can use helping as a way to maintain or establish dominance. When helping occurs between members of different groups, it inherently indicates relative status (Halabi & Nadler, 2017; Nadler, 2002, 2016). In this approach, needing help communicates low status and offering help communicates high status. Substantial evidence has now accrued for the two main propositions of this model. First, some high status people do help in order to assert their power (Halabi et al., 2008; Nadler & Chernyak-Hai, 2014; Nadler, Harpaz-Gorodeisky, & Ben-David, 2009). Second, some recipients show aversive outcomes after receiving aid, such as self-esteem threat and reduced self-confidence (Fisher, Harrison, & Nadler, 1978; Fisher & Nadler, 1974; Halabi, Nadler, & Dovidio, 2011; Nadler & Fisher, 1986; Nadler, Fisher, & Streufert, 1976; Nadler & Halabi, 2006). Relatedly, the Strategic Outgroup Helping model developed by van Leeuwen (2017) proposes that sometimes people help others out of concern for their own group interests, rather than (or as well as) out of concern for the beneficiary group. The model outlines eight strategic reasons for outgroup helping: power (to demonstrate or maintain the group’s high status position), warmth (managing impressions of the group as kind), competence (demonstrating the group has important skills), meaning (reaffirming the meaningfulness of the group identity), collective guilt (making reparations for the group’s past bad behavior), collective pride (celebrating the group’s past good behavior), inclusion (expressing solidarity and strengthening bonds with subgroups), and distinctiveness (signaling one’s group is better than another group). Empirical evidence supports many of the claims of this model. Power is a stronger motive when the status relations between groups are unstable (see also Nadler, 2002; Nadler et al., 2009). The image of one’s group can be bolstered by helping, which may enhance group distinctiveness (van Leeuwen & Harinck, 2016). 7 Helping can also be a meaningful way to communicate group identity, especially when it is under threat. For example, people are more likely to help outgroup disaster victims when their national identity is threatened, as long as the help they could offer was highly relevant to their national identity (van Leeuwen, 2007). The evidence presented above shows how helping can be an intergroup process. Although these approaches have not yet been systematically applied to the study of charitable giving, they are informative and relevant. In particular, this work shows just how important it is to consider the characteristics of both the donor and the beneficiary—not only in isolation but also in interaction. The work summarized above also points to the way contexts can make certain identities or group concerns more or less relevant. Salient characteristics within the intergroup context may affect the relations between particular donors and beneficiaries in important ways. Chapter 4 comprises two studies that test how government policy on foreign aid can influence private giving to international beneficiaries through shifting perceptions of national giving norms. Chapter 5 presents two experiments that test how public advocacy on contentious social issues influences whether or not people will donate to particular beneficiaries that are highlighted by their own and opposing groups as victims. Triadic Nature of Charitable Giving Understanding giving as dyadic (between donors and beneficiaries) helps explain donor responses. However, charitable giving is also triadic. There is a third party relevant for understanding the psychology of giving: the fundraiser. Until now, the literature has focused mostly on donors and rarely on fundraisers (Breeze, 2017a, 2017b). However, any attempt to understand giving that neglects the fundraiser will be incomplete. To date, little is known about what makes individual fundraisers successful. Research shows that personal appearance, professionalism, and incentives for asking can influence giving responses (Barasch, Berman, & Small, 2016; Bull & Gibson-Robinson, 1981; Landry et al., 2006; L. R. Levine, Bluni, & Hochman, 1998; List & Price, 2009; Sargeant & Hudson, 2008; West & Jan Brown, 1975). Fundraising organizations are sometimes more successful when they are perceived to be effective (Iyer, Kashyap, & Diamond, 2012; Katz, 2018; Scharf & Smith, 2016; J. R. Smith & McSweeney, 2007) and trustworthy (e.g., Beldad, Snip, & van Hoof, 2014; Naskrent & Siebelt, 2011; Sargeant & Hudson, 2008; Sleesman & Conlon, 2016; Thomas et al., 2015), although these relationships are not always found (Berman, Barasch, Levine, & Small, 2018; Katz, 2018; Silvergleid, 2003; Szper & Prakash, 2011; Tremblay-Boire & Prakash, 2017; Wiepking, 2010). Two studies in Chapter 6 examine the importance of fundraisers for peer-to-peer giving. Using observed donations (behavioral data), these studies show that campaigns that champion the fundraiser have more success than campaigns that highlight the beneficiary charity. These studies

8 highlight both the importance of the fundraiser as an actor, and the importance of relational considerations (e.g., between the fundraiser and the donor) which have not yet been systematically examined. Triadic relationships have previously been theorized as relevant to intergroup relations (Zagefka, 2018b) and social change (Subašić, Reynolds, & Turner, 2008), but have not yet been considered in relation to charitable giving. I elaborate this novel theoretical position in Chapter 7, in which I review over 300 articles from the interdisciplinary literature to generate the charitable triad model of charitable giving. Contributions of the Thesis A great deal of research has been done to identify who gives to charity and to try to understand why people might do so. However, the evidence base contains many inconsistent findings that suggest untested factors could moderate the reported associations. Further, why and when different giving motives take precedence over others has not previously been theorized. This thesis demonstrates the ways that donor decisions are influenced by beneficiaries, provides a framework for organizing motives (self- vs. other-oriented), and shows how different motives can be activated by different beneficiaries. It also examines the embeddedness of the donor-beneficiary dyad within groups and contexts, drawing liberally from the helping literature for relevant theory and data. Group process theories (Nadler, 2002; Tajfel, 1981; Turner et al., 1987; van Leeuwen, 2017) demonstrate two key points in relation to helping behavior. First, that identities matter—especially that beneficiaries are most likely to be helped when they share an identity with the donor. Second, that people may also help dissimilar others for strategic reasons. Yet little work has specifically applied these theories to charitable giving, a unique form of helping. When these theories have been applied it has almost exclusively been to understand disaster giving (M. Levine & Manning, 2013; M. Levine et al., 2005; M. Levine & Thompson, 2004; Zagefka, 2018a; Zagefka & James, 2015; Zagefka et al., 2013; Zagefka et al., 2011; Zagefka et al., 2012). Emergency contexts may activate different psychological processes than traditional helping contexts (Saucier et al., 2005), making it important to understand how group processes apply to a broader range of charitable giving contexts. In this thesis I do just that, showing across a range of charity types that shared identities between donors and beneficiaries are strong motives in giving and that donors give to charities that align with their group’s priorities, particularly when supporting certain types of charities. Further, I demonstrate the ways that social information from the wider context—such as norms, government policy, and political advocacy—can influence the dyadic relationships between donors and beneficiaries.

9 Finally, I propose that charitable giving is uniquely triadic—most giving decisions involve a donor, a beneficiary, and a fundraiser. Until now, little research has actively investigated the way characteristics of fundraisers inform donor decisions, let alone the way relationships between fundraisers and particular donors or beneficiaries could influence outcomes. This thesis demonstrates the importance of fundraisers working as champions of causes in peer-to-peer giving contexts. Finally, I present a broad review of the interdisciplinary literature to generate a new model of charitable giving that conceptualizes giving as triadic, relational, and contextualized. Thesis Overview This thesis comprises six manuscripts—which include five surveys, a thematic analysis, three experiments, and a conceptual review—that present a research program on the triadic and contextualized nature of charitable giving. The studies presented in this thesis show that three other factors (beyond donors) influence charitable giving: beneficiaries, fundraisers, and social contexts. Beneficiaries Matter Chapters 2 and 3 both investigate the role of beneficiaries and how particular types of causes or beneficiaries elicit donations from particular types of donors. Chapter 2 reports two cross- sectional studies conducted with Australian donors. By looking at donors’ preferred and supported charities, I show how social categories commonly associated with giving—gender, age, religiosity, and political orientation—are not universal predictors of charity support. Instead, these categories predict patterns of charity preferences that align with group priorities. Social Identity Theory (Tajfel, 1981; Turner et al., 1987) is used to interpret differences in the patterns of support. Chapter 3 reports the results of a thematic analysis of responses from 1,849 donors around the world who were asked why they support their favorite charity. Results demonstrate an important self/other dichotomy in giving: support of some types of charities is more likely to be motivated by the donor (a self-orientation), while support of other types of charities is more likely to be motivated by the beneficiary (an other-orientation). Results also show that donors often give to beneficiaries that share important identities; and that beneficiaries are often constructed as being needy and powerless. Combined results from the three studies reported in chapters 2 and 3 show that beneficiaries matter—who gives depends in part on who receives. Results also highlight the way both identities (Tajfel, 1981; Turner et al., 1987) and status relations between donors and beneficiaries (Nadler, 2002) inform giving responses. Contexts Matter Chapters 4 and 5 investigate how the social context in which giving occurs can influence charitable behavior. Chapter 4 reports results of one survey and one experimental study about perceptions of foreign aid. Participants who perceived national spending on foreign aid to be high 10 reported a greater willingness to give to charity because they perceived more people in their national group also gave (i.e., they thought the descriptive giving norm was higher). Chapter 5 reports results from two experiments about rage donations that show how public advocacy on contentious social issues can influence donor emotions (specifically anger), change the degree of identification the donor feels for highlighted beneficiaries and, more broadly, change donors’ willingness to financially support charities focused on issues that are important to them. Results reported in these two chapters demonstrate how social contexts—social norms, government policy, and advocacy—influence charitable giving by changing the dynamics between donors and beneficiaries. Fundraisers Matter Chapter 6 presents evidence that fundraisers also influence charitable outcomes. Across two samples of online peer-to-peer fundraisers, results show that the actions fundraisers take impact how much money they raise. In the peer-to-peer context, donors may be more motivated to give to the fundraiser than to the cause in question (i.e., the beneficiary). These results demonstrate that fundraisers matter—who gives depends in part on who asks and how—and support the central thesis that charitable giving is triadic in nature. Results also show how contexts, in this case the giving context, can change the nature of the relationships between actors in the charitable triad. Understanding How Donor Decisions Are Influenced by the Charitable Triad Drawing together findings from the nine empirical studies reported in Chapters 2-6 and more than 300 articles from the interdisciplinary literature, Chapter 7 puts forward a novel conceptual model for the study of charitable giving. Giving decisions are proposed to be influenced by a triad of actors—donors, beneficiaries, and fundraisers—who, in turn, are influenced by salient characteristics of their wider social context. The charitable triad model helps clarify diverse findings and orients the field toward future research. Finally, I put forward several testable propositions generated by the model. Although each chapter contains a discussion in relation to the research included there, Chapter 8 presents a general discussion of the findings, meaning, and contribution of the thesis. In this final chapter, I integrate the threads of the thesis and draw out specific conclusions and contributions. I also discuss the strengths and limitations of the program of research and provide suggestions for future research directions.

11 Chapter 2. Identifying (Our) Donors: Towards a Social Psychological Understanding of Charity Selection in Australia

This chapter tests whether the charity target (as a proxy for type of beneficiary) affects who gives to charity. I specifically examine how four donor characteristics commonly associated with the propensity to give to charity—gender, age, religiosity, and political orientation—predict giving to different categories of charity, namely religious, international, animal welfare, health, and public benevolent organizations. Results show that patterns of support align with the priorities of important donor identities, demonstrating that the question of who gives (i.e., the donor) cannot be answered without also considering who receives (i.e., the beneficiary). Further, results indicate that donors and beneficiaries influence giving in interactive ways. It is therefore important to move beyond consideration of donor and/or beneficiary characteristics, to evaluate donor-beneficiary relationship dyads.

Citation Chapman, C. M., Louis, W. R. & Masser, B. M. (2018). Identifying (our) donors: Toward a social psychological understanding of charity selection in Australia. Psychology & Marketing, 1- 10. doi: https://doi.org/10.1002/mar.21150

Contributions The work presented in this chapter was conducted by myself under the guidance of my research supervisors Professors Winnifred Louis and Barbara Masser. The data was collected by a fundraising consultancy called More Srategic. I negotiated access to the data sets, conceived the project and formulated the focal research question, ran all analyses, interpreted the results, and drafted the manuscript. Throughout the process, my advisors checked my analyses, confirmed my interpretations of the findings, and reviewed several drafts of the manuscript.

12 Identifying (Our) Donors: Towards a Social Psychological Understanding of Charity Selection in Australia

Abstract Gender, age, religiosity, and political orientation are often associated with a propensity to give to charity. However, these broad associations mask inconsistencies that are not yet understood. Just as identity plays an important role in shaping consumer choices generally, donors’ identities could explain diverging associations between demographic social categories and the types of charities supported. Two studies, with confirmed workplace giving donors (N = 675) and a community sample of self-reported donors (N = 376) in Australia, provide evidence that associations significantly vary across categories of charity. Specifically: older donors are more likely to support religious and health charities; religious donors are more likely to support religious, welfare, and international organizations but less likely to support animal charities; and politically conservative donors are less likely than liberal donors to donate to international organizations. The findings are interpreted through the lens of identity, with a focus on how group priorities and relevant norms may affect charity selection. Results have implications for non-profit marketing practice, including targeting, channel selection, and framing of fundraising appeals.

Keywords: charitable giving; identity; consumer behavior; preferences; donations.

13 Introduction Non-profit organizations provide services in areas as diverse as education, sports, religion, environment, arts, social services, and health. While research continues to examine donor motivations (e.g., Bekkers & Wiepking, 2011b) and profiles (e.g., Casale & Baumann, 2015), it remains largely silent on the question of donor choices. There are an abundance of non-profits to select from—including more than 55,000 registered in Australia (ACNC, 2018), over 165,000 in the United Kingdom (Keen & Audickas, 2017), and in excess of 1.4 million in the United States (McKeever, 2015). Understanding not just if someone will give but which causes they will support is therefore an essential task for non-profit marketers. However, charity selection has rarely been investigated explicitly and remains largely untheorized. The current paper addresses this gap by presenting two studies with community samples of donors that demonstrate how supporter profiles vary across charity targets. Results are discussed in light of psychological theories of identity and suggest donor identities influence charity selection in systematic ways. Identifying Donors Philanthropic research often uses demographic social categories to identify probable donors (e.g., Casale & Baumann, 2015). Because group membership on the basis of gender, religion, or age is often observable and accessible via secondary sources, these social categories are commonly used by non-profit organizations to segment supporters and target campaigns (Andreasen & Kotler, 2008). Research shows that women are generally more likely to donate to charity than men (e.g., Mesch et al., 2011); people are increasingly likely to give as they age (e.g., M. A. Steinberg, Crow, Cain, & Milford, 2005); and religious people give more to charity (e.g., Brooks, 2003b). These associations, however, may not be universal. Researchers are yet to actively engage with the question of how and why donors choose to support certain charities and neglect others. However, by considering research that examines supporters of specific causes or compares a discrete collection of charity missions, evidence emerges that demographic associations actually vary widely depending on the charity in question. For example, women are more likely to support animal charities than men (Piper & Schnepf, 2008; Srnka, Grohs, & Eckler, 2003), while men are more likely to support sports (Piper & Schnepf, 2008) and political organizations (Showers, Showers, Beggs, & Cox, 2011). Increasing age has been associated with greater support of religious, welfare, and health charities (R. N. James & Sharpe, 2007; Nilsson, Erlandsson, & Vastfjall, 2016; Srnka et al., 2003; Wiepking, 2010), while younger donors show greater support of environmental and animal charities (National Australia Bank; NAB, 2014). Finally, more religious donors show disproportionate support for religious organizations (Forbes & Zampelli, 2013; Helms & Thornton, 2012; Hill & Vaidyanathan, 2011), as well as international and welfare charities (Casale & Baumann, 2015; Wiepking, 2010). This 14 highlights an important direction for research in non-profit marketing. We must ask not just who gives to charity but who gives to which types of charity, and why? Despite the evidence for diverging charity preferences summarized above, researchers have rarely engaged with the question of charity selection directly (see only Breeze, 2013; Wiepking, 2010). As a result charity selection, and the psychological mechanisms that drive it, remains largely untheorized. Identity concerns may help explain donors’ charity choices. Identity and Charity Selection Consumers both communicate desired and avoid undesired identities through product choices and gift giving (Berger & Heath, 2007; K. White & Argo, 2009). Identities thus appear to play a significant role in consumer decision-making. Indeed, marketing researchers have been encouraged to examine the influence of identity on consumer behavior (Aaker & Akutsu, 2009; Oyserman, 2009; Reed, 2002). Psychological theories of identity can shed light on donor decision- making. Social Identity Theory (Tajfel, 1981) proposes that people hold both individual (personal) and group-based (social) identities that are malleable and become activated in different contexts. The social groups that a person belongs to shape their self-definition and prescribe their attitudes and behavior as group members (Turner et al., 1987). When a given identity is salient, perceptions of what other people in one’s group approve of or do (i.e., group norms) have been shown to influence a range of socially desirable behaviors (see, for example, Cialdini et al., 1990). An emerging body of research on charitable giving finds that identities and associated norms affect donor choices such as the decision to donate or the value of gift (Charnysh et al., 2015; Croson et al., 2009; Hysenbelli et al., 2013; Kessler & Milkman, 2018; Zagefka et al., 2013) as well as the effectiveness of charity marketing strategies (Park & Lee, 2015). However, research has not yet explored how identities affect charity selection. Identities that are salient, important to the consumer, relevant to the object being considered, and that assist the consumer to select between product choices are most likely to influence consumer choices (Reed, 2002). When an identity is perceived as relevant in a given context, it can influence the range of responses that are considered appropriate (Turner et al., 1987). This may include decisions about which charities or beneficiaries are prioritized and the kinds of help that are offered. As such, an essential first step for non-profit marketers is to understand which identities are salient in different charitable domains. The Current Research To date, the question of how and why donors choose to support certain charities over others has remained largely untested and untheorized. The literature discussed above suggests that donors who belong to different social categories tend to support different types of charity. From a social 15 identity perspective, donors’ identities would be expected to influence which types of organizations they choose to support through their charitable giving, with different identities being relevant to different types of giving. This article takes a first step towards understanding the role that identity plays in charity selection by testing whether the previously documented relationships between broad social categories (or identities) and charitable giving are consistent across all categories of charity. The primary hypothesis is that the nature of associations between social categories and giving will differ as a function of the type of charity being considered:

H1: Different social categories will predict support for different types of charity. The current research is the first to explicitly test the role of social categories in charity selection. As such, the authors make no specific predictions as to the nature of the associations. However, based on Social Identity Theory (outlined above) the following exploratory hypothesis is proposed:

H2: Associations between social categories (groups) and types of charity will reflect group priorities and values. Study 2.1 Secondary data were analyzed to identify the preferred charities of 675 confirmed donors. Respondents were free to name any charity and were not prompted by either brand or mission, ensuring they had complete liberty to spontaneously name the charities they supported. Named charities were later coded into categories for analysis using the national charity register. It was hypothesized that members of different demographic social categories—based on gender, age group, and religion—would preferentially support different categories of charity (H1), in ways that reflected group priorities (H2). Material and Methods Participants and Procedure In 2015, 6,000 Australians who had donated to charity through a workplace giving program in the previous year were emailed an invitation to participate in market research. Data were subsequently anonymized and shared with the researchers for analysis. In total, 1,159 (i.e., 19% of those approached) voluntarily completed the survey and 821 named the main charity they supported. An additional 146 participants were excluded from analyses due to incomplete data on the focal predictors: three participants did not declare their gender and a significant minority (17%) chose not to answer the question about religious identification.1 The final sample therefore

1Given the exploratory nature of Hypothesis 2, we have chosen to be conservative and exclude incomplete responses from analyses. Mean-replacement is not recommended for two

16 comprised 675 active workplace giving donors, of whom 405 were female and 270 were male. Participants ranged in age from 18 to 74 years, with 12% aged 18-29, 27% aged 30-39, 32% aged 40-49, 29% aged 50-64, and less than 1% aged over 65 years. Participants reported low levels of religious identification (M = 2.03, SD = 1.26, on a 5-point scale) with almost half (48%) indicating that they were not at all religious. Key measures of interest for the current study are outlined below. Measures Participants identified their gender (coded female = 1, male = -1), age bracket (“Which age bracket do you fall into?”, coded under 18 = 1, 18-29 = 2, 30-39 = 3, 40-49 = 4, 50-64 = 5, and 65 to 74 = 6) and religiosity (“Please indicate for each description how like you that person is: A religious person”, 1 = not at all like me, 5 = just like me). Participants were also asked to name their principal charity (“What is the main charity, or most recent, charity you supported through workplace giving?”), using a free response format. Responses were then coded into categories using publically searchable data on the Australian Charities and Not-for-profits Commission website (ACNC, 2016). Each named charity was coded for each category under which the charity had registered and whether or not they served communities overseas (each coded 0 = is not, or 1 = is in this category). Hybrid charities—those with multiple missions and beneficiary groups—were accounted for in the data as charities could nominate multiple categories. More detail about charity categories is included in Appendix B. Results and Discussion The four most frequently mentioned charity categories were selected for analysis: international; health; welfare; and public benevolent institution (PBI; a category that could include, for example, hospices, disability services, and aged care). In addition, religious and animal charities were included because previous research has highlighted these categories (e.g., Breeze, 2013; Hill & Vaidyanathan, 2011). Over half (56%) of participants named a PBI, 37% a health, 31% an international, 30% a welfare, 12% an animal, and 11% a religious charity.2 Means, standard deviations, and zero-order correlations between all variables are presented in Table 1. reasons. First, because of the large minority of participants with missing data (17% of responses), introducing mean-replacement would lower the variability in the dependent measure excessively, producing a potential problem of kurtosis. Second, because respondents tended to either be high or low in religiosity, the average does not represent participants well conceptually. However, when analyses were run using mean-replacement for the missing values of participants who saw the question but elected not to respond, an identitical pattern of results was returned. The only change observed was that the association between gender and giving to welfare charities (Exp(B) = 1.20, p = .046) became non-significant (Exp(B) = 1.16, p = .088). 2 Charities could register under multiple sub-types, resulting in overlap between some categories. Analyses were first run controlling for alternative categories of charities. Inclusion of these

17 Six binary logistic regression analyses (presented in Table 2) were conducted to examine how gender, age, and religiosity were associated with preferring a charity that was registered under the international, health, welfare, religious, PBI, and animal categories respectively. The model significantly predicted respondents’ preference for international (2(3) = 24.70, p < .001, Nagelkerke R2 = .05), welfare (2(3) = 16.21, p = .001, Nagelkerke R2 = .03), religious (2(3) = 26.26, p < .001, Nagelkerke R2 = .08), and animal charities (2(3) = 30.08, p < .001, Nagelkerke R2 = .08). Taken together, however, gender, age, and religion did not help explain preference for either health charities (2(3) = 0.89, p = .829, Nagelkerke R2 < .01) or PBIs (2(3) = 4.76, p = .190, Nagelkerke R2 < .01). Inspection of the coefficients showed that female donors were significantly less likely than male donors to prefer an international charity, Exp(B)3 = 0.84, p = .037, 95% CI [0.71, 0.99], but significantly more likely to prefer an animal charity, Exp(B) = 1.68, p < .001, CI [1.26, 2.24], or welfare charity, Exp(B) = 1.20, p = .046, CI [1.00, 1.42]. Older donors were significantly more likely to prefer a religious charity, Exp(B) = 1.58, p = .001, 95% CI [1.20, 2.07], and significantly less likely to prefer an international charity, Exp(B) = 0.78, p = .003, CI [0.66, 0.92], or animal charity, Exp(B) = 0.64, p = .001, CI [0.50, 0.83]. More religious donors were significantly more likely to name an international charity, Exp(B) = 1.26, p < .001, 95% CI [1.11, 1.43], welfare, Exp(B) = 1.25, p = .001, CI [1.10, 1.42], or religious charity, Exp(B) = 1.42, p < .001, CI [1.18, 1.70], as their preferred charity, and significantly less likely to name an animal charity, Exp(B) = 0.64, p = .001, CI [0.50, 0.83]. This data from a sample of active donors that documents real-world charity preferences provides preliminary evidence that the relationships typically observed between social categories and charitable giving vary across charity targets (supporting H1). However, by restricting the study to workplace giving, analyses omitted retiree-aged donors, a category shown to include some of the most generous donors (M. A. Steinberg et al., 2005). Furthermore, as only 19% of targeted donors completed the survey and a significant portion of the respondents (17%) returned missing data on religiosity, the sample is unlikely to be fully representative of the wider donor population. In addition, the study did not assess political orientation, which is another donor characteristic that has

controls did not substantively change results however, so the direct associations are presented here for simplicity. More detail about categories and observed overlap is provided in the Chapter Appendix at the end of this chapter. 3 The odds ratio, Exp(B), indicates how much the likelihood of the participant naming a charity in the target category increases or decreases for each unit change of the explanatory variable and is scale dependent. For example, an odds ratio of 1.20 indicates a 20% increase while 0.80 indicates a 20% decrease in the odds of naming a charity in the target category. 18 Table 1. Descriptives and Zero-order Correlations Between Demographic Social Categories and Charity Category Preference Variable M SD 1 2 3 4 5 6 7 8

1. Female (-1/1) 0.20 0.98 *** 2. Age (category) 3.79 1.00 -.14 * † 3. Religiosity 2.03 1.26 -.10 .06 † * *** 4. International 0.31 0.46 -.08 -.10 .14 ** 5. Health 0.37 0.48 -.02 .02 -.03 -.10 † ** *** 6. Welfare 0.30 0.46 .07 -.04 .12 .19 -.05 ** *** *** *** 7. Religion 0.11 0.31 .03 .13 .15 .29 .18 .29*** *** *** 8. PBI 0.56 0.50 -.04 .05 .07 .20 -.21 .21*** .22*** 9. Animal 0.12 0.32 .14*** .00 -.14*** -.19*** -.21*** -.20*** -.13** -.32*** Note. †p < .10; *p < .05; **p < .01; ***p < .001 (2-tailed); charity categories are coded 0/1. Listwise N = 675

19 Table 2. Binary Logistic Regressions with Demographic Social Categories Predicting Preference for Different Charity Types International Health Welfare Religion Benevolent Animal Exp(B) Exp(B) Exp(B) Exp(B) Exp(B) Exp(B) Individual Predictors Female (-1/1) .84* .95 1.20* 1.24 .95 1.68*** Age .78** 1.01 .90 1.58** 1.07 1.11 Religiosity 1.26*** .95 1.25** 1.42*** 1.12† .64**

Model Fit Model Chi Square 24.70*** 0.89 16.21** 26.26*** 4.76 30.08*** Cox & Snell R2 .04 .00 .02 .04 .01 .04 Nagelkerke R2 .05 .00 .03 .08 .01 .08 Correct Classification Overall 67% 63% 70% 89% 56% 88% Donor 8% 0% 4% 42% 100% 0% Non-donor 97% 100% 99% 98% 0% 100% † * ** *** Note. p < .10; p < .05; p < .01; p < .001

20 been associated with both religiosity and charitable giving (e.g., Forbes & Zampelli, 2013; Van Lange, Bekkers, Chirumbolo, & Leone, 2012). These limitations are addressed in the second study. Study 2.2 Using a wider community sample of donors, Study 2.2 again investigated how social categories based on gender, age, and religiosity were associated with support for a range of charity categories. Several amendments were made to the design in order to address the limitations of the first study. First, political orientation (the preference for more conservative or more liberal political policies) was included as a potential predictor of support. Religiosity often correlates with conservatism (e.g., Sibley & Bulbulia, 2015; Van Lange et al., 2012) and Brooks (2003b) argued that the relationship between religiosity and giving may be at least partly explained by political identity. Yet studies including political orientation as a predictor of charity have either failed to find significant differences between liberals and conservatives or have produced contradictory results (Mesch et al., 2011; Sibley & Bulbulia, 2015; Van Lange et al., 2012). These inconsistent results suggest that associations between political identity may vary across charity targets. Early evidence supports this contention: conservatives report giving more to religious causes than people with other political affiliations (Forbes & Zampelli, 2013); international charities may receive more support from progressives than conservatives (Nilsson et al., 2016; Wiepking, 2010); and donors are more likely to support charities whose missions align with guiding moral foundations relevant to their political identity (Winterich et al., 2012). Study 2.2 therefore included political orientation as a fourth social category expected to be associated with charity selection. Next, giving outside of the workplace was also considered. This allowed donors to consider charities they support via all channels and ensured that retiree-aged donors were not excluded from the sample. Finally, participants were compensated for their time to reduce the likelihood of returning missing data on key measures, such as religiosity. By addressing the sampling concerns raised in Study 2.1 and adding political orientation as a predictor of charity support, Study 2.2 provided a more rigorous test of the hypothesis that associations between social categories and charitable giving will vary as a function of the category of charity being considered (H1), and in ways that reflect group priorities (H2). Material and Methods Participants and Procedure In 2015, 743 market research panelists from South Australia completed an online survey about charitable giving. Of these, 60% stated that they had made a donation of money to a charity in the last 12 months and were asked to name up to 10 charities they had donated to. Of the 86% of donors who named at least one charity that could be coded, 9 participants (2%) returned missing data on both the religiosity and political orientation measures and were therefore excluded from 21 analyses. The final sample (N = 376) included 177 males and 199 females, ranging in age from 18 to over 65 years (17% aged 18-29, 13% aged 30-39, 14% aged 40-49, 34% aged 50-64, and 23% aged over 65 years). On average, participants were low in religiosity (M = 2.10, SD = 1.22) with 44% indicating they were not at all religious. Regarding political orientation, 53% indicated they were politically more progressive and 47% indicated they were more conservative. Participants were recruited though a market research panel and completed a 15-minute online survey in exchange for payment of approximately AUD$5. Data was subsequently anonymized and shared with the researchers for analysis. Key measures of interest for the current study are outlined below. Measures Gender, age, and religiosity were measured as per Study 2.1. Political orientation was assessed with a single forced-choice item (Politically I am more to the left or Politically I am more to the right, coded conservative/right = 1 or progressive/left = -1). Participants were asked to name up to 10 charities that they had supported to in the last 12 months (“Which charities have you donated to in the past 12 months? Please list all those you have made monetary donations to including any direct debits or credit card donations”), using a free response format. The coding procedure was identical to that employed for Study 2.1. Results and Discussion Participants reported donating to an average of 2.5 charities (SD = 1.99) in the previous 12 months. Means, standard deviations, and zero-order correlations between all variables are presented in Table 3. Six binary logistic regression analyses (presented in Table 4) were conducted to examine whether gender, age, religious, and political social categories were differentially associated with support of different types of non-profits.4 The model significantly predicted support for international (2(4) = 15.84, p = .003, Nagelkerke R2 = .06), religious (2(4) = 19.32, p = .001, Nagelkerke R2 = .07), and animal charities (2(4) = 9.74, p = .045, Nagelkerke R2 = .04), but only marginally predicted support for health organizations (2(4) = 9.20, p = .057, Nagelkerke R2 = .03). Taken together, gender, age, religion, and political categories did not significantly explain support of welfare charities (2(4) = 5.81, p = .210, Nagelkerke R2 = .02) or PBIs (2(4) = 4.55, p = .336, Nagelkerke R2 = .02). Inspection of the coefficients showed that no unique associations were observed between gender and support of any category of charity. Older donors were significantly more likely to support health charities, Exp(B) = 1.23, p = .008, 95% CI [1.05, 1.42], and also showed a trend

4 As in Study 1, controlling for support of other categories of charity did not affect the pattern of results and therefore the direct associations are reported for simplicity. 22 Table 3. Descriptives and Zero-order Correlations Between Demographic Social Categories and Charity Categories Supported Variable M SD 1 2 3 4 5 6 7 8 9

1. Female (-1/1) 0.06 1.00 ** 2. Age (category) 4.32 1.39 -.14

3. Religiosity 2.10 1.22 -.05 .05

4. Conservative (-1/1) -0.05 1.00 -.02 .05 .08 ** * 5. International 0.49 0.50 -.06 .02 .16 -.11 * 6. Health 0.52 0.50 .04 .13 -.07 .01 -.06 * *** 7. Social Welfare 0.26 0.44 -.10 .05 .12 .01 .31 .07 * *** 8. Religious 0.39 0.49 -.06 .11 .19 -.04 .16** -.20*** .11* 9. PBI 0.85 0.36 -.08 .08 .04 .02 .35*** .00 .16** .27*** 10. Animal 0.18 0.39 .20 .00 -.15** -.02 -.05 -.14** -.10* -.07 -.21*** Note. *p < .05; **p < .01; ***p < .001 (2-tailed); charity categories are coded 0/1. Listwise N = 376

23 Table 4. Binary Logistic Regressions with Demographic Social Categories Predicting Support of Different Charity Types International Health Welfare Religion Benevolent Animal Exp(B) Exp(B) Exp(B) Exp(B) Exp(B) Exp(B) Individual Predictors Female (-1/1) .90 1.11 1.00 .91 .82 1.04 Age 1.02 1.23** 1.08 1.16† 1.13 1.02 Religiosity 1.32** .89 1.23* 1.38*** 1.08 .69** Political Orientation .78* 1.03 .99 .88 1.05 .97

Model Fit Model Chi Square 15.84** 9.19† 5.81 19.32** 4.55 9.74* Cox & Snell R2 .04 .02 .02 .05 .01 .03 Nagelkerke R2 .06 .03 .02 .07 .02 .04 Correct Classification Overall 58% 59% 75% 64% 85% 82% Supporter 58% 73% 0% 23% 100% 0% Non-supporter 58% 44% 100% 90% 0% 100% † * ** *** Note. p < .10; p < .05; p < .01; p < .001

24 towards supporting religious charities, Exp(B) = 1.16, p = .064, CI [0.99, 1.36]. More religious donors were more likely to support international charities, Exp(B) = 1.32, p = .002, 95% CI [1.11, 1.57], welfare, Exp(B) = 1.23, p = .028, CI [1.02, 1.48], and religious charities, Exp(B) = 1.38, p < .001, CI [1.16, 1.64], but significantly less likely to support animal charities, Exp(B) = 0.69, p = .004, CI [0.53, 0.89]. Conservative donors were significantly less likely than liberal donors to support international charities, Exp(B) = 0.78, p = .017, 95% CI [0.63, 0.96]. A survey of the categories of charity supported by a wide community sample of donors replicated the key finding of Study 2.1: that donors’ social group memberships are associated more strongly with supporting some types of charity than others. Using a paid panel to recruit participants, no substantial missing data was returned on religion, and all significant results from Study 2.1 pertaining to religion were replicated. General Discussion

Two studies of charity preferences among active donors supported the hypothesis (H1) that charitable giving is not a universal response. Instead, different identities are associated with supporting different types of charities. Thus, who gives to charity depends in part on the type of charity in question: a point that has not been systematically tested before now. Some consistent patterns emerged across samples of both workplace giving (Study 2.1) and self-reported (Study 2.2) donors: older donors are more likely to support religious organizations; more religious donors are more likely to support international, welfare, and religious causes, but less likely to support animal charities; and politically conservative donors are less likely to support international charities. These patterns support the notion that donors give in ways that reflect the priorities of groups they belong to (H2), which is elaborated below. However, several inconsistencies also emerged across the results of these two studies, highlighting the importance of understanding how identity and other socio-contextual factors affect donor decision-making. Large-scale category memberships such as those employed here— specifically gender, age, religion, and political orientation—are especially important sources of social identity (Turner et al., 1987), and previous research suggests people give to causes that are relevant to their priorities and values (Bekkers & Wiepking, 2011b; Sargeant, 1999; Srnka et al., 2003). Indeed, we put forward the argument that charitable giving could be understood as a group process, with charity selection influenced by the priorities and norms related to salient donor identities. This identity-based approach helps us to understand the emergent patterns in charity support, as articulated below. Older donors were more likely to support religious charities (corroborating R. N. James & Sharpe, 2007; Wiepking, 2010). Older people are more likely to participate in religious activities, independent of their religiosity per se (Sargeant, 1999). It is therefore proposed that they give more

25 to religious causes for three inter-related reasons: higher exposure to and solicitation by religious causes; more religious social networks; and supportive social norms. In an extended community sample (Study 2.2) older donors were also found to be more likely to support health charities (consistent with Nilsson et al., 2016; Srnka et al., 2003). As health problems increase with age, this association may reflect older donors’ donations to organizations that their (age) group members are most likely to benefit from, or simply memorial giving as friends and family pass away. Thus, older donors may give to religious charities because of the way their social groups are integrated with religious activities. Futher, they may give to health charities because they see their own social group as directly benefiting from such gifts. These propositions remain to be tested in future research. In both studies, religiosity was associated with an increased likelihood of supporting religious, welfare, and international organizations (supporting Casale & Baumann, 2015; Wiepking, 2010), and reduced likelihood of supporting animal charities. To the authors’ knowledge the latter relationship has not been previously reported. Support of religious charities can be most clearly understood in terms of identity: more religious donors contribute to causes that directly spread the values associated with their group. The other associations can also be understood in terms of group norms and values. Given Australian national demographics, the religious respondents were likely to be primarily Christian. Dominant beliefs of Christianity help explain religious respondents’ greater support of charities that serve the vulnerable and needy, and tendency to prioritize human beneficiaries over animals (McLaughlin, 2014). As a limitation, however, the measure of self- report religiosity employed did not identify specific faith backgrounds. Further, Australia is relatively less religious than many other countries. According to the latest census data, 30% of Australians identify with no religion (ABS, 2017). In comparison, just 15% of Americans claim no religious ties (US Census Bureau, 2015). While the results obtained here are theorized to be driven by identity concerns that should cross borders, it is possible that results will not translate to all contexts. Future research should investigate these associations in other national contexts and actively consider the potential role of specific religious affiliations. Donors who identified with a conservative political orientation were less likely to support international charities (supporting Nilsson et al., 2016; Wiepking, 2010). Potentially this can be explained by values and tendencies more likely to be shared by people who identify as conservative—such as stronger nationalism (van Der Toorn, Nail, Liviatan, & Jost, 2014), more prejudice against outgroups (e.g., Webster, Burns, Pickering, & Saucier, 2014), or justification of inequality (Jost, Glaser, Kruglanski, & Sulloway, 2003). Results therefore suggest that donors choose charities that reflect the worldviews that are priorities of their political identities. A more

26 granular examination of different types of international charity, however, would allow marketers to identify any types of international aid that conservatives would support. Finally, mixed support was found for diverging gender associations. In Study 2.1, men were more likely than women to nominate an international charity and less likely to nominate an animal or welfare organization as their preferred. However, these associations were not replicated in Study 2.2, where no significant associations were observed between gender and the charity categories supported. It should first be noted that those categories of charity previously found to be supported more by males (e.g. sports, politics; see Piper & Schnepf, 2008; Showers et al., 2011) are not included as sub-types in the Australian charity register and, therefore, could not be analyzed. Nonetheless, gender differences in giving may be more likely for confirmed donors (Study 2.1) than self-reported donors (Study 2.2); or may play out more strongly in selection of preferred charity (Study 2.1) than all charities supported (Study 2.2). Alternatively, gender may simply not be a very salient identity in the context of charitable giving, resulting in weaker and more inconsistent gender effects. Contributions to Theory and Practice Overall, the results highlight the importance of understanding the identities that inform consumer choices in the charitable domain. The evidence presented here suggests that identities do matter and that charitable choices may reflect the priorities of relevant social identities. As such, this paper has highlighted a missing piece of the donor psychology puzzle: the role identity plays not just in motivating overall charitable giving, but in delineating which charities an individual chooses to support. The evidence presented here—a systematic comparison of identities in relation to charity selection—shows that identities related to age, religion, and political affliations appear to structure giving in meaningful ways. Gender, however, may not be a meaningful identity in charitable contexts. Future research is needed to determine which other identities matter and in which charitable domains. This new knowledge about the role identities play in charity selection can aid non-profit marketers in three key ways. First, identity research will help fundraisers understand which identities a particular type of charity should make salient in campaign materials to uplift response rates. Second, marketers can prioritize identity-relevant channels based on identities core to a particular charity’s donor base. Finally, once better understood, identities and associated norms and values can be used to help charities frame fundraising appeals to ensure they resonate with donor priorities. Strengths, Limitations, and Future Directions The current studies are unique in that they surveyed active donors and asked them to name their favorite charities without prompt. Each charity was coded using objective data available in the 27 local charity register, in which charities self-nominate their charitable purposes. This data allowed for overlap between charity missions, a strength given that non-profits are becoming increasingly hybrid (Hasenfeld & Gidron, 2005). Demographic categories were used to understand associations between social categories and charitable choices. Demographics provide a practical method to segment donors and therefore have applied value for non-profit marketers. Using social categories as a stand-in for identity is nonetheless problematic because identity is complex, subjective, and relies primarily on an individual’s self-categorization as a group member (Turner et al., 1987). Future research will benefit from employing measures designed to capture both degrees of identification (how important the group is for the individual) and the norms attached to those groups (how much the group is perceived to support this type of charity), in order to test the identity processes proposed here. Specifically, understanding both explicit (self-reported) and implicit (behavioral) relationships between identities and charity targets will be important. Qualitative research can be employed to unravel the complex ways that donor identities may interact with beneficiary identities to motivate charity responses. Experimental work is also needed to evaluate the potential causality of the associations observed here. Finally, building a richer landscape of the charitable sector, which includes perceived normative targets of giving attached to different social identities will be essential to moving research on identity in non-profit marketing forward. Conclusions This paper makes an important contribution to the psychological understanding of donor behavior, by presenting clear evidence from two community samples of donors showing how associations between social categories and charitable giving vary as a function of the target charity considered. This suggests that different identities motivate support of different types of charity: an intuitive point yet one that has rarely been emphasized in the literature. The research contributes to a growing inter-disciplinary movement seeking to empirically test how social concerns affect consumer behavior in the charitable domain. The data highlight that charitable giving is not only an individual tendency, but a social response prompted by an interaction between aspects of the donor and the beneficiary. These findings have implications for fundraising practice, especially donor recruitment and appeal framing.

28

Chapter 3. Self and Other Orientations in Charitable Giving: Thematic Analyses from a Global Donor Survey

This chapter applies a social identity lens to the question of donor motivation. Building off results of Chapter 2—which showed that donors’ social categories inform their charity preferences in meaningful ways—for this chapter I surveyed a diverse group of donors from around the world and asked them why their favorite charity is important to them. The analyses consider which identities are most relevant to charitable giving and how the dynamics between particular donors and particular beneficiaries influence giving decisions. Results provide more evidence that donors have different motives and responses for different kinds of beneficiaries, further highlighting the importance of studying donor-beneficiary dyads.

Citation Chapman, C. M., Masser, B. M., & Louis, W. R. (under review). Self and other orientations in charitable giving: Thematic analyses from a global donor survey. Manuscript submitted for review to the European Journal of Social Psychology, 6 November 2018.

Contributions The work presented in this chapter was developed and written by myself under the guidance of my research supervisors Professors Winnifred Louis and Barbara Masser. Specifically, I designed the survey, identified the research question and methodology, conducted all analyses, interpreted the findings, and wrote the manuscript. My advisors provided feedback on the draft survey, reviewed my coding after 10% was completed, and critically reviewed several drafts of the manuscript.

29

Self and other orientations in charitable giving: Thematic analyses from a global donor survey

Abstract Multiple motives for charitable giving exist, but there is no clarity about when and why different motives dominate. The social identity approach argues that ingroup-outgroup (i.e., self- other) distinctions help us navigate the social world, and identities also influence giving. A deductive thematic analysis of why participants see their favorite charity as important (N=1,849 from 117 countries) shows how people use their own and others’ identities to explain their charitable preferences. Nine sub-themes were generated under two overarching themes: Self and Other. Donors were more likely to explain giving to religious and research charities in relation to the self, but giving to social service, animal, or international charities in relation to the other. We show giving is a social process that depends on the particular intersection between donor and beneficiary. Results highlight which of multiple identities structure giving and suggest certain beneficiaries more often attract outgroup donations.

Keywords: charitable giving, identity, charity preference, donor motivation, social identity.

30

Introduction Why do people give to charity? This question has been systematically studied by researchers across the disciplines of philanthropy, economics, and social psychology. A range of motives have been identified, including altruism, reputation, values, identities, and strategic group interests (e.g., Aaker & Akutsu, 2009; Bekkers & Wiepking, 2011b; Nadler, 2002). From a social identity perspective (Tajfel, 1981; Turner et al., 1987), self-other (or ingroup-outgroup) distinctions emerge naturally in social decision-making, and a core process motivating charitable giving is that individuals give to promote the priorities of their social groups (e.g., Chapman, Louis, et al., 2018). However, little is known about which of donors’ multiple identities are most salient in charitable giving and why. The current research applies a social identity lens to interpret the way people explain the importance of certain charities over others. Although various motives have been documented, to our knowledge no research has yet attempted to map various motives across large-scale samples of donors and target charities. Such a mapping has applied implications for fundraisers, and also theoretical implications for better understanding the content of identities and the salience of particular identities over others in this important domain of social life. Below we review the existing literature on motives for charitable giving, argue that giving can be understood as a social process using the social identity perspective to explain self-other considerations, and then present the results of thematic analyses from a global survey of donors explaining why certain charities are important to them. In particular, we seek to understand how donors use conceptions of self and other to make sense of their giving choices. Results highlight the motives and identities that are most relevant for giving to different categories of charity or beneficiaries, advancing theory by considering how multiple identities influence social action. Our findings also have important implications for how charitable organizations communicate with donors and promote their causes. Multiple Motives for Giving Charitable giving refers to the act of donating money to organizations that benefit people beyond the donor’s own family (Bekkers & Wiepking, 2011b). We might therefore expect that charity is motivated primarily by an altruistic concern for some other (i.e., the beneficiary). Yet charitable giving may also accrue benefits to the self (i.e., the donor). A range of other-oriented motives have been highlighted in the literature, which explain giving as motivated by concerns for beneficiaries. For example, awareness of need (knowing donations are needed), altruism (concern about the wellbeing of the less fortunate), efficacy (whether or not the donation will make a difference), and empathy (the ability to understand what other people are thinking or feeling) have all been highlighted as common other-oriented motives for giving (Bekkers & Wiepking, 2011b; Konrath & Handy, 2017; Sargeant & Woodliffe, 2007b). 31

Many motives for giving also appear to be self-oriented, driven by the interests of the donor rather than the beneficiary. For example, donors may give for egoistic reasons (seeking to gain reputation or praise), to gain material or psychological benefits (e.g. tax rebates or a good feeling), or to enact important values in the world (Bekkers & Wiepking, 2011b; Konrath & Handy, 2017; Sargeant & Woodliffe, 2007b). While these motives stress the interests of the individual donor, people may also give in order to promote the strategic interests of their important social groups. For example, donors sometimes give to restore a threatened group identity, to make their own group look good in front of others, or to demonstrate the power of their group over beneficiaries (Nadler, 2002, 2016; Nadler & Chernyak-Hai, 2014; Nadler et al., 2009; van Leeuwen, 2007, 2017; van Leeuwen & Harinck, 2016). We propose a self/other dichotomy as an overarching framework for organising these diverse giving motives: some motives are oriented more towards the personal or social self (e.g., seeking benefits or reputational rewards, enacting values, or demonstrating positive group attributes), while other motives are oriented more towards some other (e.g. altruistic or empathic concern, awareness of need, or efficacy). Such a dichotomy illustrates how charitable giving is a group process and suggests that identities may structure giving in meaningful ways. Self-Oriented Versus Other-Oriented Giving Charitable giving is inherently a social process. At a minimum, any gift occurs in intersection between a donor and a beneficiary. Different combinations of donors and beneficiaries, therefore, may promote giving through different mechanisms. As such, understanding both the self (donor) and the other (beneficiary) are important when seeking to understand whether or not a gift will be made, and why or why not. The social identity approach (Tajfel, 1981; Turner et al., 1987) theorizes that humans use a process of categorization to navigate the complexity of the social world. This leads us to think of both self and other in terms of groups. Individuals have a personal identity (sense of I), yet the groups they belong to can also be seen as extensions of the self, creating social identities (sense of we). Self-other distinctions (or, using social identity terminology, ingroup-outgroup distinctions) are a normal and natural by-product of social categorization. When targets are categorized as ‘other’ they are seen as more interchangeable with one another, perceived less favorably, and are less likely to be helped than ingroup targets (Tajfel, 1981; Turner et al., 1987). It therefore follows that donors may be more willing to give when they perceive the beneficiary to be similar to them, particularly when a shared identity is salient. Donors’ identities do affect charitable giving (Aaker & Akutsu, 2009). For example, donors may strategically use their national identity to articulate why they do or do not help certain groups (Stevenson & Manning, 2010), and priming donor and community identities in fundraising appeals 32 can increase charitable gifts (Kessler & Milkman, 2018). Identifying with a charity can lead peer- to-peer fundraisers to put in more effort to raise money for the cause (Chapman, Masser, & Louis, 2018), and identifying with the beneficiaries via some shared identity increases the likelihood that help will be offered (Charnysh et al., 2015; T. K. James & Zagefka, 2017; M. Levine et al., 2005). These identity motives also emphasize that shared and relational identities that link the self (donor) to the other (beneficiary) are motivating for donors. Despite this strong evidence of ingroup favoritism in charity allocations, donors still often give to beneficiaries that fall outside the traditional boundaries of their groups. This includes obvious outgroup helping, such as donations to international appeals and support of animal welfare. But it also includes donations to outgroup members who nonetheless fall within the superordinate borders of one’s nation or species. As we have seen, people may help others for either altruistic or strategic concerns. However, it remains unclear when people are most likely to help beneficiaries from other groups. The Current Research The current research seeks a deeper understanding of how donors use concepts of self and other to construct their relationships with charities. A variety of reasons for charitable giving have been identified in the literature. To date, however, research has not theorized or tested the contexts in which self- or other-oriented motives, or motives based on shared identities, may emerge. Exploratory in nature, but informed by the social identity approach (Tajfel, 1981; Turner et al., 1987), the present research asks how donors use conceptions of their personal and social identities as well as beneficiary identities to explain their giving to the charities that are most important to them. Though important to understand, charity preferences have rarely been studied explicitly (but see Breeze, 2013; Chapman, Louis, et al., 2018; Neumayr & Handy, 2017; Wiepking, 2010). The research therefore also seeks to understand how self- and other-oriented explanations may vary across different types of charities. This study is the first to use a large-scale global sample to examine the self-reported motives for different kinds of charitable giving. While the sample is one of convenience, as elaborated in the discussion, the present qualitative analysis brings a richness and depth in allowing diverse donors to explain in their own words why they support their favorite charity. The research is also theoretically innovative in three key ways. First, informed by the social identity approach, it considers giving motives under a unifying structure of self- vs. other-orientations. Second, it considers which of multiple donor and beneficiary identities structure giving and when. Third, it considers the characteristics of the beneficiary that increase the likelihood that help will be received from outgroup donors.

33

Method Procedure Individuals took part in an online study in exchange for partial course credit in an introduction to psychology massive open online course. These free online courses tend to attract participants who are older than traditional University students, from diverse corners of the globe, and already employed but taking the course for curiosity or career development (Christensen et al., 2013). Participation was voluntary as students could select from a range of studies or choose to participate in none at all. During a 30-minute survey they answered questions about their own charitable giving. Participants first named the charity that was most important to them personally (“Now, thinking just about the charity or not-for-profit organization that is most important to you personally [whether you have donated to them recently or not], please answer the following questions: What is the name of this organization?”). They then indicated the charity category (“Which category best describes the work of [nominated charity]?”), with 15 response options based on the United Nations non-profit reporting guidelines (UN Statistics United Nations Statistics Division, 2003). Finally, participants responded to an open-ended question asking why that charity was important to them (“In your own words, why is [nominated charity] important to you?”). Demographic questions regarding gender, age, and nationality were asked at the end of the survey and were optional. On completion, participants were fully debriefed as to the purpose and hypotheses of the study. Participants In total 1,887 participants had valid responses to the focal question of why their nominated charity was important to them. Nine participants asked for their data to be deleted after debriefing, 12 were excluded because their response was not written in English, and 17 duplicates were removed. Of the final pool of 1,849 participants, 968 identified as female, 419 as male, 9 as other, and 453 did not indicate their gender. Respondents ranged in age from 10 to 78 years (M = 32.21, SD = 14.21) and represented 117 different nationalities. The most common nationalities were American (n = 251), Australian, (n = 196), and Indian (n = 99). Further, 63% of respondents indicated that they had personally made monetary donations to charity in the previous 12 months. Analysis Survey data were imported into NVivo 11 and thematic analyses were conducted on responses to the question: “In your own words, why is [nominated charity] important to you?”. Responses ranged from 1 to 181 words in length, and most were relatively short (M = 23.32 words, SD = 20.95). To explore the ways that participants used self and other orientations, we conducted a deductive analysis, informed by the social identity approach (Tajfel, 1981; Turner et al., 1987). Although the Self and Other themes were explored in the data on theoretical grounds, sub-themes 34 were generated inductively. This method followed the guidelines for thematic analysis in psychology put forward by Braun and Clarke (2006). Initial coding was conducted blind to the charity category to reduce expectation bias. The first author completed all analyses, with two co- authors (also blind to category) reviewing codes after approximately 10% of the data (n = 200) had been coded. Results Nine sub-themes, nested under the overarching themes of Self and Other, were generated by the analyses. These themes are presented in Table 5. Below we summarize each theme and sub- theme and provide examples. Where examples are given, the participants’ full response is included for context. Responses are verbatim, with spelling and grammatical issues retained. Substantial overlap between sub-themes is present in the data, with some responses tapping multiple sub- themes. Where necessary for clarity, the relevant coded portion is italicized. Attribution is given with the participant’s nationality, gender, age, and the category of their nominated charity. Self-Oriented Giving Motives We first explored how donors construct some charities in relation to their sense of self. Informed by the social identity approach (Tajfel, 1981; Turner et al., 1987), the self is conceptualized to include both personal identities (“I”) and social identities (“we”). Whether talking about social identities, values and beliefs, benefits, suffering, or shared identity, the sub- themes outlined below describe how donor motives for giving are sometimes self-oriented. Social identities. The most common thread within the ‘self’ theme was that participants simply explicitly identified one of their own social identities in explaining their connection with the charity. For example:

I, myself, am a veteran and have had friends who have been wounded in Afghanistan [emphasis added] (American male, 26; advocacy).

As I am diabetic, the sooner I get a cure the more help I can be and people can be to working on other charities [emphasis added] (Canadian male, 31; research).

However, social identities were more often implied. For example, a participant talking about prayer or God is highlighting the importance of their religious identity to their giving preference:

Guides me spiritually. Gives me a feeling of connection to God (nationality unknown, Female, 29; religious).

35

Table 5. Description and prevalence of themes and sub-themes generated from the analysis of donors use of self- and other-oriented explanations of their charity preferences Theme Description Frequency Sub-theme n % Self Charity constructed in relation to personal ("I") or social ("we") identities 834 45% Social Identities Donor evoked their social identities in explaining the importance of the charity 600 32% Values and Beliefs The charity's mission aligned with the donor or their group's values or beliefs 226 12% Benefits Donors or their groups have benefited from the charity's services 159 9% Suffering Donors or people in their groups have suffered from something the charity addresses 156 8% Shared Identity Donors share an identity with the beneficiaries 148 8% Other Charity constructed in relation to beneficiary ("they") identities 1,094 59% Beneficiary Identities Donor evoked beneficiary identities in explaining the importance of the charity 991 54% Power Donors perceive the beneficiaries to be low in power or unable to care for themselves 143 8% Value Donors perceive the beneficiaries to be valuable and important 130 7% Neediness Donors give in response to perceived need 108 6% Note. N = 1,849. Participants responses have been coded into any theme they evoked, thus themes overlap.

36

Alternatively, mentioning ‘our country’ or the name of their nation suggests that national identity is relevant to the giving decision:

Because it fights for the people's civil rights in America, which, under Trump, is crucial right now [emphasis added] (American male, 76; advocacy).

Family identities were frequently evoked. Participants either mentioned their family explicitly, or evoked family identity through relationship labels like mother, son, sister, or grandfather. For example,

It is a army welfare group . I always stand with my army because they are the one who are there at border because of which my family is safe [emphasis added] (Indian male, 19; social services).

In addition to religious, geographic, and family identities, diverse other identities were evoked, including organizational, friendship, human, professional, alumni, and identities related to illness or suffering (see Table 6 for summary of the social identities mentioned by participants). Values and beliefs. Another way that participants explained their giving choices in relation to personal and social selves was to talk about values and beliefs. This sub-theme constructed important charities as speaking for the donor in the world—promoting what the donor and their group believed in or valued. In many of these cases, the focus was on the mission of the charity rather than its beneficiaries. That mission was seen as aligned with the donor’s priorities. Sometimes the connections were very explicit. For example:

I value animal rights (German female, 25; animal).

It corresponds with my ideology of life (Pakistani female, 22; social services).

Because they spread the word o God around the world (Ecuadorian female, 24; religious).

At other times, the values and beliefs were suggested by the passion and urgency in the donors’ responses. For example,

The stigma of mental illness has gone on long enough and I for one am tired of it. Bringing awareness to the very real disorders of the mind is of the utmost importance in finally gaining a solid treatment and someday a cure for these disorders. Until people understand that mental illness is not something that is chosen or can just happen after a bad day, then it will go on being shadowed by

37

Table 6. Types and frequencies of social identities spontaneously evoked by participants in relation to their most important charity Social Identity Description Example Mentions Family Family identity evoked directly or Because I lost both my Mum and my Aunt to Cancer. I have also watched my 152 indirectly through mention of Step-Mum and her Mum both battle various types of Cancer over the years. family roles The more I can support them in anyway the quicker we can be to finding a cure or more successful treatments.

Geographic National, regional, or other Because in Mexico there are no animal rights, or the law is just ignored, and 139 geographic identities evoked we are living in a extremely violent environment, specifically in Mexico, so, this violence is translated to any other part of our life. Organization Participant identifies directly with I am part of [Charity] board and I really like the friendly atmosphere and 90 the organization in question inclusiveness. Religious Religious identity is specifically Because I am a Seventh-Day Adventist and [Charity] is an SDA based 87 mentioned or evoked indirectly organization. Friends Participant mentions friendship I, myself, am a veteran and have had friends who have been wounded in 52 groups or acquaintances Afghanistan. Human Participant evokes their identity as Because I believe this charity does a lot of work in supporting our wild 47 a human animals, nature and the world we live in. We often forget that there are other things just as important if not more than our human lives and it's good to get involved to keep yourself grounded and remember what is important.

Health Participant mentions identities I have been suffering from an eating disorder for the past 4 years. [Charity] 28 based on physical or mental health helps sufferers like me. They also raise awareness for eating disorders which issues they suffer from I find hugely important as in my experience we as a society dont really understand eating disorders and this creates a lot of problems for sufferers.

Professional Work or professional identities As a medical practitioner, it is the charity that directly affwcts my day to day 22 evoked activities due to the far reaching support they offer to the Kenyan Health System.

38

Alumni Alumni or student identities My alma mater and an incredible staff. 11 evoked Economic Participant evoked current or past I do not like to see children suffer, who are not able to afford their own food 7 identities related to class or and clothes. I grew up living in poverty and it was a really difficult point in economic standing my life. Pet Owner Participant mentions their identity I think how we treat animals reflects our humanity. Please I have a dog from 8 as a pet owner [Charity] and the thought of her without a family breaks my heart. Disabled Participant evokes their identity as because they are major advocates for my independence as a disabled adult 6 a disabled person Survivor Participant identities as a survivor I'm a survivor of rape. 5 LGBT Participant mentions their gender- The organization is on the forefront of educating the public and fighting for 4 or sexual-orientation identity equality for the LGBTQ community. As a member of that community, the work they do has an impact on my life directly. Orphan Participant identities as an orphan Because I was once a foster child and I understand the experience that kids 4 or former foster child entering the system have to go through. Gender Participant evokes their gender [Charity] is important to me because, they help a lot of women like myself to 3 identity obtain affordable medical help when needed and if any diagnosed with cancer, they do all in their power to help as quickly and efficiently as possible. Age Age group identities evoked I am a senior. 3 Racial Racial identities evoked AS AN AFRICAN AMERICAN I THINK ASSOCIATION WITH 3 [CHARITY] IS AN OPPORTUNITY ADD TO BLACK PRESPECTIVES Vegetarian Vegetarian or vegan identities i became a vegetarian on my way to becoming vegan because i dont want to 2 evoked contribute to animal cruelty anymore Total 673 Note. Relevant coded sections have been italicized for clarity.

39

what are deemed "real" illnesses only because one can see them (English/American male, 37; health).

Benefits. Participants also described giving to charity because they had benefited from the services of the charity, or expected to benefit in the future. Sometimes the benefits had accrued to them personally:

It helped me get clean and into a different environment (American female, 31; legal).

I have mental health issues and [Charity Name] have helped me in various ways (British female, 22; health).

It meets my spiritual needs, some of my social needs, and gives me some focus in my retirement (Canadian male, 67; religious charity).

Other times the benefits fell to others within groups that were important to them, serving their social, rather than personal identities. For example,

[Charity Name] is important to me as my youngest sister was diagnosed with autism when she was very young and being able to use this charities resources has been very helpful to our family and in helping my sister grow and develop [emphasis added] (Australian female, 19; social services).

Suffering. Participants sometimes described charities as being important because of their focus on something that the donor had suffered from personally:

[Charity Name] is important to me because I had a child born prematurely (American female, 34; research).

It deals with a medical condition I have had all my life (American female, 56; research).

Sometimes participants’ responses indicated that they cared about charities that addressed something that their groups, or important people within them, had suffered from. For example,

My family has a history of cancer and we've already lost two members due to nose cancer and brain cancer. Donating to Cancer Research helps fund the research for this illness [emphasis added] (Malaysian female, 25; research).

40

I have a daughter with disabilities (Jamaican female, 42; social services).

Shared identity. A final sub-theme existed at the intersection of self and other, where beneficiaries were drawn towards donors through a shared sense of self. Respondents mentioned that they shared an identity with the beneficiary in question, and saw commonality between one of their social identities and the person in need. The most frequent shared identities were geographic, gender, or as sufferer of some form of illness. Donors sometimes explicitly mentioned the commonality:

[Charity Name] is important to me because, the kids there are very poor. Having went through poverty myself, I felt that I am somehow connected with those Children (Rwandan female, 23; social services).

[Charity Name] deals with issues that has a direct impact on my color, class, and culture of people (American male, 38; advocacy).

Other times shared identity was implied by the connection between the beneficiary’s identity and the donor’s demographics. For example, women helping women:

[Charity Name] is important to me because it gives women access to safe healthcare (American female, 35; health).

Given that the connection was not always explicitly made, it is possible that this sub-theme is even more prevalent in donor motivation than captured by the coding. A number of references to beneficiary groups may implicitly have named a shared identity. Nonetheless, a significant number of participants actively discussed shared identity as a motivating factor in their giving. Other-Oriented Giving Motives We next explored how participants described charities in relation to some beneficiary, or ‘other’. Relating as they do to beneficiary identities, power, value, and neediness, the sub-themes described below demonstrate the ways in which participants construct their motives for giving in relation to the identities of beneficiaries. Beneficiary identities. The most prevalent sub-theme pertaining to the other was when participants simply evoked or named the beneficiary group. Whether the beneficiaries were, for example, children, animals, the poor, the sick, or people overseas, there was usually little comment made beyond simply naming them:

41

It helps take care of orphans (Ugandan female, 37; social services).

It helps refuges around the world (Palestinian male, 33; international).

To potect the environment around the world (Thai male, 42; environment).

It supports people with cystic fibrosis (Canadian female, 29; health).

Protection of animals (American female, 62; animal).

Give food and cloth to the poor (Venezuelan male, 36; social services).

The simplicity of such responses implies that the respondents perceived inherent value in helping such beneficiaries. Donors seemed to feel that some beneficiaries—most notably children and animals—are inherently valuable and worthy of care. Thus, while value was only occasionally mentioned explicitly (see value section below), it seemed to be a core feature of other-oriented motives, whereby participants did not feel the need to explicitly articulate any purpose or value beyond merely naming the beneficiary group. Further, it is illustrative to note not only the beneficiaries most likely to be mentioned—children, animals, poor people, people in other countries, or people who were sick and dying—but also to consider which types of beneficiaries are rarely mentioned, including addicts, offenders, and people in the LGBT community (see Table 7 for summary of all beneficiary groups mentioned). Power. Participants sometimes talked about helping because the beneficiaries lacked power. They explicitly indicated that those ‘others’ are vulnerable, helpless, disadvantaged, or unable to look after themselves:

They are helpless, unable to speak (Italian Polish, 49; animal).

Charity for children is important as children do not understand why is their situation different than others. They do not know how to tackle their problems (Polish female, 26; health).

Value. Many participants supported charities that they perceived to be helping beneficiaries that the donor explicitly valued or felt were important. There was a sense that certain beneficiaries were important and deserved good things, and that this could be freely asserted:

42

Table 7. Categories of beneficiaries mentioned by respondents when explaining why they support their most important charity Beneficiaries Example Frequency Children and Youth [Charity] help disadvantaged children to get the best out of education, by 437 supporting them financially. Animals They help protect animals - mainly cats and dogs through Thailand and other 187 parts of Asia. They are also strong advocates for stopping the cat and dog meat trade. I strongly agree with what they stand for and think they're a great company. Poor People raise money for and help the poor. 126 People in Other Countries They have been helping people in need around the world especially those in the 123 third world. No matter their religion, race, gender etc. They are very selfless. People Who Are Sick or Dying [Charity] is important to me because they are helping fight cancer and improving 103 the lives of those with cancer. Women and Girls [Charity] is important to me because it gives women access to safe healthcare. 59 Families IT HELPS VULNERABLE FAMILIES 50 Disabled People Cares for mentally and/or physically disabled children until adulthood and even 42 longer if necessary. Homeless People because it helps the homeless and provides shelter and food. 35 The Environment to potect the environment around the world. 34 Victims of Violence or Conflict it helps stop sex trafficing and supports people who have experienced sex 29 trafficing People Suffering from Mental Illness They do a variety of essential, incredible work for people who are dealing with a 26 range of mental health issues. Elderly People Old aged people with no family to support require a lot of assistance for 22 everything.

43

Disaster Victims It helps people locally with basic provisions, emergency shelters and food, 17 hygiene kits in event of disasters anywhere in the world Refugees Providing legal assistance to refugees and asylum seekers 16 Veterans It helps veterans who are experiencing hardships due to their military experiences 10 LGBT People They work to protect the rights of LGBT people. 9 Addicts I think that it is an excellent program that has helped many people over the years 6 to return to the work force after their recovery from alcoholism and/or drug addiction. Offenders Its main objective is to rehabilitate Ex-offenders and reintegrate them back into 4 the community to live responsible lives thereby creating peaceful society. Other I believe in the work they are doing to spread the gospel and support persecuted 14 Christians.

Note. Other included categories that were each mentioned by fewer than 3 people, namely Christians, minorities, educators, Michael Jackson fans, hunters, single people, Chinese immigrants, gypsy people, people with unusual appearances, prostitutes, and Sadhus. Some respondents mentioned multiple beneficiary groups. Relevant coded sections have been italicized for clarity.

44

Because I love animals (particularly dogs) and believe they all deserved to be loved and treasured (Papua New Guinean/Thai female, 18; animal).

The take care of the less privilege children and I love to help people and children are the future of our nation and they should have the best (Nigerian male, 33; social services).

It can help me to support the people and art I like (Chinese male, 46; social services).

Neediness. Finally, sometimes beneficiaries were described in terms of their neediness. For some donors, it appeared that the existence of need was reason enough to help:

Helps to save people in need (Kazakh male, 20; health).

Is one of the organizations in Honduras that really help, financially, to people who need it (Honduran male, 25; social services).

Prevalence of Self/Other Orientations Across Categories of Charity Above we have described emergent themes in the way participants explain their giving in relation to their own or beneficiaries’ identities, or an explicitly shared identity. However, the prevalence of self- versus other-orientations varied considerably when considering the category of charity that the donor was talking about. The prevalence of themes and sub-themes varied across all fifteen categories of charity. Here we present and discuss results only for those categories mentioned by at least 5% of participants. In Table 8 we present the prevalence of participants who explained their charity preferences in terms of only self, only other, in relation to themes of both self and other, or in relation to neither self nor other. Table 9 summarizes the results across sub- themes. It is not our intention to interpret minor differences between categories, but instead to elaborate briefly on the emergent patterns in the data, which may be of interest. Religious and research charities were more likely than other types of charity to be explained only in relation to the donors’ personal and social identities, suggesting giving was more self- oriented. Donors to religious charities stressed their religious identities and talked about charity as a way of reflecting and promoting their religious groups’ values and beliefs. Most of the research charities mentioned worked on medical research, particularly cancer research. Donors to these organizations explained their giving preferences with reference to their own suffering or that of important group members, in particular family members who suffered or died from cancer. Family identity was frequently evoked in reference to such charities.

45

Table 8. Relative prevalence of use of Self only (“I/we”), Other only (“they”), or Self and Other (“I/we” and “they”) themes in explanations of giving to charities in different categories Theme All Frequency by Category of Charity Categories Social Services Health Animal International Education Religious Research Self only 27% 17% 27% 5% 18% 20% 62% 56% Other only 41% 52% 37% 62% 46% 43% 10% 18% Self and Other 18% 22% 18% 33% 16% 24% 12% 9% Neither 14% 9% 18% 0% 20% 13% 15% 18% n (category) 1,849 423 365 190 182 140 125 120 Note. N = 1,849. Participants responses have been coded into any theme they evoked, thus themes overlap. Only categories nominated by more than 5% of sample are represented in the Table.

46

Table 9. Relative prevalence of Self/Other themes and sub-themes in explanations of giving to charities in different categories Theme All Frequency by Category of Charity Sub-theme Categories Social Services Health Animal International Education Religious Research Self 45% 39% 45% 38% 35% 44% 74% 65% Social Identities 32% 28% 36% 25% 17% 26% 68% 61% Values and Beliefs 12% 7% 5% 19% 16% 16% 20% 3% Benefits 9% 8% 8% 3% 4% 14% 21% 6% Suffering 8% 4% 16% 0% 1% 1% 0% 54% Shared Identity 8% 12% 9% 1% 7% 9% 6% 4% Other 59% 72% 52% 95% 61% 65% 22% 26% Beneficiary Identities 54% 64% 49% 91% 53% 59% 20% 25% Power 8% 11% 2% 24% 7% 6% 4% 2% Value 7% 10% 3% 26% 3% 6% 0% 1% Neediness 6% 9% 5% 3% 9% 6% 2% 0% n (category) 1849 423 365 190 182 140 125 120 Note. N = 1,849. Participants responses have been coded into any theme they evoked, thus themes overlap. Only categories nominated by more than 5% of sample are represented in the Table.

47

Animal, social services, and international charities were more likely than other organizations to be discussed with reference to beneficiaries, suggesting giving was more other-oriented. Animals and children were two categories of beneficiary highly valued by donors, and the focus on helping these groups was illustrated by the prevalence of the beneficiary identities sub-theme, as well as ideas about value and power. Those who gave to international charities often highlighted the beneficiary as being poor, in another country, and in need. Finally, animal and education organizations were more likely than other types of charity to be motivated by the intersection of self and other. Many donors explained their motives for giving to animal charities as an intersection between their care for and valuing of animals (other), a reflection of deeply held values to do so (self), and the interplay between human (self) and animal (other) identities. Donors to the education sector also talked about the value of helping children (other) and the benefit to “us” as humans in wider society (self). Discussion Thematic analyses of a large-scale, global survey of donors revealed both self- and other- oriented explanations for charity preferences. Nine sub-themes were generated from the data. When talking about giving in relation to the self, donors mentioned social identities, values and beliefs, benefits, suffering, and shared identity. In contrast, when giving was constructed in relation to the other, donors highlighted beneficiary identities, power, value, and neediness. Overall, other- oriented motives were more prevalent than self-oriented motives. However, the prevalence of themes varied substantially depending on the category of charity being considered. Of particular note, identities (social and shared) were more frequently mentioned motives for giving than those that have traditionally been highlighted in the literature, such as values, benefits, and need (Bekkers & Wiepking, 2011b). These findings provide a framework for making sense of the diverse donor motivations previously identified in the literature (see especially Aaker & Akutsu, 2009; Bekkers & Wiepking, 2011b; Konrath & Handy, 2017; Sargeant & Woodliffe, 2007b). When giving was explained in relation to the self, donors highlighted positive aspects including values and benefits received (see also Bekkers & Wiepking, 2011b), and sometimes mentioned their own experiences of suffering (see also Kearns, Muldoon, Msetfi, & Surgenor, 2017; Taylor & Hanna, 2018). Donors also frequently named or implied specific social identities when explaining their giving. These results are in line with the view that identities influence charitable decisions (Aaker & Akutsu, 2009; Chapman, Louis, et al., 2018; Kessler & Milkman, 2018). The social identity approach (Tajfel, 1981; Turner et al., 1987) proposes that people have a range of identities, which may be more or less important in different contexts. But not all ingroups (or outgroups) are equally likely to receive aid. Until now, research has rarely considered which of multiple donor identities may be most

48 relevant in giving contexts. Previous research has focused mostly on geographic, gender, or broad donor identities (Charnysh et al., 2015; Croson, Handy, & Shang, 2010; Kessler & Milkman, 2018; M. Levine & Thompson, 2004; Zagefka et al., 2013). The current results identify religious, family, organizational, friendship group, human, and health-related identities as also relevant. However, the findings also are clear that donors’ identity motives matter more in some contexts than others. When giving was explained in relation to the other, donors also focussed on identities—this time those of the beneficiary. When describing the beneficiary, themes of value (I/we like them), power (they are powerless), and neediness (they are in need) were common. Social Identity Theory (Tajfel, 1981) proposes that people show a strategic preference towards people within their own groups. From this perspective, when helping people outside the group, it makes sense that the types of ‘other’ most likely to receive aid would be those who are valued, pose little threat, and need help the most (see also Cuddy et al., 2007; Nadler, 2002). Results confirm that status differences promote giving to outgroup targets (see also Halabi & Nadler, 2017; Nadler, 2002, 2016) and highlight particular beneficiaries—animals, children, the poor—as especially likely to evoke a helping response in the absence of self-oriented motives to give. Although giving to help these targets is framed by donors as other-oriented, it is nonetheless possible that vulnerable, needy, or highly valued beneficiaries also receive help because offering such help aligns with strategic group motives—for example, allowing the donor to restore their group’s esteem, maintain their high status position, or signal to observer that the group is kind or generous (see van Leeuwen, 2007, 2017; van Leeuwen & Harinck, 2016). Indeed, helping beneficiaries who are widely valued or visibly needy would allow self-interested donors to accrue the greatest social benefits. Future research should consider explicitly when some beneficiary identities attract norms of helping while others— arguably equally needy—do not; and also, how beneficiaries can attract help in the absence of group interests or when helping contradicts the interests of the donor’s salient social group. Finally, some giving was explicitly explained as an intersection between the donor and the beneficiary. Donors motivated by self-other unions talked about shared identities, supporting the social identity perspective (Tajfel, 1981; Turner et al., 1987). Shared and superordinate identities have previously been shown to promote giving and other social actions (Charnysh et al., 2015; T. K. James & Zagefka, 2017; M. Levine et al., 2005; L. G. E. Smith, Gavin, & Sharp, 2015; Zagefka et al., 2013). Yet again, the present results provide a richer understanding of when shared identities may and may not elicit gifts. Certain kinds of beneficiaries appear to be easier to incorporate into the self. Donor responses show that shared identities were most often related to gender, nationality, and identities born of suffering. Attempts to construct other shared identities may not be as effective. Future research should consider why and when shared identities are likely to form. It

49 will also be important for future research to identify what attributes of ingroups (compared to outgroups) are mostly likely to promote ingroup generosity. Both donor (self) and beneficiary (other) identities strongly influence giving decisions. This self-other distinction builds upon Breeze (2013), who argued that donors use binary distinctions to help them navigate the charitable domain. Breeze highlighted human-animal and domestic- international as two important binaries relevant to giving. The current research adds self-other as a broad, important framework to understand how donors construct their relationships with charities. Future research could test whether self/other orientations have implications for the type of help that is offered. An important question is whether different frames for the same charity target would affect likelihood, size, and motives for donations. For example, would other-focussed frames motivate more palliative support to alleviate others’ suffering, and self-focussed frames motivate more empowering, political forms of support to benefit targets framed as ingroups (see Louis et al., under review; Thomas & McGarty, 2018)? The data support the view that charitable giving is a social process that takes place at the intersection between donors and beneficiaries (see also Andreoni & Payne, 2013; Chapman, Louis, et al., 2018). Participants construct charities differently, depending on the beneficiary that those charities are perceived to represent. Religious and research charities, for example, are more strongly constructed in relation to the self. On the other hand, animal, welfare, and international charities are instead constructed in relation to the beneficiary (the other). In participants’ spontaneous discourse, these different framings appear to be associated with different motives for giving. Future research could also explore experimentally how donor behavior or the importance of different motives vary as a function of manipulated self-other frames. Practical Implications Results highlight that donors use identities (related to both self and other) to explain their giving choices, with family, geographic, organizational, religious, and even human identities found to be especially relevant. Furthermore, self and other are relevant in different contexts. Charities focussed on religion and research, for example, would most effectively engage donors by stressing focal identities (religious and family, respectively) and connecting to relevant group values. Charities working for children, animals, and those living in poverty, on the other hand, would do better to focus their appeals on the targets in need rather than the people with the wallets. Strengths, Limitations, and Future Directions Strengths of the current study include the size and breadth of the participant sample—1,849 participants from 117 countries. Although students in a large online psychology course are not representative of the broader community, such sampling begins to move beyond the traditionally WEIRD research samples (western, educated participants, from industrialized, rich, and democratic 50 countries; Henrich, Heine, & Norenzayan, 2010) to ensure a range of voices are present in the data. Further, the use of thematic analysis allows rich insight into the self- and other-oriented motives of donors, and how they vary as a function of charity category, at a scale not yet seen within the literature. Two limitations, however, warrant consideration. Donors tend to stereotype “charity” as meaning helping the poor and needy (Breeze, 2013). When we asked donors to think of the charitable organization that was most important to them, they may have been more likely to think of an organization that fits that stereotype and less likely to spontaneously think of political organizations, sports teams, or their child’s school, for example. As a result, our data may have disproportionately sampled descriptions of giving preferences that fit the charity stereotype, and thus inflated the prevalence of other-oriented giving compared to a general question such as “what groups do you support financially, if any?” A second, related limitation is the focus on donors’ most important charity alone, which may have resulted in an over-estimate of self-oriented motives. By using qualitative methods and diverse participants to understand donors’ motives, we highlight a number of ways identities are used to inform charitable choices. Future work should build on this foundation by evaluating all giving, not only support of stereotypical or favorite charities. Further exploration of the specific identities evoked in different giving contexts, and how the prevalence of self/other orientations may vary as a function of demographics, would also be useful directions for research. Conclusion Thematic analyses of a global survey (1,849 participants from 117 countries) revealed both self- and other-orientations at play in charitable giving. Participants describe giving to charity for self-oriented motives (social identities, values and beliefs, benefits, suffering, and shared identity), other-oriented motives (beneficiary identities, power, value, neediness), or both. The prevalence of self/other giving orientations varied across different categories of charities and identities were mentioned as often or more often than motives more traditionally associated with giving, such as values, benefits, and need. The data highlight the importance for charitable organizations to understand when and how identities may form, in order to link their beneficiary with donors effectively. Such an understanding could help organizations to understand when their appeals fail to mobilize new donor groups, and how to succeed. On a deeper level, the paper contributes to the literature on both charitable giving and group processes by showing which among multiple identities are most likely to structure giving. Many of the social identities highlighted by the current analysis have not yet been tested empirically in relation to giving. Further, results demonstrate the contexts and conditions in which ingroup vs. outgroup targets are most likely to be cared for. By suggesting some of the moderators—beneficiaries' power, needs, and likeability, for

51 example—that may promote norms of outgroup helping, the paper speaks to the increasing interest in understanding and theorising how individuals experience the multiple group world and decide whom it is most important to help.

52

Chapter 4. Norm Divergence Model of Prosocial Influence: Government Aid Both Crowds In and Crowds Out Citizens’ Charitable Giving Via Diverging Descriptive and Injunctive Norms

The two studies presented in this chapter highlight the importance of considering social context when investigating charitable giving. These studies demonstrate two important points. First, the studies show that government policy about foreign aid can affect private citizens’ charitable giving. Second, the research suggests that foreign aid may do so in part by changing perceptions of what other people in the national group do or approve of. These findings both illustrate the influence of salient aspects of the social context and, in particular, suggest that contextual cues can influence the relationships between donors and particular beneficiaries (in this case people overseas).

Citation Chapman, C. M., Louis, W. R., & Masser, B. M. (in prep). Norm divergence model of prosocial influence: Government aid both crowds in and crowds out citizens’ charitable giving via diverging descriptive and injunctive norms. Manuscript in preparation.

Contributions The work presented in this chapter was conducted by myself under the guidance of my research supervisors Professors Winnifred Louis and Barbara Masser. Specifically, I designed the survey in Study 4.1 and experiment in Study 4.2, ran data collection, analysed the data and interpreted the results, and wrote the manuscript. My advisors guided me throughout the process. They reviewed the survey and experimental design, helped me interpret the results, and provided feedback on the manscript. Emma Thomas, Nikki Harré, and Joanne Smith also provided feedback on an earlier version of this manuscript.

53

Norm divergence model of prosocial influence: Government aid both crowds out and crowds in citizens’ charitable giving via diverging descriptive and injunctive norms

Abstract Government funding of charitable organizations sometimes crowds out (reduces), sometimes crowds in (increases), and sometimes has no effect on private charitable giving. We propose the norm divergence model to help explain these contradictory findings. Government spending may provide a cue for what other citizens do (increasing the descriptive norm), but may simultaneously reduce the responsibility citizens feel to respond individually (decreasing the injunctive norm). Two studies—in Australia (N = 189) and the United States (N = 290)—demonstrated that foreign aid spending influences citizens’ international giving via national giving norms. Strong evidence was found for crowding in via an increased descriptive norm. This effect was replicated across national contexts and on donation likelihood, willingness, and behavior. Modest evidence of simultaneous crowding out via a reduced injunctive norm was also found, though inconsistently. Neither normative pathway was moderated by strength of national identification. Implications for policy-makers and charities are discussed.

Keywords: crowding out; charitable giving; norms; donations; foreign aid.

54

Introduction How much sway do politicians hold? Can their public rhetoric and policy decisions influence the choices we make in everyday life? Economists have long argued that they can, at least when it comes to charitable giving. According to the crowding out hypothesis (see Roberts, 1984; Warr, 1982), rational donors will reduce their voluntary contributions to charities dollar-for- dollar in response to government funding. While there are good theoretical grounds for expecting public funding to crowd out private giving, in practice pure crowding out is not found. Instead, evidence varies along a spectrum from partial crowding out (e.g., A. A. Payne, 1998) through no relationship (e.g., De Wit, Bekkers, & Broese van Groenou, 2017) to crowding in, whereby an increase in public funding is met with a corresponding increase in donor support (e.g., Khanna & Sandler, 2000). This paper proposes a novel explanation for this variability—namely that government spending affects private giving through two competing normative pathways, which we call the norm divergence model. We argue that government funding increases the sense that the group gives (the descriptive norm), thereby increasing private giving; but that government funding also decreases the injunctive pressure on individual group members to give (the injunctive norm), thus simultaneously decreasing private giving. The contextually-determined strength of these relative pathways may explain the array of effects found in the crowding out literature. Below we summarize the known evidence for and against crowding out, build a case for national norms being the pathways through which government policy affects private giving to international charities, and then present and discuss two empirical studies testing our hypotheses. While we test the model here in the context of international aid and donation decisions, we see the norm divergence model as being broadly relevant to the crowding out literature, and to all prosocial contexts where decisions of salient group members may both increase helping via descriptive normative signals, and decrease helping via reduced injunctive normative pressure. Crowding Out or Crowding In? For decades, economists have asked whether government funding of public goods displaces (crowds out) or stimulates (crowds in) private donations. In the context of charity, it is theorized that ‘rational’ donors (i.e., those who care only about the provision of a service) will reduce their private voluntary contributions dollar-for-dollar as the government allocates more to meet the need (Roberts, 1984; Warr, 1982). In practice, however, perfect crowding out does not occur. Empirical evidence of crowding out has been mixed. Some studies have found partial crowding out, whereby government funding reduces private giving to some extent but the net financial gain for the charity is still positive (see Andreoni & Payne, 2011; A. A. Payne, 1998; W. O. Simmons & Emanuele, 2004). Others have found the opposite, where government funding 55 actually crowds in, or increases, private giving (see Herzer & Nunnenkamp, 2013; Heutel, 2014; Hughes, Luksetich, & Rooney, 2014; Khanna, Posnett, & Sandler, 1995; Khanna & Sandler, 2000; Okten & Weisbrod, 2000). Still other studies have shown no significant relationship between public and private giving (see Brooks, 2003a; De Wit et al., 2017), suggesting that the two may be independent. Indeed, a meta-analysis of 422 findings from 70 studies found 262 instances of partial crowding out and 160 of crowding in (De Wit & Bekkers, 2017). In sum, the evidence base for crowding out is mixed. We aim to extend the literature on crowding out by testing it in the specific context of giving to international charities. Organizations serving communities overseas stake a large claim to both public and private funds. In 2015, over $31 billion dollars were allocated by the United States government to foreign aid (OECD, 2016) and a further $32 billion were donated by American citizens to organizations working on international affairs (McKeever, 2015). Few studies have as yet considered crowding out across different charity types (cf. De Wit et al., 2017; Khanna et al., 1995; Okten & Weisbrod, 2000). Those that included international nonprofits as a category of analysis have observed either crowding in (Herzer & Nunnenkamp, 2013) or no significant association between public and private giving (De Wit et al., 2017; Khanna et al., 1995). Yet government spending on foreign aid seems to influence private attitudes to international giving. Mixed-methods case studies that matched individual-level panel survey to media archives in the Netherlands over a period of 2002-2014 showed that highly publicized funding cuts sometimes influence private giving to international charities (De Wit et al., 2017). The authors document that Oxfam suffered from a series of substantial and widely publicized government budget cuts over the period. In their case, a trend of crowding in was observed, with private donations decreasing after declining government support.5 Further research is warranted to understand the ways that government spending may affect private giving in the context of international charity. For our purposes, international giving is also a valuable context in which to test the group processes that we theorize may explain inconsistent results in past research. In the international aid context, groups are likely to be salient (see Thomas, McGarty, & Mavor, 2010). Donors belong to one national group and beneficiaries belong to another. We therefore anticipate that national group identities and their associated norms may be particularly influential in international giving decisions.

5 ‘Crowding in’ generally refers to the positive influence of government funding, whereby it increases private giving. Nonetheless, another implication of donors following government cues is that that a decrease in government funding can decrease private giving. 56

Perceived Group Norms Affect Attitudes and Behavior Social norms provide cues to group members about what is expected in a given context (e.g., Cialdini et al., 1990). Perceptions of group norms can influence a range of socially desirable attitudes and behaviors (Jacobson, Jacobson, & Hood, 2015; Kallgren, Reno, & Cialdini, 2000; McDonald, Fielding, & Louis, 2013; Nolan, Schultz, Cialdini, Goldstein, & Griskevicius, 2008; Reno, Cialdini, & Kallgren, 1993; Terry & Hogg, 1996)—including charitable giving (Agerström, Carlsson, Nicklasson, & Guntell, 2016; Bartke, Friedl, Gelhaar, & Reh, 2017; Croson et al., 2009; J. R. Smith & McSweeney, 2007). We propose that government policies may influence perceptions of national group norms, which in turn affect private giving. Two categories of norms are usefully distinguished—descriptive and injunctive (Cialdini et al., 1990). Descriptive norms reflect the prevalence of a behavior and highlight what most people typically do, while injunctive norms reflect what most people think is the right thing to do. Although perceptions of descriptive and injunctive norms are generally associated (Eriksson, Strimling, & Coultas, 2015), they remain theoretically distinct (Cialdini et al., 1990). Both descriptive norms (Agerström et al., 2016; Bartke et al., 2017; Croson et al., 2009; McDonald et al., 2013; Nolan et al., 2008) and injunctive norms (Kallgren et al., 2000; Reno et al., 1993) can influence behavior independently; although it is important to also consider them jointly as they sometimes pull in different directions (Cialdini et al., 1990; Jacobson, Mortensen, & Cialdini, 2011). The Current Research In light of a contradictory evidence base for the crowding out hypothesis, we seek to test crowding out in one specific charity context—that of international aid. We theorize that government policy that relates to international (intergroup) relations will make national identity salient and that, therefore, government policy on foreign aid will cue citizens to perceive group norms for international giving. We argue that actions of group leaders—in this case the elected government—may influence perceived group norms in two competing ways. When salient group members take public action, it should influence perceptions of what the group as the whole does. In this way, government aid spending should increase the national descriptive norm of international charitable giving. Thus, descriptive norms are the proposed mechanism through which government spending can crowd in private donations. However, the very same action could simultaneously reduce private donations. Our reasoning stems from evidence that groups respond to perceived need differently than individuals. Indeed, people are less helpful in a group than as individuals (e.g., Maruyama, Fraser, & Miller, 1982). This phenomenon is explained by a perceived diffusion of responsibility among 57 group members (Darley & Latane, 1968). However, such diffusion does not occur evenly across the group. Some group members—especially people in leadership positions—are perceived to be more responsible than others (Baumeister, Chesner, Senders, & Tice, 1988; Forsyth, Zyzniewski, & Giammanco, 2002). For example, leaders may be expected to act on behalf of the group (substituting for other group members’ actions) as well as leading by example (promoting other group members’ action). The actions of such leaders may thus influence perceptions of responsibility differently within the group itself. When leaders “step up” to their responsibility—in this case by sending government aid overseas—this may actually decrease the perceived responsibility felt by group members to act themselves, resulting in reduced injunctive pressure from the group for individuals to send private donations too. Thus, injunctive norms are the proposed mechanism through which government spending can crowd out private donations. To summarize, we propose that foreign aid will affect private giving through two competing normative pathways: the norm divergence model. First, when the government gives more, it will provide a cue about the national descriptive norm for international charitable giving, which can be used as a guide for group members’ private decision-making. However, simultaneously, increased government spending may reduce injunctive pressure on group members to respond individually. We test both pathways in the model. Finally, national identification is expected to moderate both normative pathways. According to the social identity approach (see Tajfel, 1981; Turner et al., 1987), when individuals self-define as group members (thinking of themselves as part of a “we” rather than an independent “I”) they are motivated to act out their group’s norms. Individuals are therefore most likely to be influenced by perceived norms when they identify more strongly with the group in question (see also J. R. Smith & Louis, 2008; Terry & Hogg, 1996). We therefore hypothesize the following:

H1: Government spending on foreign aid will crowd in private giving to international charities via the perceived descriptive norm.

H2: Government spending on foreign aid will simultaneously crowd out private giving to international charities via the perceived injunctive norm.

H3: National identification will moderate the effects of both descriptive and injunctive norms on private giving, with stronger effects for high identifiers. The current research answers calls to look at crowding out for different charity types (e.g., De Wit & Bekkers, 2017) by testing crowding out specifically in the international charity context (see also Herzer & Nunnenkamp, 2013). To our knowledge, this is the first research to (1) test simultaneous crowding out and crowding in (proposed by Brooks, 2003a; De Wit et al., 2017; Hughes et al., 2014; W. O. Simmons & Emanuele, 2004; but not yet tested) and (2) to evaluate

58 norms as potential mediators for these effects. By integrating the economic literature on crowding out with a social psychological analysis of normative influence, these studies are poised to answer an important applied question as well as to contribute to the literature on prosocial behavior and giving. Study 4.1 Study 4.1 evaluated the relationships between government aid and private giving and the potential mediating role of perceived norms, in a sample of Australian students. Two competing pathways were predicted—crowding in of private donations via the descriptive norm and crowding out via the injunctive norm. Both pathways were expected to be stronger for those who more strongly identified with the national group. Participants Two-hundred and thirty-four students in Australia voluntarily completed an online survey reporting their charitable attitudes and behavior in exchange for course credit or a small chocolate. After debriefing, eight participants opted to withdraw their data from the study, and a further 37 participants who failed to provide the correct response to at least one of three attention checks (e.g., “For this item, please select very unlikely”) were excluded. The final sample (N = 189) included 134 females and 55 males, ranging in age from 17 to 51 years (M = 20.03, SD = 5.01). Most participants (86%) identified their nationality as Australian. Approximately one-third (37%) reported that they had made a donation of money to any type of charity in the previous three months. This is substantially lower than the national giving rate (Charities Aid Foundation, 2017b), a point we return to in the discussion. Procedure Participants were recruited either through an online research participation system in exchange for partial course credit or through volunteer requests distributed at lectures or via email. Sample size was limited by time constraints, with data collection running until the end of semester. Participation was voluntary and anonymous, and participants were free to withdraw at any time without penalty. The topic under study was not communicated in advance but participants were fully debriefed at the end of the survey. Measures assessing charitable behavior and attitudes, perceived levels of government spending on aid, Australian giving norms, and socio-demographic and individual difference information were included. Measures of interest for the current study are summarized below. Measures Perceptions of aid. A single item measured participants’ perceptions of international aid levels (“Thinking about the Australian government, what percentage of the budget do you believe is spent on international aid (voluntary donations to other countries) each year?”, response scale 0- 59

100%). Participant raw scores ranged between 0-70% with a strong positive skew (M = 9.25, SD = 12.14). Despite this predominance of low values, participants grossly overestimated foreign aid spending, which at the time was less than 0.3% of gross national income (OECD, 2016). Following Friedline, Masa, and Chowa (2015), an Inverse Hyperbolic Sine transformation was applied to raw data to reduce the influence of extreme outliers. Analyses presented below use these transformed values.6 Norms. Participants indicated perceived norms using single-items adapted from McDonald, Fielding, and Louis (2014), including both injunctive norms (“Rate how a typical Australian would respond to the statement ‘giving to charity is the right thing to do’”, 1 = strongly disagree, 7 = strongly agree) and descriptive norms (“Indicate what percentage of Australians you believe made donations of money to charity in the last year”, 0-100%). Identification. Participants’ national identification was assessed by three items adapted from Ellemers and colleagues (1999) to capture participants’ degree of identification as an Australian (“I identify with other Australians”, “I am like other Australians”, “Being Australian is an important reflection of who I am”, 1 = strongly disagree, 7 = strongly agree). Scores were averaged with higher scores indicating stronger identification, α = .88. Donation likelihood. Self-reported donation likelihood was used a measure of personal charitable giving. Although a social desirability bias is found in self-reports related to charitable giving, such self-reports are strongly correlated with actual giving (Bekkers & Wiepking, 2011a). Participants indicated their likelihood of making a donation of money in the next 3 months to an international charity with a single item (“Thinking about the next 3 months, how likely will you be to make a donation of money to each of the following types of charitable organization?: International [including poverty reduction, disaster relief, human rights]”, 1 = Not at all likely, 7 = Extremely likely). Results Means, standard deviations, and zero-order correlations between the variables are reported in Table 10. Overall, participants were moderately identified with their fellow citizens (M = 4.76, SD = 1.47), perceived that others endorsed charitable giving strongly (M = 5.63, SD = 1.12) but that less than half actually gave to charity in any given year (M = 43.80, SD = 19.72). They reported a moderate likelihood of donating to international charity in the next 3 months (M = 4.16,

6 With untransformed values retained, the same pattern of results were observed. However, the overall model significance changed to p = .053 and the observed indirect effect via the descriptive norm pathway was also weaker, 95% CI [.003, .02].

60

SD = 2.04). Participants who thought the government spent more on aid also perceived more Australians gave to charity (higher descriptive norm). Those more identified with other Australians also perceived significantly higher injunctive and descriptive norms for giving. Those perceiving higher descriptive norms also reported higher likelihood of personally donating. No other significant associations were observed.

Table 10. Descriptive statistics and zero-order correlations between perceived aid, norms, identification, and donation likelihood M SD 1 2 3 4 1 Perceived aid (0-100%) 9.25 12.14 ** 2 Descriptive norm (0-100%) 43.80 19.72 .23 3 Injunctive norm (1-7) 5.63 1.12 -.07 .12 4 Identification (1-7) 4.76 1.47 .04 .21** .17* 5 Donation likelihood (1-7) 4.16 2.04 .00 .20** .07 -.10 Note. N = 186 (Listwise). Perceived aid has been transformed for the analyses using Inverse Hyperbolic Sine to account for skewness; however, mean and standard deviation reported in this Table are untransformed values for clarity. *p < .05; **p < .01

To test the hypothesis that government aid levels indirectly influence private giving via perceived national giving norms, mediation analyses with 5000 bootstraps was conducted using the PROCESS macro for SPSS (Hayes, 2012; models 4 and 14). Descriptive and injunctive norms were tested as parallel mediators, with identification proposed to moderate both pathways. The model is summarized in Figure 1.

Figure 1. Results of the moderated mediation analysis of the effect of perceived aid levels on international donation likelihood via norms, moderated by identification Note. Values represent standardized path coefficients. Dashed lines indicate non-significant pathways. *p < .05; **p < .01 61

Overall, the model explained 7% of the observed variation in donation likelihood, F(6,179) = 2.15, p = .050.7 Although there was no direct association between government aid and donation likelihood, a significant positive indirect pathway emerged via the descriptive norm, IE = .08, SE = .05, 95% CI [.02, .20]. The indirect pathway via injunctive norms was not significant, IE = -.01, SE = .01, 95% CI [-.06, .01]. Neither mediation pathway was qualified by strength of national identification, ps > .856; however, overall higher national identifiers reported less likelihood of donating to international charities, ß = -.16, p = .034. Discussion Study 4.1 provides the first evidence that government aid may affect private giving via norms. The overall relationship between perceived government aid levels and citizens’ likelihood to give to international charity was not significant. This finding corroborates prior work showing no significant relationship between public and private spending in in the international aid context (De Wit et al., 2017; Khanna et al., 1995), but contradicts Herzer and Nunnenkamp (2013), who found crowding in for international charities. Crowding in occurred indirectly, however. Government aid had a positive indirect effect on likelihood of donating to international charity by increasing the perceived descriptive norm

(supporting H1). Government aid levels were positively associated with perceptions that other citizens gave to charity (descriptive norm) which, in turn, was associated with reporting a greater likelihood of personally donating to international charity. To our knowledge, the relationship between government aid and perceived giving norms has not been demonstrated before. The relationship between descriptive norms and private giving, however, corroborates previous work showing descriptive norms influence charitable giving (Agerström et al., 2016; Bartke et al., 2017; Croson et al., 2009).

No evidence of simultaneous crowding out was found in Study 4.1 (no support for H2). Though the direction of the indirect effect was negative as predicted, the effect was not significant. Thus, no support was found for the notion that government aid crowds out private giving via a reduced injunctive norm. Although previous research has shown that injunctive norms can influence socially desirable behaviors (e.g., Kallgren et al., 2000; Reno et al., 1993; J. R. Smith & McSweeney, 2007), no significant relationship was found between injunctive norms and donation likelihood in this sample of Australian students.

7 One outlier more than three standard deviations from prediction was identified when predicting descriptive norms, and a further four when predicting injunctive norms. These outliers were retained in the analyses. When they are excluded there are no substantive changes to the individual results, although all effect sizes are increased and the overall variance accounted for slightly increases, F(6,176) = 2.57, R2 = .08, p = .021. 62

Contrary to expectations (H3) and previous research showing norms are more influential when people identify with the group in question (e.g., J. R. Smith & Louis, 2008; Terry & Hogg, 1996), neither indirect pathway was qualified by strength of national identification. There was an unexpected direct effect of identification: participants who more strongly identified as Australian reported a lower likelihood of donating to international charities. Though not a focus of the study, this relationship may simply reflect ingroup favoritism (Brewer, 1999), or could possibly be explained by a third factor: political conservatism. Prior research has demonstrated that political conservatives tend to be higher in nationalism (van Der Toorn et al., 2014) and are also less likely to donate to international charities (Chapman, Louis, et al., 2018; Nilsson et al., 2016). Study 4.1 had several limitations to acknowledge, however. First, a student sample may not be appropriate for research on charitable giving, as young people tend to be among the least generous members of the community (e.g., M. A. Steinberg et al., 2005). Approximately one-third of participants reported having donated to charity in the previous three months. This giving rate is considerably less than the national rate: in a nationally representative telephone survey, 68% of Australians reported they had donated money in the previous month (Charities Aid Foundation, 2017b). Next, sample size may not have delivered sufficient power to detect effects if they existed. Recruitment was limited by pragmatic concerns (the end of semester) rather than a priori power analyses. Furthermore, the cross-sectional design employed precludes inference of causality. Although mediation was found by norms, it is feasible that the influence goes in the opposite direction: participants who give more may be likely to infer positive descriptive norms and positive foreign aid (the false-consensus effect; Ross, Greene, & House, 1977). Finally, there was arguably a mismatch between the measurement of norms and the focal dependent variable. Although we were interested in the likelihood of donating to international charities specifically, perceived norms were measured in relation to charitable giving in general. Previous research has shown that variables are most powerful when their measurement is specific to the context (Fishbein & Ajzen, 1974). It is therefore possible that the observed effects of norms were dampened by this methodological oversight. Study 4.2 Study 4.2 was designed to replicate the findings of Study 4.1 with a more robust method and sample and in a new national context—that of the United States. To overcome the acknowledged limitations of Study 4.1, we used a larger community sample, employed an experimental design, and measured perceived norms specific to international giving. In addition, personal giving response was assessed through both a self-report item and a behavioral measure. All hypotheses remained the same as Study 4.1. Government aid was expected to influence private giving via norms in two competing ways, the norm divergence model: crowding in via the

63

descriptive norm (H1) and simultaneously crowding out via the injunctive norm (H2). Both indirect pathways were expected to be stronger for participants who strongly identified with the nation (H3). Method We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study. Participants Two-hundred and ninety-two Americans took part in an online experiment relating to perceptions of, and responses to, foreign aid policy. Sample size was determined by an a priori power analysis determined by 80% power to detect small (sr2 = .03) interactive effects. A 10% buffer was added to the required sample of 264 to allow for any potential exclusions. All participants passed the attention check (“To show you're paying attention, please select "totally agree" for this item”); however, data were excluded from two participants who guessed the true purpose of the study. The final sample (N = 290) included 149 females, 140 males, and one person identifying as non-binary. Participants ranged in age from 20 to 74 years (Mage = 38.37, SD = 11.17). All participants identified their nationality as American. Approximately three-quarters (76%) reported having made donations of money to charitable organizations in the previous 12 months. This rate is comparable to that reported for the United States in the World Giving Index (i.e., 62% having donated in the previous month; Charities Aid Foundation, 2017b). Procedure Participants were recruited online via the Amazon Mechanical Turk platform. The advertisement invited participants to a study about citizens’ attitudes towards government policy and social issues. First, participants completed a 1-minute screening survey containing demographic data, for which they received payment of USD$0.05. Only participants identifying their nationality as American were invited to participate in the full 15-minute study. Of 295 Americans invited to take part in the complete study, 292 (99%) chose to do so, and received a bonus payment of USD$2.00 for their time. First, participants completed three individual difference measures. They were then randomly allocated to read information about the United States’ position on foreign aid: framed either to seem relatively high (n = 146), or low (n = 144). Aid manipulations are summarized below. After reading about foreign aid, participants completed two manipulation check items, reported their willingness to send a personal donation to help people in other countries and preferences for how this donation would be used, and were given the opportunity to earn a potential real donation to help people overseas as a behavioral measure. Next, participants answered questions about perceived norms and their attitudes towards foreign aid. Finally, they were

64 debriefed as to the true purpose of the study, the United States’ position on foreign aid, and given the opportunity to withdraw their data. Aid Manipulation In both aid conditions, the true position of United States spending on foreign aid was presented (based on data sourced from OECD, 2016). In order to manipulate perceptions, however, data was presented either as an absolute value (more than $31 billion, which was expected to be perceived as high) or as a percentage of gross national income (less than 0.2%, which was expected to be perceived as low). In both conditions, participants saw the United States’ rankings compared to other OECD countries, either appearing to be one of the most generous in absolute spending (high aid condition) or one of the least generous by percentage of income (low aid condition). Finally, to further strengthen the manipulation, quotes were presented in both conditions from four political leaders: two Democrats (Barack Obama and Hillary Clinton) and two Republicans (Donald Trump and George W. Bush). All quotes were real, sourced from media outlets or official sources. In the high aid condition all leaders were quoted to suggest they supported spending foreign aid (e.g., George W. Bush: “American is a generous nation with a moral responsibility to do our part to help relieve poverty and despair”). Alternatively, in the low aid conditions quotes were selected to suggest that leaders wanted to keep aid spending low (e.g., George W. Bush: “Oftentimes we’re well-intentioned when it comes to foreign help, but the money never makes it to the people that we’re trying to help”). Thus, in the high aid condition, America was positioned as a leader in foreign aid spending with leaders of both major parties supportive of aid. In the low aid condition, by comparison, America was positioned as spending little on foreign aid, with leaders wishing to keep spending low. Measures Focal measures for the current study are outlined below, and the full questionnaire is available on request. Manipulation check: aid spending. A single item (i.e., “What percentage of the federal budget do you believe America spent on aid to other countries in 2015?”) served to check the effect of the manipulation on perceptions of foreign aid spending. Manipulation check: leader support. A single item (i.e., “In general, do you think America’s leaders approve or disapprove of foreign aid?”, 1 = Strongly disapprove, 7 = Strongly approve) evaluated the effect of the manipulation on perceptions of leadership support for foreign aid. Perceived norms. Perceptions of American norms for giving to international charities were assessed. One item measured the perceived Descriptive Norm (i.e., “In the last 12 months, what percentage (0-100%) of Americans do you believe made donations of money to charities that help 65 people in other countries?”). A second item measured the perceived Injunctive Norm (i.e., “Would the average American agree or disagree that giving to international charity is the right thing to do?”, 1 = Totally disagree, 7 = Totally agree). For both items, higher values represent stronger norms for international giving. Identification. Participants’ national identification assessed by the same three items used in Study 4.1, adapted from Ellemers and colleagues (1999). Scores were averaged with higher scores indicating stronger identification, α = .93. Donation willingness. Participants reported their personal willingness to donate in response to a single question (“How willing would you be to send a donation of your personal money to help people in need overseas?”, 1 = Extremely unwilling, 7 = Extremely willing). Donation behavior. Participants were given the opportunity to answer a series of math and word problems to go into a draw to win a real donation of up to $100 for an international charity of their selection (four large, well-reputed international charities were presented—CARE, World Vision, Unicef, and Catholic Relief Services—with the option to specify another charity if desired). They were informed that participation was voluntary and time spent on the task would not be compensated. Participants could answer as many or as few problems as they wished, and could continue with the survey at any time. One response was randomly selected at the end of the study, with each correct answer that they had completed earning a $1 donation for the selected charity.8 Participants answered an average of 27.92 questions (SD = 35.70, range = 0 - 99), with many (55%) electing not to answer any. Analyses presented below therefore use donation behavior that has been transformed using the Inverse Hyperbolic Sine method, which accounts for skewness while maintaining zero values (see Friedline et al., 2015).9 This transformation method therefore allows the retention of data from those who opted not to complete any puzzles. Results Manipulation Checks Independent t-tests revealed that participants in the high aid condition perceived a significantly larger percentage of the United States’ gross national income to be spent on aid (M = 7.07, SD = 5.80) compared to participants in the low aid conditions (M = 0.46, SD = 0.51), p < .001, Cohen’s d = 1.61. Likewise, participants in the high aid condition perceived American political leaders to approve of government spending on foreign aid (M = 5.86, SD = 1.07) more than participants in the low aid condition (M = 3.59, SD = 1.42), p < .001, d = 1.81. Results indicate that manipulation of aid was effective. There was only a moderate association between perceptions of

8 This offer was real. The draw at the end of the study resulted in a $95 donation to CARE. 9 No substantive changes to the model are observed if untransformed values are retained. 66 aid spending and leader support (r = .40, p < .001), and these manipulation checks were retained as measured variables in the model below. Preliminary Analyses Means, standard deviations, and zero-order correlations are reported in Table 11. Overall, participants perceived Americans to be supportive of making international donations (injunctive norm [1-7]; M = 5.09, SD = 1.10) but relatively unlikely to do so (descriptive norm [0-100]; M = 23.55, SD = 18.67). This highlights the international aid context as a possible source of conflicting norms, a point we return to later. Perceptions of descriptive and injunctive norms were weakly positively associated (r = .22, p < .001). Participants who thought more aid was being sent, also thought more Americans gave to international charity (higher descriptive norm). Perceiving leaders supported aid, that Americans gave to international charities (higher descriptive norm), and that Americans endorsed such giving (higher injunctive norm) were all positively associated with willingness to make a personal donation. However, only perceiving a higher descriptive norm was associated with donation behavior. People who reported they were willing to donate also showed more effort behaviorally (r = .37, p < .001). National identification was not significantly associated with norms or giving responses. Focal Analyses Moderated serial mediation analysis was conducted using a series of models in the PROCESS macro for SPSS (Hayes, 2012; models 4, 14, and 6). The effects of the manipulation of aid context on both donation willingness and behavior were expected to be mediated by dual pathways—via descriptive and injunctive norms. Both pathways were expected to be stronger for high national identifiers. Full results of the analysis are reported in Table 12. For simplicity, a summarized results model is also presented in Figure 2, which includes only the focal pathways. Crowding in via descriptive norm. When the manipulated aid context was high, participants perceived more money being spent by the nation on aid (ß = .65, p < .001) which, in turn, led to thinking more Americans gave money to international charity (ß = .25, p = .001). That, in turn, led to greater personal willingness to donate (ß = .13, p = .038) and more effort exerted to earn a potential donation (ß = .17, p = .007). Bootstrapping analysis with 5000 resamples revealed significant positive indirect effects of the aid manipulation via perceived aid levels and descriptive norm on both donation willingness, B = .04, SE = .02, 95% CI [.01, .08], and donation behavior, B = .06, SE = .03, CI [.02, .13].

67

Table 11. Descriptive statistics and zero-order correlations between perceived aid, leader support, norms, identification, and donation responses M SD 1 2 3 4 5 6 1 Perceived aid (0-20%) 3.82 5.31 *** 2 Perceived leader support (1-7) 4.75 1.69 .40 ** 3 Descriptive norm (0-100%) 23.55 18.67 .18 .10 *** 4 Injunctive norm (1-7) 5.09 1.10 -.02 .12 .22 5 Identification (1-7) 4.97 1.47 .01 -.06 .07 .05 6 Donation willingness (1-7) 3.93 1.84 .05 .21*** .17** .19** -.05 7 Donation behavior (0-100) 27.92 35.70 .05 .12* .19** .10 .09 .37*** Note. N = 287 (Listwise). Donation behavior has been transformed for the analyses using Inverse Hyperbolic Sine to account for skewness; however, mean and standard deviation reported in this Table are untransformed values for clarity. *p < .05; **p < .01; ***p < .001

68

Table 12. Results of moderated serial mediation analysis of the effects of government aid on donation responses via perceived norms and moderated by national identification Perceived Descriptive Injunctive Donation Donation aid (ß) Norm (ß) Norm (ß) Willingness (ß) Behavior (ß) Manipulated aid (High) .65*** -.18† .19* .00 .05 Perceived leader support -.04 .09 .04 .20* .10 Perceived aid .25*** -.18* -.06 -.07 Descriptive norm (DN) - .21*** .13* .17** Injunctive norm (IN) .21*** - .15* .05 Identification (ID) -.05 .09 DNxID .09 .09 INxID -.03 -.09

R2 .39*** .08*** .07*** .10*** .07** Note. N = 287. Donation behavior has been transformed using Inverse Hyperbolic Sine to account for skewness. †p = .05; *p < .05; **p < .01; ***p < .001

69

Crowding out via injunctive norm. The perception that the government spent more on aid also led participants to perceive that the average American would agree less that giving to international charities was the right thing to do (ß = -.18, p = .015). However, those who perceived a higher injunctive norm reported a greater willingness to make a personal donation (ß = .15, p = .012) but showed no difference in charitable behavior (ß = .05, p = .394). Bootstrapping analysis with 5000 resamples revealed a significant negative indirect effect of the aid manipulation via perceived aid levels and the injunctive norm on donation willingness, B = -.03, SE = .02, CI [-.09, - .01], but not on donation behavior, B = -.01, SE = .02, CI [-.06, .01]. Identification. National identification neither directly predicted donation responses, nor moderated the effect of descriptive or injunctive norms, ps > .126. Consideration of leader support. In the analyses above, perceived spending was the focal mediator (controlling for leader support). While perceived aid spending and leader support were positively correlated, the strength of the association was weaker than anticipated and each had different effects on private giving. Leader support had no influence on norms, a direct positive effect on donation willingness, but none on behavior. Further, unexpected suppression effects were found, as shown in Table 12. We return to these findings in the discussion, but the key point is that responses to government aid and responses to the four leaders’ support for aid were different, with the responses to the leaders more variable and less predictable. Summary. Overall, the proposed model significantly explained 10% of the observed variance in willingness to make a donation, F(8,278) = 3.70, p < .001, and 7% of the variance in behavior, F(8,279) = 2.73, p = .006.

Figure 2. Summary of results of the moderated mediation analysis of the effect of manipulated aid levels on international donation willingness and behavior via perceived aid levels and norms Notes. Values represent standardized path coefficients. Dashed lines indicate non-significant pathways. Full results are reported in Table 12. *p < .05; **p < .01; ***p < .001

70

Discussion Study 4.2 replicates the key finding of Study 4.1 in a second national context, using a larger community sample and more robust design. Government foreign aid policy can influence citizens private giving via norms. Strong support was found for government spending crowding in private donations via the descriptive norm (H1) and modest support for crowding out via the injunctive norm (H2). Indirect effects were found on both private donation willingness (via both types of norms) and donation behavior (only via descriptive norm). No support, however, was found for strength of national identification moderating these indirect effects (H3). To our knowledge, this is the first study to show that perceived government aid spending can have competing effects on descriptive and injunctive norms. The flow on effects from norms to donation willingness and behavior corroborate previous work showing that norms can influence socially desirable attitudes and behaviors (e.g., Agerström et al., 2016; Cialdini et al., 1990; McDonald et al., 2013; Nolan et al., 2008). Overall, these effects were stronger on willingness than behavior (corroborating J. R. Smith & Louis, 2008), although there was a positive association between the two measures. As in Study 4.1, the normative pathways were no stronger for high than low identifiers (contradicting H3 and J. R. Smith & Louis, 2008; Tajfel, 1981; Terry & Hogg, 1996). However, the unexpected negative direct effect of national identification on international charitable giving found in Study 4.1 was not replicated in Study 4.2. Study 4.2 employed a larger, community sample that was more demographically similar to the the general donor population. Furthermore, the use of an experimental design and measurement of specific norms matched to target behavior builds confidence in the results. An important direction for future study, however, would be to manipulate aid spending and leader support orthogonally to tease apart their effects. In this study, it appeared that responses to the leaders’ support for aid and to government aid spending were different, with the former more variable and less predictable. This heterogeneity may be an artefact of the polarizing nature of the four leaders considered (i.e., Bush, Clinton, Obama, and Trump), or may reflect a real difference in responses to the motives attributed to the different leaders (Thomas, Amiot, Louis, & Goddard, 2017). General Discussion This paper presents evidence that government policy affects citizens’ behavior via diverging norms. Across two studies, we show that foreign aid spending influences private giving to international charities by changing perceptions of the national giving norms. Specifically, we demonstrate that government spending increases private giving via the perception that more citizens give (H1: crowding in via descriptive norm). We also find partial support for the notion that government spending simultaneously reduces private giving via the perception that citizens endorse international giving less (H2: crowding out via injunctive norm). Neither normative pathway was

71

moderated by strength of national identification (contrary to H3). These findings shed light on the contradictory evidence base for the crowding out hypothesis: opposing normative processes may be at play, with overall effects dependent on the relative salience of the injunctive or descriptive norm. While the present data support this contention in the context of international aid, norm divergence processes could be at work in crowding out and prosocial contexts more broadly. Put simply, leaders’ prosocial actions may often both motivate us to act (via descriptive normative influence) and decrease the impetus to act (via reduced injunctive pressure on individual group members). To our knowledge, the present studies provide the first evidence of simultaneous crowding out and in, and are the first studies to test group norms as a mechanism for such effects. Crowding In and Crowding Out: Support for the Norm Divergence Model Overall, we find strong evidence for government spending on foreign aid crowding in donations to international charity (see also Herzer & Nunnenkamp, 2013). We demonstrate that this effect occurs through the mechanism of descriptive norms. When the government spends more on foreign aid, it increases the perception that the group gives to international charity. In turn, this increased descriptive norm increases private donation responses. We demonstrate this indirect effect on donation likelihood (Study 4.1), willingness, and behavior (both Study 4.2). Descriptive norms have previously been shown to influence a range of socially desirable behavior, including charitable giving (Agerström et al., 2016; Bartke et al., 2017; Croson et al., 2009; McDonald et al., 2013; Nolan et al., 2008). The current study confirms and extends that evidence base by demonstrating that the perceived norms of group members themselves can be influenced by institutional signals (i.e., government policy and leader rhetoric) and can explain observed instances of crowding in. At the same time, the data show that government spending may simultaneously decrease private giving to international charities by diffusing the perceived responsibility for individual group members to act too. When the government steps in to meet a need, there may be less pressure for citizens to also respond, and thus a reduction in the perceived injunctive norm to give. In the second study, using an experimental design, a larger sample size, and matching norms to target response, we find the hypothesized indirect pathway of crowding out via injunctive norms is significant on donation willingness but not behavior. Injunctive norms have been shown to influence behavior in some contexts (e.g., littering; Kallgren et al., 2000; Reno et al., 1993) but not others (e.g., student issues; J. R. Smith & Louis, 2008). In the case of donating to international charity, they did not. It is possible that these weaker effects on injunctive norms may be due to a restriction of range in terms of need. The perception of need for foreign aid may be so high, and the social approval of giving so whole hearted, that our manipulation of government spending made little 72 difference. A test of the norm divergence model in a context where need is more contained may yield stronger results for the crowding out pathway. Alternatively, these weaker effects may reflect the more effortful nature of conforming to injunctive (vs. descriptive) norms (Jacobson et al., 2011; Jacobson, Mortensen, Jacobson, & Cialdini, 2015). Future research could specifically test diffusion of responsibility as the mediator of a declining injunctive norm in response to increased government spending, or assess perceived effort in conforming to different national giving norms. Finally, although we find some evidence for the injunctive norm crowding out pathway, it is possible that other factors may motivate people to rush in and fill the void left by policy changes (or stand back when enough is being done). In both studies, people perceived that other citizens endorsed international giving but did not necessarily give themselves. Previous research has shown that conflicted norms can be demotivating for some people (e.g., McDonald et al., 2013; J. R. Smith & Louis, 2008). However, groups and individuals can sometimes be mobilized to act by the perception that norms have to change (e.g., L. G. E. Smith, Thomas, & McGarty, 2015). For example, even high identifiers may ignore existing norms if they think they could benefit the group by doing so (Packer, 2008). Thus, the perception of normative conflict may mobilize some people (McDonald et al., 2013, 2014; Plows et al., 2017; L. G. E. Smith, E. F. Thomas, et al., 2015). So– called “reluctant altruists”, for example, are most willing to act when they perceive that others will not (Veldhuizen, Kort, Ferguson, & Atsma, 2012). An interesting avenue for future research will be to examine how perceiving unfavorable norms may in fact shape donor identities and motivate them to give more. A final consideration is that donors to international charities may actually respond differently to changing policies and perceived norms than the average citizen (see related discussions in Schultz, Nolan, Cialdini, Goldstein, & Griskevicius, 2007; Thomas, McGarty, Reese, Berndsen, & Bliuc, 2016). Future work should therefore endeavor to study donors specifically. The Impact of Identification Norms are generally most influential for individuals who identify with the group (J. R. Smith & Louis, 2008; Tajfel, 1981; Terry & Hogg, 1996; Turner et al., 1987). Thus, it was expected that both pathways in the norm divergence model would be accentuated for participants who identified more strongly with the nation. No evidence of this was found. One reason may be that, although national identity may be implied in intergroup helping contexts (e.g., sending foreign aid), identity was not made salient in our studies. Given that foreign aid spending influenced national group norms, national identity must have been activated to some extent. Nonetheless, alternative identities may also be at work that obscured the impact of national identity alone. For example, making the international context salient or presenting leaders’ photos and quotations may have instead cued political identities as Republican or Democrat; or making giving salient may have cued religious identities. Both political and religious identities have been associated with giving 73 responses (see Chapman, Louis, et al., 2018). Future research should measure the salience of alternative identities explicitly, or specifically cue national identity to investigate its influence on the normative pathways observed here. Strengths, Limitations, and Future Directions The research presented here has a number of strengths. We evaluated a social psychological model of normative influence that can explain past contradictory crowding out and in findings. A novel crowding in pathway via descriptive norms was replicated across two national contexts. Further, the experimental manipulation of aid to establish causality, and the combination of self- report and behavioral donation measures both contribute to the robustness of the findings reported here. Two limitations also warrant discussion, however. First, the sample size for Study 4.1 was determined primarily by pragmatic concerns, with students recruited until the end of semester. As such, Study 4.1 was likely underpowered. The sample size for Study 4.2, however, was determined a priori and both replicated the crowding in pathway and detected the hypothesized crowding out pathway. Future research should be attentive to power concerns. Also, the evidence presented here shows the effect of relatively short exposures to changing aid policies on citizens’ charitable behavior. In actuality, policies change often—sometimes gradually, sometimes drastically. A fruitful direction for future study would be to investigate longitudinally how real-world policy changes are associated with national giving rates, and evaluate whether these can be explained by changing norm perceptions. One factor that influences the size and direction of crowding out effects is the type of charity in question (Andreoni & Payne, 2011; De Wit et al., 2017; Khanna et al., 1995; Okten & Weisbrod, 2000). Different psychological processes may motivate giving to different categories of charity (Chapman, Louis, et al., 2018), and researchers have called for research designed to specifically investigate how crowding out changes across charity contexts (De Wit et al., 2017). While the present research shows the norm divergence model is useful for predicting the impact of foreign aid on private donations to international charities, it would be interesting in future research to manipulate the charity type explicitly and examine how attributes of that charity context affect the relative strength of the two normative paths. Practical Implications Study 4.1 shows that the public is largely misinformed about aid. In actuality, foreign aid is very low (less than 0.5% of national income for most countries; OECD, 2016) but people perceive it to be very high (average estimates of 10% in the student sample and 4% in the community sample). Study 4.2 shows that perceptions of aid can be easily influenced by the way information is presented, with people in the low aid condition accurately estimating aid levels but participants in 74 the high aid condition grossly overestimating spending. How policy is framed clearly matters. Given this information is demonstrated to affect private giving decisions, politicians and media outlets must be very careful about the way they present information and the potential flow-on effects. Furthermore, the evidence presented here that norms, particularly descriptive norms, influence giving decisions can be used by organizations to mobilize support. While an intuitive point, it is also important: creating a sense that others in the group are giving should increase peoples’ willingness to give too. Conclusion Government policies can influence perceived national norms. Such norms, in turn, influence charitable giving decisions. In particular, increased foreign aid spending at the national level can provide cues about the national group’s descriptive norm (i.e., what we do) around international giving, which can increase citizens’ private giving. However, when the government spends more, this may simultaneously reduce the perceived injunctive norm (i.e., what we should do) and subsequently reduce citizens’ willingness to give personally. These simultaneous, oppositional pathways evoked by the public decisions of group representatives therefore promote crowding in and crowding out at the same time. Demonstrating the existence of competing normative pathways, the research and norm divergence model presented here help explain the range of contradictory findings in the economic and psychological literatures on the crowding out hypothesis. Testing the utility of the model in other prosocial contexts is an important direction for future research.

75

Chapter 5. Rage Donations: Examining Charitable Giving as a Form of Collective Action Motivated by Anger

This chapter provides further evidence that salient characteristics of the social context influence charitable giving. Two experiments show the ways that political advocacy on contentious social issues affect both willingness to make personally congruent donations and donation behavior. In particular, these studies show that political advocacy can affect donor emotions (anger), feelings of group efficacy, and identification with particular beneficiaries. All of these, in turn, affect donation choices.

Citation Chapman, C. M., Mirnajafi, Z., Masser, B. M., & Louis, W. R. (under review). Rage donations: Examining charitable giving as a form of collective action motivated by anger. Manuscript submitted to Personality and Social Psychology Bulletin, 19 December 2018.

Contributions The work presented in this chapter was developed by me and Zahra Mirnajafi under the guidance of my research supervisors Professors Winnifred Louis and Barbara Masser. I collected data for Study 5.1, analyzed all data reported in this chapter, and wrote the manuscript. The entire team participated in survey design and Zahra Mirnajafi collected data for Study 5.2. My three co- authors provided critical feedback on earlier drafts of the manuscript.

76

Rage donations: Examing charitable giving as a form of collective action motivated by anger

Abstract As documented by the media, ‘rage donations’ are made in response to public statements or policy directions that provoke outrage. Anger has rarely been tested as a motive for giving but is a known driver of collective action. Across two experiments—in the contexts of debates about racial discrimination (N = 219) and abortion (N = 221)—we examined rage donations empirically and tested the possible underlying mechanisms of anger, efficacy, and identification. When exposed to advocacy that opposed their own views, participants experienced more anger. This, in turn, predicted their willingness to make cause-congruent donations, but not their giving behavior. Effects of efficacy were inconsistent, while identification with perceived victim groups was consistently associated with giving responses but not consistently influenced by advocacy. Results suggest rage donations are an emerging form of ally collective action that can be provoked by opponent advocacy.

Keywords: charitable giving, donations, collective action, anger, identity

77

Introduction After taking office in 2017, U.S. President Donald Trump announced an executive order to ban visitors from six Muslim-majority countries (Trump, 2017). In response, people donated $24 million to the American Civil Liberties Union in just one week; seven times as much as they had received in the entire previous year (Stack, 2017). Later that year, a right-wing politician in Australia tweeted criticism of a school’s ‘do-it-in-a-dress’ fundraiser for girl’s education in Africa, calling it “absurd” and “gender bending.” In response, donations flooded in from around the world—with the school raising over $180,000 (their target was $900; BBC, 2017). These are examples of a new phenomenon: ‘rage donations’ (Fetters, 2017). The concept of rage donations has been discussed in the media, and fundraising practitioners have sought ways to capitalize on the motivating power of rage (Fetters, 2017; Teson, 2017). Yet it remains uncertain whether or not anger underpins such donations. Anger has rarely been considered a motive for charitable giving, but is known to promote collective action (e.g., van Zomeren, Postmes, & Spears, 2008). In this paper, we present two experiments testing for rage donations (collective giving provoked by anger). We also test the mechanism of anger against two other established motives of collective action: efficacy and identification. Anger and Charitable Giving A variety of emotions have been shown to affect charitable giving, including gratitude, guilt, and sadness (Basil, Ridgway, & Basil, 2008; Hibbert, Smith, Davies, & Ireland, 2007; Liang, Chen, & Lei, 2016; Ma et al., 2017; Small & Verrochi, 2009; Zhou et al., 2012). Anger, however, has only occasionally been considered as an emotion with prosocial outcomes. Mixed evidence has been found for a role of anger in giving. For example, one study found no relationship between experiencing anger while visiting the U.S. Department of Motor Vehicles—where waiting times can be long and frustration is common—and choosing to register as an organ donor (Siegel et al., 2016). Feeling anger about the actions of others is associated with increased moral courage—willingness to take action at great personal risk—but is not associated with general helping (Kayser, Greitemeyer, Fischer, & Frey, 2010). Although moral outrage does not affect how much study prize money people will donate to charity, it can promote justice-related actions like political participation (Van de Vyver & Abrams, 2015). Similarly, evoking anger through recalling autobiographical events can increase people’s reported willingness to make hypothetical donations, but only to charities that have a focus on restoration of justice (van Doorn, Zeelenberg, & Breugelmans, 2017). The evidence is therefore inconclusive but suggests that anger can motivate giving under certain conditions. The studies described above examine the effects of anger that was unrelated to the cause in question—prompted, for example, by autobiographical recall (e.g., van Doorn et al., 2017), 78 emotion-enducing videos (Van de Vyver & Abrams, 2015), or other experiences unrelated to the act of giving (Siegel et al., 2016). Rage donations, however, emerge as a specific response to a specific provocation, with emotion and response linked to a common, collective issue. Anger and Collective Action Anger is proposed as one of three key motives for collective action (Thomas, Mavor, & McGarty, 2012; van Zomeren et al., 2008). According to the Social Identity Model of Collective Action (SIMCA; van Zomeren et al., 2008), identification with a group is a prerequisite of experiencing group-based anger over perceived injustice and perceiving the group to be effective in making a difference. Anger, efficacy, and identification are key drivers of collective action to benefit one’s own group (e.g., Thomas et al., 2012; van Zomeren, Kutlaca, & Turner-Zwinkels, 2018; van Zomeren et al., 2008; van Zomeren, Spears, Fischer, & Leach, 2004). However, groups can also act as allies of disadvantaged groups when spurred by violations of their moral convictions—deeply held beliefs about what is right—to act on behalf of perceived victims of social injustice (van Zomeren, Postmes, Spears, & Bettache, 2011). The examples of rage donations—such as the response of U.S. citizens to immigration bans and the responses of the Australian public to transphobic rhetoric— suggest rage donations are typically a form of ally collective action. As such, identification with perceived victim groups (rather than an ingroup) is likely to influence collective action responses. The Current Research The current research tests rage donations as a form of collective action, with anger, efficacy, and identification as potential mechanisms motivating donations after provocation. Given that people prefer to give to causes that reflect their own values (Chapman, Louis, et al., 2018; Chapman, Masser, & Louis, under review), individuals are expected to donate more to charities that promote their stance on contested social issues (vs. charities that promote the opposition agenda; H1). Based on SIMCA, we test three possible mechanisms through which exposure to advocacy from the opposition side promotes giving to cause-congruent donations—through increased anger (H2a), increased perceptions of efficacy (H2b), and increased identification with the ingroup’s victim group and/or decreased identification with the outgroup’s victim group (H2c). Anger is implied to be the dominant pathway promoting rage donations. However, it is also possible that efficacy and identification could play a role in promoting collective giving in response to provocation. Advocacy from an opponent group may highlight donations as a symbolically or instrumentally effective way to oppose their agenda. Thus, efficacy could be a viable mediator. Alternatively, advocacy from one’s own side may highlight a victim group and increase identification with them. Advocacy from the opposing side, on the other hand, may highlight a 79 different victim group but inadvertently promote disidentification with them as the donor rejects not only the opponent but their whole agenda. Thus, identification could also be a viable mediator. The current research makes several important contributions to the literature. First, it is the first empirical investigation of the new social phenomenon called ‘rage donations.’ Second, it asks if and how political advocacy can influence responses in relation to social issues. Third and finally, it applies SIMCA to understand rage donations, thus examining charitable giving as a form of collective action. Study 5.1 Study 5.1 was conducted in the context of debate over whether or not athletes should be allowed to kneel during the U.S. national anthem. Starting in 2016, NFL player Colin Kaepernick refused to stand during the pre-match national anthem in protest against racial discrimination and police brutality. Others followed Kaepernick’s lead. Study 5.1 was an online experiment with American participants during December 2017, when some form of active protest was being engaged in by at least 23 players from 10 of the 32 NFL teams (ESPN, 2018). At that time, both President Trump and Vice-President Pence had publicly stated their strong opposition to kneeling during the national anthem (Landler, Belson, & Haberman, 2017) and Kaepernick had recently been named the ‘Citizen of the Year’ by GQ magazine (Editors of GQ, 2017). The volatility of the NFL kneeling debate suggested it was an appropriate context in which to test the existence and mechanisms of rage donations. Method We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study (J. P. Simmons, Nelson, & Simonsohn, 2012). The methods, hypotheses, and analyses for this study were pre-registered on the open science framework (https://osf.io/8q4fm/?view_only=df718bc2cfef4924bb291da91eede455). Participants Two hundred and twenty-eight American residents took part in an online experiment in December 2017. The sample size was determined by an a priori power analysis for 90% power to detect medium (f = .25) effects, with a 10% buffer to allow for any potential exclusions. Given no research has previously examined rage donations, we considered a medium effect would be the minimum effect worth detecting to establish the phenomenon. One person asked for their data to be deleted after debriefing and a further 5 participants were excluded for giving an incorrect answer to an attention check question (“To show you're paying attention, please select "totally agree" for this item”). In addition to these planned exclusions, two additional participants were excluded because they reported technical glitches that compromised their responses and one for failing the

80 comprehension check on all five attempts. The final sample (N = 219) included 130 females and 89 males. Participants ranged in age from 19 to 73 years (Mage = 40.50, SD = 13.09). Procedure Participants were recruited online via the Amazon Mechanical Turk platform. The advertisement invited participants to a study of attitudes and emotions about social issues. First, participants completed a 1-minute screening survey assessing demographic information and asking their stance on a range of current social issues, for which they received payment of USD$0.05. Only participants who either opposed or supported NFL players kneeling during the national anthem were invited to participate, while people without clear views on the issue were screened out. Among respondents to the screening survey, 49% held pro-kneeling views, 29% anti-kneeling views, and 22% had no clear views on the issue. Quotas ensured that equal numbers of participants held each view (n = 114 pro-kneelers and n = 114 anti-kneelers). The full experiment took approximately 5 minutes to complete, for which participants were compensated with a bonus of USD$0.70. First, participants reported how strongly they felt about NFL players kneeling during the national anthem and the degree to which they identified with others who held the same position as them on the issue. Participants were then randomly allocated to read public tweets that either advocated support of players kneeling, opposed players kneeling, or to a control condition in which they read no tweets. Advocacy manipulations are summarized below. After reading the tweets, participants completed a comprehension check to ensure they had understood the content of the tweets (i.e., “What was the key message being advocated in the tweets you just read?: Players who kneel during the national anthem are disrespecting the nation”; or “Players who kneel during the national anthem are protesting racism and police brutality”). If participants selected the incorrect response they were looped back to read the tweets again and had five chances to answer correctly. Of the 219 participants that passed the comprehension check, 5 required two attempts, 2 required three attempts, and all others passed on the first attempt. Once the comprehension check was successfully completed, participants indicated their emotional response to the tweets, their state anger and contempt for the advocates who wrote the tweets, identification with the advocates and with perceived victim groups (i.e., people of color or police and soldiers), and perceived group efficacy in relation to the issue. Next, participants reported how willing they were to make a personal donation to charities that were supporting either side of the kneeling issue and then had the opportunity to allocate a real donation. Finally, they completed demographic questions and were debriefed.

81

Advocacy Manipulation In each of the two experimental advocacy conditions, eight public tweets were presented. All tweets were real and publicly visible. In the pro-kneeling condition, all tweets supported NFL players who kneeled in protest during the national anthem and took the perspective that such actions were a fair protest for systemic racial injustice and police brutality against people of color. In the anti-kneeling condition, all tweets opposed players kneeling and took the perspective that such actions were disrespectful to the American flag and to those who served the nation such as police officers and the military. In the control condition, no tweets were presented but, when answering questions that followed, participants were instructed to think about any tweets or media discussion they may have been exposed to that discussed kneeling during the anthem. For the analyses reported below, the advocacy conditions were coded in relation to the participant’s own stance on the issue. For example, for an anti-kneeling participant, exposure to advocacy that endorsed kneeling was coded as opposing (the views of the participant; +1); exposure to advocacy that was against kneeling was coded as supporting (the views of the participant; -1); and no exposure was coded as 0. Two contrasts were created for the purpose of regression analysis. To examine the effects of exposure to any advocacy (vs. control), the two advocacy conditions were each coded +1 and the control was coded -2. To zero in on the effect of opposing (vs. supporting) advocacy, opposing advocacy was coded +1, supporting advocacy -1, and the control condition as 0. Measures Focal measures for the current study are outlined below. Except where otherwise noted, all items were measured on 7-point Likert scales: 1 = Strongly disagree, 7 = Strongly agree, and scales were averaged with higher scores indicating more of the construct in question. Anger. Two items adapted from Tausch and colleagues (2011) measured participants anger about the advocacy (“I was outraged about what they said” and “I felt angry at what they said”), r = .91, α = .95. Efficacy. Two items adapted from van Zomeren and colleagues (2011) measured the degree to which participants felt they could effect change in relation to the kneeling issue (“I think together individuals can make a difference on this issue” and “I think together we can successfully change things in relation to this issue”), r = .85, α = .92. Identification. Two items adapted from van Zomeren and colleagues (2011) measured identification with perceived victim groups (e.g., “I identify with [those who serve and protect, like soldiers and police / people of color]”; rs = .87 and .78, α = .93 and .88, respectively). As noted above, the pro-kneeling advocacy highlighted people of color as the victim group while the anti- kneeling advocacy highlighted soldiers and police as the victim group. Victim groups were 82 therefore coded in relation to the participants’ own stance on the issue. For pro-kneelers, people of color were the ingroup’s perceived victim group while soldiers and police were the outgroup’s perceived victim group. For anti-kneelers, this coding was reversed. Donation willingness. Participants were asked “How willing would you be to make a donation of your personal money today to an organization that: works toward social justice for minorities” (pro-kneeling cause; 1 = Not at all willing, 7 = Extremely willing), and “works to ensure all people respect the national anthem” (anti-kneeling cause). Responses were coded in relation to the participants’ own stance. Thus, for pro-kneelers, the pro-kneeling cause was the congruent cause and the anti-kneeling cause was the incongruent cause; and vice versa for anti-kneelers. Donation behavior. Despite evidence that people overreport socially desirable behaviors (e.g., Bekkers & Wiepking, 2011a), most research on collective action and charitable giving has assessed intentions or other self-report measures (Tausch et al., 2011; van Zomeren et al., 2008). We therefore incorporated a behavioral measure into the study. Participants were advised that, in addition to their compensation for participation (USD$0.75), the researchers were willing to make a USD$0.25 donation on their behalf to a charity of their selection. As a behavioral measure for the study, participants read short descriptions of seven organizations before indicating which they wished their donation to be given to. All donation choices were subsequently honored by the researchers. Two organizations reflected priorities of the pro-kneeling cause (i.e., Black Lives Matter and the National Association for the Advancement of Colored People), two reflected the priorities of anti-kneeling cause (i.e., Blue Lives Matter and The American Legion), and three organizations were unrelated to the kneeling debate (i.e., ASPCA, United Way, and UNICEF). Participants could elect not to have a donation made on their behalf but did not have the option of keeping the money themselves. Choices were coded based on congruence with the participant’s stance on the kneeling issue. Thus, for a person that was pro-kneeling, donations to pro-kneeling causes were coded 1 (for congruence), and all other options were coded as 0. For participants who were anti-kneeling, donations to anti-kneeling causes were coded 1 and all other options were coded 0. Results Means, standard deviations, and zero-order correlations between all variables are reported in Table 13. Personal stance on the kneeling issue was significantly associated with several variables. People who supported players kneeling also reported greater efficacy (r = .34, p < .001) and identification with the other side’s victim group (r = .16, p < .001) but lower identification with their own side’s victim group (r = -.34, p < .001) than did people who opposed kneeling. In

83

Table 13. Means, standard deviations, and zero-order correlations for Study 5.1 (NFL Players Kneeling) M SD 1 2 3 4 5 6 7 8 1 Pro-kneeling stance (-1, 1) 0.00 1.00 2 Opposing advocacy (-1, 0, 1) 0.00 0.81 -.01 3 Exposure to advocacy vs control (1, -2, 1) -0.04 1.43 .01 .00 4 Anger 3.84 2.10 -.01 .53*** -.22** 5 Efficacy 5.56 1.44 .24*** -.01 -.06 -.09 6 Identification: ingroup’s victims 5.33 1.44 -.34*** .07 -.08 .16* .03 7 Identification: outgroup’s victims 4.13 1.68 .16*** .03 .04 -.01 .23*** .20** 8 Donation willingness: congruent cause 4.53 2.10 .34*** .09 .03 .20** .14* .14* .13 9 Donation behavior: congruent cause (0, 1) 0.30 0.46 -.02 -.02 .02 .02 .05 .19** .04 .14* Note. N = 217 (Listwise). One person elected not to respond to the identification questions. *p < .05, **p < .01, ***p < .001

84 addition, pro-kneelers reported a greater willingness to donate to congruent charities than anti- kneelers did (r = .34, p < .001). As expected, exposure to opposing advocacy was associated with greater anger than exposure to supporting advocacy (r = .53, p < .001). However, exposure to any advocacy was associated with reduced anger relative to the control condition (r = -.22, p < .001), suggesting that when unprompted people spontaneously think of opposing advocacy. To account for these relationships, participant stance and the overall effect of advocacy (versus control) were both included as control variables in the analyses that follow. H1: Collective Giving Participants reported greater willingness to donate to a cause that was congruent with their stance on the kneeling issue (M = 4.53, SD = 2.10) than a cause that was incongruent with their stance (M = 2.63, SD = 2.02), t(217) = 10.82, p < .001. When it came to selecting a charity to receive a real donation, 67 participants selected a congruent charity while only 5 selected an incongruent charity.10 H2: Rage Donations and Other Explanations for Collective Giving Mediation analyses were conducted using the lavaan package for R (Rosseel, 2012). Anger, efficacy, and identification with both ingroup and outgroup victim groups were tested as parallel mediators of the effect of opposing versus supporting advocacy on donation willingness and behavior. Participant stance, the effects of exposure to advocacy versus control, and the interactions of advocacy and stance were included as control variables. The model explained 25% of the variance in willingness to make a congruent donation, F(9, 208) = 7.69, p < .001, and 9% of the variance in actual donation behavior, F(9, 208) = 2.29, p = .018. Results of the analyses are presented in Table 14 and focal results are summarized in Figure 3. Exposure to opposing (vs supporting) advocacy did not have direct effects on either participants’ willingness to make congruent donations or their actual donation choices, and the interaction of stance and advocacy on both donation measures were not significant, |ßs| < .14, ps > .071. Tests of the indirect effects are reported below. Anger. Exposure to opposing advocacy had a significant indirect effect on willingness to donate to a congruent cause via anger, IE = .33, SE = .10, 95% CI [.13, .53]. However, this indirect effect was not found on donation behavior, IE = .01, SE = .02, 95% CI [-.04, .05]. When exposed to opposing advocacy on a contested social issue, participants experienced more anger (ß = .53, p < .001) which, in turn, enhanced their willingness to make a cause-congruent donation (ß = .23, p = .002) but did not affect their actual donation choices (ß = .02, p = .767).

10 A total of 147 participants (67%) did not select a charity that was relevant to the kneeling issue: 60 donated to the ASPCA, 37 to Unicef, 31 to the United Way, and 19 preferred not to donate. 85

Table 14. Regressions for Study 5.1 (NFL Players Kneeling) Mediator (ß) DV: Donations (ß) Anger Efficacy Ingroup ID Outgroup ID Willingness Behavior Opposing (vs supporting) advocacy .53*** -.01 .06 .04 -.04 -.04 Advocacy exposure (vs control) -.22*** -.06 -.08 .04 .09 .04 Pro-kneeling (vs anti-kneeling) stance -.01 .23*** -.34*** .16* .40*** .09 Opposing advocacy x stance .11 .08 -.05 .10 -.14 -.03 Advocacy exposure x stance .00 .02 .01 .02 -.02 .03 ** Anger .23 .02 Efficacy .05 .07 *** *** Identification: ingroup’s victims .23 .27 ** Identification: outgroup’s victims .07 -.21

Model R2 .34*** .06* .13*** .04 .25*** .09* Note. N = 217 *p < .05; **p < .01; ***p < .001

86

Figure 3. Mediation analysis for Study 5.1 (NFL players kneeling) Note: controls for exposure to advocacy (vs control), stance (pro-life vs pro-choice), and interactions between advocacy and stance. N = 217.

87

Efficacy. No indirect effects of opposing advocacy were found via efficacy on either donation willingness, IE = .00, SE = .01, 95% CI [-.02, .02], or behavior, IE = .00, SE = .00, 95% CI [-.01, .01]. Identification. Participants who identified more with their ingroup’s victim group reported a greater willingness to donate to a congruent cause (ß = .23, p < .001) and were also more likely to actually donate to a congruent cause when given the chance (ß = .27, p < .001). In addition, participants who identified more with the outgroup’s victim group were significantly less likely to donate to a congruent cause (ß = -.21, p = .003), despite expressing no difference in their willingness to do so (ß = .07, p = .306). However, exposure to opposing advocacy did not influence identification with the perceived victim groups of either the ingroup (ß = .06, p = .320) or the outgroup (ß = .04, p = .592). Therefore, no indirect effects of opposing advocacy on donation responses were found via either identification with the ingroup’s victim group, IE on willingness = .04, SE = .04, 95% CI [-.04, .12] and IE on behavior = .01, SE = .01, 95% CI [-.01, .03], or identification with the outgroup’s victim group, IE on willingness = .01, SE = .01, 95% CI [-.02, .03] and IE on behavior = .00, SE = .01, 95% CI [-.02, .01]. Discussion In response to NFL players kneeling during the national anthem, people preferred to donate to causes that were congruent with their personal stance (supporting H1). Exposure to opposing advocacy provoked anger which, in turn, enhanced willingness to make congruent donations: a rage donation effect (supporting H2a). However, provocation did not affect donation behavior. Further, opponent advocacy did not influence participants’ perceived efficacy or identification with victim groups (no support for H2b and 2c). The rage effect was not found on behavior, perhaps because so many participants (68%) chose to donate to unrelated causes. We therefore sought to replicate the methodology and analyses of Study 5.1 in a new context and with a narrower range of giving options. Study 5.2 Abortion remains a fiercely contested social issue in many societies around the world. We accessed public tweets on either side of a debate about the 2018 United States H.R.36 Pain-Capable Unborn Child Protection Act, which was introduced to ban abortion after 20 weeks (Congress.gov, 2018). In addition, we refined the behavioral measure from Study 5.1 by reducing the number of charity options from eight to four. We specifically removed options to donate to help animals and children, which are attractive beneficiaries for many donors (Chapman, Masser, et al., under review).

88

Method We report how we determined our sample size, all data exclusions (if any), all manipulations, and all measures in the study (J. P. Simmons et al., 2012). The methods, hypotheses, and analyses for this study were pre-registered on the open science framework (https://osf.io/qg29d/?view_only=31a7ce8540fe43eaaaf52c586974898c). Participants The online experiment was run in July 2018. The sample size of 228 was determined by an a priori power analysis based on 90% power to detect medium (f = .25) effects, with a 10% buffer to allow for any potential exclusions. Three participants asked for their data to be deleted and 4 were excluded for failing the attention check (“To show you're paying attention, please select "strongly agree" for this question”). The final sample (N = 221) included 121 females, 99 males, and 1 participant who identified as gender fluid. Participants ranged in age from 18 to 75 years

(Mage = 35.61, SD = 12.50). Procedure The procedure was identical to Study 5.1 except screening was conducted on the basis of stance on the abortion issue (pro-life vs pro-choice). Among respondents to the screening survey, 64% held pro-choice views, 28% pro-life views, and 9% had no clear views on the issue (numbers do not add to 100% because of rounding). Quotas ensured that equal numbers of participants held each view (n = 114 pro-choice and n = 114 pro-life). Advocacy manipulations are summarized below. Advocacy Manipulation Eight tweets from either side of the abortion legislation debate were selected. All tweets were real and publicly visible. In the pro-life condition, all tweets advocated for the protection of unborn babies and most used the campaign hashtag #TheyFeelPain. In the pro-choice condition, all tweets advocated for the reproductive rights of women and used the campaign hashtag #StopTheBans. In the control condition, no tweets were presented but, when answering questions that followed, participants were instructed to think about any tweets or media discussion they may have been exposed to that discussed abortion. Measures The same anger, r = .87, α = .93, and efficacy, r = .84, α = .91, measures outlined in Study 5.1 were used. The victim group highlighted by pro-life tweeters was unborn babies (e.g., “I feel strong ties with unborn babies”; r = .86, α = .92), while the victim group highlighted by pro-choice advocates was women seeking abortions (e.g., “I identify with women accessing abortions”; r = .93, α = .97). For the donation behavior measure, participants could select between donating to the National Pro-Life Alliance (pro-life cause), Planned Parenthood (pro-choice cause), the Multiple 89

Sclerosis Society, or the United Way (two neutral causes). Again, choices were coded based on congruence with the participant’s stance on the abortion issue (1= congruent, 0 = incongruent). Results Means, standard deviations, and zero-order correlations between all variables are reported in Table 15. In the context of abortion, participant stance was not associated with anger, efficacy, or identification. However, people who held a pro-life stance were less likely to donate to a congruent cause than participants with a pro-choice stance (r = -.27, p < .001). As in Study 5.1, being exposed to an opposing advocacy was associated with greater anger than being exposed to a supporting advocate (r = .33, p < .001). Once more, however, exposure to any advocacy was associated with less anger than exposure to no advocacy (r = .28, p < .001), suggesting that participants were likely thinking about opposing advocacy in the control condition. Participant stance and exposure to advocacy (versus control) are both controlled for in the analyses that follow. H1: Collective Giving As predicted, participants reported a greater willingness to donate to a cause that was congruent with their stance on the abortion issue (M = 4.91, SD = 2.11) than to a cause that was incongruent with their stance (M = 2.17, SD = 1.78), t(218) = 14.81, p < .001. When donation behavior was observed, 134 participants selected a congruent charity, while only 15 selected an incongruent charity.11 H2: Rage Donations and Other Explanations for Collective Giving Identical mediation models as Study 5.1 were tested using the lavaan package for R (Rosseel, 2012). The model explained 25% of the variance in willingness to make a congruent donation, F(9,209) = 7.86, p < .001, and 29% of the variance in actual donation behavior, F(9,211) = 9.49, p < .001. Results of the analyses are presented in Table 16 and focal results are summarized in Figure 4. Exposure to opposing (vs supporting) advocacy did not have direct effects on either participants’ willingness to make congruent donations or their actual donation choices, and there were no significant interactions of stance and advocacy on donation measures, |ßs| < .13, ps > .076. Tests of the indirect effects are reported below. Anger. As predicted by the focal rage donations hypothesis, exposure to opposing advocacy had a significant indirect effect on willingness to donate to a congruent cause via anger, IE = .18, SE = .06, 95% CI [.06, .31]. Again, however, this indirect effect was not found on donation behavior, IE = .02, SE = .01, 95% CI [-.01, .04]. When exposed to opposing advocacy on a contested social issue, participants experienced more anger (ß = .32, p < .001) which, in turn,

11 Of the remaining participants, 69 (31%) selected a neutral charity and 3 (1%) did not answer. 90

Table 15. Means, standard deviations, and zero-order correlations for Study 5.2 (Abortion) M SD 1 2 3 4 5 6 7 8 1 Pro-life (vs pro-choice) stance (1, - 1) -0.03 1.00 2 Opposing advocacy (1, 0, -1) -0.02 0.82 .02 3 Exposure to advocacy vs control (1, -2, 1) 0.00 1.42 -.01 -.02 4 Anger 3.74 1.88 .06 .33*** -.28*** 5 Efficacy 5.42 1.47 .03 -.21** .03 -.07 6 Identification: ingroup’s victims 4.92 1.69 .05 -.13* -.05 .14* .06 7 Identification: outgroup’s victims 2.53 1.61 -.04 .16* .05 .13 .00 -.12 8 Donation willingness: congruent cause 4.91 2.11 -.03 -.05 -.02 .20** .27*** .36*** -.14* - 9 Donation behavior: congruent cause (0, 1) 0.61 0.49 .27*** -.17* .05 -.02 .05 .32*** -.30*** .24*** Note. N = 220 (Listwise). *p < .05, **p < .01, ***p < .001

91

Table 16. Regressions for Study 5.2 (Abortion) Mediators (ß) DV: Donations (ß) Anger Efficacy Ingroup ID Outgroup ID Willingness Behavior Opposing (vs supporting) advocacy .32*** -.21** -.14* .17* .00 -.09 Advocacy exposure (vs control) -.27*** .03 -.05 .05 .05 .09 Pro-life (vs pro-choice) stance .05 .03 .05 -.05 -.07 -.28*** Opposing advocacy x stance -.22** .12 .19* -.02 -.10 .13 Advocacy exposure x stance .02 .05 .00 .05 -.03 -.05 ** Anger .20 .08 *** Efficacy .28 .02 *** *** Identification: ingroup’s victims .30 .30 * *** Identification: outgroup’s victims -.13 -.28

Model R2 .22*** .06* .05 .04 .25*** .29*** Note. N = 220 *p < .05; **p < .01; ***p < .001

92

Figure 4. Mediation analysis for Study 5.2 (Abortion) Note. Controls for exposure to advocacy (vs control), stance (pro-life vs pro-choice), and advocacy x stance interactions. N = 220.

93 enhanced their willingness to make a cause-congruent donation (ß = .20, p < .001), although it did not affect their actual donation choices (ß = .08, p = .257). The effect of opposing advocacy on anger was qualified by a significant interaction with participant stance (ß = -.22, p = .003). Simple slopes analysis revealed that the effect of opposing advocacy on anger was stronger for pro-choice participants (ß = .52, p < .001) than pro-life participants (ß = .14, p = .145). However, as shown in Table 16, participant stance did not moderate the impact of advocacy on either willingness or action. Efficacy. Unexpectedly, a significant negative indirect effect of opposing advocacy was found on donation willingness via efficacy, IE = -.15, SE = .06, 95% CI [-.26, -.04]. Participants exposed to opposing advocacy about abortion reported a reduced sense of efficacy (ß = -.21, p = .002). Perceived efficacy, in turn, was positively associated with donation willingness (ß = .28, p < .001) but not behavior (ß = .02, p = .762). Thus, the indirect effect was not found on behavior, IE = .00, SE = .01, 95% CI [-.02, .01]. Identification. Participants exposed to opposing advocacy also unexpectedly identified less with their ingroup’s victim group (ß = -.14, p = .035) and more with the outgroup’s victim group (ß = .17, p = .013). The effect of opposing advocacy on identification with the ingroup’s perceived victims was qualified by an interaction with participant stance (ß = .19, p = .020). Simple slopes revealed that, for pro-life participants, exposure to opponent advocacy decreased their identification with their ingroup’s victim group (i.e. unborn babies; ß = -.29, p = .003) but the same effect was not found for pro-choice participants (ß = .02, p = .872). Those who identified more with their ingroup’s victim group expressed a greater willingness to donate to a congruent cause (ß = .30, p < .001) and were more likely to actually make a congruent donation (ß = .30, p < .001). Also, those who identified more with the outgroup’s victim group reported reduced willingness to make a congruent donation (ß = -.13, p = .036) and were also less likely to do so when given the chance (ß = -.28, p < .001). Nonetheless, neither of the indirect pathways via identification with the ingroup’s victim group were found to be significant, IE on willingness = -.11, SE = .06, 95% CI [-.22, .01] and IE on behavior = -.02, SE = .01, 95% CI [-.05, .00]. There was, however, a significant indirect effect via identification with the outgroup’s victim group on behavior, IE = -.09, SE = .04, 95% CI [-.17, -.02], but not on willingness, IE = -.06, SE = .03, 95% CI [-.13, .00].

Discussion Study 5.2 replicated the key findings of Study 5.1. First, people showed strong preferences to give to causes that align with their personal stance on contested social issues (supporting H1). Second, being exposed to opposing advocacy increased willingness to make cause-congruent

94 donations because of anger (supporting H2a). The second study also identified mechanisms that may undermine the impulse to rage donate after exposure to opponents’ messages: a lower sense of efficacy, and increased identification with the outgroup’s perceived victims. These nuances are discussed below. General Discussion The current research is the first to test empirically the phenomenon of rage donations— collective giving in response to advocacy from opponents that provokes anger. Across two studies we found that people donate to causes aligned with their personal beliefs (supporting H1). Exposure to opposing advocacy had an indirect effect on preference for making cause-congruent donations via anger (supporting H2a) but not consistently via efficacy (H2b) or identification (H2c). The finding that donors prefer to give to causes that align with their personal preferences is consistent with past research (see also Berman et al., 2018; Chapman, Louis, et al., 2018; Chapman, Masser, et al., under review). People were more willing to donate to cause-congruent organizations when they were angered by opponents’ messages. Thus, while anger may not be relevant to traditional giving contexts (Siegel et al., 2016), it appears to be associated with politicized giving (see also Van de Vyver & Abrams, 2015; van Doorn et al., 2017). It has previously been shown that generosity can be prompted by exposure to moral exemplars (e.g., Nook, Ong, Morelli, Mitchell, & Zaki, 2016). The current results show the opposite can also be true—that generosity can be prompted by exposure to those perceived to be moral opponents. We demonstrate here—for the first time—that provocative advocacy from an opponent can anger people and increase their willingness to partake in collective giving—i.e., to rage donate. Yet does this willingness flow through to action? Collective action researchers have called for more research measuring behavioral outcomes (e.g., Tausch et al., 2011; van Zomeren et al., 2008) because existing evidence mostly relates to intentions. Despite the consistency across both studies of the rage donation effect on willingness, in neither study did anger predict behavior. This may reflect a real disconnect between willingness and action that research has failed to delineate previously. However, it is also possible that our experimental paradigm did not fully capture the essence of rage donations, which typically emerge spontaneously in response to new provocation. The issues we examined—athletes kneeling and abortion—were social debates that had lasted months or years at the time of data collection. Behavioral responses motivated by anger may burn like a flash fire—with impulsive behaviors most likely to be observed immediately after the first provocation—and be difficult to elicit once the initial flames of fury have subsided. In both studies, the full SIMCA model—anger, efficacy, identification—strongly predicted willingness to make cause-congruent donations (though Study 5.1 efficacy did not). This highlights the need to consider charitable giving not just as an individual act, but also a collective one (see also

95

Louis et al., under review; Thomas et al., 2016). Although collective action and charitable giving have largely been studied in isolation, they may share many psychological characteristics. Future research should apply collective action and other social psychological theories to further understand charitable behavior. Only identification with the ingroup’s victim group and disidentification with the outgroup’s victim groups predicted donation behavior. The strong role for identification with the victim group is consistent with previous research (Chapman, Louis, et al., 2018; Chapman, Masser, et al., under review; Louis et al., under review; van Zomeren et al., 2011). Significant positive correlations were observed between willingness and action in each study (see Table 13 and Table 15: .14 and .24, respectively), but their modest size highlights the need for future research to examine moderators of this relationship. Effects of participant stance emerged in both studies. Though inconsistent across contexts, stance influenced anger, perceptions of efficacy, identification with victim groups, and donation responses. Opinion-group identities have previously been linked to collective action (Bliuc, McGarty, Reynolds, & Muntele, 2007; Thomas & McGarty, 2009). Our results highlight how different identities may have different norms for social action (see also Thomas et al., 2012). In the second study, exposure to opponent advocacy depressed people’s sense of efficacy and identification with their ingroup’s victim group and simultaneously increased identification with the outgroup’s victim group. These findings are interesting in and of themselves, as one of few studies to show effects of political messaging on opponent groups (see also Louis, Duck, Terry, & Lalonde, 2010). Here the findings highlight that caution is warranted for fundraising practitioners seeking to capitalize on rage donations: anger may not translate from willingness to action, while exposure to opponent messages may disempower and demobilize supporters. Strengths, Limitations, and Future Directions The present research examines rage donations empirically, using an experimental design that incorporated real-world tweets on current social debates and captured donation behavior. Research on both charitable giving and collective action has relied largely on self-report measures. Publishing results that highlight the potential disconnect between willingness and behavior is therefore of vital theoretical and applied importance Although we find evidence for rage donations, we acknowledge that our design may not fully capture the phenomenon. Given both studies were conducted in the context of existing social debates, our findings likely underestimate the true effect size of the initial rage response. It would not be easy to capture the spontaneous fury that characterizes real-world rage donations. Three fruitful hypotheses to explore, in our view, are that donations will more likely follow willingness

96 when the provocation is more egregious or unexpected; when the time interval to respond is shorter or immediate; and when the particular charity is saliently working against opponents. Conclusion Across two experiments using social debates about racial discrimination and abortion, we examine an emerging form of ally collective action: rage donations. When exposed to advocacy from vocal opponents, people experience anger, which promotes a willingness to donate to causes that promote people’s own views on the issue. Across both studies, SIMCA mechanisms—anger, efficacy, and identification—promoted willingness to donate. Donation behavior was influenced most strongly by identification with perceived victim groups. Results demonstrate three things. First, the value of considering charitable giving through the lens of collective action models. Second, the importance of opponent as well as supporter advocacy in fundraising. And third, specific mechanisms which may be targeted by practitioners or in future research.

97

Chapter 6. The Champion Effect in Peer-to-Peer Giving: Successful Campaigns Highlight Fundraisers More Than Causes

This chapter investigates the role of the fundraiser in determining charitable gifts. Although donations rarely occur without an ask being made (Bryant et al., 2003), fundraisers are not often studied (Breeze, 2017a, 2017b). This chapter investigates how the actions that fundraisers take can influence how much money is raised for a cause (or beneficiary). Results demonstrate that, in the context of peer-to-peer giving at least, fundraisers may be more important considerations than beneficiaries. This highlights the importance of (a) understanding the donor-fundraiser dyad and (b) testing how different giving contexts may affect the relationships between donors, fundraisers, and beneficiaries.

Citation Chapman, C. M., Masser, B. M., & Louis, W. R. (2018). The champion effect in peer-to-peer giving: Successful campaigns highlight fundraisers more than causes. Nonprofit and Voluntary Sector Quarterly, Online First, 1-21. doi:10.1177/0899764018805196

Contributions The work presented in this chapter was conducted by myself under the guidance of my research supervisors Professors Winnifred Louis and Barbara Masser. Data were collected by a fundraising consultancy called More Strategic. I negotiated access to the data sets, formulated the research question, ran and interpreted the analyses, and wrote the manuscript. Throughout the process, my advisors reviewed results and helped me to interpret them, and provided critical feedback on several drafts of the manuscript.

98

The Champion Effect in Peer-to-Peer Giving: Successful Campaigns Highlight Fundraisers More Than Causes

Abstract Online peer-to-peer giving is an emerging charity context that has rarely been investigated. Using a unique combination of survey and behavioral data from 1,647 online peer-to-peer fundraisers (whom we call ‘champions’), we tested empirically the influence of different best practices on fundraising success in this novel giving context. Across two samples, we found the fundraiser’s identification with the cause led them to engage in more best practice actions, which in turn led to greater fundraising success. However, not all actions were equally influential. Actions that made the champion salient—namely those relating to solicitation and those that signaled the fundraiser was highly invested in their campaign—were the strongest predictors of fundraising success, together explaining 28 times the variance accounted for by actions signaling charity efficacy. Thus, fundraisers will have more success by championing themselves than by promoting the charity in question: a finding with important applied and theoretical implications.

Keywords: fundraising; charitable giving; peer-to-peer; social networks; best practice.

99

Introduction New media and social networking are transforming the way people learn about and contribute to charitable causes (e.g., Guo & Saxton, 2014). However, little research has evaluated how our increasingly networked world influences charitable giving. Two notable exceptions show that online giving is at least partially driven by social network effects that are not typically observed in traditional giving contexts (Saxton & Wang, 2014; Scharf & Smith, 2016). However, it is not yet clear how the influence of social networks becomes manifest. Using behavioral and self-report data from two large samples of online peer-to-peer fundraisers, we analyze the best practice actions of fundraisers to test empirically four mechanisms known to drive charitable giving more broadly and evaluate their relative influence in predicting peer-to-peer fundraising outcomes. These are: identification, solicitation, signaled investment (evoking reputation), and signaled efficacy (see Aaker & Akutsu, 2009; Bekkers & Wiepking, 2011b). We propose that success in the peer-to-peer fundraising context is influenced more by the champion than by the charity—a phenomenon we call the Champion Effect. Thus, while peer-to- peer fundraisers may be motivated by their connection with the cause, their donors are most likely to give because of their connection with the fundraiser (see also Scharf & Smith, 2016). We argue that, if champions are key determinants of fundraising success, then fundraisers will succeed to the degree that they make it clear that success is important to them. Below we outline the existing evidence base in relation to peer-to-peer giving, discuss several mechanisms that may influence giving within peer networks, and then test empirically the relative roles of fundraisers’ identification, solicitation, signaled investment, and signaled efficacy in determining fundraising outcomes. We later discuss how this evidence supports the notion that champions strongly influence peer-to-peer giving. Finally, we highlight the implications for nonprofit practitioners. Charitable Giving Within Social Networks Peer-to-peer fundraising harnesses the social connections of charity supporters to promote causes within social networks. Individual fundraisers become advocates for their favorite causes by asking friends, family, and colleagues for donations on behalf of a charity, often as sponsorship of fundraisers’ participation in endurance or symbolic events. In an increasingly networked society, where consumers are more likely to trust peer endorsement than traditional marketing communications (Miller, 2009), leveraging peer-to-peer networks is likely to become an increasingly important component of nonprofit fundraising success. Mechanisms that promote charitable giving more broadly are documented (see review by Bekkers & Wiepking, 2011b), including a long-standing body of research on traditional philanthropy that acknowledges the instrumental role that fundraisers themselves play in campaign success (e.g., Breeze, 2017b; Tempel, Seiler, & Burlingame, 2016). To date, however, the specific 100 factors at play in the peer-to-peer context have received less research attention. Two key studies, summarized below, show that giving in networked environments is influenced by the relationships between fundraisers and their networks. Saxton and Wang (2014) evaluated the online fundraising success of 66 of the largest nonprofits in the United States. By comparing data scraped from charities’ Facebook Cause pages to information obtained from those organizations’ tax submissions, the authors demonstrated that the size of charities’ social media fan base was positively associated with the amount of money they raised on their Facebook Cause pages, but the organizations’ efficiency ratings were not. They concluded that online fundraising success is determined more by social network effects than traditional mechanisms like efficiency. A second study specifically tested network effects in the peer-to-peer context. Scharf and Smith (2016) analyzed data from 35,571 online peer-to-peer fundraisers in the United Kingdom. They found that, after controlling for relevant fundraiser demographics and charity event factors, the number of Facebook friends a fundraiser had was associated with the amount of money they raised. Specifically, fundraisers with larger online social networks received a greater number of donations but smaller average gifts. The authors argued that results reflected a ‘local’ public good that must be provided by the particular social group. These patterns of response indicate that donors are motivated by what the authors call ‘relational altruism’: they give because they care about the fundraiser and they perceive that the fundraiser cares about how much money they raise. The two studies outlined above provide evidence that social network size influences online fundraising outcomes and suggest that peer-to-peer donors may be motivated by relational altruism (Scharf & Smith, 2016). These studies, however, do not speak to the way fundraisers may harness that motivation by signaling how much they care. We propose that the actions which fundraisers decide to take in their effort to raise money may determine their outcomes, and that different fundraising actions signal different priorities or degrees of care. In turn, the signals that these actions give evoke different degrees of responsiveness in donors. Outside of the academic domain, the charity sector itself has highlighted particular ‘best practice’ actions that individual fundraisers should take. Best practices are those actions that fundraising professionals identify as being most effective in raising money. For example, industry reports suggest that fundraisers who send more emails to their networks and who tell a story about why they are fundraising are more successful (Braiterman & Masterson, 2015; DonorDrive, 2017). Blogs targeted at fundraising practitioners suggest diverse tactics, including setting a low initial target, personalizing the fundraising page, asking for specific donation values, sending targeted emails, and sharing fundraising pages via social media (e.g., Classy, 2017; Francis, 2017). Within the scholarly research literature, however, there is little empirical evidence that these tactics work

101 more effectively than alternatives. We aim to build an evidence base not only for what works in online peer-to-peer fundraising, but for why it works. To do so, we must first understand the mechanisms that promote giving in general. Mechanisms Driving Peer-to-Peer Giving A range of mechanisms promote charitable giving more broadly (see Bekkers & Wiepking, 2011b for review). Four of these—solicitation, reputation, efficacy, and identification—may help us understand the determinants of successful peer-to-peer fundraising as well. Below we discuss each potential mechanism in turn as well as how it may relate to charitable giving within peer networks. Solicitation, or the simple act of being asked to donate, is a major driver of charitable giving (Bekkers & Wiepking, 2011b). Indeed, most donations are given in response to a request (Bryant et al., 2003). Solicitation may be a particularly strong mechanism in peer-to-peer giving because most donors to online fundraising pages are already in the social network of the fundraiser (A. Payne, Scharf, & Smith, 2014), and the closeness of the relationship between the donor and solicitor influences success. In traditional giving contexts, being solicited by an acquaintance rather than a stranger increases both the likelihood of donating and the value of the gift (Meer, 2011). Being asked by a family member or friend appears to be especially powerful in motivating response (Scharf & Smith, 2016). Solicitation is therefore expected to be a powerful determinant of peer-to- peer fundraising success because the person asking is likely to be known to and valued by the donor. Reputation refers to the social consequences of making (or refraining from making) a donation (Bekkers & Wiepking, 2011b). Social concerns about gaining status and avoiding shame are inherent in charitable decisions (Bénabou & Tirole, 2006). People are more generous when their charitable donations are visible to others, especially people that matter to them (Alpízar & Martinsson, 2013; Bereczkei, Birkas, & Kerekes, 2007; Satow, 1975; Soetevent, 2005). If reputation—specifically maintaining a positive relationship with the fundraiser—is a key consideration for donors choosing whether or not to respond to charitable solicitation, actions that make the fundraiser and their personal motivation salient should improve fundraising outcomes. The more the fundraiser cares about a particular charity, the more important it may be for the donor to support them through donating to ‘their’ cause. We propose that fundraising targets, which have been shown to affect donor responses (A. Payne et al., 2014; S. Smith, Windmeijer, & Wright, 2015), are one way that fundraisers may signal their level of investment in the outcome. Targets are usually not linked to any specific funding need and the funds raised are passed to the charity regardless of whether or not the target is met. Therefore, the use of targets in online peer-to-peer contexts likely signals the fundraiser’s motivation and how much they care about the charity in 102 question (see also Scharf & Smith, 2016). Other actions that may highlight the champion have been emphasized as good practices by charities, but to our knowledge remain untested until now. These include personalization of the fundraising page, and sharing personal stories or connection with the cause (Classy, 2017; Francis, 2017). We propose that actions that highlight champions and signal their investment may be strong determinants of success in peer-to-peer fundraising contexts, even compared to actions highlighting charity efficacy. Information that communicates efficacy, or the perception that donations will actually make a difference to the relevant cause, can encourage donors to make charitable contributions (Bekkers & Wiepking, 2011b). The desire to personally make a difference is theorized to be a key consideration among philanthropists (Duncan, 2004), and there is evidence that people who think donations are more likely to help the needy report greater intentions to donate (J. R. Smith & McSweeney, 2007). Although field experiments show that information about effectiveness does not always affect donation likelihood, donors believe they are more likely to give to effective charities (Parsons, 2007; Scharf & Smith, 2016). However, efficacy has been demonstrated to be less influential in driving donations on social media, as noted earlier (Saxton & Wang, 2014), and may not be as important a consideration as simple personal preferences (Berman et al., 2018). If the champions themselves are key determinants of fundraising outcomes in the peer-to-peer domain, the perceived efficacy of the charity in question may play less of a role in motivating giving than the perceived importance of the campaign to the fundraiser. Thus, a friend or family member may give to support a fundraiser’s charity effort without caring especially about the charity’s cause or perceiving the charity as an effective agent. Finally, identities have been demonstrated to influence charitable decisions, especially whom we choose to help (e.g., Blinded, 2018; Wiepking, 2010). As mentioned above, we expect donors to be more motivated by their connection with the fundraiser than the cause. However, the degree to which the fundraiser is identified with their selected charity should also influence outcomes, insomuch as identification motivates them to exert effort to achieve fundraising success. People are more willing to help when they identify with the individual or group in need (e.g., M. Levine & Thompson, 2004; Zagefka et al., 2013). Extrapolating out, we reason that fundraisers who identify more with their selected cause are likely to do more to help the charity. The Current Research Building on previous work that demonstrates social network effects in online giving (Saxton & Wang, 2014; Scharf & Smith, 2016), we propose that giving contexts may change the relative importance of fundraising techniques. The peer-to-peer domain makes the social network salient, and different techniques may therefore be effective in peer-to-peer vs. traditional giving contexts. We aim to test empirically several mechanisms known to influence charitable giving in traditional 103 contexts in order to assess their relative roles in motivating online peer-to-peer gifts. These are: the fundraisers’ identification with the charity in question (identification), efforts to ask for donations (solicitation), the degree to which the fundraiser signals investment in the outcome (evoking reputation), and the degree to which the fundraiser signals the efficacy of the charity (efficacy). To do so, we look to fundraising best practices highlighted by charities and assess their comparative influence on fundraising success. While fundraisers themselves may be motivated by their identification with the cause they have nominated, we propose it will be the actions they choose to take in fundraising that will primarily determine their fundraising outcomes. If relational altruism motivates donors in peer-to- peer contexts (see Scharf & Smith, 2016), any action that signals the importance of fundraising success to the fundraiser will be particularly influential in reaching fundraising targets. Such actions include asking for donations, setting an ambitious target, and making their own identity and motivation a key component of their campaign. Specifically, we hypothesize the following:

H1: Fundraisers’ identification with the cause, their solicitation efforts, signaled investment, and signaled efficacy will all positively predict amount raised.

H2: Fundraiser actions (solicitation, signaled investment, and signaled efficacy) will mediate the relationship between fundraiser identification and amount raised.

H3: Solicitation and signaled fundraiser investment, both of which make the champion salient, will be stronger predictors of amount raised than signaled charity efficacy. The current research answers calls to create stronger links between researchers and practitioners (Bushouse & Sowa, 2012) by testing best practices identified by fundraising professionals. To our knowledge, this is the first research to (1) examine the relationships between self-reported fundraiser identification, best practices taken, and behavioral fundraising outcomes (i.e., dollars actually raised), and (2) to evaluate empirically the relative explanatory power of best practices on fundraising outcomes in the emerging online peer-to-peer domain. By evaluating the efficacy of various fundraising practices in this new giving context, the research can contribute to evidence-based practice while providing a more nuanced understanding of the ways that charity contexts may influence the psychological mechanisms underpinning the gift. Method Secondary data from Australian online peer-to-peer fundraisers were analyzed to investigate the influence of fundraisers’ identification with the cause, solicitation practices, signaled investment, and signaled charity efficacy in fundraising success. To increase confidence in the results obtained (see Asendorpf et al., 2013; Brandt et al., 2014; Schmidt, 2009), analyses were conducted on survey responses from peer-to-peer fundraisers at two separate timepoints.

104

Participants were surveyed in both 2013 and 2014 after taking part in a large Australian fun run. When taking part in this event, fundraising for a charity is optional and fundraisers nominate a beneficiary charity of their own selection. All runners who fundraised via everydayhero (N2013 =

5,609 and N2014 = 5,502), the official online fundraising platform for the event, were invited after the event to complete a survey about their fundraising experiences. More Strategic, a fundraising consultancy, designed the survey and collected the data via the Qualtrics survey platform on behalf of everydayhero. Data were subsequently anonymized and shared with the researchers free of charge for scholarly purposes. The researchers received no compensation for their work with this data set. Participants In 2013, 1,040 (19%) of fundraisers voluntarily responded to the survey, of whom 768 (74%) completed all measures of interest for this study and are included in the analysis. Respondents were majority female (59%), with 31% male and 10% preferring not to disclose their gender. Their age ranged from under 18 to over 65 years, with 29% aged under 30, 27% aged 30- 39, 20% aged 40-49, 14% aged over 50, and the remaining 10% preferring not to disclose their age. Collectively, they fundraised for 273 different charities. In 2014, 1,180 (21%) voluntarily responded. Of those, a total of 878 (74%) completed all measures of interest for this study and are included in the analysis. Two-thirds (66%) of respondents identified as female and 34% as male. Six participants (i.e., less than 1%) selected not to disclose their gender. Participants’ ages ranged from under 18 to over 65 years, with 32% aged under 30, 27% aged 30-39, 23% aged 40-49, and 18% aged over 50. Collectively, participants were fundraising for a total of 223 different charities. Measures Measures outlined below represent only a subset of those administered in the full survey, which is available on request. All measures were identical across the two years, except identification. Fundraiser identification. In 2013, participants indicated the degree to which they identified with the particular charity they were fundraising for by agreeing or disagreeing with three statements (“Supporting them is an important part of who I am”, “It is a cause I have always been passionate about”, and “They are the most important cause I support”, all coded 0 = disagree, 1 = agree). A supporter identity scale was created by averaging scores, with higher scores indicating greater identification with their selected charity,  = .62. In 2014, respondents indicated the degree to which they agreed or disagreed with 5 statements (“I am really passionate about their work”, “They are the most important cause I support”, “Giving to them is an important part of who I am”, “I will try and live my life in a way that supports this cause”, and “I would proudly wear their T 105 shirt when going out to be associated with them”, 1 = strongly disagree, 5 = strongly agree). A supporter identity scale was again created by averaging scores,  = .89).12 Fundraiser actions. A list of 14 actions identified by More Strategic and everydayhero to be best practices in peer-to-peer fundraising were included in the survey. Fundraisers indicated which, if any, of the actions they had performed in the course of their fundraising efforts (each scored 0 = no, 1 = yes). One item (“Set myself a challenging target”) was excluded in favor of using the behavioral measure of target value that was pre-loaded into the survey, which is described below. A principal components analysis on the remaining 13 items from the 2013 sample revealed a 3-factor solution using either eigenvalues greater than 1 or the scree plot, which accounted for 44% of the variance. Three items cross-loaded greater than .30 on two factors (see Table 17). In the 2014 follow-up sample, two of the same three items cross-loaded on two factors and were therefore removed from analyses.13 The social media post item (“Posted links and messages on my Facebook and other social media”), however did not cross-load but instead loaded at .57 on the signaled investment factor. This item was therefore retained for the 2014 sample. Identified cross- loading items were omitted and the analysis was re-run with oblimin rotation as factors were theorized to be correlated. Factor scores were computed by regression and saved for inclusion in the analyses below.14 Table 17 reports the final items and coefficients for the three factors, operationalizing the theorized constructs of solicitation, signaled investment, and signaled efficacy, which explained 50% of the variance in the 10 items retained in 2013 and 47% of the 11 items retained in 2014. Fundraising target. The target that fundraisers set was preloaded into the data in categorical form. For the purpose of analysis, value categories were coded in relation to the everydayhero default setting. Participants were coded 0 if they had selected or left the default target of $700 in place, 1 if they had raised their target above the default, and -1 if they had lowered their target. Funds raised. The actual funds raised by each fundraiser were preloaded into the survey from the everydayhero platform, allowing analysis of behavioral rather than self-reported data. Funds were reported in Australian dollars and cents.

12 One original item (“I would like to encourage others to support them by posting messages on my social media profile”) was excluded from the identification scale due to lack of construct clarity. However, an identical pattern of results is returned if this item is retained. 13 The two items that loaded on both solicitation and signaled investment were “Directly asked for donations using social media like Facebook” (loaded .41 and .48 respectively) and “Personally thanked every donor” (.36 and .40 respectively). 14 Given our factors contained relatively few items, all dichotomous, unit weighted scales would be less reliable and we employed factor scores for the analyses. Factor scores create a latent construct that can be used for analyses, which weights each item’s influence according to its factor loading. 106

Table 17. Item wording and factor loadings for fundraising best practice questions (2013)

Oblimin rotated factor loadings Item Item wording Solicitation Investment Efficacy Emailed everyone Emailed everyone I could (not just everyone I was willing to) .53 -.15 .07 Email reminders Followed up my initial email with a reminder .64 .09 -.06 Asked in person Reminded people to donate when I met them socially .55 .06 .05 Updated page Made changes to the standard Supporter Page provided by the charity .02 .74 -.05 Shared reasons Shared the reasons why I care about this cause on my Supporter Page .10 .67 .10 Uploaded photo Uploaded a personal photo to my Supporter Page -.08 .76 .01 Donation impact Told people what their donation could achieve for [Charity Name] -.03 .00 .81 Fundraising impact Told people what reaching my target could enable [Charity Name] to do .00 -.05 .81 Shared charity info Shared information from the charity with people I have asked .08 .15 .60 Specific donation Suggested a specific donation amount (for example $67 will allow the Charity to do xyz) -.01 .04 .39

Cross-loading items excluded from final solution Social media post Posted links and messages on my Facebook or other social media pages .38 .55 -.22 Social media ask Directly asked people for donations using social media such as Facebook .55 .37 -.06 Thanked Personally thanked every donor .44 .32 -.02 Note. Factor loadings over .30 appear in bold.

107

Results Means, standard deviations, and bivariate correlations for both data collection periods are summarized in Table 18. In both years, fundraisers actioned, on average, approximately half of the best practices (see Appendix C for more detail). Furthermore, roughly one-third of respondents respectively set a higher target, lower target, or chose or left the default target setting. Respondents raised a total of $753,764 for charity in 2013 (M2013 = $981.46, SD2013 = $2,380.97) and $885,804 in 2014 (M2014 = $1,007.74, SD2014 = $1,602.27). The amounts raised, however, ranged from $5 to $50,230 with a strong positive skew (i.e., most people raised small amounts). Values were log transformed in order to meet the assumptions of normality for regression modeling. As shown in Table 18, all predictors were positively associated with the amount of funds raised (rs > .12, ps < .01). Most predictors were significantly associated with one another, though all collinearities between factor scores were low (rs < .32). The analyses below report results based on factor scores for solicitation, signaled investment, and signaled efficacy rather than the individual actions. The use of aggregated scores is preferred here for parsimony and in order to remove potential instability in the model caused by multicollinearity. Results with individual actions may be of particular interest to fundraising practitioners, however. Details of the prevalence and relative influence of individual actions are therefore reported in Appendix C. Hierarchical multiple regressions (summarized in Table 19) were conducted to regress log- transformed donations raised on fundraiser identification, solicitation, signaled fundraiser investment, and signaled charity efficacy. Identification with the nominated charity was entered in Step 1 to assess fundraiser motivation. Fundraisers’ solicitation factor score was entered at Step 2, while their signaled investment score and fundraising target were entered at Step 3 to assess the influence of making it apparent that their fundraising was important to them. Finally, signaled efficacy was entered at Step 4 to assess its unique impact over and above the champion-relevant signals.15 2013 Fundraisers In the 2013 survey, fundraiser identification explained 2% of the variance in funds raised, F ch.(1,766) = 16.31, p < .001. Fundraisers who identified more strongly with the nominated charity raised more money, ß = .14, p < .001. Solicitation actions significantly explained an additional 8% of variance, F ch.(1,765) = 65.11, p < .001. As expected, fundraisers who asked for donations through more channels raised significantly more money, ß = .28, p < .001. Signaled investment explained an additional 20% of the variance in amount raised, F ch.(2,763) = 107.48, p < .001.

15 Selected order of entry corresponds to theorized temporal sequence. 108

Table 18. Descriptive statistics and zero-order correlations between fundraising predictors and amount raised in 2013 (below the diagonal) and 2014 (above the diagonal) 2013 M (SD) 2014 M (SD) 1 2 3 4 5 6 1 Fundraiser identification 0.67 (0.35) 3.92 (0.78) - .09** .15*** .07* .20*** .12** 2 Solicitation 0.02 (1.00) 0.01 (1.00) .15*** - .20*** .32*** .26*** 32*** 3 Signaled investment 0.03 (1.00) 0.01 (0.99) .10** .13*** - .15*** .22*** .26*** 4 High target 0.13 (0.84) 0.08 (0.79) .14*** .22*** -.01 - .19*** .49*** 5 Signaled efficacy 0.01 (1.01) 0.01 (1.00) .23*** .25*** .21*** .18*** - .23*** 6 Raised (log) 981.46 (2380.97) 1008.54 (1603.00) .14*** .30*** .17*** .49*** .24*** -

N2013 = 768; N2014 = 878 (Listwise) Note. Solicitation, signalled investment, and signalled efficacy use factor scores from Principal Components Analysis. Fundraiser identification was measured on a 0-1 scale in 2013 and a 1-5 scale in 2014. Target coded -1 = less than $700; 0 = default value $700, 1 = more than $700. Mean and standard deviation reported for Raised are untransformed. *p < .05, **p < .001, ***p < .001

109

Table 19. Hierarchical regressions with fundraisers’ identification and actions as predictors of funds raised in 2013 and 2014 2013 Amount Raised: Log (ß) 2014 Amount Raised: Log (ß) Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4 Fundraiser identification .14*** .10** .04 .02 .12*** .09** .05 .04 Solicitation .28*** .17*** .16*** .31*** .18*** .17*** *** *** *** *** Signalled investment .15 .13 .17 .16 *** *** *** *** High target .44 .43 .41 .41 ** * Signalled efficacy .09 .06

R2 ch. .02*** .08*** .20*** .01** .01*** .09*** .20*** < .01* Model R2 .10*** .30*** .30*** .11*** .31*** .31***

Note. N2013 = 768; N2014 = 878 * ** *** p < .05; p < .01; p < .001

110

Fundraisers who signaled investment through personalization actions, ß = .15, p < .001, and setting a higher fundraising target, ß = .44, p < .001, raised significantly more money. Finally, actions that communicated the efficacy of the charity explained an additional 1% of variance, F ch.(1,762) = 7.47, p = .006, with fundraisers who shared more information about charity efficacy also raising significantly more, ß = .09, p = .006. Taken together, the full model explained 30% of the observed variance in dollars raised, F(5,762) = 66.98, p < .001. In the final model, all predictors remained significant except for fundraiser identification, which became non-significant once fundraising actions were accounted for. Bootstrapping analyses conducted in R using the lavaan package (Rosseel, 2012) with 1000 resamples confirmed that the actions fundraisers took (relating to solicitation, signaled investment, and signaled efficacy) fully mediated the relationship between their identification with the charity and their fundraising success, combined IE = .39, SE = .07, 95% CI = [.26, .52].16 Results of the mediation model are presented in Figure 5.

Figure 5. Mediation model showing how fundraiser’s identification with the cause predicts fund’s raised indirectly via their best practice actions of solicitation, signaled investment, high target, and signaled efficacy (2013 sample) Note. N = 768. Standardized betas are reported.

2014 Fundraisers In the 2014 survey, the same pattern of results were returned (see Table 19). Fundraiser identification, entered at Step 1, explained just 1% of the variance in amount raised, F ch.(1,876) =

16 Examination of unique effects showed each indirect pathway was significant, via: solicitation, IE = .08, SE = .01, 95% CI [.03, .12]; signaled investment, IE = .04, SE = .02, CI [.01, .08]; high target, IE = .20, SE = .05, CI [.10, .31]; and signaled efficacy, IE = .07, SE = .03, CI [.01, .12]. 111

12.65, p < .001. Fundraisers who identified more strongly with their nominated charity raised significantly more money, ß = .12, p < .001. The inclusion of solicitation factor scores at Step 2 significantly improved the model, explaining an additional 9% of the variance, F ch.(1,875) = 93.29, p < .001. Fundraisers asking through more channels raised significantly more money, ß = .31, p < .001. Entered at Step 3, variables related to fundraiser investment significantly explained an extra 20% of the variance, F ch.(2,873) = 126.53, p < .001. Fundraisers who signaled their investment by personalizing their campaign, ß = .17, p < .001, and setting a higher target, ß = .41, p < .001, raised significantly more money. Finally, signaled efficacy factor scores explained less than 1% of additional variance, F ch.(1, 872) = 4.39, p = .036. Fundraisers who shared information about the efficacy of the charity in question raised significantly more money, ß = .06, p = .036. Taken together, the full model explained 31% of the observed variance in amount raised, F(5,872) = 79.40, p < .001. All variables remained significant in the final model except for fundraiser identification, which was no longer significant once fundraiser actions were included. Bootstrapping analyses conducted in R using the lavaan package (Rosseel, 2012) confirmed that the actions fundraisers took (relating to solicitation, signaled investment, and signaled efficacy) fully mediated the relationship between their identification with the charity and their fundraising success, combined IE = .13, SE = .03, 95% CI = [.07, .19].17 Discussion Across two field samples with online peer-to-peer fundraisers, we find support for all hypotheses related to the notion that fundraising outcomes within peer networks are influenced strongly by the fundraising champions themselves. Specifically, across both samples, we find that fundraisers who identified more with their selected charity, and who took more actions to solicit donations, signal their personal investment, and signal the efficacy of the charity, raised more money (supporting H1). The fundraiser’s greater identification with the cause was associated with taking more actions, which in turn was associated with greater fundraising success (supporting H2). All actions are not equal, however. Asking for donations through more channels and, especially, actions that signaled the fundraisers’ personal investment in the outcome, were stronger predictors of fundraising success than actions that signaled the efficacy of the charity in question (supporting

H3), respectively explaining at least 8 and 20 times the variance in amount raised. While these findings will be intuitive to practitioners and to many scholars, they have not been demonstrated empirically before. To our knowledge, these are the first studies to (1) combine

17 Again, all unique indirect pathways of identification on amount raised were all positive, via: solicitation, IE = .02, SE = .01, CI [.004, .04]; signaled investment, IE = .04, SE = .01, CI [.02, .06]; high target, IE = .05, SE = .02, CI [.004, .09]; and signaled efficacy, IE = .02, SE = .01, CI [.00, .04]. 112 self-reported actions with behavioral outcomes in the peer-to-peer domain, and (2) evaluate the relative effectiveness of fundraising best practices in this emerging charity context. Findings are of both theoretical and practical importance as they highlight how effective fundraising actions change in the new, increasingly important online peer-to-peer context. Furthermore, using best practices identified by fundraising practitioners helps to overcome the research-practice divide (see Bushouse & Sowa, 2012) by ensuring that actions have relevance to nonprofits themselves. The focal hypothesis that the champion would be an important determinant of fundraising success in the online peer-to-peer context was strongly supported. In both samples, actions that made the fundraiser and their investment salient—such as uploading a photo, personalizing their fundraising page, articulating their reasons for fundraising, and setting a high target—were those most strongly associated with fundraising outcomes. Though we do not test donor motivations, these findings are consistent with Scharf and Smith’s (2016) assertion that people give in peer-to- peer contexts because they care about the fundraiser and they know the fundraiser cares about raising money. In our view, the current results show how fundraisers can effectively harness the power of such relational altruism by signaling their commitment to potential donors. They also echo evidence from research on traditional giving contexts that individual fundraisers are important for campaign success (Breeze, 2017b; Tempel et al., 2016). Results show a strong Champion Effect in the peer-to-peer domain, with the fundraiser themselves being a key component of fundraising success. Fundraisers who were more identified with their selected charity raised more money, apparently because they put more effort into their fundraising and performed a greater number of best practice actions. Because our data were collected after fundraising was complete, we cannot be certain that identification led to success via actions. It is also possible that fundraisers who raised more money later felt more identified with the cause. However, our contention that identification with the cause led fundraisers to exert more effort to raise money aligns with previous work showing that people who are more identified with an individual or group are willing to do more to help them (e.g., M. Levine & Thompson, 2004; Zagefka et al., 2013). In addition, we identify concrete mechanisms through which identities can affect outcomes in the charitable domain: the best practice actions fundraisers take. As hypothesized, solicitation—or asking for donations through more channels—was shown to be an important contributor to fundraising success (corroborating Bekkers & Wiepking, 2011b; Breeze, 2017b; Bryant et al., 2003). In the fundraising literature, it has been shown that people are more likely to give when they are asked to donate by someone known to them, and that donors tend to respond more favorably to fundraisers who are close to them (Meer, 2011; Scharf & Smith, 2016). In the peer-to-peer domain, fundraisers are almost certainly known to the donors (A. Payne

113 et al., 2014). We argue throughout that signaling investment may make reputation salient to the donor. One important way that fundraisers signal their investment is to ask more often, and through more channels. In this way, solication may amplify the perceived relational consequences— whether positive or negative—of the donor’s response. Indeed, within the wider literature, concerns about reputation have been observed to affect charitable responses (Alpízar & Martinsson, 2013; Bekkers & Wiepking, 2011b; Bénabou & Tirole, 2006; Bereczkei et al., 2007; Satow, 1975; Soetevent, 2005). In the peer-to-peer context, where relationships between fundraiser and donor are personal, solicitation factors may be especially powerful. Beyond mere solicitation, fundraisers can also draw attention to themselves (and indirectly highlight potential reputational consequences) by signaling personal investment in their campaign. It has been argued before that peer-to-peer donors give, at least in part, because they care about the fundraiser, they understand the fundraiser cares about the cause, and they want the fundraiser to succeed (Scharf & Smith, 2016). We propose that signaling investment in the campaign evokes reputation because it makes clear that the fundraiser will be paying attention to donor responses. Results here show that actions that signal investment are indeed strong determinants of fundraising success. Setting a high target was the strongest unique predictor of fundraising success. We interpret this finding as showing that people who are more identified with their cause also set a higher fundraising target, indicating that targets are determined at least in part by fundraiser motivation (see also Scharf & Smith, 2016). Fundraisers could, however, adjust their targets during the campaign, and the current data cannot differentiate between those who set a high initial target and those who set a lower initial target but raised it as their campaign progressed. It must also be acknowledged that targets may also be determined by such pragmatic concerns as perceived wealth and size of the network in question. That is to say, in addition to their personal investment, fundraisers surely consider how many people they know well enough to ask for a donation, and the relative resources those people hold, when determining an appropriate target. Future research would therefore benefit from controlling for network size and resources to assess the impact of the champion effect over and above these practical considerations. When solicitation and signaled investment were considered in unison, they accounted for almost 30% of the observed variation in fundraising success. This effect size is substantial. To contextualize, it is comparable both to the combined predictive power of household income, debt, and demographic make up on the size of household charitable donations (28%; Hughes & Luksetich, 2008), and to the combined impact of fundraising expenditure, price of giving, organizational age, and social network size on the value of donations received on Facebook Cause pages (30%; Saxton & Wang, 2014). Regardless of their motives, champions who both ask for

114 donations through more channels (solicitation) and ensure their social networks know how much they care about the outcome (signaling investment) are those who raise the most money. Although we distinguish solicitation and signaled investment, it is hard to disentangle the effects of these two mechanisms in the peer-to-peer domain because the solicitor is known to the donor (A. Payne et al., 2014) and the very act of asking signals investment. As mentioned previously, both types of action make champions salient and, we argue, both evoke reputation. It would be interesting to explore their joint operation or interactions in future research. The perceived efficacy of a charity generally influences fundraising success (Bekkers & Wiepking, 2011b; J. R. Smith & McSweeney, 2007), and as hypothesized, signaling charity efficacy was associated with greater donations in the two samples here. Yet the role of signaled efficacy was less important in promoting fundraising success than other factors (echoing Berman et al., 2018; Saxton & Wang, 2014). After accounting for champion-related actions—solicitation and signaled investment—promoting the efficacy of the charity or donations explained no more than 1% of extra variance in the amount raised.18 These results support the idea that champions are more important drivers of fundraising outcomes in the peer-to-peer domain than charities are. Nonetheless, it is also possible that donors may simply be relying on sources other than the fundraiser for information about charity effectiveness. Future research could ask if, when, and how donors seek effectiveness information in the peer-to-peer domain. Successful peer-to-peer campaigns highlight fundraisers more than causes: a phenomenon we call the Champion Effect. On a theoretical level, these findings highlight the apparent motivational duality of peer-to-peer fundraising: it may be that in peer-to-peer contexts the fundraiser and donor perceive different targets as the beneficiaries of the altruistic response. Donors appear to be focused on giving to the fundraiser (Scharf & Smith, 2016). On the other hand, fundraisers who take more best practice actions are those who are more strongly identified wih the charity, suggesting fundraisers are focused on the charity as the beneficiary of their actions. The fundraiser’s pivotal role as a champion of the cause can be enhanced according to the actions that they take. How to most effectively equip the champion with best practices that signal their investment emerges as an exciting direction for researchers and practitioners to explore. Strengths, Limitations, and Future Directions Strengths of the current research include the use of behavioral outcome data, large community samples of actual fundraisers, practitioner-led survey development, and the close

18 Due to the nature of the hypotheses, efficacy actions were considered in the final step of the model. When entered before solicitation and signaled investment, signaled efficacy still explained just 4% of variance compared to at least 24% explained by champion-relevant factors. 115 replication of methodology and results across two samples. However, several limitations warrant mention. First, response rates were relatively low, with only 19-21% of invited fundraisers choosing to participate in the research. We therefore cannot rule out the possibility that results may only reflect the experiences of a subset of highly motivated fundraisers who are particularly passionate and dedicated. Second, because fundraisers were free to select the beneficiary charity for their fundraising, a vast array of different charities were included in the data (273 in 2013 and 223 in 2014). Research has shown that some charity types are more effective in fundraising via social media (Saxton & Wang, 2014) and that different types of people are motivated to give to different types of charities (Blinded, 2018; Wiepking, 2010). Future research should therefore consider how the type of charity may moderate the relative importance of the mechanisms evaluated here. Third, the data were collected exclusively in Australia. Given that some cultures exhibit different patterns of giving (Charities Aid Foundation, 2017b), it will be important to also test the Champion Effect in different cultural contexts. Finally, although the model tested here explains substantial variation in the amount of funds raised, it does not explain all of it. As previously mentioned, practical concerns such as the size and wealth of fundraisers’ social networks surely matter. Further, factors such as awareness of need, altruism, norms, values, and prestige have all been demonstrated to influence charitable giving in traditional contexts (see reviews by Bekkers & Wiepking, 2011b; Sargeant & Woodliffe, 2007b). Future research on peer-to-peer giving would benefit from studying the impact of such factors in networked contexts. Evidence presented here suggest that donors are influenced by the perceived investment of the fundraiser more than by the efficacy of the charity. An interesting avenue for future research will be to understand the boundary conditions of this effect. In particular, research should investigate whether there are certain issues or causes that are so polarizing that donors would be unwilling to support them, regardless of the enthusiasm of the fundraiser. Applied Implications Overall, evidence presented in this article supports the assertion that peer-to-peer giving is influenced by a champion effect, where campaign success is determined more by actions highlighting fundraisers than actions highlighting causes. Two factors are of particular importance: asking through more channels (solicitation) and fundraisers’ signaling the importance of the outcome to them personally (signaled investment). Fundraisers themselves are likely to be motivated by the cause in question (A. Payne et al., 2014) and may therefore select tactics aligned to their own motives (e.g., promoting the effectiveness of the charity in achieving its mission) while potentially neglecting tactics that could motivate others (e.g., promoting themselves and their connection with the cause). Our data speak to this phenomenon. Therefore, charities should 116 intervene and educate individual fundraisers to help them to prioritize their efforts. Fundraisers should be encouraged to ask for donations through as many channels as possible and to ensure that their campaigns are personalized with photos, high targets, and articulation of their investment and motives. Conclusion In two large community samples of online peer-to-peer fundraisers, using behavioral outcome data, we found that fundraiser identification, asking for donations through more channels, signaling personal investment in fundraising success, and highlighting the efficacy of the charity were all significantly associated with raising more money. However, actions that highlighted the fundraiser themselves (solicitation and signaled investment) accounted for substantially more variance in fundraising outcomes than the individual or charity factors. Results highlight that fundraising best practices will depend on the giving context. We demonstrate an important “Champion Effect” in online peer-to-peer fundraising and suggest that efforts to equip the fundraiser with tools to convey their own personal connection with the cause will lead to greater success than efforts to highlight the effectiveness of the charity or its overarching mission. Fundraising practitioners should therefore evaluate the communications, toolkits, and other support they provide to individual fundraisers with the relative importance of these motivations in mind.

117

Chapter 7. The Charitable Triad: How Donors, Beneficiaries, and Fundraisers Influence Charitable Giving

The ideas presented in this chapter represent the culmination of my thinking over the course of my PhD. The review was conducted over many years as I delved into diverse literatures on charitable giving. Based on this broad reading of the interdisciplinary literature and my own research findings, this chapter also presents my novel theoretical model for conceptualizing charitable giving as triadic. These ideas build directly off my own evidence (1) of the importance of studying donors and beneficiaries in dyads (Chapters 2 and 3), (2) showing how social contexts affect giving responses (Chapters 4 and 5), and (3) demonstrating the pivotal role of the fundraising in charitable decisions (Chapter 6).

Citation Chapman, C. M., Louis, W. R., Masser, B. M., & Thomas, E. F. (in prep). The charitable triad: How donors, beneficiaries, and fundraisers influence charitable giving. Manuscript in preparation.

Contributions I personally developed the ideas presented in this chapter, conducted the literature review, and wrote the manuscript. My research supervisors Professors Winnifred Louis and Barbara Masser, and my research collaborator Associate Professor Emma Thomas (a co-author on this paper), also contributed to this research by critically reviewing and commenting on previous versions of this manuscript.

118

The charitable triad: How donors, beneficiaries, and fundraisers influence charitable giving Cassandra M. Chapman, Winnifred R. Louis, Barbara M. Masser, & Emma F. Thomas

Abstract Charitable giving is the act of donating money to organizations that benefit others. Research on giving has primarily focused on the characteristics of donors and has generated inconsistent findings. In contrast, we propose that charitable decisions are determined by the charitable triad— donors, beneficiaries, and fundraisers—and that failure to consider any of these actors obscures the true psychology of giving. We review over 300 articles from psychology and from the interdisciplinary literature to generate the charitable triad model, conceptualizing charitable giving as triadic, relational, and contextualized. We argue that charitable decisions are influenced by all three actors, the relationships between them, and salient aspects of the wider social context (including perceived norms, audience, issue salience, giving context, and funding sources). The charitable triad model highlights gaps in current knowledge and explains previous inconsistencies. We close with five novel propositions generated by the model to stimulate future research. Keywords: charitable giving; fundraising; philanthropy; helping; prosocial behavior.

119

Introduction In 2017, American citizens donated $410 billion to charitable organizations (Giving USA, 2018a). These charities and nonprofits are tasked with addressing some of the world’s most pressing problems. Whether trying to cure cancer, protecting human and animal rights, caring for the sick or elderly, or advocating for the environment, charities are critical to the functioning of human society. And most charities rely on donations to achieve their social goals. Understanding the psychology of charitable giving—how, why, and when an individual will give their money away to further a charitable cause—is therefore of vital importance. Charitable giving is the act of donating money to organizations that help non-kin others (Bekkers & Wiepking, 2011b). For decades, researchers have posed questions about such giving and have gathered data to understand it. Economists, for example, ask why rational donors voluntarily give away their money, going against apparent self-interest. Researchers and practitioners of philanthropy ask who gives, and why, and how much. Marketers focus on techniques that organizations can use to solicit funds. Psychologists and sociologists investigate social contexts and group dynamics that may influence prosocial behavior. Each of these bodies of work advance our understanding of charitable giving in important ways, approaching the act from different perspectives—studying the what (money), the who (donors or beneficiaries), the how (solicitation), or the where and when (context). Yet each perspective on its own leaves our understanding incomplete. Like the parable of the blind men encountering an elephant for the first time, each approach looks at one aspect of the phenomenon and describes what it sees: the blind man touching the trunk likens an elephant to a snake, while the blind man touching the leg likens it to a pillar. Similarly here, our understanding of charitable giving is incomplete because there has been little systematic integration of findings across disciplines (economics, marketing, philanthropy, sociology, psychology) or perspectives (who, why, when, and how much). In one sense these disciplines have made good progress in understanding charitable giving but, on closer inspection, the evidence base is also riddled with inconsistencies. Consider the impact of income on giving, as an example. Generally, as expected, people with more money are seen to give more to charity (e.g., Choi & DiNitto, 2012; Hughes & Luksetich, 2008; Korndörfer et al., 2015). However, poorer donors give a greater share of their income (e.g., Bennett, 2012; Piff et al., 2010) and sometimes no relationship between income and giving is found (Neumayr & Handy, 2017; Shier & Handy, 2012). So, despite the obvious connection between having money and being able to share it with others, the relationship between income and charitable giving remains unclear. As we shall see, inconsistences such as these emerge in relation to almost all known predictors of giving.

120

To solve these problems, we review existing evidence from diverse disciplines and generate a holistic model of charitable giving. The charitable triad model clarifies inconsistencies by proposing that relationships between donor characteristics, such as wealth, and giving will critically depend on three things: characteristics of the beneficiary (i.e., what type of cause is being considered), characteristics of the fundraiser (i.e., what kind of person or organization is asking for donations), and characteristics of the social context (e.g., who is watching, what is normative, or who else if giving). Without considering the entire triad, the model proposes, the factors influencing giving will remain unclear and inconsistent. A key insight of the charitable triad model is that charitable decisions are influenced not only by individual differences, nor only by dyadic interactions, but instead by a triad of actors— donors, beneficiaries, and fundraisers—who, in turn, are influenced by the features of their wider social context. Our review reveals stark asymmetries in the focus of prior research, which has favored understanding the characteristics of donors over those of beneficiaries and, especially, fundraisers. We focus this conceptual review on charitable giving. Psychology as a discipline has spent decades examining prosocial behavior, or actions intended to benefit others (see, for example, Dovidio et al., 2006; Eagly & Crowley, 1986; Fischer et al., 2011; Ma et al., 2017; Penner et al., 2005; Saucier et al., 2005; Shariff et al., 2015). However, there is evidence that different forms of prosocial action have different psychological underpinnings. The helping literature, for example, focuses mostly on interpersonal contexts when bystanders respond to individuals in need (e.g., Eagly & Crowley, 1986; Stürmer et al., 2006), although there is also growing interest in understanding helping as an intergroup process (e.g., M. Levine & Manning, 2013; Nadler, 2002; van Leeuwen, 2017; Zagefka & James, 2015). However, we argue that giving decisions are distinct from helping decisions, partly because of the triadic nature of the phenomenon—giving is often a response to a third party asking on behalf of a beneficiary. Further, within the giving literature there is evidence of different motives for giving money compared to time, blood, or other body parts (Chell, 2016; Clary & Snyder, 1999; Clary et al., 1998; Konrath & Handy, 2017; L. Lee & Piliavin, 1999). The focus of this review, therefore, is specifically on charitable giving: the psychological processes that underpin both whether a donation of money will be made and, if so, how much will be given and to whom. However, the potential generalizability of the model to other helping and giving contexts is addressed briefly in the discussion. The Charitable Triad: Donors, Beneficiaries, and Fundraisers Some donors may be inherently charitable. Others may be moved to give to certain beneficiaries. Still others may be motivated in response to specific fundraisers. In this article, we propose that charitable giving can be influenced by all three members of the charitable triad:

121 donors, beneficiaries, and fundraisers. While donors hold the purse-strings, they may give (or fail to give) because of: (1) their own characteristics; (2) characteristics of the beneficiary in question; (3) characteristics of the person asking for donations; (4) the dyadic relationships among these three actors; or (5) the unique triadic combination of donor, beneficiary, and fundraiser characteristics. We also propose that salient information from the wider social context in which giving occurs may amplify or dampen the impact of any of these characteristics or relationships. Figure 6 represents the conceptual relationships between actors in the charitable triad and highlights the asymmetry in research attention—brought to light by the review that follows—on different elements of the triad.

Figure 6. The charitable triad model: charitable decisions are influenced by characteristics of the three actors, but also the interactions between particular donors, beneficiaries, and fundraisers within a wider social context Note. The shading and lines indicate the relative size of each body of research at the time of writing: darker shading or thicker lines indicate a stronger research focus in the previous literature.

Previous broad surveys of charitable giving have primarily focused on donors (Bekkers & Wiepking, 2011c; Konrath & Handy, 2017; Wiepking & Bekkers, 2012), fundraisers (Sargeant & Woodliffe, 2007a), or both (Andreoni & Payne, 2013; Bekkers & Wiepking, 2011b), or have focused on a particular type of giving (like disaster giving; Zagefka & James, 2015). The charitable triad model integrates these perspectives in order to understand the ways that each actor influences the others. Further, the model goes beyond specific types of giving by actively considering the role of social context. 122

The charitable triad model contributes to research on charitable giving in four key ways. First, it highlights how two of the three critical actors in giving—beneficiaries and fundraisers—are understudied. Second, it explains previous inconsistencies in evidence by proposing that dyadic relationships between actors change how the particular characteristics of any given actor impact giving. Third, it further proposes that each of these dyadic relationships are also influenced by the third actor in the triad. Fourth, it highlights specific contextual factors that may further change the roles of and relationships between the three actors. Below, we briefly review over 300 articles from the disciplines of psychology, sociology, philanthropy, marketing, and economics to survey the existing evidence of who gives (donor- centered research), who receives (beneficiary-centered research), who asks (fundraiser-centered research), and dyadic interactions between each. We critically examine this evidence in relation to the charitable triad model, identifying asymmetries in research attention and highlighting contradictions in evidence that application of the triadic approach could clarify. Next, we integrate knowledge of how contextual factors may influence each of the actors in the triad and their relative roles. Finally, five testable propositions derived from the model are presented, offering future directions for research on charitable giving and other forms of prosocial action. Donors as Agents of Charity The most common approach to understanding charitable giving—for researchers and practitioners alike—has been to seek to understand the psychology of the donor (e.g., Aaker & Akutsu, 2009; Bekkers & Wiepking, 2011c; Breeze, 2013; de Oliveira, Croson, & Eckel, 2011; Thomas et al., 2017; Thomas et al., 2016; Wiepking & Bekkers, 2012). Although businesses and foundations can also be donors, individuals make almost 80% of all charitable donations (Giving USA, 2018a). Below we review the existing evidence base in relation to donors, including the role of demographics, stable individual differences, and other donor characteristics. Throughout we discuss the donor characteristics that are associated with both donor status (i.e., who is most likely to give to charity at all) and donor value (i.e., who is most generous). Demographics Demographics are practical tools that can be used by nonprofits for campaign targeting and personalization (Andreasen & Kotler, 2008). Those characteristics used to identify and segment donors or predict giving levels include gender, age, wealth, education, and family status. A close review of the demographic literature highlights contradictions that are explainable by consideration of the full charitable triad (see Figure 6). Gender. In general, women are more likely to give to charity than men and are more generous when they do give (Bègue, 2014; Herzog & Yang, 2018; Leslie et al., 2013; Mesch et al., 2011; Mesch et al., 2015a; Mesch et al., 2006; Rajan et al., 2009; Tremblay-Boire & Prakash, 123

2017). Women are also more diverse in their giving, tending to spread their gifts across a larger number of charities than men (Andreoni et al., 2003; Mesch et al., 2015b; Piper & Schnepf, 2008). However, sometimes men have been found to make larger average donations than women (e.g., Everatt, Habib, Maharaj, & Nyar, 2005). Age. As people age, they are more likely to give to charity (Bègue, 2014; Bekkers & Wiepking, 2011c; Herzog & Yang, 2018; McGregor-Lowndes & Crittall, 2014; Mesch et al., 2006; Midlarsky & Hannah, 1989; Neumayr & Handy, 2017; Rajan et al., 2009; Sibley & Bulbulia, 2015; M. A. Steinberg et al., 2005; Wiepking & James, 2013), though age is not always a significant predictor of being a donor (Tremblay-Boire & Prakash, 2017). In addition, research has found both that older donors give more on average (Bennett, 2003; M. A. Steinberg et al., 2005) and that younger donors do (Midlarsky & Hannah, 1989, Study 1). Income. Unsurprisingly, wealthier donors seem more likely to give to charity and may give more in absolute terms (Choi & DiNitto, 2012; Herzog & Yang, 2018; Hill & Vaidyanathan, 2011; Hughes & Luksetich, 2008; Korndörfer et al., 2015; McGregor-Lowndes & Crittall, 2014; Mesch et al., 2006; Rajan et al., 2009; Sibley & Bulbulia, 2015; Tremblay-Boire & Prakash, 2017; Vaidyanathan et al., 2011; Wiepking & Bekkers, 2012; Wiepking & Maas, 2009). However, poorer donors give a greater proportion of their income to charity (Bennett, 2012; Breeze, 2006; Piff et al., 2010; Wiepking, 2007). Moreover, sometimes no relationship between income and giving is found (Neumayr & Handy, 2017; Shier & Handy, 2012). Education. People who have attained higher levels of formal education give more to charity (Bekkers & Wiepking, 2011c; Forbes & Zampelli, 2013; Manesi et al., 2018; Mesch et al., 2006; Rajan et al., 2009; Rooney et al., 2005). In one study, for example, people with college degrees were 58% more likely to be donors than people without college degrees (Herzog & Yang, 2018). Yet sometimes more educated people are found to be less likely to donate (Tremblay-Boire & Prakash, 2017) and one study of Austrian donors found that education level did not influence the likelihood of giving (Neumayr & Handy, 2017). Family status. Married people are generally more likely to give to charity than single people (Forbes & Zampelli, 2013; Mesch et al., 2015a; Mesch et al., 2006; Neumayr & Handy, 2017; Rajan et al., 2009; Rooney et al., 2005), but this trend has not always been observed (see Herzog & Yang, 2018; Wiepking & Bekkers, 2012). Single households are also less generous, giving 39% less on average than married households (Eagle, Keister, & Read, 2017). Further, having children changes the way people give to charity. Parents of newborns give less but increase their giving as their children age (Einolf, 2018; Wiepking & Bekkers, 2012). The effects of family status also depend on prior marital status and gender (Einolf & Philbrick, 2014; Mesch et al., 2006).

124

Individual Differences in Giving Demographics are frequently used to predict giving, both independently and in interaction (Andreoni et al., 2003; Eagle et al., 2017). Such effects are frequently inconsistent, suggesting that, consistent with the insights of the charitable triad, there are other factors that themselves impact on the relationship between the donor and the beneficiary. Also, although they may be pragmatic for campaign targeting, demographics tell us little about the psychological mechanisms that promote giving. Individual psychological differences, however, can tell us more about mechanisms. Individuals differ in terms of their view of the world and their place in it. Do they show concern for the welfare of others? Do they perceive a competitive jungle? Do they have a higher purpose? Below we discuss individual differences known to be relevant to giving—empathy, religiosity, political orientation, ideologies, and values. These approaches have value in understanding who is most likely to give to charity and which donors will be most generous. Empathy. Empathy refers to the ability to understand and respond appropriately to the thoughts and feelings of other people (Baron-Cohen, 2011). Overall, there is strong evidence that people higher in empathic concern give more to charity (Batson et al., 2002; Bekkers & Ottoni- Wilhelm, 2016; Davis, 1983; Kim & Kou, 2014; Luengo Kanacri et al., 2016; Mesch et al., 2011; Neumayr & Handy, 2017; Wiepking & Maas, 2009; Zhou et al., 2012). However, a cross-cultural study of 63 countries found that average country-level empathy was not associated with the percentage of people from that country who reported having made donations in the previous month (Chopik, O’Brien, & Konrath, 2017). The impact of empathy on giving may operate within cultural norms, highlighting the need to consider the wider social context. Religiosity. Broadly speaking, religion promotes charity. All religions endorse the act of giving to people less fortunate and this emphasis is borne out in the rates of giving among the religious. Religious people are between 25% and 53% more likely to report being donors than secular people (Brooks, 2003b; Helms & Thornton, 2012; Hill & Vaidyanathan, 2011; Rajan et al., 2009; Stavrova & Siegers, 2014; Wiepking et al., 2014). Religious people are also more generous than their secular counterparts (Forbes & Zampelli, 2013). Two studies—conducted in the United States and New Zealand—both found religious people reported giving over $750 more per year to charities than non-religious people (Helms & Thornton, 2012; Sibley & Bulbulia, 2015). However, religious donors are not universally generous but strongly favor donating to causes affiliated with their religious groups (e.g., Chapman, Louis, et al., 2018; Forbes & Zampelli, 2013; Helms & Thornton, 2012; Hill & Vaidyanathan, 2011), highlighting how the charity that is fundraising influences who gives. Political Orientation. In many western countries, people can identify their political orientation, or the degree to which they see themselves as endorsing more conservative or more

125 liberal political opinions. However, it is unclear whether political orientation influences charitable giving in systematic ways. Some studies have shown that conservatives are less likely than liberals to give to charity (Forbes & Zampelli, 2013; Manesi et al., 2018), while another study found the opposite (Sibley & Bulbulia, 2015). Sometimes no clear relationship is found between political orientation and giving (Herzog & Yang, 2018; Thomsson & Vostroknutov, 2017; Tremblay-Boire & Prakash, 2017; Yen & Zampelli, 2014). We expect that both the beneficiary and the fundraiser may change the relationship between political orientation and donation responses. For example, conservatives are less likely to give to international beneficiaries (Chapman, Louis, et al., 2018) but may be moved to do so when foreign aid is positioned as serving the national interest or when solicited by a respected peer. Ideologies. Two particular ideologies have been studied in relation to charitable giving: social dominance orientation and right-wing authoritarianism. Social dominance orientation is the degree to which individuals endorse a worldview of competition and group-based hierarchy (Pratto, Sidanius, Stallworth, & Malle, 1994). Generally speaking, people higher in social dominance orientation are less generous and less willing to help others (M. A. Brown, 2011; Freeman et al., 2009; Halabi et al., 2008; Pratto et al., 2013). Individuals high in right-wing authoritarianism endorse strong leaders and traditional values (Altemeyer, 1981). There is considerable overlap between social dominance orientation and right-wing authoritarianism, mostly in terms of both being associated with a conservative political orientation and prejudice against racial, ethnic, gender, and sexual minority groups (Altemeyer, 2004; Dallago, Cima, Roccato, Ricolfi, & Mirisola, 2008; Wilson & Sibley, 2013). Nonetheless, the respective relationships with charitable giving diverge. There is no consistent evidence for a relationship between right-wing authoritarianism and giving—though some studies have shown that people higher in right-wing authoritarianism are less likely to give, at least to some types of targets (Blogowska & Saroglou, 2011; Perry, Paradies, & Pedersen, 2015; Reese, Proch, & Cohrs, 2014). Values. Donors often give because they feel that the charity in question aligns with their personal values and is promoting their vision for a better world (Bennett, 2003; Chapman, Masser, et al., under review). Schwartz (1994) found that people all around the world consistently say that ten key values dominate their life outlook. Of these ten, universalism (care for the welfare of all people; Schwartz, 1994) is most important for charitable giving, and those higher in this value tend to give more to charity (Joireman & Duell, 2007; Schwartz, 1994). Other Donor-Centric Factors that Influence Giving While the individual differences outlined above tell us something about the donor’s broad orientation toward the world, they tell us little about the donor as a dynamic entity—their self- conception, focus, or interactions with others and how these influence donors’ decision-making 126 from moment to moment. Other donor-centric factors that have been shown to affect giving are perceived benefits and reputation, emotions, and identity. Benefits and reputation. For some donors, giving may be more like a transaction or a type of commercial exchange. People are more likely to give to a charity if they have benefited in the past or stand to benefit in the future (Chapman, Masser, et al., under review). Some donors give with the expectation of a particular reward: a tax deduction (e.g., Fack & Landais, 2010), a thank you gift (Newman & Jeremy Shen, 2012), or perhaps the promise of returns in the afterlife (Thornton & Helms, 2013). Andreoni (1990) also argues that some donors receive an emotional kick-back (a “warm glow”) for giving (see also Aknin et al., 2013; Korenok, Millner, & Razzolini, 2013; Ottoni-Wilhelm, Vesterlund, & Xie, 2017). Benefits may also come in the form of reputational rewards. In general, people that are generous are evaluated favorably by others (Fehrler & Przepiorka, 2013). For this reason, one potential benefit that motivates giving may be to secure or protect one’s reputation (Bénabou & Tirole, 2006). People may give to secure reputational rewards (i.e., to have people think well of them) or to avoid reputational punishment (i.e., to avoid having people think ill of them). Some donors will be more responsive to rewards than others. Likewise, helping some beneficiaries will be more (or less) likely to earn the donor reputational rewards than helping other beneficiaries, a point we elaborate later. Emotions. In addition to the good feeling that sometimes comes from giving, various emotions have been demonstrated to affect charitable giving. Giving can be enhanced when the donor feels nostalgic (Zhou et al., 2012), grateful (Ma et al., 2017), or determined (Liang et al., 2016). Emotions can also determine the type of action that is taken. For example, people who feel sympathy for beneficiaries are more likely to respond with donations than with activism (Thomas & McGarty, 2018; Thomas, McGarty, & Mavor, 2009b). So-called “bad” emotions are also instrumental in giving. Some donors report giving to alleviate guilt (Basil et al., 2008; Hibbert et al., 2007). Sadness increases giving, especially to vaguely defined groups of victims presented as statistics (Kandrack & Lundberg, 2014; Small & Verrochi, 2009). And anger can also promote giving (Chapman, Mirnajafi, Louis, & Masser, under review; Van de Vyver & Abrams, 2015; van Doorn et al., 2017), though not in all circumstances (Kayser et al., 2010; Siegel et al., 2016). There is little systematic comparison of these emotions, however, or understanding of when and how they arise and when they fail to motivate action. Further, a range of emotions that may be relevant to giving remain to be systematically tested. These include fear, curiosity, joy, awe, regret, and compassion. Identities. Individuals can identify with social or opinion-groups that they belong to or with roles they play in society (Bliuc et al., 2007; McCall & Simmons, 1978; Tajfel, 1981; Turner et al.,

127

1987). Identities have been theorized to motivate charitable giving (Aaker & Akutsu, 2009; Chapman, Louis, et al., 2018; M. Levine & Manning, 2013). Donors seem to give to causes that reflect the priorities of their important social groups (Chapman, Louis, et al., 2018; Thomas, Cary, Smith, Spears, & McGarty, 2018; Thomas, Smith, et al., 2018), and are more likely to donate when their help will be directed to people who belong to their own group (Charnysh et al., 2015; M. Levine et al., 2005; M. Levine & Thompson, 2004). Role identities also matter: people are more likely to give who identify more strongly as a donor (Kessler & Milkman, 2018; K. M. White, Poulsen, & Hyde, 2017), or who are higher in moral identity—the self-importance of being seen as a moral person (Aquino & Reed, 2002; Saerom Lee et al., 2014; Winterich et al., 2009; Winterich et al., 2012). Summary Some types of donors are more likely to donate to charities and are typically more generous when they do so. Donor demographics help to understand who gives to charity and how much they give. Individual differences between people may help us to understand the psychological mechanisms underpinning charitable giving; these include empathy, religiosity, political orientation, ideologies, and values. Even taking such stable differences into account, situation-dependent aspects of the donor such as benefits and reputation, emotions, and identities can further affect both the likelihood that someone will give to charity. Despite a great deal of research energy being devoted to the question of who gives, the consideration of donor characteristics above has not produced a conclusive answer: there are many discrepancies and important caveats in the evidence discussed. We propose that donor-centric approaches have been limited to the extent that they have failed to consider how the targets of giving (beneficiaries), solicitors of gifts (fundraisers), and wider social context may affect who gives and how much. Beneficiary-Centered Approaches to Understanding Charitable Giving A key consideration in whether help will be offered is the degree to which the potential beneficiary is considered to be worthy of care. Worthiness can be determined by a range of factors, including the degree to which they are perceived to be in need, responsible for their situation, valued, and identifiable. Need Need has often been studied in relation to the donor, with a focus on whether or not the potential donor is aware of need (see overview by Bekkers & Wiepking, 2011b). Yet need is really something related to the beneficiary: does the target look like they need help? Greater perceived need is associated with greater giving (Bennett & Kottasz, 2000; Chapman, Masser, et al., under review; Leeuwen & Wiepking, 2012; Loseke & Fawcett, 1995; Polonsky et al., 2002; Zagefka et 128 al., 2012). A focus on need is pragmatic. Why would a donor part with their money unless it was needed? However, this focus on need can also influence who receives help because some needs are more visible than others. Responsibility Relatedly, some beneficiaries are judged to be responsible for their situation, while others are seen as innocent. When the beneficiary is perceived to be responsible for their predicament, they may be judged less worthy of aid and therefore be less likely to elicit donations (Fong & Luttmer, 2011; Leeuwen & Wiepking, 2012; Loseke & Fawcett, 1995; Rudolph, Roesch, Greitemeyer, & Weiner, 2004). People may explain global poverty, for example, as variously caused by the poor themselves, natural causes, corrupt governments, civil conflict, or international exploitation (Bolitho, Carr, & Fletcher, 2006). The causal attributions made, in turn, shape donation behavior but can also be seen as narrative constructions (or post-hoc explanations for behavior; Harper, 1996). Donors are more likely than non-donors to make situational attributions about the causes of poverty (governments, conflict, exploitation, nature, vs. blaming the poor themselves), further showing that giving is suppressed when beneficiaries are held responsible (Campbell, Carr, & Maclachlan, 2006; Carr & Maclachlan, 1998). This responsibility effect holds even for disasters: those that have human causes elicit less help than those with natural causes, an effect which is explained in part by perceptions that the victims are to blame (Zagefka et al., 2011). To some extent, this prejudice against responsibility implies that “worthy” beneficiaries are those who are stereotyped as lower in agency (see Cuddy et al., 2007), and perceived competence does indeed have an inverse relationship with receiving help. One study of Chinese people’s willingness to help Japanese earthquake victims found that, to the extent that victims were perceived to be competent, people were less concerned for them (Sun, Zagefka, & Goodwin, 2013). Perceived value Some lives are valued more than others (Johansson-Stenman & Martinsson, 2008). Beneficiaries that are seen to be warm (and not capable) are the most likely to garner active help (Cuddy et al., 2007). Other types of beneficiaries may be less likely to elicit sympathy from potential donors (Body & Breeze, 2016), meaning some types of beneficiaries get more help than others. For example, charities supporting children, animals, and sick people may be particularly popular (Chapman, Masser, et al., under review; Radley & Kennedy, 1992), while charities supporting refugees, offenders, and prostitutes may be unpopular (Body & Breeze, 2016). When beneficiaries are dehumanized—seen as more like animals or machines than like humans—donors feel less empathy for them and are less likely to help (Andrighetto, Baldissarri, Lattanzio, Loughnan, & Volpato, 2014). Donors sometimes prefer to help beneficiaries who they perceive are 129 trying to help themselves (Bennett & Kottasz, 2000; Burgoyne, Young, & Walker, 2005). However, these beneficiaries must be failing in their attempt, or they won’t be perceived as being needy. Identifiability We help identifiable victims more than anonymous ones (Dickert et al., 2016; Genevsky et al., 2013; Jenni & Loewenstein, 1997; Kogut & Ritov, 2005; Seyoung Lee & Feeley, 2016). That means that a named beneficiary receives more support than a beneficiary in the abstract. For example, people were as much as 78% more willing to contribute to help a child who was identified by their name, age, and a photograph than to help an anonymous child (Kogut & Ritov, 2005; Study 1). Further, donors are more generous to identifiable victims: giving as much as twice as much to help identified victims than victims presented as mere statistics (Small & Loewenstein, 2003; Small, Loewenstein, & Slovic, 2007). Thus, framing an appeal to help a single, identifiable target should be more effective in raising money than framing the appeal to help anonymous targets, although this is not always found (Erlandsson, Björklund, & Bäckström, 2017; Lesner & Rasmussen, 2014). For example, one study in India found the effect did not explain participants’ willingness to donate to stigmatized (lower-caste) recipients (Deshpande & Spears, 2016). A second study showed that the identifiable victim effect occurs only when donors are temporally or socially close to the donation target (Ein-Gar & Levontin, 2013). A meta-analysis identified a number of moderators of the identifiable victim effect: the effect is stronger when a single child victim is presented, who is suffering from poverty, not responsible, and who is depicted in a photograph (Seyoung Lee & Feeley, 2016). In combination these findings suggest that the effect of beneficiaries depend also on their relationship with the potential donor in question, a point we return to below. Relatedly, individuals receive more support than groups of beneficiaries: a phenomenon called psychic numbing (Slovic, 2007). One person is more motivating in eliciting aid than many, while millions of victims are overwhelming. In reality, numbing seems to set in with victim groups as large as just two. One study showed that people will actually donate less to help two children than they would have donated to help either one of them individually (Slovic, 2007). This preference for helping individuals over groups can be attenuated when the multiple targets are perceived as forming a single entity—such as a family (R. W. Smith, Faro, & Burson, 2013). Summary A range of factors determine whether or not a beneficiary is worthy of being helped. To maximize their chances, beneficiaries must be clearly in need but not responsible for their predicament. They should be trying to help themselves, but not be too competent. They should be valuable and likeable, identifiable and individual. These preferences also impact who is helped in 130 important ways. The needs of some beneficiaries are less visible than the needs of others, while some suffering looks worse. For example, people give more to disasters that have a greater number of fatalities but do not consider the number of affected survivors, despite the latter being the actual beneficiaries of aid (Evangelidis & Van den Bergh, 2013). Beneficiary factors also help us understand preferences for helping animals and children (typically low in agency) and sick people (typically low in perceived responsibility). Not only do such preferences impact who will receive donations, they also shape the way charities communicate about beneficiaries. These communications, in turn, may further accentuate beneficiary stereotypes. Donor-Beneficiary Dyad: Who Gives Depends on Who Will Receive The charitable triad model shows that donors and beneficiaries do not influence giving in isolation. Instead, there is an interaction—particular donors decide whether or not to give to particular beneficiaries. Such donor-beneficiary dyads have been explored in the context of other forms of prosocial behavior (e.g., Nadler & Halabi, 2006; Penner et al., 2005; Stürmer & Siem, 2017). Below we review the existing evidence for how the relationships between donors and beneficiaries influence charitable gifts; specifically considering relationships of proximity, knowledge, shared identities, and power. Proximity Proximity influences charitable giving. People give more to help beneficiaries that live near them (Erlandsson et al., 2017; T. K. James & Zagefka, 2017). Similarly, having social proximity to a cause can increase giving (Ein-Gar & Levontin, 2013). For example, among donors who had given to illness charities in the last year, members of sexual minority groups were almost 8 times more likely to have donated to HIV and AIDS causes than members of sexual majority groups (Allman et al., 2014). Further, having a friend who has suffered increases generosity towards victims of the same misfortune (Small & Simonsohn, 2008). Proximity can also be hypothetical. Zagefka (2018a), for example, found that people were more likely to donate to disaster appeals when the disaster happened in a place they had previously wished to visit. Knowledge. Donors will help beneficiaries that they have special knowledge about. Such knowledge may be from personal experience, such as having had a friend or family member suffer from the same, or vicarious experiences, such as having met people that have suffered (Burgoyne et al., 2005; Polonsky et al., 2002; Small & Simonsohn, 2008). The impact of knowledge is also causal. Experimental evidence also shows that merely presenting facts about a location can increase the likelihood that beneficiaries in that place would receive help from foreign donors (Zagefka et al., 2013). Shared identities. One of strongest predictors of giving between particular donors and beneficiaries is whether or not they share an identity (Prendergast & Maggie, 2013). In a global 131 qualitative study of donors’ reasons for giving, shared identities were explicitly named as their motive by 8% of the 1,849 participants (Chapman, Masser, et al., under review). The kinds of shared identities most frequently mentioned in that study were geographic, gender, and identities based on an experience of illness. In another study of University alumni donations, 68% of donors explained their giving in terms of their alumni identity (Stephenson & Bell, 2014). Bequest donors also report giving because they identify with the beneficiary group (Sargeant & Shang, 2011). And there is evidence from interpersonal contexts that empathy motivates helping to ingroup rather than all targets (Stürmer et al., 2006; Stürmer, Snyder, & Omoto, 2005). The role of shared identity has also often been shown for geographic identities. Many donors prefer to give to people in their national group or local community (e.g., Erlandsson et al., 2017; T. K. James & Zagefka, 2017) and donors may actually think of charities in terms of a binary distinction between domestic and international beneficiaries (Breeze, 2013; Stevenson & Manning, 2010). Donors show a strong preference towards helping their own: a nationally representative survey of 13,125 Canadians, for example, found that 81% gave exclusively to Canadian charities compared to the less than 1% who gave exclusively to charities in other countries (Rajan et al., 2009). Many studies have shown that a huge proportion of donor funds are allocated to charities serving local national beneficiaries—94% of all donations in the United States (Giving USA, 2018b) and 90% of donations in the United Kingdom (Charities Aid Foundation, 2017a) are directed to domestic beneficiaries. Such preferences can be further shaped by donor dispositions. For example, people high in social dominance orientation, right-wing authoritarianism, and political conservatism demonstrate an aversion to helping racial outgroups (Perry et al., 2015; Webster et al., 2014). Sometimes donors are likely to help beneficiaries outside the traditional boundaries of their own group. People give more to traditional outgroups when the beneficiaries are framed as sharing an important inclusive or superordinate identity. For example, English people were more willing to help European disaster victims when their European identity was primed than when their British identity was primed (M. Levine & Thompson, 2004). Likewise, Hindus reported a greater willingness to help Muslim disaster victims when their shared Indian national identity was made salient (Charnysh et al., 2015). These findings demonstrate a strong preference to give to people who share important identities, and show that changes in identity salience can affect giving decisions. Helping outgroups can also sometimes be done for strategic reasons, to benefit the ingroup (Hopkins et al., 2007; van Leeuwen, 2017). For example, people may be more generous to outgroups (English) when they perceive that their own group (Scottish) is stereotyped as stingy— suggesting a strategic motive for helping dissimilar others (Hopkins et al., 2007). Similarly, in a

132 series of studies, van Leeuwen and colleagues demonstrated that when their national identity is threatened people give to help beneficiaries in other countries as a way to restore group-esteem or positive distinctiveness (van Leeuwen, 2007; van Leeuwen & Harinck, 2016; van Leeuwen & Mashuri, 2012). Overall, the proximity of the beneficiary to the donor appears to influence the likelihood of charitable giving. In at least one way, however, beneficiaries who are further from the donors fare better: in terms of power. Power Any helping that occurs between members of different groups involves status differentials (Halabi & Nadler, 2017; Nadler, 2002, 2016). Almost by definition, beneficiaries must be perceived to be needy in order to receive help. However, neediness communicates low status. Likewise, helping communicates high status: one cannot help someone without having more of what is needed (in this case money) than the beneficiary. As such, beneficiaries by definition must be lower status than donors. The relationship between helping and status has been studied extensively in the social psychological literature (Halabi et al., 2008; Halabi, Dovidio, & Nadler, 2016; Halabi & Nadler, 2017; Nadler, 2002, 2016; Nadler & Chernyak-Hai, 2014; Nadler & Halabi, 2006; Nadler et al., 2009). High status beneficiaries are less likely to be the targets of prosocial responses (Van Doesum, Tybur, & Van Lange, 2017); and receiving help can be an aversive experience for some beneficiaries who experience self-esteem threat (Fisher & Nadler, 1974; Fisher, Nadler, & Whitcher-Alagna, 1982). From this point of view, beneficiaries who do not share group membership with donors will ‘benefit’ from being lower status—in the limited sense that they are more likely to receive financial support from outgroup donors. Beneficiaries perceive help from high status groups as a power move (Halabi et al., 2016). Further, low-status group members who identify strongly with their group do not like receiving help from high-status groups, especially when group relations are unstable and the form of help offered encourages dependency (Halabi et al., 2011; Nadler & Halabi, 2006). Finally, although we have seen that donors are more likely to give when the beneficiary is similar to them, some beneficiaries actually experience more negative outcomes (such as self-esteem threat and reduced self- confidence) when they receive aid from others who are more similar (Fisher et al., 1978; Fisher & Nadler, 1974; Nadler & Fisher, 1986; Nadler et al., 1976). Summary Generally speaking, donors give more to beneficiaries who are proximal, who they have knowledge about or experience with, or with whom they share important identities. When, however, beneficiaries fall beyond this circle of care, they are more likely to be helped if their group is perceived to be lower in status than the donor’s own group. 133

Though largely unexamined to date, there are likely other important ways that the donor- beneficiary interaction influences charitable giving. For example, people high in social dominance orientation may be more motivated by strategic group concerns, therefore giving only to low-status beneficiaries and only when they perceive likely ingroup benefits. Indeed, certain beneficiaries may more readily highlight these benefits to the donor. On the other hand, people high in empathy may be especially motivated by extremely needy or identified victims. The threshold of neediness and perceived responsibility may also change depending on whether the target beneficiary is in the donor’s salient ingroup or outgroup. These hypotheses warrant investigation. Research considering donor-beneficiary dyads provides a more nuanced view of giving than one that considers only the donor. According to the charitable triad model, however, it still leaves out an essential component of the charitable interaction: the fundraiser. Fundraisers Promote Charitable Giving Donors rarely give to charity without being asked (Bryant et al., 2003) and generally beneficiaries do not ask for themselves. The ask is an essential component of the gift and draws our attention to a hitherto largely neglected component of the charitable triad: the fundraiser. Fundraisers are the ones asking for money—sometimes called solicitors because they solicit funds. The role of fundraisers in charitable giving has often been deemphasized in favor of donors (Breeze, 2017a, 2017b). Yet being asked is a predictor of both the likelihood and value of charitable gifts (Andreoni, Rao, & Trachtman, 2017; Wiepking, 2007), highlighting the important role that the fundraiser plays. Consideration of the fundraiser demonstrates that giving decisions are influenced by a triad of actors (see Figure 6). Although not all prosocial behavior is triadic—sharing and helping, for example, are usually dyadic—charitable giving is almost always a response to a third party asking on behalf of the beneficiary. For this reason, the fundraiser is an essential actor in the charitable triad and, as we shall see, the actor that has been most neglected in research. One complexity here is that fundraisers can be both individuals and/or organizations. Think of your friend running a marathon to raise money for charity or the person who stops you at the mall to ask you to sponsor a child. In this situation, the ask is made by an individual but on behalf of a charitable organization. Indeed, in many of the countries in which charitable giving is studied (predominantly in the West) one of the primary ways that charities serve their beneficiaries is by soliciting funds on their behalf. Some large international organizations in developed countries even exist with the sole purpose of raising money. Below we review the existing literature on the role of fundraisers in charitable giving—both individuals as fundraisers and charities as fundraisers.

134

Individual Fundraisers When people think about fundraising, they may spontaneously think about people on street corners shaking buckets or standing next to a display of sponsor children at the mall. In other words, people think about individual fundraisers. Yet, as Breeze (2017a) highlights, the role that fundraisers play in charitable giving is rarely studied. The small literature on this topic has been qualitative in nature and largely focused on the experiences of major donor fundraisers who develop one-on-one relationships with high-value donors and philanthropists (Alborough, 2017; Breeze & Jollymore, 2017; Dean & Wood, 2017). Successful fundraisers know how to use emotional levers when soliciting funds (Dean & Wood, 2017). Major donor fundraisers work as trusted brokers who build the donor’s confidence in the charity and help the donor to understand the charity world (Breeze & Jollymore, 2017). They see themselves as mediators between the donors and the beneficiaries (Alborough, 2017). In short, fundraisers are inherently aware of the triadic nature of charitable giving. Less is known about how fundraiser characteristics may enhance or inhibit their effectiveness. Fundraisers are often women (Dale, 2017) and female fundraisers are more successful if they are physically attractive (Landry et al., 2006; West & Jan Brown, 1975). Further, fundraisers who are well dressed may have more success than those that are poorly dressed, though the evidence is mixed (Bull & Gibson-Robinson, 1981; L. R. Levine et al., 1998). Donors recruited by fundraisers that they perceived to be professional, knowledgeable, and helpful are more likely to continue supporting the cause (Sargeant & Hudson, 2008). In the context of door-to-door fundraising in the United States, Caucasian fundraisers raise more money than minority fundraisers, regardless of the race of the donor (List & Price, 2009); the authors reasoned that donors are less trusting of fundraisers who belong to historically marginalized groups. In addition to physical attributes, the reason an individual is asking appears to matter. Experimental evidence suggests that people give less to fundraisers who have been incentivized, even when those donors do not know about the incentives (Barasch et al., 2016). This suggests that incentivized fundraisers may appear less sincere which, in turn, may limit the number of people who will respond or the value they donate when they do give. Although individual fundraisers play a pivotal role in mediating between donors and beneficiaries, research has rarely investigated the characteristics of individual fundraisers that make them successful. The limited research that has been done has dwelt primarily on physical attributes and inferred motivations. Little theorizing has yet been done as to why or how these fundraiser characteristics matter. More research, however, has interrogated the characteristics that make fundraising organizations successful.

135

Organizations as Fundraisers Charitable organizations themselves are the fundraisers that we are most likely to come into contact with: via TV ads, billboards, radio ads, and direct mail. It is the brand that is doing the asking, and the one we need to trust. Below we review the roles that perceived efficacy of the organization, trustworthiness, mission, and other organizational characteristics play in charitable giving. Efficacy. Charities that are effective are those that make an impact on the cause they represent. Generally speaking, charities that are perceived to be making the greatest impact should receive more funding (Duncan, 2004). Although donors are the ones who perceive and respond to the effectiveness, the quality lies in the charity itself. Does the charity make some difference in the world? Does it help its beneficiaries in meaningful ways? Donors say that the effectiveness of the charity is an important consideration in deciding whether or not they will donate (Katz, 2018; Scharf & Smith, 2016) and report greater intentions to give when the think their donation will make a difference (Iyer et al., 2012; J. R. Smith & McSweeney, 2007). Some popular campaigns—such as the effective altruism movement (effectivealtruism.org, 2018; Singer, 2009)—even promote giving to organizations that are perceived to be especially effective. Other studies, however, have found that charity effectiveness is relatively unimportant in determining gifts (Berman et al., 2018; Chapman, Masser, et al., 2018; Saxton & Wang, 2014; Tremblay-Boire & Prakash, 2017). A series of studies designed to test the impact of effectiveness ratings on charity preferences found that people prioritize subjective preferences over rational data on effectiveness in allocating donations (Berman et al., 2018). Further, two separate longitudinal analyses of charity watchdog ratings and funding in the United States found that changes in charity effectiveness ratings do not influence donor support (Silvergleid, 2003; Szper & Prakash, 2011). Charitable effectiveness may therefore be a consideration in giving but its importance may depend on the particular donor or beneficiary in question. Trustworthiness. While effectiveness relates to perceptions of a charity’s competence, trust relates to perceptions of a charity’s integrity. It is generally assumed that charities must be trusted to receive funds. Many studies find a link between giving and trust: whether generalized trust (believing that most people can be trusted; e.g., Bekkers, 2003; Brooks, 2005; Evers & Gesthuizen, 2011; Herzog & Yang, 2018; Neumayr & Handy, 2017; Wiepking & Maas, 2009), trust or confidence in the charitable sector (e.g., Alhidari, Veludo-de-Oliveira, Yousafzai, & Yani-de- Soriano, 2018; Bourassa & Stang, 2016; Osili, Hirt, & Raghavan, 2011), or trust in specific organizations (e.g., Beldad et al., 2014; Naskrent & Siebelt, 2011; Sargeant & Hudson, 2008; Sleesman & Conlon, 2016; Thomas et al., 2015). Trusting charities in general is positively associated with giving more to charities; and people give more to specific charities if they trust

136 them. Yet occasionally no significant relationship has been found between trust and giving (Katz, 2018; Tremblay-Boire & Prakash, 2017; Wiepking, 2010). Anecdotally, many people believe that trust is an integral component in giving. However, on-the-ground evidence is mixed. For example, after the international development agency Oxfam was involved in a highly-publicized scandal about sexual misconduct by some high-ranking staff members, Oxfam UK reported at least 7,000 regular donors had cancelled their gifts (Elgot & McVeigh, 2018). Yet the Charities Aid Foundation UK reported that the scandal had had no discernable impact on the public’s level of trust in charities in general (Charities Aid Foundation, 2018). Such patterns of trust suggest that trustworthiness is most relevant to decisions at the individual charity level. Alternatively, trust may matter for some donors but not others; or trust in the organization may be a concern only when serving certain beneficiaries. Mission. The charity’s mission is an important consideration in the success of their fundraising. This mission often highlights specific beneficiaries but may also highlight methods. For example, the environmental organizations Greenpeace and Sea Shepherd both have similar missions to protect ocean wildlife and habitats. The latter, however, uses more militant methods. On the other hand, some organizations have the same methods but different missions—such as two competing political parties. When considering whether a gift will be made both the mission and the method may be important considerations. Some sympathizers are reluctant to take part in causes that they perceive to have a negative public image (Klas, Zinkiewicz, Zhou, & Clarke, 2018; Stuart, Thomas, & Donaghue, 2018). The mission of the fundraising organization will influence giving; a point we return below to when discussing the donor-fundraiser dyad. Other organizational characteristics. Well-known charities have greater fundraising success than obscure ones (Kinsbergen & Tolsma, 2013). Donors also pay attention to efficiency ratios—the percentage of donations that are used to cover overheads—with donors reporting greater willingness to donate to more efficient charities (Gneezy, Keenan, & Gneezy, 2014; Kinsbergen & Tolsma, 2013; Michel & Rieunier, 2012; Tinkelman & Mankaney, 2007). However, efficiency information does not affect the value of giving (Ryazanov & Christenfeld, 2018). The marketing tactics and channels charities use also affect giving (e.g., Barasch et al., 2016; Jensen, King, & Carcioppolo, 2013; Newman & Jeremy Shen, 2012; Schlegelmilch, Love, & Diamantopoulos, 1997), although these reflect choices fundraisers make rather than characteristics of the fundraisers themselves. Summary The literature on fundraiser characteristics illustrates an asymmetry in the focus of research. Very little research has focused on fundraisers as essential components of charitable decisions. The emphasis has been on understanding how charities-as-fundraisers can succeed rather than 137 determining the factors that make individuals-as-fundraisers effective. Traditionally, charities fundraised through broad, mass market channels like TV, radio, and print, or via branded direct mail. These days, however, more personalized channels of recruitment and solicitation are becoming normative. Face-to-face, door-to-door, and telephone fundraising are popular methods of fundraising and each requires an interaction between an individual fundraiser (as a representative of the charity) and a potential donor (Jay, 2001; Sargeant & Hudson, 2008; Sargeant & Jay, 2004; Sargeant & Kähler, 1999). Social media interactions around causes and peer-to-peer fundraising have also increased, meaning more and more people are being asked to give by people they know personally (Higgins & Lauzon, 2003; Lucas, 2017; Miller, 2009). These changes necessitate a refocus on the psychology of the fundraiser, and especially a new consideration of the way donors and fundraisers interact to determine fundraising outcomes. Donor-Fundraiser Dyad: Who Gives Depends on Who Asks A giving exchange usually occurs between a fundraiser and a potential donor. The beneficiary is unlikely to be present but is inferred or brought to life in the process (by the fundraiser). Although we have reviewed evidence that details how characteristics of the donor or the fundraiser individually contribute to the giving process, in reality these rarely occur in isolation. Instead particular fundraisers interact with particular donors, sometimes winning donations and sometimes walking away empty-handed. Below we discuss the existing literature that considers both fundraisers and donors in this interaction—focusing on the closeness of the relationship and alignment of priorities—before highlighting the gaps in our current understanding of this dyad. Closeness of the relationship. People give more when asked by people that they are personally close with (Meer, 2011; Scharf & Smith, 2016). Strong evidence of this phenomenon comes from studies of the effectiveness of peer-to-peer fundraisers. These are people who do not work for a charity but voluntarily fundraise on an organization’s behalf, such as people taking part in fun runs or organizing bake sales. One survey of more than 19,000 peer-to-peer donors found that 87% said they always gave when asked by a family member, 67% when asked by a friend, 48% when asked by a colleague, and just 9% when asked by a professional fundraiser (Scharf & Smith, 2016). Further evidence from the peer-to-peer fundraising domain suggests that social concerns, such as maintaining a positive relationship with the fundraiser, may be stronger drivers of giving than the perceived efficacy of the charity in question (Chapman, Masser, et al., 2018). People can also relate to each other at the level of groups and their associated identities. Similar to the way that donors are more likely to give to beneficiaries who share an identity with them, we would expect that donors may also be more likely to give to fundraisers who share an identity with them. Although one study found no effect of donors sharing gender or racial group membership with fundraisers (List & Price, 2009), another study found that football fans donated

138 more money to fundraisers stationed outside a football stadium if those fundraisers were wearing the same team colors as the fan (Platow et al., 1999). However, research has yet to systematically examine if and when shared identities between donors and fundraiser affect giving. Alignment of priorities. A donor will be more motivated to support charities-as- fundraisers that advance the donor’s priorities. Thus, different donors will be mobilized by different charity missions. A great deal of evidence supports the claim that different donors are attracted to different charities: women are more likely to give to animal charities while men are more likely to support sports organizations (Neumayr & Handy, 2017; Piper & Schnepf, 2008; Srnka et al., 2003); older people give more to welfare, religious, and health charities while younger people give to environmental groups and animal welfare (Chapman, Louis, et al., 2018; R. N. James & Sharpe, 2007; NAB, 2015; Nilsson et al., 2016; Srnka et al., 2003; Wiepking, 2010); and people who are more religious show preferences toward religious, international, and welfare charities but are less likely to give to animal welfare or environmental causes (Casale & Baumann, 2015; Chapman, Louis, et al., 2018; Choi & DiNitto, 2012; Forbes & Zampelli, 2013; Helms & Thornton, 2012; Hill & Vaidyanathan, 2011; Neumayr & Handy, 2017; Schnable, 2015; Showers et al., 2011; Wang & Graddy, 2008; Wiepking, 2010). Finally, politically progressive and conservative donors also show different patterns of charity preferences (Chapman, Louis, et al., 2018; Nilsson et al., 2016; Perry et al., 2015; Vaidyanathan et al., 2011; Van Lange et al., 2012; Wiepking, 2010; Winterich et al., 2012). Summary A gift is more likely to be made when donors and fundraisers share priorities or when they have existing relationships. Yet very little work has explicitly investigated the donor-fundraiser dyad. An opportunity exists for both cross-sectional and experimental research to understand how the unique interaction of particular donors with particular fundraisers determines whether or not a gift will be made. In addition, we must consider the ‘silent’ party in the interaction—the beneficiary—and how their relationship with the fundraiser may influence that fundraiser’s success. Fundraiser-Beneficiary Dyad: Who Asks Determines If They Receive What do we know about the way fundraisers and beneficiaries interact to influence giving? Very little. Figure 6 draws our attention to the dearth of research on fundraiser-beneficiary dyads. There is some evidence that people find it harder to say no to a fundraiser who is either a beneficiary themselves or has a personal relationship with beneficiaries (Ratner, Zhao, & Clarke, 2011; Ratner, Zhao, & Miller, 2009). For example, a mother of a differently-abled child who is fundraising for a child disability organization may be more successful than a paid fundraiser. This effect may be a signal of authenticity. In recent years, negative publicity has emerged to condemn the work of so-called chuggers, or “charity muggers” (Anonymous, 2016; Walker, 2016). These are 139 incentivized, paid charity fundraisers who lack any meaningful relationship with the beneficiaries, and work on contract to recruit donors either on street curbs, malls, or at the donor’s front door. Indeed, fundraisers who are incentivized may raise less money (Barasch et al., 2016). Growing discomfort about fundraising tactics and mistrust of charity messages may make fundraisers with personal relationships with the beneficiary seem more genuine and therefore more successful. However, very little research has yet examined the fundraiser-beneficiary dyad. Putting Charitable Giving in Context We have reviewed the existing literature on charitable giving from three perspectives: that of donor, beneficiary, and fundraiser. Beyond characteristics of these three actors in isolation, we have also demonstrated that charitable giving is influenced by the dyadic interactions between them (donor-beneficiary, donor-fundraiser, and fundraiser-beneficiary). However, each of these dyadic interactions would also, in turn, be influenced by the broader social context in which those interactions occur (see Figure 6). Answering calls to consider the social context in which charity occurs (Hibbert & Horne, 1996), we now turn our attention to the wider context and how it may influence the actors in the triad and their relationships with each other. The charitable triad model highlights five key factors that demonstrate the importance of considering the broader circumstances in which a charitable gift is made: social norms, audience, issue salience, giving context, and other funding sources. We discuss evidence and theorize further ways that such contextual factors may influence the triad. Perceived Norms Identities have been discussed above, both in respect to donor identities and the way shared identities between donors and either beneficiaries or fundraisers can impact donation decisions. Identities also bring with them norms (Terry & Hogg, 1996). Social norms provide guidance to group members about what is expected or typical in a particular situation (Cialdini et al., 1990). Norms are generally specific to a particular social group, such as national norms, gender norms, or academic norms, which prescribe what citizens, men or women, and academics do, respectively. Such norms may also have different expectations for the form, strength, and timing of care that is given (see M. Levine & Manning, 2013; Thomas, McGarty, & Mavor, 2009a). Evidence abounds that norms influence charitable giving (Agerström et al., 2016; Bartke et al., 2017; Chapman, Louis, & Masser, in prep; Croson et al., 2009, 2010; Croson & Shang, 2008; Hysenbelli et al., 2013; Kashif, Sarifuddin, & Hassan, 2015; Nook et al., 2016; J. R. Smith & McSweeney, 2007; S. Smith et al., 2015). Norms can be communicated as social information about what others do. For example, a study of passengers on a German streetcar found that people were more likely to give if they observed an initial donor give first (Ebeling, Feldhaus, & Fendrich, 2017). Likewise, in the United States, callers to a public radio appeal gave larger donations when 140 they were informed that a previous caller had made a larger gift (Croson et al., 2009). Specifically, each $1 increase in the size of the previous caller’s stated gift was met with a corresponding 16 cent increase in the donation pledged (Croson & Shang, 2008). Donations can be influenced by the make-up of coins and notes in a donation box (Martin & Randal, 2008) or previous donations on an online fundraising page (S. Smith et al., 2015), explicit recommendations for donation value (Alpízar & Martinsson, 2012), and information about the size of donations received from other people (Croson & Shang, 2008). And adding information about local giving norms to standard appeals can as much as double donations (Agerström et al., 2016). These social norms are especially powerful when associated with an important identity. In the public radio experiment mentioned above, the effect of normative information was stronger when the former caller was presented as sharing the same gender as the donor (Croson et al., 2010; Shang, Reed, & Croson, 2008). Another study shared information with potential Italian donors about the average donations of either Italian or German donors (Hysenbelli et al., 2013). When the average gift was high and associated with the ingroup (Italians), 98% of participants elected to donate to the appeal; when the normative information was associated with the outgroup (Germans), however, only 75% elected to donate. Political identity can also influence the impact of norms. Political conservatives are more likely to conform to norms of a salient group identity than political liberals are (Kaikati, Torelli, & Winterich, 2014). In other words, normative information interacts with the donor characteristic of identity to determine whether and how the norm will have an influence on giving. Though untested, it is also possible that different groups have different normative targets of giving—norms that communicate not just if we help, but also who we help (see also Chapman, Louis, et al., 2018; M. Levine & Manning, 2013). If so, contextual information about norms could in fact influence who gives (donors), who receives (beneficiaries), and also the relationships between particular donors and beneficiaries. Audience When it comes to charitable giving, having an audience matters. People are more generous when their gifts are visible to others (Alpízar & Martinsson, 2013; Bereczkei et al., 2007; Bull & Gibson-Robinson, 1981; Reinstein & Riener, 2012; Satow, 1975; Soetevent, 2005). Even if the gift itself is not visible, just knowing someone is watching when the giving decision is being made can influence the response. In a series of experiments at a national park in Costa Rica, Alpízar and Martinsson (2012, 2013) found that people were more likely to make a voluntary donation on entry when they arrived in groups or if there was someone present at the entrance. Even short exposures to images of eyes can increase generosity (Sparks & Barclay, 2013; Vaish, Kelsey, Tripathi, & Grossmann, 2017). Such findings point to one conclusion—being seen increases giving.

141

Nonetheless, the very same processes may reduce giving to non-normative or stigmatized targets, when privacy might increase the likelihood that giving will occur. Audience effects are more profound for men, people from higher social classes, and people higher in narcissism (Böhm & Regner, 2013; Konrath, Ho, & Zarins, 2016; Kraus & Callaghan, 2016), suggesting that some people, or perhaps some groups of people, are more sensitive than others to the presence of an audience. An interaction with the donor characteristic of reputation is probable—to the extent that donors care about maintaining or promoting their reputation, audiences will affect them more. Audiences may also improve adherence to group norms, further accentuating the influence of norms on giving (e.g., Postmes, Spears, Sakhel, & de Groot, 2001; S. D. Reicher, Spears, & Postmes, 1995). It is also feasible that some beneficiaries will benefit more from audience effects than others—those that are likeable, broadly considered valuable, or clearly in a lower status position may be those most likely to receive help when others are watching. Likewise, charities that are highly trusted, perceived to be effective, those with socially-desirable missions, or individual fundraisers that are especially popular are those who may be extra successful when the giving interaction is visible to an audience. Such propositions remain to be tested. Issue Salience The current salience of an issue can draw attention from both donors and fundraisers to new or existing beneficiaries. When a disaster strikes, the media coverage can accentuate the neediness of certain beneficiaries (Simon, 1997). As such, issue salience can influence donors and fundraisers by mobilizing them to care about particular beneficiaries. Sometimes new beneficiaries emerge during a crisis, such as when the Boxing Day tsunami in 2004 killed hundreds of thousands of people in South-east Asia. Other times a political or media moment brings salience to existing beneficiaries, as when the heartbreaking image of drowned 3-year old Alan Kurdi brought international attention (and donor support) for the ongoing Syrian refugee crisis in Europe (see Slovic, Västfjäll, Erlandsson, & Gregory, 2017; L. G. E. Smith, McGarty, & Thomas, 2018). Indeed, salience is most often influenced by media coverage of current events and political debate (e.g., Lobb, Mock, & Hutchinson, 2012) and social media interactions (e.g., Thomas, Cary, et al., 2018). Quantifying the effect of issue salience, one study found that each extra minute of nightly news coverage of the 2004 tsunami lifted online donations to five humanitarian organizations by 20% the following day (P. H. Brown & Minty, 2008). Giving Context Different giving contexts are also proposed to influence the relationships in the charitable triad. People show different patterns of giving when donating through their workplace giving programs (Osili et al., 2011), giving on social media and through peer-to-peer campaigns 142

(Chapman, Masser, et al., 2018; Saxton & Wang, 2014; Scharf & Smith, 2016), or leaving a bequest (R. N. James, 2016; R. N. James & Routley, 2016; Wiepking, Scaife, & McDonald, 2012). These examples show how the context of giving, such as workplace, events, or within social networks, can change the way donors respond to fundraisers and the types of beneficiaries that are more likely to receive help. Although it is clear that the general setting or context of the charitable act influences giving, it is not clear exactly why and how different giving contexts affect donation outcomes. Funding Sources The state of existing funding can influence charitable giving. A prolific literature in economics has studied the impact of government funding on private giving. The crowding out hypothesis proposes that rational donors will reduce their giving dollar-for-dollar in response to government funding (see Roberts, 1984; Warr, 1982). Sometimes crowding out is observed—with increased government funding reducing private donations to some extent (Andreoni & Payne, 2011; A. A. Payne, 1998; W. O. Simmons & Emanuele, 2004). Other times, research shows evidence of crowding in—with increased government funding increasing private donations (Herzer & Nunnenkamp, 2013; Heutel, 2014; Hughes et al., 2014; Khanna et al., 1995; Khanna & Sandler, 2000; Neto, 2018; Okten & Weisbrod, 2000). Other studies have found no significant effects either way (Brooks, 2003a; De Wit et al., 2017). To date it remains unclear what causes these diverging responses. It may be that government funding is a contextual factor that has different effects for different donors, beneficiaries, and fundraisers, but at present the specific relationships are unclear. For some donors, the government funding may communicate relevant norms about giving, including whom to help. For some fundraising organizations, outside funding may enhance their trustworthiness or perceived effectiveness. Thus, other funding sources may influence the charitable triad in important ways. Summary As social creatures, humans live and act in social contexts. Any discussion of the psychology of charitable giving therefore requires discussion of the wider context in which the gift is made. We have reviewed the literatures on a range of contextual factors that influence giving. Social norms can promote giving when they are made salient and aligned with a relevant identity. When the charitable interaction is observed by others, donors behave differently than when their actions are private. The salience of an issue and the giving context in which the gift is made can influence the charitable decision, as can the presence and size of other funding sources. We have also discussed how each of these contextual cues may influence the triadic actors in different ways. Now that we have reviewed the evidence base on charitable giving in relation to the charitable triad, we summarize our detailed model of giving and present testable propositions and future directions for enquiry. 143

The Charitable Triad Model: Charitable Giving as Triadic, Relational, and Contextual At the start of this article, we proposed the charitable triad model, which conceptualizes charitable giving as a social process between donors, beneficiaries, and fundraisers. In Figure 6 we summarized the charitable triad model with reference to the literatures reviewed above. Large gaps have been identified. We know a lot about donors but less about beneficiaries, and almost nothing about fundraisers. Research on the donor-beneficiary dyad sheds light on the importance of examining the interactions between actors, yet research has rarely considered the donor-fundraiser and fundraiser-beneficiary dyads. To date, no single study has attempted to consider all three actors in unison. Several aspects of the model already have substantial evidence bases. First, research supports the notion that some people, because of their individual characteristics, values, ideology, identities, or demographics are more likely to be donors and more generous when they do give. The charitable triad model qualifies this point by proposing that donor characteristics are not sufficient to promote giving; donor choices will also depend on other salient actors in the triad. Second, research confirms the idea that some types of beneficiaries are deemed worthier of care and therefore garner more support. Third, there is evidence that beneficiaries who are perceived to be similar to donors (e.g., those who share identities) are more likely to receive aid. Despite the significant attention given to identifying donor and beneficiary characteristics, a range of factors have yet to be comprehensively tested. We suggest further examination of how giving is influenced by donors’ narcissism (Kang & Lakshmanan, 2018; Konrath et al., 2016), machiavellianism (Bereczkei, Birkas, & Kerekes, 2010), belief in a just world (Bègue, 2014; Campbell et al., 2006), and compassion (Leiberg, Klimecki, & Singer, 2011; Lim & Desteno, 2016; Lupoli, Jampol, & Oveis, 2017). Based on what we know from allied literatures (e.g., prejudice; Dovidio, Hewstone, Glick, & Esses, 2010), untested beneficiary factors that may also contribute to perceived worthiness include beneficiary gender, age, race, species, attractiveness, and displays of emotion. Finally, in addition to nationality or superordinate regional identities (Charnysh et al., 2015; M. Levine & Thompson, 2004), recent research points to identities related to gender, family, occupation, and suffering as particularly relevant to giving (e.g., Chapman, Masser, et al., under review; Lim & Desteno, 2016). Propositions Generated by the Model To address the gaps highlighted in the review above, we outline five novel propositions generated by the charitable triad model. These remain largely untested. We therefore provide suggestions that may help to guide future research.

144

Proposition 1. Dissimilar beneficiaries may receive aid if there are strategic interests in helping them for the donor (e.g. benefits, reputational concerns, or audience) or the donor’s group (e.g., hierarchy maintenance, impression management). In addition to a preference that donors show for helping beneficiaries perceived to be similar to them, some dissimilar beneficiaries may also receive support. Individual donors may be motivated by personal interests—reputational rewards or other benefits—to give to dissimilar others, especially when they have an audience and when the beneficiary is positively valued by their group. For donors who care about their group’s standing in the social hierarchy, we have seen that helping can be a strategic way to signal status and/or a way to communicate positive group stereotypes (e.g., Nadler, 2002; van Leeuwen, 2017). We further propose that this response relies on particular combinations of donors (high status or high identifiers) and beneficiaries (needy, stereotypical, low status, visible). In particular, low status beneficiaries that do not challenge the donor’s group position in the hierarchy will be those most likely to be helped by strategic high- status donors. Proposition 2. Individual fundraisers who are perceived to be authentic (e.g., moral, competent, and warm) will elicit more donations. There is little research investigating what makes individual fundraisers successful. Limited work points the roles of personal attractiveness and motivation (e.g., Barasch et al., 2016; Landry et al., 2006; L. R. Levine et al., 1998). Qualities such as gender, age, social status, race, emotion expression, professionalism (vs volunteerism), incentive levels, and experience remain to be tested. Each of these could feasibly influence the perceived authenticity of the fundraiser, or more broadly impact whether or not they are perceived as moral, warm, or competent (as above; see Cuddy et al., 2007; Dovidio et al., 2010). Such evaluations may come from external markers. For example, women, older people, and the religious may all be perceived to be higher in warmth and empathy. It is possible that fundraisers from these groups could be more successful, perhaps because of perceived authenticity. Likewise, emotional displays such as being visibly moved by the plights of beneficiaries when sharing their stories may enhance giving by promoting perceptions of warmth or morality. One study has shown that fundraiser incentives can depress donor generosity even when donors are ignorant of the incentive (Barasch et al., 2016). This finding should be replicated in naturalistic fundraising contexts to understand the difference in effectiveness between volunteer fundraisers those who earn a salary or incentives and, especially, to identify why unpaid fundraisers may be more successful. Proposition 3. Organizational fundraisers that are perceived to be higher in competence, integrity, and benevolence will be more successful.

145

Organizations also seem to be more successful in fundraising when they are trusted or seen as moral. The role of efficacy and efficiency seems more complicated. There are still many unknowns about what makes an organization an effective fundraising agent. We propose, however, that perceived competence (i.e., efficacy), integrity (i.e., trustworthiness), and benevolence (doing right by donors and/or beneficiaries) will be focal concerns given their role in trust formation within organizational contexts (see Mayer, Davis, & Schoorman, 1995). Proposition 4. Fundraisers that are aligned with either donors or beneficiaries will be more successful, and those aligned with both will be especially successful. Understanding who is asking, and how they relate both to who is being helped and who is potentially giving will be important. We propose that, for a fundraiser to be effective, they ideally should be aligned either with the donor or the beneficiary (i.e., by sharing a common identity or relationship of trust). If they are related to the donor—as a friend, colleague, or family member, for example—they are more likely to garner support, and also if they share a visible and important identity with the potential donor. Fundraisers who are dissimilar from the donor may still be successful when they are seen as aligned with the beneficiaries—when they are similar to or have close experience with the people that will receive the care. We expect that these two effects of alignment will be additive. As a hypothetical example, when a mother of a differently-abled child is soliciting funds for a disability charity from another mother, she will be more successful than if her child was not differently-abled or if the potential female donor was not a mother. Yet it is also possible that same mother of a differently-abled child might be most effective soliciting funds for a disability charity from a man—here, the fundraiser’s relationship with the beneficiary signals her authenticity and the intergroup dynamic between the fundraiser and the donor simultaneously makes salient the donor’s relative high status, which can be affirmed through the act of giving. Proposition 5. The unique interaction between particular donors, particular beneficiaries, and particular fundraisers determines both whether or not a gift will be made and the value of the gifts. The utility of the charitable triad model rests on four novel aspects. First, we propose that all three actors influence donation outcomes; yet two of the three—beneficiaries and fundraisers— are understudied. Second, we propose that dyadic relationships between actors moderate the impact of particular characteristics of any given actor upon giving, explaining some of the inconsistencies in previous work. Third, we propose that each dyadic relationship is further influenced by the third actor in the triad. Fourth, we have identified contextual factors that also impact on the giving decision interactively. We embed these four novel aspects of the model in a comprehensive literature review. Most important, in our view, is the alignment between the actors, whether in

146 terms of shared identity, power dynamics, or role relationships. Donors who identify with both beneficiary and fundraiser (albeit for different reasons) will be most moved to give. Such an effect should be enhanced when the fundraiser is seen as an authentic agent of the beneficiary. It may be that a meaningful interaction between any one of the dyads (as proposed above) will be enough to elicit a gift, but the third actor will determine the size of the gift. For example, empathic concern may motivate a donor to help children in distress overseas when asked. The donor will give because they feel the suffering of those needy beneficiaries and are motivated to respond. However, the donor may give twice as much when asked by Oxfam than when asked by Save The Children (for example), if they perceive Oxfam to be higher in integrity. Examining these three-way interactions between the actors of the charitable triad will require a new and more complex methodology (see also Subašić et al., 2008; Zagefka, 2018b). We focus on charitable giving narrowly because we see it as distinct from dyadic forms of prosocial behavior, such as sharing, kindness, and interpersonal helping. Yet the charitable triad model may also be generative in understanding other forms of giving, such as time, blood, and body parts. We expect that similar propositions could be relevant for these areas of study, but that the magnitude of effects may differ. For example, volunteering motives suggest that people give time for different reasons than they give money: they are more likely to be motivated by self- enhancement concerns such as feeling important (Clary & Snyder, 1999). Donor characteristics may therefore be especially important in determining volunteer decisions. Blood donation agencies frequently stress the beneficiary when communicating with donors, despite evidence suggesting the agency itself (in its role as blood-raiser) is a key factor in success (Healy, 2000). Donations of gametes (i.e., eggs and sperm) are often motivated by the altruistic disposition of the donor (Svanberg et al., 2012), while donations of organs are usually exchanges between donors and beneficiaries who are very close (usually partners or family members; Hyde & White, 2009). All these findings illustrate how the charitable triad will be broadly relevant to other forms of prosocial giving, yet may operate in ways that require elaboration for each context. Conclusion We reviewed over 300 articles from the psychology and interdisciplinary literature on charitable giving to synthesize isolated bodies of work that focus on donors, beneficiaries, fundraisers, and/or contexts. Based on this critical review we present a new model to conceptualize giving. The charitable triad model argues that a triad of actors determine donation decisions—the donor, the beneficiary, and the fundraiser. Rather than study each in isolation, we must examine how they relate to one another and are shaped by the wider context to influence charitable giving. We put forward five testable propositions for the ways that actors in the charitable triad may be

147 influenced by each other and the wider context to determine charitable outcomes. Future research should endeavor to test these propositions, especially seeking to consider the entire triad at once.

148

Chapter 8. General Discussion

This program of research applies social psychological theory to the study of charitable giving, examining how intergroup dynamics and social contexts affect giving responses and highlighting the triadic nature of charitable giving. Chapters 2 and 3 show that beneficiaries influence both who gives and what motivates donor gifts and also highlight the ways that particular donors are motivated to give to particular beneficiaries because of shared identities or their group’s priorities and values. Chapters 4 and 5 demonstrate how salient characteristics of the social context, such as norms, government policy, and political advocacy, can influence donor responses and change the way donors relate to beneficiaries. Chapter 6 establishes that fundraiser actions influence donor responses in the peer-to-peer context, providing evidence that fundraisers affect giving. Building off the research presented in Chapters 2-6, Chapter 7 presents a novel theoretical model to understand charitable giving: the charitable triad. Drawing on my own research and over 300 articles from the interdisciplinary literature, the charitable triad model provides a framework to make sense of inconsistent findings, highlights gaps in existing knowledge, and proposes that the relationships between particular donors, beneficiaries, and fundraisers—dyadically and triadically, within social contexts—affect charitable giving. Taken together, the five surveys, three experiments, thematic analysis, and conceptual review presented in this thesis show that charitable giving is triadic, relational, and contextualized. That is, charitable decisions are informed by a triad of actors (donor, beneficiary, and fundraiser), the relationships between them, and salient characteristics of the wider social context. Donors: Understanding Who Gives and Why Chapter 2 demonstrates that four characteristics typically associated with giving—being a woman, older, religious, and politically liberal—are not universally related to giving. Instead, and consistent with the social identity perspective (Tajfel, 1981; Turner et al., 1987), they are associated only with giving to categories of charity that reflect the priorities and concerns of the particular social category or identity. These findings align with previous work showing how different types of donors give to different causes (e.g., Neumayr & Handy, 2017; Srnka et al., 2003; Wiepking, 2010) and also extend that work by theorizing and demonstrating how these associations relate to donors’ social identities. Although research has demonstrated that people are more willing to help those who share their national or regional identity (e.g., Charnysh et al., 2015; M. Levine & Thompson, 2004; Zagefka et al., 2013, Study 3), many other types of social identities have yet to be considered in reference to charitable giving. The results reported in Chapter 2 confirm that identities based on gender, age, religion, and political orientation also structure giving in meaningful ways. Chapter 3

149 further identifies family identities and identities born of suffering as relevant to giving. Consistent with Social Identity Theory (Tajfel, 1981), donors appear to support charities which reflect the priorities and values of their social groups. These results highlight how understanding donor psychology is assisted by also considering the beneficiary. Beneficiaries: Understanding Who Is Most Likely to Receive and Why Certain types of beneficiaries are more likely to receive help than others (e.g., Cuddy et al., 2007; Small & Loewenstein, 2003; Zagefka & James, 2015). Data presented in Chapter 3 provide further evidence that beneficiaries who are valued and perceived to be needy and powerless are more likely to be helped. These results also extend previous research by showing how different kinds of beneficiaries elicit support from different kinds of donors (see also Chapter 2) and, in particular, elicit support for different reasons. Chapter 3 puts forward a self/other dichotomy in donor motivation: support of certain categories of charity (representing certain kinds of beneficiary) is more likely to be motivated by self-oriented concerns (such as social identities, benefits, and values) while support of charities in other categories is more likely to be motivated by other- oriented concerns (such as neediness, powerlessness, and perceived value). A binary distinction between egoistic (self-interested) and altruistic (other-interested) motives has long been discussed in relation to helping behavior (see, for example, Batson, 1991; Maner & Gailliot, 2007; Schroeder, Dovidio, Sibicky, Matthews, & Allen, 1988), and the self/other dichotomy proposed here clearly relates to those concepts. Chapter 3 reports that support of religious and research organizations is more frequently self-oriented, while support of animal, international, and welfare organizations is more frequently other-oriented. Although previous work suggests that donors can either be motivated by concerns about the self or other, results presented here show that self and other motives sometimes overlap, particularly when donors give because of shared identities (see also M. Levine et al., 2005; M. Levine & Thompson, 2004; Platow et al., 1999). These results progress theory about donor motivation by showing which types of giving are most motivated by self vs. other concerns. They also further elaborate the processes by which social identities promote giving, especially demonstrating the importance of overlap between self and other as a motivating force in charitable behavior. Chapter 3 also discusses the way that power dynamics between donors and beneficiaries (in terms of relative status) inform what stereotypical beneficiaries look like and who is worthy of care (see also M. A. Brown, 2011; Fisher et al., 1978; Halabi et al., 2011; Nadler et al., 2009). Consistent with the Intergroup Helping Relations as Status Relations model (Nadler, 2002), people generally highlighted outgroup beneficiaries that were needy and vulnerable. These findings extend evidence for Nadler’s model by moving beyond the traditional consideration of helping behavior in

150 two-group contexts to show that status differences also inform charitable giving among a diverse sample of international donors self-defining their preferred beneficiaries. Beyond stereotypes and power, we also theorize that different donor groups may consider different beneficiary groups to be normative targets of giving (see also M. Levine & Manning, 2013). If true, this would mean the motive to give could critically depend both on which donor identity were salient at the time and on whether or not the beneficiary group in question were considered to be normative targets of charity for that particular group. These ideas remain to be tested. Social Contexts Influence Giving Chapters 2 and 3 examine the donor-beneficiary dyad, or how the relationships between particular donors and particular beneficiaries inform giving. The research presented in Chapters 4 and 5 then go on to test the influence that information in the social context has on charitable giving. It has been demonstrated before that making inclusive identities salient can influence donors’ willingness to help beneficiaries when they are reframed as being ingroup members (Charnysh et al., 2015; M. Levine & Thompson, 2004). We extend knowledge of how salient identities influence giving by focusing on the role of advocacy (from opinion-group members and national representatives) and group norms. Norms have previously been demonstrated to affect charitable giving (Agerström et al., 2016; Bartke et al., 2017; Croson et al., 2009; J. R. Smith & McSweeney, 2007). Results presented in Chapter 4 show how national giving norms, particularly those communicating what other citizens do (the descriptive norm), impact private decisions about donations. However, these results also go further by showing how salient information in the social context—in this case, information about the government’s policy on foreign aid spending—can influence responses by affecting perceptions of these norms. Although previous research has demonstrated that government spending can impact private giving (see De Wit & Bekkers, 2017 for a meta-analysis), the mechanisms through which it does so have rarely been studied and norms have not previously been tested as an explanation. While this research demonstrates a highly interesting finding if reliable, the consistency of the effects is unclear at this stage. Following reviewers’ advice after submission of the manuscript in Chapter 4, a third study (not presented in this thesis) was conducted aiming to replicate and extend the research. However, the manipulations in the third study unexpectedly had no effect. This null finding may have been caused by a controversy that erupted over the Trump administration’s proposed foreign aid budget the day before data collection (Tillerson, 2018), potentially overwhelming the attempt to manipulate perceptions of bipartisan support for higher or lower aid. Further research, which will either consider the role of partisanship or seek a less

151 partisan context in which to test the model, will be needed to give confidence in the norm divergence model before publication. Chapter 5 also highlights how salient social information affects private charitable responses. In particular, results from two experimental studies show that political advocacy via social media can affect both people’s willingness to give (i.e., to become a donor) and their selection of recipient charities (charity preferences). Consistent with the Social Identity Model of Collective Action (van Zomeren et al., 2008), these studies show that anger, efficacy, and identification all motivate collective giving. In these studies, supporting or opposing advocacy affected giving responses by influencing the potential donor’s emotion (anger), perceived efficacy on the issue, and identification with beneficiary groups highlighted as victims in the advocacy. These findings test empirically the phenomenon of rage donations—an emerging form of collective action where people are motivated to give in response to advocacy that angers them. The results show that charitable giving can—in some contexts—be a form of ally collective action (see also Louis et al., under review), and that political advocacy can affect charitable giving intentions. When advocacy is effective at changing the donor’s identification with beneficiaries, it also affects donation behavior. Fundraisers: Understanding Who Is Most Effective at Asking for Donations and Why Previous research has focused mostly on donors (e.g., Bekkers & Wiepking, 2011c; Casale & Baumann, 2015; Konrath & Handy, 2017; Leslie et al., 2013) and occasionally on beneficiaries (e.g., Fong & Luttmer, 2011; Halabi et al., 2011; Seyoung Lee & Feeley, 2016; Lesner & Rasmussen, 2014; Nadler & Halabi, 2006; Zagefka et al., 2012). Research presented in this thesis, too, has considered charitable giving as dyadic—examining relationships between donors and beneficiaries. Yet charitable giving usually occurs in response to an ask (Bryant et al., 2003), highlighting that fundraisers can affect the decisions of potential donors. Indeed, scholars have called for a stronger research focus on fundraisers (Breeze, 2017a). In Chapter 6, I presented data from two studies of peer-to-peer fundraisers that consider how the fundraisers’ identification with the charity they nominate and the actions they take affect their success. Fundraisers’ identification with the cause predicts their success—fundraisers who are more identified raise more money. The effect of identification has not been systematically studied in relation to fundraisers. However, as previously discussed, donor’s identification with beneficiaries motivates giving (M. Levine & Thompson, 2004; Zagefka et al., 2013) and Social Identity Theory (Tajfel, 1981) anticipates that people will be more energized to act for targets that they identify with. Indeed, the actions fundraisers took fully explained the effect of their identification on successfully eliciting donations. That is, fundraisers who identified more with the charity in question engaged in more actions to raise money which, in turn, resulted in greater

152 fundraising success. Finally, the classes of actions that made the fundraisers themselves more salient—asking through more channels, personalizing their fundraising page, and sharing their personal motives—explained considerably more variance in fundraising outcomes than did actions that made the beneficiary charity salient. These results highlight the importance of the fundraiser and suggest, though rarely considered, that relationships between donors and fundraisers can strongly influence giving responses (see also Meer, 2011; Platow et al., 1999; Scharf & Smith, 2016). The Charitable Triad Above I have shown that donors, beneficiaries, and fundraisers all matter—each actor informs giving individually and in combination. Fundraisers, in particular, are rarely considered in charitable giving research (Breeze, 2017a), despite their role being pivotal and unique to charitable giving as a form of prosocial behavior. While interpersonal sharing, helping, and acts of kindness are typically dyadic, financial donations are commonly elicited by third parties. In some contexts there may even be stigmas or barriers that prevent beneficiaries from directly requesting donations. The trajectory of research and theorizing presented in this thesis has culminated in the generation of a new theoretical model of charitable giving: the charitable triad. Presented in Chapter 7, the charitable triad model proposes that charitable giving is triadic, relational, and contextualized. That is, donor decisions are informed by a triad of actors (donors, beneficiaries, and fundraisers), the dyadic and triadic relationships between them, and salient cues from the social context. In Chapter 7, I organize the existing interdisciplinary literature in relation to the charitable triad model and generate novel propositions to guide future research on charitable giving from the triadic perspective. These propositions include that fundraisers who are aligned with either donors or beneficiaries will be more successful and that the unique combination of actors in the charitable triad will determine both the likelihood that a gift will be made and the value of that gift. Many of the ideas presented in Chapter 7 remain to be tested. My own future research and, hopefully, that of other scholars can generate new evidence for (or against) the charitable triad model. A Word on Measurement Research presented in this thesis includes both subjective (see Chapters 2, 3, 4, and 5) and objective measures of charitable giving (see Chapters 4, 5, and 6). Prior research on charitable giving has often relied on self-reported measures, such as extent of giving in the past year (see, for example, De Wit & Bekkers, 2016; Helms & Thornton, 2012; Rajan et al., 2009), future giving intentions (e.g., Bennett, 2013; J. R. Smith & McSweeney, 2007), willingness to give (e.g., Zagefka, 2018a; Zagefka et al., 2013), and likelihood of giving (e.g., M. Levine & Thompson, 2004). Although work comparing self-reported past giving with objective measures has found a positive correlation between the two, direct comparisons show that over half of people over-report

153 their giving when asked to recall it (Bekkers & Wiepking, 2011a; Z. Lee & Sargeant, 2011). For this reason, in addition to traditional subjective measures such as self-reported charity preferences (Chapters 2 and 3), likelihood to donate (Chapter 4), and willingness to donate (Chapters 4 and 5), behavioral measures have been included throughout this thesis. I designed a behavioral measure for use in online experiments (see Study 4.2), which allows donors to spend time earning a potential donation for charity. Several other researchers have already employed the measure with success (e.g., Gulliver, Schultz, Solly, & Chapman, in prep; Oelrichs, 2018). In the two studies reported in Chapter 5, I also gave participants a 25c donation which they could allocate between a selection of charities. Finally, in both samples reported in Chapter 6 I used transactional data scraped from a fundraising database to assess actual behavioral outcomes recorded in the fundraising platform. It is a strength of the current research that a range of behavioral measures were employed. However, it is also notable that in the studies that included both subjective and objective measures (Studies 4.2, 5.1, and 5.2) different patterns of explanation for giving responses were observed across the two measures (see also Gulliver et al., in prep; Nilsson et al., 2016). Such results suggest that attitudes, intentions, and behavior are not always aligned. It will be important to include behavioral outcomes in future charity research alongside traditional self-report measures, to examine the congruence or incongruence of the measures and to theorize the patterns that are found. Implications for Practice The research presented in this thesis has a number of important implications for the way charities solicit funds for the causes they represent. Although fundraising practice often focuses on identifying who is more likely to donate in order to target campaigns most efficiently (e.g., Schlegelmilch et al., 1997; Srnka et al., 2003), my own and others’ research shows that who gives will depend critically on who receives. Charities must therefore consider the beneficiaries they represent when assessing who is most likely to give. Results presented in Chapters 2 and 3 in particular highlight the way different donors are willing to support different beneficiaries in ways that reflect their groups and priorities. Next, results show that social contexts influence donors and their relationships with certain beneficiaries. Charities must consider salient aspects of the social context and how such characteristics may affect donors. For example, charities may wish to highlight supportive norms, cultivate audiences, highlight issues or particular beneficiaries, or draw attention to supportive advocacy or government policy. Finally, results highlight how important fundraisers (whether individuals or organizations) are in the giving process. Not much is yet known about what makes fundraisers effective. However, when designing fundraising strategy, charities must consider the characteristics of both their ambassadors and their brand, and how they relate to both donors and beneficiaries. Chapters 6 and 7 provide theory and evidence to endorse this view of fundraisers as critical actors in determining giving outcomes.

154

Strengths, Limitations, and Future Directions This thesis comprises a mixed-methods approach (surveys, experiments, qualitative analyses, and literature review) and uses diverse analytic techniques to understand charitable giving as a social process. Sampling 5,466 people from 117 countries—including student, community, and donor samples—this program of research represents a broad range of experiences and perspectives. Further, throughout this thesis, efforts have been made wherever possible to assess both subjective (self-report) and objective (behavioral) aspects of charitable giving. However, a number of limitations should be acknowledged. The data presented in Chapters 2 and 6 were obtained from practitioner collaborators who designed the questionnaires and collected data prior to my involvement. Although it is a strength to use existing data, especially that which captures behavioral outcomes, analyzing data that was not originally designed to answer the research questions presents some challenges. As disclosed in the chapters, these data sets also contained a range of other measures that were not considered or analyzed for the current program of research but are available upon request. In light of growing awareness of the problems of selective reporting and the current move toward more transparent and open research practices (Hesse, 2018; J. P. Simmons et al., 2012), later studies (those conducted since late 2017) were preregistered on the open science framework (https://osf.io). These include both studies reported in Chapter 5 and the additional study that was recently run to supplement those reported in Chapter 4. Materials and data will both be made publicly available when they are published. Moving forward, my research will be conducted in line with open science principles. In Chapters 3, 4, and 5, I sampled from students and Mechanical-Turk workers for convenience. However, other studies recruited from paid panels (Chapter 2) and confirmed donors or fundraisers (Chapter 2 and 6), which affords more confidence that responses represent those of the wider community. Nonetheless, data collection (apart from Chapter 3) were exclusively from Australian and American participants. Given that some cultures exhibit different patterns of giving (Charities Aid Foundation, 2017b), it will be important to also confirm findings of the current program of research in different cultural contexts. Further, this research suffers from the same flaw as much of psychology—a predominance of WEIRD samples (i.e., participants from western, educated, industrialized, rich, and democratic countries; Henrich et al., 2010). Although Chapter 3 analyzed responses from donors from 117 different countries, future work should endeavor to recruit more diverse participants. Research included in this thesis examined the charitable responses of individual donors— people who give a little to charity each year. Such donors contribute 80% of all charitable gifts (Giving USA, 2018b), and are therefore important to understand. Nonetheless, it is likely that the psychological processes at play in large-scale philanthropy (people making very large gifts),

155 bequest giving (people leaving gifts to charities in their Wills), and foundation and corporate giving (organizations making gifts) may be different. Results presented in this thesis may not generalize to those unique forms of giving and future research should assess how the charitable triad applies to corporate, legacy, and major giving. For example, bequest donors may weight trustworthiness of the fundraising organization as especially important because they will not be around to oversee the disbursement of funds. Major donors, on the other hand, have more one-on-one contact with individual fundraisers and may therefore be more strongly influenced by characteristics of those fundraisers or their relationships with them. Finally, I have argued throughout this thesis that charitable giving is a distinct form of prosocial behavior. In particular, I distinguish giving from other, dyadic forms of prosociality (e.g., helping, sharing, kindness). Further, I propose that psychological processes in financial giving are different from those of giving time, objects, blood, or body parts (see also Chell, 2016; Clary & Snyder, 1999; Clary et al., 1998; Konrath & Handy, 2017; L. Lee & Piliavin, 1999). However, many scholars treat helping and prosocial behavior as umbrella terms and study any form—e.g., volunteering, donations, interpersonal helping—as examples (e.g., Penner et al., 2005; Stürmer et al., 2005). Whether or not charitable giving is distinct from other forms of prosocial action is ultimately an empirical question. I therefore urge scholars to apply the charitable triad model to other forms of prosocial behavior to investigate how the processes proposed in this thesis may (or may not) apply to other forms of helping and prosociality. For example, volunteers are more likely to be motivated by self-enhancement concerns than financial donors (Clary & Snyder, 1999). Donor characteristics may therefore be especially important in determining volunteer decisions. Also, blood donors have no chance to elect the beneficiary of their gifts. Beneficiary characteristics may therefore not be relevant in understanding blood donation responses. The key contribution of this thesis is to theorize the triadic nature of giving. As the culmination of a multi-year program of research, I have not yet empirically tested the model presented in Chapter 7. The empirical studies presented in Chapter 2-6 and over 300 articles support the notion that charitable giving is triadic, relational, and contextualized. Yet this key proposition remains to be tested. Triadic relations have previously been theorized as relevant to intergroup processes and Zagefka (2018b) recently proposed a methodology for testing triadic relations in social psychological research. Future research should apply such methods to test the charitable triad using experimental techniques that systematically vary the relationships between donors, beneficiaries, and fundraisers to examine the potential three-way interaction on giving. In my view, the first place to start is in terms of congruent vs. incongruent identities between the actors. Investigating these relations will be the primary focus of my continued research.

156

Conclusion This thesis presents evidence from 5 surveys, 3 experiments, a thematic analysis, and a conceptual review that demonstrates how charitable giving is triadic, relational, and contextualized. That is, charitable responses are determined by a triad of actors (donors, beneficiaries, and fundraisers), the relationships between them, and salient features of the social context. In this thesis, I bring diverse interdisciplinary streams of research together and shed new light on the psychology of charitable giving through the application of social psychological theories—especially Social Identity Theory (Tajfel, 1981; Turner et al., 1987) and the Intergroup Helping Relations as Status Relations model (Nadler, 2002). Although previous research has predominantly asked who gives and why (focusing largely on donors), results of the empirical research presented in this thesis highlight that beneficiaries, fundraisers, and social contexts also affect giving. Identities are demonstrated to be particularly relevant to charitable giving, with shared identities between actors in the triad being especially motivating for donors. In addition to the empirical evidence, this thesis presents a new theoretical model of charitable giving—the charitable triad—and puts forward novel propositions for how the actors relate to one another and inform giving. The charitable triad model sets a new agenda for both future research on charitable giving and fundraising practice.

157

References

Aaker, J. L., & Akutsu, S. (2009). Why do people give? The role of identity in giving. Journal of Consumer Psychology, 19(3), 267-270. doi:10.1016/j.jcps.2009.05.010 ABS, A. B. o. S. (2017). 2016 Census data reveals “no religion” is rising fast. Retrieved from http://www.abs.gov.au/AUSSTATS/[email protected]/mediareleasesbyReleaseDate/7E65A1445405 51D7CA258148000E2B85?OpenDocument ACNC. (2018). Australian Charities and Not-for-profts Commission website. Retrieved from http://www.acnc.gov.au Agerström, J., Carlsson, R., Nicklasson, L., & Guntell, L. (2016). Using descriptive social norms to increase charitable giving: The power of local norms. Journal of Economic Psychology, 52, 147-153. doi:10.1016/j.joep.2015.12.007 Aknin, L. B., Barrington-Leigh, C. P., Dunn, E. W., Helliwell, J. F., Burns, J., Biswas-Diener, R., . . . Norton, M. I. (2013). Prosocial spending and well-being: Cross-cultural evidence for a psychological universal. Journal of Personality and Social Psychology, 104(4), 635-652. doi:10.1037/a0031578 Alborough, L. (2017). Lost in translation: A sociological study of the role of fundraisers in mediating gift giving in non-profit organisations. International Journal of Nonprofit and Voluntary Sector Marketing, 22(4), e1602. doi:10.1002/nvsm.1602 Alhidari, I. S., Veludo-de-Oliveira, T. M., Yousafzai, S. Y., & Yani-de-Soriano, M. (2018). Modeling the effect of multidimensional trust on individual monetary donations to charitable organizations. Nonprofit and Voluntary Sector Quarterly. doi:10.1177/0899764017753559 Allman, D., Calzavara, L., Worthington, C., Tyndall, M., Adrien, A., Walters, M., . . . Jones, M. K. (2014). Charitable Giving for HIV and AIDS: Results from a Canadian National Survey. PLoS ONE, 9(8), 1-10. doi:10.1371/journal.pone.0103184 Alpízar, F., & Martinsson, P. (2012). Paying the price of sweetening your donation: evidence from a natural field experiment. Economics Letters, 114(2), 182-185. doi:10.1016/j.econlet.2011.10.008 Alpízar, F., & Martinsson, P. (2013). Does it matter if you are observed by others? Evidence from donations in the field. The Scandinavian journal of economics, 115(1), 74-83. doi:10.1111/j.1467-9442.2012.01744.x Altemeyer, B. (1981). Right-Wing Authoritarianism. Winnipeg, Canada: University of Manitoba Press.

158

Altemeyer, B. (2004). Highly Dominating, Highly Authoritarian Personalities. The Journal of social psychology, 144(4), 421-448. doi:10.3200/SOCP.144.4.421-448 Andreasen, A. R., & Kotler, P. (2008). Strategic marketing for nonprofit organizations (7th ed.). Upper Saddle River, NJ: Pearson Education. Andreoni, J. (1990). Impure altruism and donations to public goods: A theory of warm-glow giving. The Economic Journal, 100(401), 464-477. doi:10.2307/2234133 Andreoni, J., Brown, E., & Rischall, I. (2003). Charitable Giving by Married Couples: Who Decides and Why Does It Matter? The Journal of Human Resources, 38(1), 111-133. doi:10.2307/1558758 Andreoni, J., & Payne, A. A. (2011). Is crowding out due entirely to fundraising? Evidence from a panel of charities. Journal of Public Economics, 95(5), 334-343. doi:10.1016/j.jpubeco.2010.11.011 Andreoni, J., & Payne, A. A. (2013). Charitable giving. Handbook of Public Economics, 5, 1-50. doi:10.1016/B978-0-444-53759-1.00001-7 Andreoni, J., Rao, J. M., & Trachtman, H. (2017). Avoiding the Ask: A Field Experiment on Altruism, Empathy, and Charitable Giving. Journal of Political Economy, 125(3), 625-653. doi:10.1086/691703 Andrighetto, L., Baldissarri, C., Lattanzio, S., Loughnan, S., & Volpato, C. (2014). Human‐itarian aid? Two forms of dehumanization and willingness to help after natural disasters. British Journal of Social Psychology, 53(3), 573-584. doi:10.1111/bjso.12066 Anonymous. (2016). The secret life of a chugger: most of us are motivated by money, not charity. The Guardian. Retrieved from https://www.theguardian.com/commentisfree/2016/apr/04/secret-life-chugger-charity- fundraiser Aquino, K., & Reed, A. (2002). The self- importance of moral identity. Journal of Personality and Social Psychology, 83(6), 1423-1440. doi:10.1037//0022-3514.83.6.1423 Asendorpf, J. B., Conner, M., de Fruyt, F., De Houwer, J., Denissen, J. J. A., Fiedler, K., . . . Wicherts, J. M. (2013). Recommendations for increasing replicability in psychology. European Journal of Personality, 27(2). p.108-119(2), 108-119. Barasch, A., Berman, J. Z., & Small, D. A. (2016). When payment undermines the pitch: On the persuasiveness of pure motives in fundraising. Psychological Science, 27(10), 1388-1397. doi:10.1177/0956797616638841 Baron-Cohen, S. (2011). Zero Degrees of Empathy: A New Theory of Human Cruelty and Kindness. London: Penguin Books.

159

Bartke, S., Friedl, A., Gelhaar, F., & Reh, L. (2017). Social comparison nudges—Guessing the norm increases charitable giving. Economics Letters, 152, 73-75. doi:10.1016/j.econlet.2016.12.023 Basil, D. Z., Ridgway, N. M., & Basil, M. D. (2008). Guilt and giving: A process model of empathy and efficacy. Psychology & marketing, 25(1), 1-23. doi:doi:10.1002/mar.20200 Batson, C. D. (1991). The Altruism Question: Toward a Social-psychological Answer. Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Batson, C. D., Chang, J., Orr, R., & Rowland, J. (2002). Empathy, attitudes, and action: Can feeling for a member of a stigmatized group motivate one to help the group? Personality & Social Psychology Bulletin, 28(12), 1656-1666. doi:10.1177/014616702237647 Baumeister, R. F., Chesner, S. P., Senders, P. S., & Tice, D. M. (1988). Who's in charge here? Group leaders do lend help in emergencies. Personality and Social Psychology Bulletin, 14(1), 17-22. doi:10.1177/0146167288141002 BBC. (2017). 'Dress day' donations pour in after politician's criticism. BBC News. Retrieved from https://www.bbc.com/news/world-australia-41356928 Bègue, L. (2014). Do just‐world believers practice private charity? Journal of Applied Social Psychology, 44(1), 71-76. doi:10.1111/jasp.12201 Bekkers, R. (2003). Trust, Accreditation, and Philanthropy in the Netherlands. Nonprofit and Voluntary Sector Quarterly, 32, 596-615. doi:10.1177/0899764003258102 Bekkers, R., & Ottoni-Wilhelm, M. (2016). Principle of Care and Giving to Help People in Need. European Journal of Personality, 30(3), 240-257. doi:10.1002/per.2057 Bekkers, R., & Wiepking, P. (2011a). Accuracy of self-reports on donations to charitable organizations. Quality and Quantity, 45(6), 1369. Bekkers, R., & Wiepking, P. (2011b). A literature review of empirical studies of philanthropy: Eight mechanisms that drive charitable giving. Nonprofit and Voluntary Sector Quarterly, 40(5), 924-973. doi:10.1177/0899764010380927 Bekkers, R., & Wiepking, P. (2011c). Who gives? A literature review of predictors of charitable giving Part One: Religion, education, age and socialisation. Voluntary Sector Review, 2(3), 337-265. Beldad, A., Snip, B., & van Hoof, J. (2014). Generosity the Second Time Around. Nonprofit and Voluntary Sector Quarterly, 43(1), 144-163. doi:10.1177/0899764012457466 Bénabou, R., & Tirole, J. (2006). Incentives and prosocial behavior. American Economic Review, 96(5), 1652-1678. doi:10.1257/aer.96.5.1652

160

Bennett, R. (2003). Factors underlying the inclination to donate to particular types of charity. International Journal of Nonprofit and Voluntary Sector Marketing, 8(1), 12-29. doi:10.1002/nvsm.198 Bennett, R. (2012). Why urban poor donate: a study of low-income charitable giving in London. Nonprofit and Voluntary Sector Quarterly, 41(5), 870-891. doi:10.1177/0899764011419518 Bennett, R. (2013). Elements, causes and effects of donor engagement among supporters of UK charities. International Review on Public and Nonprofit Marketing, 10(3), 201-220. doi:10.1007/s12208-013-0100-1 Bennett, R., & Kottasz, R. (2000). Emergency fund‐raising for disaster relief. Disaster Prevention and Management: An International Journal, 9(5), 352-360. doi:doi:10.1108/09653560010361393 Bereczkei, T., Birkas, B., & Kerekes, Z. (2007). Public charity offer as a proximate factor of evolved reputation-building strategy: an experimental analysis of a real-life situation. Evolution and Human Behavior, 28(4), 277-284. Bereczkei, T., Birkas, B., & Kerekes, Z. (2010). The presence of others, prosocial traits, machiavellianism: A personality × situation approach. SOCIAL PSYCHOLOGY, 41(4), 238- 245. doi:10.1027/1864-9335/a000032 Berger, J., & Heath, C. (2007). Where consumers diverge from others: Identity signaling and product domains. Journal of Consumer Research, 34(2), 121-134. doi:10.1086/519142 Berman, J. Z., Barasch, A., Levine, E. E., & Small, D. A. (2018). Impediments to effective altruism: The role of subjective preferences in charitable giving. Psychological Science, 29(5), 834- 844. doi:10.1177/0956797617747648 Blinded. (2018). This citation has been blinded for review. Bliuc, A.-M., McGarty, C., Reynolds, K., & Muntele, D. (2007). Opinion-based group membership as a predictor of commitment to political action. European Journal of Social Psychology, 37(1), 19-32. doi:10.1002/ejsp.334 Blogowska, J., & Saroglou, V. (2011). Religious Fundamentalism and Limited Prosociality as a Function of the Target. Journal for the Scientific Study of Religion, 50(1), 44-60. doi:10.1111/j.1468-5906.2010.01551.x Body, A., & Breeze, B. (2016). What are ‘unpopular causes’ and how can they achieve fundraising success? International Journal of Nonprofit and Voluntary Sector Marketing, 21(1), 57-70. doi:10.1002/nvsm.1547 Böhm, R., & Regner, T. (2013). Charitable giving among females and males: an empirical test of the competitive altruism hypothesis. Journal of Bioeconomics, 15(3), 251-267. doi:10.1007/s10818-013-9152-x

161

Bolitho, F. H., Carr, S. C., & Fletcher, R. B. (2006). Public thinking about poverty: why it matters and how to measure it. International Journal of Nonprofit and Voluntary Sector Marketing, 12(1), 13-22. doi:10.1002/nvsm.220 Bourassa, M. A., & Stang, A. C. (2016). Knowledge is power: why public knowledge matters to charities. International Journal of Nonprofit and Voluntary Sector Marketing, 21(1), 13-30. doi:10.1002/nvsm.1537 Braiterman, A., & Masterson, S. (2015). The 2014 Blackbaud peer-to-peer fundraising study. Retrieved from http://charity.everydayhero.com.au/wp-content/uploads/2015/07/2014-peer- to-peer-fundraising-study-1.pdf Brandt, M. J., Ijzerman, H., Dijksterhuis, A., Farach, F. J., Geller, J., Giner-Sorolla, R., . . . van 'T Veer, A. (2014). The replication recipe: What makes for a convincing replication? Journal of Experimental Social Psychology, 50, 217-224. doi:10.1016/j.jesp.2013.10.005 Braun, V., & Clarke, V. (2006). Using thematic analysis in psychology. Qualitative Research in Psychology, 3(2), 77-101. doi:10.1191/1478088706qp063oa Breeze, B. (2006). Robin Hood in reverse: Exporing the relationship between income and charitable giving. Retrieved from London: http://eprints.lse.ac.uk/29223/1/VSWP3_Breeze.pdf Breeze, B. (2013). How donors choose charities: The role of personal taste and experiences in giving decisions. Voluntary Sector Review, 4(2), 165-183. doi:10.1332/204080513X667792 Breeze, B. (2017a). IJNVSM special edition: Understanding fundraising. International Journal of Nonprofit and Voluntary Sector Marketing, 22(4), e1608. doi:10.1002/nvsm.1608 Breeze, B. (2017b). The New Fundraisers: Who Organises Charitable Giving in Contemporary Society? Bristol: Policy Press. Breeze, B., & Jollymore, G. (2017). Understanding solicitation: Beyond the binary variable of being asked or not being asked. International Journal of Nonprofit and Voluntary Sector Marketing, 22(4), n/a-n/a. doi:10.1002/nvsm.1607 Brewer, M. B. (1999). The psychology of prejudice: Ingroup love and outgroup hate? Journal of Social Issues, 55(3), 429-444. doi:10.1111/0022-4537.00126 Brooks, A. C. (2003a). Do government subsidies to nonprofits crowd out donations or donors? Public Finance Review, 31(2), 166-179. Brooks, A. C. (2003b). Religious faith and charitable giving. Policy Review(121), 39-50. Brooks, A. C. (2005). Does social capital make you generous? Social Science Quarterly, 86(1), 1- 15. doi:10.1111/j.0038-4941.2005.00287.x Brown, M. A. (2011). The power of generosity to change views on social power. Journal of Experimental Social Psychology, 47(6), 1285-1290. doi:10.1016/j.jesp.2011.05.021

162

Brown, P. H., & Minty, J. H. (2008). Media coverage and charitable giving after the 2004 tsunami. Southern economic journal, 75(1), 9-25. doi:10.2307/20112025 Bryant, W., Jeon-Slaughter, H., Kang, H., & Tax, A. (2003). Participation in philanthropic activities: Donating money and time. Journal of Consumer Policy, 26(1), 43-47. Bull, R., & Gibson-Robinson, E. (1981). The Influences of Eye-Gaze, Style of Dress, and Locality on the Amounts of Money Donated to a Charity. Human Relations, 34(10), 895-905. doi:10.1177/001872678103401005 Burgoyne, C. B., Young, B., & Walker, C. M. (2005). Deciding to give to charity: a focus group study in the context of the household economy. Journal of Community & Applied Social Psychology, 15(5), 383-405. doi:10.1002/casp.832 Bushouse, B. K., & Sowa, J. E. (2012). Producing knowledge for practice: Assessing NVSQ 2000- 2010. Nonprofit and Voluntary Sector Quarterly, 41(3), 477-513. doi:10.1177/0899764011422116 Campbell, D., Carr, S. C., & Maclachlan, M. (2006). Attributing “third world poverty” in Australia and Malawi: A case of donor bias? Journal of Applied Social Psychology, 31(2), 409-430. doi:10.1111/j.1559-1816.2001.tb00203.x Carr, S. C., & Maclachlan, M. (1998). Actors, Observers, and Attributions for Third World Poverty: Contrasting Perspectives from Malawi and Australia. The Journal of social psychology, 138(2), 189-202. doi:10.1080/00224549809600370 Casale, D., & Baumann, A. (2015). Who gives to international causes? A sociodemographic analysis of U.S. donors. Nonprofit and Voluntary Sector Quarterly, 44(1), 98-122. doi:10.1177/0899764013507141 Chapman, C. M., Louis, W. R., & Masser, B. M. (2018). Identifying (our) donors: Toward a social psychological understanding of charity selection in Australia. Psychology & marketing, 35(12), 980-989. doi:https://doi.org/10.1002/mar.21150 Chapman, C. M., Louis, W. R., & Masser, B. M. (in prep). Norm divergence model of prosocial influence: Government aid both crowds in and crowds out citizens’ charitable giving via diverging descriptive and injunctive norms. Manuscript in preparation. Chapman, C. M., Masser, B. M., & Louis, W. R. (2018). The champion effect in peer-to-peer giving: Successful campaigns highlight fundraisers more than causes. Nonprofit and Voluntary Sector Quarterly, Online First, 1-21. doi:10.1177/0899764018805196 Chapman, C. M., Masser, B. M., & Louis, W. R. (under review). Self and other orientations in charitable giving: Thematic analyses from a global donor survey. Manuscript under review. Chapman, C. M., Mirnajafi, Z., Louis, W. R., & Masser, B. M. (under review). Rage donations: Examining charitable giving as a form of collective action. Manuscript under review.

163

Charities Aid Foundation. (2017a). CAF UK Giving 2017: An overview of charitable giving in the UK. Retrieved from https://www.cafonline.org/docs/default-source/about-us- publications/caf-uk-giving-web.pdf Charities Aid Foundation. (2017b). World Giving Index 2017: A Global View of Trends. Retrieved from https://www.cafonline.org/docs/default-source/about-us- publications/cafworldgivingindex2017_2167a_web_210917.pdf?sfvrsn=ed1dac40_10 Charities Aid Foundation. (2018). Levels of trust remain consistent despite recent reports about charities [Press release]. Retrieved from https://www.cafonline.org/about-us/media-office- news/levels-of-trust-remain-consistent-despite-recent-reports-about-charities Charnysh, V., Lucas, C., & Singh, P. (2015). The ties that bind: National identity salience and pro- social behavior toward the ethnic other. Comparative Political Studies, 48(3), 267-300. doi:10.1177/0010414014543103 Chell, K. W. (2016). Giving and sharing: The predictors and outcomes of online donor appreciation. (PhD thesis), Queensland University of Technology. Retrieved from https://eprints.qut.edu.au/98002/1/Kathleen_Chell_Thesis.pdf Choi, N. G., & DiNitto, D. M. (2012). Predictors of time volunteering, religious giving, and secular giving: implications for nonprofit organizations. Journal of Sociology and Social Welfare, 39(2), 93-120. Chopik, W. J., O’Brien, E., & Konrath, S. H. (2017). Differences in Empathic Concern and Perspective Taking Across 63 Countries. Journal of Cross-Cultural Psychology, 48(1), 23- 38. doi:10.1177/0022022116673910 Christensen, G., Steinmetz, A., Alcorn, B., Bennett, A., Woods, D., & Emanuel, E. (2013). The MOOC phenomenon: Who takes massive open online courses and why? Retrieved from SSRN: http://dx.doi.org/10.2139/ssrn.2350964 Cialdini, R. B., Reno, R. R., & Kallgren, C. A. (1990). A focus theory of normative conduct: Recycling the concept of norms to reduce littering in public places. Journal of Personality and Social Psychology, 58(6), 1015-1026. doi:10.1037//0022-3514.58.6.1015 Clary, E. G., & Snyder, M. (1999). The Motivations to Volunteer: Theoretical and Practical Considerations. Current Directions in Psychological Science, 8(5), 156-159. Clary, E. G., Snyder, M., Ridge, R. D., Copeland, J., Stukas, A. A., Haugen, J., & Miene, P. (1998). Understanding and assessing the motivations of volunteers: A functional approach. Journal of Personality and Social Psychology, 74(6), 1516-1530. doi:10.1037/0022-3514.74.6.1516 Classy. (2017). Tips for successful fundraising. Retrieved from https://www.classy.org/fundraising-tips

164

Congress.gov. (2018). H.R.36 - Pain-Capable Unborn Child Protection Act. Retrieved from https://www.congress.gov/bill/115th-congress/house-bill/36 Croson, R., Handy, F., & Shang, J. (2009). Keeping up with the Joneses: The relationship of perceived descriptive social norms, social information, and charitable giving. Nonprofit Management and Leadership, 19(4), 467-489. doi:10.1002/nml.232 Croson, R., Handy, F., & Shang, J. (2010). Gendered giving: The influence of social norms on the donation behavior of men and women. International Journal of Nonprofit and Voluntary Sector Marketing, 15(2), 199-213. doi:10.1002/nvsm.385 Croson, R., & Shang, J. (2008). The impact of downward social information on contribution decisions. Experimental economics, 11(3), 221-233. doi:10.1007/s10683-007-9191-z Cuddy, A. J. C., Fiske, S. T., & Glick, P. (2007). The BIAS map: Behaviors from intergroup affect and stereotypes. Journal of Personality and Social Psychology, 92(4), 631-648. doi:10.1037/0022-3514.92.4.631 Dale, E. J. (2017). Fundraising as women's work? Examining the profession with a gender lens. International Journal of Nonprofit and Voluntary Sector Marketing, 22(4), e1605. doi:10.1002/nvsm.1605 Dallago, F., Cima, R., Roccato, M., Ricolfi, L., & Mirisola, A. (2008). The Correlation Between Right-Wing Authoritarianism and Social Dominance Orientation: The Moderating Effects of Political and Religious Identity. Basic and Applied Social Psychology, 30(4), 362-368. doi:10.1080/01973530802502333 Darley, J. M., & Latane, B. (1968). Bystander intervention in emergencies: Diffusion of responsibility. Journal of Personality and Social Psychology, 8(4), 377-383. doi:10.1037/h0025589 Davis, M. H. (1983). Empathic concern and the Muscular Dystrophy Telethon: Empathy as a multidimensional construct. Personality and Social Psychology Bulletin, 9(2), 223. de Oliveira, A. C. M., Croson, R., & Eckel, C. (2011). The giving type: Identifying donors. Journal of Public Economics, 95(5), 428-435. doi:10.1016/j.jpubeco.2010.11.012 De Wit, A., & Bekkers, R. (2016). Exploring gender differences in charitable giving. Nonprofit and Voluntary Sector Quarterly, 45(4), 741-761. doi:10.1177/0899764015601242 De Wit, A., & Bekkers, R. (2017). Government support and charitable donations: A meta-analysis of the crowding-out hypothesis. Journal of Public Administration Research and Theory, 27(2), 301-319. doi:10.1093/jopart/muw044 De Wit, A., Bekkers, R., & Broese van Groenou, M. (2017). Heterogeneity in crowding-out: When are charitable donations responsive to government support? European Sociological Review, 33(1), 59-71. doi:10.1093/esr/jcw048

165

Dean, J., & Wood, R. (2017). “You can try to press different emotional buttons”: The conflicts and strategies of eliciting emotions for fundraisers. International Journal of Nonprofit and Voluntary Sector Marketing, 22(4), e1603. doi:10.1002/nvsm.1603 Deshpande, A., & Spears, D. (2016). Who Is the Identifiable Victim? Caste and Charitable Giving in Modern India. Economic Development & Cultural Change, 64(2), 299-321. 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. doi:10.1371/journal.pone.0148274 DonorDrive. (2017). The state of peer-to-peer fundraising 2017. Retrieved from http://www.donordrive.com/library/ Dovidio, J. F., Hewstone, M., Glick, P., & Esses, V. M. (2010). The SAGE Handbook of Prejudice, Stereotyping and Discrimination. London: SAGE Publications Ltd. Dovidio, J. F., Piliavin, J. A., Schroeder, D. A., & Penner, L. A. (2006). The Social Psychology of Prosocial Behavior. Mahwah, N.J.: Lawrence Erlbaum Publishers. Duncan, B. (2004). A theory of impact philanthropy. Journal of Public Economics, 88(9), 2159- 2180. Eagle, D., Keister, L. A., & Read, J. n. G. (2017). Household Charitable Giving at the Intersection of Gender, Marital Status, and Religion. Nonprofit and Voluntary Sector Quarterly, 47(1), 185-205. doi:10.1177/0899764017734650 Eagly, A. H., & Crowley, M. (1986). Gender and Helping Behavior: A Meta-Analytic Review of the Social Psychological Literature. Psychological Bulletin, 100(3), 283-308. doi:10.1037/0033-2909.100.3.283 Ebeling, F., Feldhaus, C., & Fendrich, J. (2017). A field experiment on the impact of a prior donor’s social status on subsequent charitable giving. Journal of Economic Psychology, 61, 124-133. doi:10.1016/j.joep.2017.03.010 Editors of GQ. (2017). Colin Kaepernick will not be silenced. GQ. Retrieved from https://www.gq.com/story/colin-kaepernick-will-not-be-silenced effectivealtruism.org. (2018). Effective Altruism website. Retrieved from https://www.effectivealtruism.org Ein-Gar, D., & Levontin, L. (2013). Giving from a distance: Putting the charitable organization at the center of the donation appeal. Journal of Consumer Psychology, 23(2), 197-211. doi:10.1016/j.jcps.2012.09.002 Einolf, C. J. (2018). Parents’ Charitable Giving and Volunteering: Are They Influenced by Their Children’s Ages and Life Transitions? Evidence From a Longitudinal Study in the United

166

States. Nonprofit and Voluntary Sector Quarterly, 47(2), 395-416. doi:10.1177/0899764017737870 Einolf, C. J., & Philbrick, D. (2014). Generous or Greedy Marriage? A Longitudinal Study of Volunteering and Charitable Giving. Journal of Marriage and Family, 76(3), 573-586. doi:10.1111/jomf.12115 Elgot, J., & McVeigh, K. (2018). Oxfam loses 7,000 donors since sexual exploitation scandal. The Guardian. Retrieved from https://www.theguardian.com/world/2018/feb/20/oxfam-boss- mark-goldring-apologises-over-abuse-of-haiti-quake-victims Ellemers, N., Kortekaas, P., & Ouwerkerk, J. W. (1999). Self-categorisation, commitment to the group and group self-esteem as related but distinct aspects of social identity. European Journal of Social Psychology, 29(2-3), 371-389. doi:10.1002/(SICI)1099- 0992(199903/05)29:2/3<371::AID-EJSP932>3.0.CO;2-U Eriksson, K., Strimling, P., & Coultas, J. C. (2015). Bidirectional associations between descriptive and injunctive norms. Organizational Behavior and Human Decision Processes, 129, 59-69. doi:10.1016/j.obhdp.2014.09.011 Erlandsson, A., Björklund, F., & Bäckström, M. (2017). Choice-justifications after allocating resources in helping dilemmas. Judgment and Decision Making, 12(1), 60-80. ESPN. (2018). NFL players who protested during national anthem in Week 16. ESPN.com. Retrieved from http://www.espn.com/blog/nflnation/post/_/id/263318/nfl-players-who- protested-during-national-anthem-in-week-16 Evangelidis, I., & Van den Bergh, B. (2013). The Number of Fatalities Drives Disaster Aid:Increasing Sensitivity to People in Need. Psychological Science, 24(11), 2226-2234. doi:10.1177/0956797613490748 Everatt, D., Habib, A., Maharaj, B., & Nyar, A. (2005). Patterns of Giving in South Africa. Official journal of the International Society for Third-Sector Research, 16(3), 275-291. doi:10.1007/s11266-005-7725-z Evers, A., & Gesthuizen, M. (2011). The impact of generalized and institutional trust on donating to activist, leisure, and interest organizations: individual and contextual effects. International Journal of Nonprofit and Voluntary Sector Marketing, 16(4), 381-392. doi:10.1002/nvsm.434 Fack, G., & Landais, C. (2010). Are Tax Incentives for Charitable Giving Efficient? Evidence from France. American Economic Journal: Economic Policy, 2(2), 117-141. Fehrler, S., & Przepiorka, W. (2013). Charitable giving as a signal of trustworthiness: Disentangling the signaling benefits of altruistic acts. Evolution and Human Behavior, 34(2), 139-145. doi:https://doi.org/10.1016/j.evolhumbehav.2012.11.005

167

Fetters, A. (2017). The rise of the 'rage-donation'. GQ. Retrieved from https://www.gq.com/story/the-rise-of-the-rage-donation Fischer, P., Krueger, J. I., Greitemeyer, T., Vogrincic, C., Kastenmüller, A., Frey, D., . . . Kainbacher, M. (2011). The bystander-effect: A meta-analytic review on bystander intervention in dangerous and non-dangerous emergencies. Psychological Bulletin, 137(4), 517-537. doi:10.1037/a0023304 Fishbein, M., & Ajzen, I. (1974). Attitudes towards objects as predictors of single and multiple behavioral criteria. Psychological Review, 81(1), 59-74. doi:10.1037/h0035872 Fisher, J. D., Harrison, C. L., & Nadler, A. (1978). Exploring the Generalizability of Donor- Recipient Similarity Effects. Personality and Social Psychology Bulletin, 4(4), 627-630. doi:10.1177/014616727800400428 Fisher, J. D., & Nadler, A. (1974). The effect of similarity between donor and recipient on recipient's reactions to aid. Journal of Applied Social Psychology, 4(3), 230-243. doi:10.1111/j.1559-1816.1974.tb02643.x Fisher, J. D., Nadler, A., & Whitcher-Alagna, S. (1982). Recipient reactions to aid. Psychological Bulletin, 91(1), 27-54. doi:10.1037/0033-2909.91.1.27 Fong, C. M., & Luttmer, E. F. P. (2011). Do fairness and race matter in generosity? Evidence from a nationally representative charity experiment. Journal of Public Economics, 95(5), 372- 394. doi:https://doi.org/10.1016/j.jpubeco.2010.07.010 Forbes, K. F., & Zampelli, E. M. (1997). Religious Giving by Individuals: A Cross Denominational Study. American Journal of Economics and Sociology, 56(1), 17-30. doi:10.1111/j.1536- 7150.1997.tb03447.x Forbes, K. F., & Zampelli, E. M. (2013). The impacts of religion, political ideology, and social capital on religious and secular giving: evidence from the 2006 Social Capital Community Survey. Applied Economics, 45(17), 2481-2490. doi:10.1080/00036846.2012.667555 Forsyth, D. R., Zyzniewski, L. E., & Giammanco, C. A. (2002). Responsibility diffusion in cooperative collectives. Personality and Social Psychology Bulletin, 28(1), 54-65. doi:10.1177/0146167202281005 Francis, S. (2017). Top 10 tips to help your peer-to-peer fundraising participants raise more. Retrieved from https://probonoaustralia.com.au/news/2017/01/top-10-tips-help-peer-peer- fundraising-participants-raise/ Freeman, D., Aquino, K., & McFerran, B. (2009). Overcoming beneficiary race as an impediment to charitable donations: Social dominance orientation, the experience of moral elevation, and donation behavior. Personality and Social Psychology Bulletin, 35(1), 72-84. doi:10.1177/0146167208325415

168

Friedline, T., Masa, R., & Chowa, G. (2015). Transforming wealth: Using the inverse hyperbolic sine (IHS) and splines to predict youth's math achievement. Social Science Research, 49, 264., 49, 264-287. 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, 33(43), 17188. Giving Australia. (2016). Individuals Giving: Overview. Retrieved from http://www.philanthropy.org.au/images/site/blog/Giving_Australia_2016_Individuals_Fact_ Sheet.pdf Giving USA. (2018a). Giving USA 2018: Americans Gave $410.02 Billion to Charity in 2017, Crossing the $400 Billion Mark for the First Time [Press release]. Retrieved from https://givingusa.org/giving-usa-2018-americans-gave-410-02-billion-to-charity-in-2017- crossing-the-400-billion-mark-for-the-first-time/ Giving USA. (2018b). See the numbers – Giving USA 2018 Infographic. Retrieved from https://givingusa.org/see-the-numbers-giving-usa-2018-infographic/ Gneezy, U., Keenan, E. A., & Gneezy, A. (2014). Avoiding overhead aversion in charity. Science, 346(6209), 632-635. doi:10.1126/science.1253932 Gulliver, R., Schultz, T., Solly, K., & Chapman, C. M. (in prep). Images in Environmental Campaigns Increase Efficacy but not Anger or Identification. Manuscirpt in preparation. Guo, C., & Saxton, G. D. (2014). Tweeting social change: How social media are changing nonprofit advocacy. Nonprofit and Voluntary Sector Quarterly, 43(1), 57-79. doi:10.1177/0899764012471585 Halabi, S., Dovidio, J. F., & Nadler, A. (2008). When and how do high status group members offer help: Effects of Social Dominance Orientation and status threat. Political Psychology, 29(6), 841-858. doi:10.1111/j.1467-9221.2008.00669.x Halabi, S., Dovidio, J. F., & Nadler, A. (2016). Help that hurts? Perceptions of intergroup assistance. International Journal of Intercultural Relations, 53, 65-71. doi:https://doi.org/10.1016/j.ijintrel.2016.05.007 Halabi, S., & Nadler, A. (2017). The Intergroup Status as Helping Relations model: Giving, seeking and receiving help as tools to maintain or challenge social inequality. In E. van Leeuwen & H. Zagefka (Eds.), Intergroup Helping (pp. 205-221). Cham: Springer International Publishing. Halabi, S., Nadler, A., & Dovidio, J. F. (2011). Reactions to Receiving Assumptive Help: The Moderating Effects of Group Membership and Perceived Need for Help. Journal of Applied Social Psychology, 41(12), 2793-2815. doi:doi:10.1111/j.1559-1816.2011.00859.x

169

Harper, D. J. (1996). Accounting for poverty: From attribution to discourse. Journal of Community & Applied Social Psychology, 6(4), 249-265. doi:10.1002/(SICI)1099- 1298(199610)6:4<249::AID-CASP369>3.0.CO2-U Hasenfeld, Y., & Gidron, B. (2005). Understanding multi-purpose hybrid voluntary organizations: The contributions of theories on civil society, social movements and non-profit organizations. Journal of Civil Society, 1(2), 97-112. doi:10.1080/17448680500337350 Hayes, A. F. (2012). PROCESS: A versatile computational tool for observed variable mediation, mod- eration, and conditional process modeling [White paper]. Retrieved from http://www.afhayes.com/public/process2012.pdf Healy, K. (2000). Embedded Altruism: Blood Collection Regimes and the European Union's Donor Population. American Journal of Sociology, 105(6), 1633-1657. doi:10.1086/210468 Helms, S. E., & Thornton, J. P. (2012). The influence of religiosity on charitable behavior: A COPPS investigation. Journal of Socio-Economics, 41(4), 373-383. doi:10.1016/j.socec.2012.04.003 Henrich, J., Heine, S. J., & Norenzayan, A. (2010). The weirdest people in the world? Behavioral and Brain Sciences, 33(2-3), 61-83. doi:10.1017/S0140525X0999152X Herzer, D., & Nunnenkamp, P. (2013). Private donations, government grants, commercial activities, and fundraising: Cointegration and causality for NGOs in international development cooperation. World Development, 46, 234-251. doi:10.1016/j.worlddev.2013.02.010 Herzog, P. S., & Yang, S. (2018). Social Networks and Charitable Giving: Trusting, Doing, Asking, and Alter Primacy. Nonprofit and Voluntary Sector Quarterly, 47(2), 376-394. doi:10.1177/0899764017746021 Hesse, B. W. (2018). Can psychology walk the walk of open science? American Psychologist, 73(2), 126-137. doi:10.1037/amp0000197 Heutel, G. (2014). Crowding out and crowding in of private donations and government grants. Public Finance Review, 42(2), 143-175. doi:10.1177/1091142112447525 Hibbert, S., & Horne, S. (1996). Giving to charity: questioning the donor decision process. JOURNAL OF CONSUMER MARKETING, 13(2), 4-13. doi:10.1108/07363769610115366 Hibbert, S., Smith, A., Davies, A., & Ireland, F. (2007). Guilt appeals: Persuasion knowledge and charitable giving. Psychology & marketing, 24(8), 723-742. doi:doi:10.1002/mar.20181 Higgins, J. W., & Lauzon, L. (2003). Finding the funds in fun runs: exploring physical activity events as fundraising tools in the nonprofit sector. International Journal of Nonprofit and Voluntary Sector Marketing, 8(4), 363-377. doi:10.1002/nvsm.226

170

Hill, J. P., & Vaidyanathan, B. (2011). Substitution or symbiosis? Assessing the relationship between religious and secular giving. Social Forces, 90(1), 157-180. doi:10.1093/sf/90.1.157 Hopkins, N., Reicher, S., Harrison, K., Cassidy, C., Bull, R., & Levine, M. (2007). Helping to improve the group stereotype: On the strategic dimension of prosocial behavior. Personality and Social Psychology Bulletin, 33(6), 776-788. doi:10.1177/0146167207301023 Hughes, P., & Luksetich, W. (2008). Income volatility and wealth: The effect on charitable giving. Nonprofit and Voluntary Sector Quarterly, 37(2), 264-280. doi:10.1177/0899764007310416 Hughes, P., Luksetich, W., & Rooney, P. (2014). Crowding-out and fundraising efforts. Nonprofit Management and Leadership, 24(4), 445-464. doi:10.1002/nml.21102 Hyde, M., & White, K. (2009). Knowing a donor and identifying as one: Determinants of people's willingness for related and anonymous living donation in Australia. Psychology, Health & Medicine, 14(5), 524-535. doi:10.1080/13548500903012855 Hysenbelli, D., Rubaltelli, E., & Rumiati, R. (2013). Others' opinions count, but not all of them: Anchoring to ingroup versus outgroup members' behavior in charitable giving. Judgment and Decision Making, 8(6), 678-690. Iyer, E. S., Kashyap, R. K., & Diamond, W. D. (2012). Charitable Giving: Even the Willing Need to be Persuaded. Journal of Current Issues & Research in Advertising, 33(1), 115-127. doi:10.1080/10641734.2012.676489 Jacobson, R. P., Jacobson, K. J. L., & Hood, J. N. (2015). Social norm perceptions predict citizenship behaviors. Journal of Managerial Psychology, 30(8), 894-908. doi:10.1108/JMP-12-2013-0408 Jacobson, R. P., Mortensen, C. R., & Cialdini, R. B. (2011). Bodies obliged and unbound: Differentiated response tendencies for injunctive and descriptive social norms. Journal of Personality and Social Psychology, 100(3), 433-448. doi:10.1037/a0021470 Jacobson, R. P., Mortensen, C. R., Jacobson, K. J. L., & Cialdini, R. B. (2015). Self-control moderates the effectiveness of influence attempts highlighting injunctive social norms. Social Psychological and Personality Science, 6(6), 718-726. doi:10.1177/1948550615578463 James, R. N. (2016). Phrasing the Charitable Bequest Inquiry. Official journal of the International Society for Third-Sector Research, 27(2), 998-1011. doi:10.1007/s11266-015-9556-4 James, R. N., & Routley, C. (2016). We the living: the effects of living and deceased donor stories on charitable bequest giving intentions. International Journal of Nonprofit and Voluntary Sector Marketing, 21(2), 109-117. doi:10.1002/nvsm.1549

171

James, R. N., & Sharpe, D. L. (2007). The "sect effect" in charitable giving : distinctive realities of exclusively religious charitable givers. American Journal of Economics and Sociology, 66(4), 697-726. James, T. K., & Zagefka, H. (2017). The effects of group memberships of victims and perpetrators in humanly caused disasters on charitable donations to victims. Journal of Applied Social Psychology, 47(8), 446-458. doi:10.1111/jasp.12452 Jay, E. (2001). The rise—and fall?—of face‐to‐face fundraising in the United Kingdom. New Directions for Philanthropic Fundraising, 2001(33), 83-94. doi:10.1002/pf.3306 Jenni, K. E., & Loewenstein, G. (1997). Explaining the "Identifiable Victim Effect". Journal of Risk and Uncertainty, 14(3), 235-257. Jensen, J. D., King, A. J., & Carcioppolo, N. (2013). Driving toward a goal and the goal-gradient hypothesis: the impact of goal proximity on compliance rate, donation size, and fatigue. Journal of Applied Social Psychology, 43(9), 1881-1895. doi:10.1111/jasp.12152 Johansson-Stenman, O., & Martinsson, P. (2008). Are some lives more valuable?: an ethical preferences approach. Journal of health economics, 27(3), 739-752. doi:10.1016/j.jhealeco.2007.10.001 Joireman, J., & Duell, B. (2007). Self-transcendent values moderate the impact of mortality salience on support for charities. Personality and Individual Differences, 43(4), 779-789. doi:10.1016/j.paid.2007.02.003 Jost, J. T., Glaser, J., Kruglanski, A. W., & Sulloway, F. J. (2003). Political Conservatism as Motivated Social Cognition. Psychological Bulletin, 129(3), 339-375. doi:10.1037/0033- 2909.129.3.339 Kaikati, A., Torelli, C., & Winterich, K. (2014). Conforming Conservatives: How Norms of Salient Social Identities Overcome 'Heartless Conservative' Tendencies. Advances in Consumer Research, 42, 59. Kallgren, C. A., Reno, R. R., & Cialdini, R. B. (2000). A focus theory of normative conduct: When norms do and do not affect behavior. Personality and Social Psychology Bulletin, 26(8), 1002-1012. doi:10.1177/01461672002610009 Kandrack, R., & Lundberg, G. (2014). On the Influence of Emotion on Decision Making: The Case of Charitable Giving. In P. Guo & W. Pedrycz (Eds.), Human-Centric Decision-Making Models for Social Sciences (pp. 57-73). Berlin, Heidelberg: Springer Berlin Heidelberg. Kang, E., & Lakshmanan, A. (2018). Narcissism and self- versus recipient-oriented imagery in charitable giving. Personality and Social Psychology Bulletin, 44(8), 1214-1227. doi:10.1177/0146167218764658

172

Kashif, M., Sarifuddin, S., & Hassan, A. (2015). Charity donation: intentions and behaviour. Marketing Intelligence & Planning, 33(1), 90-102. doi:10.1108/MIP-07-2013-0110 Katz, H. (2018). The Impact of Familiarity and Perceived Trustworthiness and Influence on Donations to Nonprofits: An Unaided Recall Study. Journal of Nonprofit & Public Sector Marketing, 30(2), 187-199. doi:10.1080/10495142.2017.1326874 Kayser, D. N., Greitemeyer, T., Fischer, P., & Frey, D. (2010). Why mood affects help giving, but not moral courage: Comparing two types of prosocial behaviour. European Journal of Social Psychology, 40(7), 1136-1157. doi:10.1002/ejsp.717 Kearns, M., Muldoon, O. T., Msetfi, R. M., & Surgenor, P. W. G. (2017). Darkness into light? Identification with the crowd at a suicide prevention fundraiser promotes well-being amongst participants. European Journal of Social Psychology, 47(7), 878-888. doi:10.1002/ejsp.2304 Keen, R., & Audickas, L. (2017). Charities and the Voluntary Sector: Statistics. Retrieved from researchbriefings.files.parliament.uk/documents/SN05428/SN05428.pdf Kessler, J. B., & Milkman, K. L. (2018). Identity in charitable giving. Management Science, 64(2), 845-859. doi:10.1287/mnsc.2016.2582 Khanna, J., Posnett, J., & Sandler, T. (1995). Charity donations in the UK: New evidence based on panel data. Journal of Public Economics, 56(2), 257-272. doi:10.1016/0047-2727(94)01421- J Khanna, J., & Sandler, T. (2000). Partners in giving: The crowding-in effects of UK government grants. European Economic Review, 44(8), 1543-1556. doi:10.1016/S0014-2921(99)00007- 0 Kim, S.-J., & Kou, X. (2014). Not all empathy is equal: how dispositional empathy affects charitable giving. Journal of Nonprofit and Public Sector Marketing, 26(4), 312-334. doi:10.1080/10495142.2014.965066 Kinsbergen, S., & Tolsma, J. (2013). Explaining monetary donations to international development organisations: A factorial survey approach. Social science research, 42(6), 1571-1586. doi:https://doi.org/10.1016/j.ssresearch.2013.06.011 Klas, A., Zinkiewicz, L., Zhou, J., & Clarke, E. J. R. (2018). “Not All Environmentalists Are Like That … ”: Unpacking the Negative and Positive Beliefs and Perceptions of Environmentalists. Environmental Communication, 1-15. doi:10.1080/17524032.2018.1488755 Kogut, T., & Ritov, I. (2005). The “identified victim” effect: an identified group, or just a single individual? Journal of Behavioral Decision Making, 18(3), 157-167. doi:doi:10.1002/bdm.492

173

Konrath, S., & Handy, F. (2017). The development and validation of the motives to donate scale. Nonprofit and Voluntary Sector Quarterly, 47(2), 347-375. doi:10.1177/0899764017744894 Konrath, S., Ho, M.-H., & Zarins, S. (2016). The strategic helper: Narcissism and prosocial motives and behaviors. Current Psychology, 35(2), 182-194. doi:10.1007/s12144-016-9417-3 Korenok, O., Millner, E. L., & Razzolini, L. (2013). Impure altruism in dictators' giving. Journal of Public Economics, 97, 1-8. doi:https://doi.org/10.1016/j.jpubeco.2012.08.006 Korndörfer, M., Egloff, B., & Schmukle, S. C. (2015). A Large Scale Test of the Effect of Social Class on Prosocial Behavior. PLoS ONE, 10(7), e0133193. doi:10.1371/journal.pone.0133193 Kraus, M. W., & Callaghan, B. (2016). Social Class and Prosocial Behavior: The Moderating Role of Public Versus Private Contexts. Social Psychological and Personality Science, 7(8), 769- 777. doi:10.1177/1948550616659120 Landler, M., Belson, K., & Haberman, M. (2017). Trump tells pence to leave N.F.L. game as players kneel during anthem. New York Times. Retrieved from https://www.nytimes.com/2017/10/08/us/politics/pence-anthem-colts.html Landry, C. E., Lange, A., List, J. A., Price, M. K., & Rupp, N. G. (2006). Toward and understanding of the economics of charity: Evidence from a field experiment. Quarterly Journal of Economics, 121(2), 747-782. Lee, L., & Piliavin, J. A. (1999). Giving Time, Money, and Blood: Similarities and Differences. Social Psychology Quarterly, 62(3), 276-290. Lee, S., & Feeley, T. H. (2016). The identifiable victim effect: a meta-analytic review. Social Influence, 11(3), 199-215. doi:10.1080/15534510.2016.1216891 Lee, S., Winterich, K. P., & Ross, W. T. (2014). I’m Moral, but I Won’t Help You: The Distinct Roles of Empathy and Justice in Donations. Journal of Consumer Research, 41(3), 678-696. doi:10.1086/677226 Lee, Z., & Sargeant, A. (2011). Dealing with social desirability bias: An application to charitable giving. European journal of marketing, 45(5), 703-719. doi:10.1108/03090561111119994 Leeuwen, M. H. D., & Wiepking, P. (2012). National campaigns for charitable causes: A literature review. Nonprofit and Voluntary Sector Quarterly, 42(2), 219-240. doi:10.1177/0899764012467084 Leiberg, S., Klimecki, O., & Singer, T. (2011). Short-term compassion training increases prosocial behavior in a newly developed prosocial game. PLoS ONE, 6(3), e17798. doi:10.1371/journal.pone.0017798

174

Leslie, L. M., Snyder, M., & Glomb, T. M. (2013). Who gives? Multilevel effects of gender and ethnicity on workplace charitable giving. Journal of Applied Psychology, 98(1), 49-62. doi:10.1037/a0029943 Lesner, T. H., & Rasmussen, O. D. (2014). The identifiable victim effect in charitable giving: evidence from a natural field experiment. Applied Economics, 46(36), 4409-4430. doi:10.1080/00036846.2014.962226 Levine, L. R., Bluni, T. D., & Hochman, S. H. (1998). Attire and Charitable Behavior. Psychological Reports, 83(1), 15-18. doi:10.2466/pr0.1998.83.1.15 Levine, M., Cassidy, C., Brazier, G., & Reicher, S. (2002). Self-Categorization and Bystander Non- intervention: Two Experimental Studies. Journal of Applied Social Psychology, 32(7), 1452- 1463. doi:doi:10.1111/j.1559-1816.2002.tb01446.x Levine, M., & Manning, R. (2013). Social identity, group processes, and helping in emergencies. European Review of Social Psychology, 24(1), 225-251. doi:10.1080/10463283.2014.892318 Levine, M., Prosser, A., Evans, D., & Reicher, S. (2005). Identity and emergency intervention: How social group membership and inclusiveness of group boundaries shape helping behavior. Personality and Social Psychology Bulletin, 31(4), 443-453. doi:10.1177/0146167204271651 Levine, M., & Thompson, K. (2004). Identity, place, and bystander intervention: Social categories and helping after natural disasters. The Journal of social psychology, 144(3), 229-245. doi:10.3200/SOCP.144.3.229-245 Liang, J., Chen, Z., & Lei, J. (2016). Inspire me to donate: The use of strength emotion in donation appeals. Journal of Consumer Psychology, 26(2), 283-288. doi:10.1016/j.jcps.2015.09.001 Lim, D., & Desteno, D. (2016). Suffering and compassion: The links among adverse life experiences, empathy, compassion, and prosocial behavior. Emotion, 16(2), 175-182. doi:10.1037/emo0000144 List, J. A., & Price, M. K. (2009). The role of social connections in charitable fundraising: evidence from a natural field experiment. Journal of economic behavior & organization, 69(2), 160- 169. doi:10.1016/j.jebo.2007.08.011 Lobb, A., Mock, N., & Hutchinson, P. L. (2012). Traditional and social media coverage and charitable giving following the 2010 earthquake in Haiti. Prehospital and Disaster Medicine, 27(4), 319-324. doi:10.1017/S1049023X12000908 Loseke, D. R., & Fawcett, K. (1995). Appealing Appeals: Constructing Moral Worthiness, 1912- 1917. The Sociological Quarterly, 36(1), 61-77.

175

Louis, W. R., Duck, J. M., Terry, D. J., & Lalonde, R. N. (2010). Speaking out on immigration policy in Australia: Identity threat and the interplay of own opinion and public opinion. Journal of Social Issues, 66(4), 653-672. doi:10.1111/j.1540-4560.2010.01669.x Louis, W. R., Thomas, E. F., Chapman, C. M., Achia, T., Wibisono, S., Mirnajafi, Z., & Droogendyk, L. (under review). Emerging research on intergroup prosociality: Group members' charitable giving, positive contact, allyship, and solidarity with others. Manuscript under review. Lucas, E. (2017). Reinventing the rattling tin: How UK charities use Facebook in fundraising. International Journal of Nonprofit and Voluntary Sector Marketing, 22(2), n/a-n/a. doi:10.1002/nvsm.1576 Luengo Kanacri, B. P., González, R., Valdenegro, D., Jiménez-Moya, G., Saavedra, P., Mora, E. A., . . . Pastorelli, C. (2016). Civic engagement and giving behaviors: The role of empathy and beliefs about poverty. The Journal of social psychology, 156(3), 256-271. doi:10.1080/00224545.2016.1148006 Lupoli, M. J., Jampol, L., & Oveis, C. (2017). Lying because we care: Compassion increases prosocial lying. Journal of Experimental Psychology: General, 146(7), 1026-1042. doi:10.1037/xge0000315 Ma, L. K., Tunney, R. J., & Ferguson, E. (2017). Does Gratitude Enhance Prosociality?: A Meta- Analytic Review. Psychological Bulletin, 143(6), 601-635. doi:10.1037/bul0000103 Maner, J. K., & Gailliot, M. T. (2007). Altruism and egoism: prosocial motivations for helping depend on relationship context. European Journal of Social Psychology, 37(2), 347-358. doi:10.1002/ejsp.364 Manesi, Z., Van Lange, P. A. M., Van Doesum, N. J., & Pollet, T. V. (2018). What are the most powerful predictors of charitable giving to victims of typhoon Haiyan: Prosocial traits, socio-demographic variables, or eye cues? Personality and Individual Differences. doi:https://doi.org/10.1016/j.paid.2018.03.024 Martin, R., & Randal, J. (2008). How is donation behaviour affected by the donations of others? Journal of economic behavior & organization, 67(1), 228-238. doi:10.1016/j.jebo.2007.08.001 Maruyama, G., Fraser, S. C., & Miller, N. (1982). Personal responsibility and altruism in children. Journal of Personality and Social Psychology, 42(4), 658-664. doi:10.1037/0022- 3514.42.4.658 Mayer, R., Davis, J., & Schoorman, F. (1995). An integrative model of organizational trust. The Academy of Management Review, 20(3), 709. doi:10.2307/258792 McCall, G. J., & Simmons, J. L. (1978). Identities and Interactions. New York: Free Press.

176

McDonald, R. I., Fielding, K. S., & Louis, W. R. (2013). Energizing and de-motivating effects of norm-conflict. Personality and Social Psychology Bulletin, 39(1), 57-72. doi:10.1177/0146167212464234 McDonald, R. I., Fielding, K. S., & Louis, W. R. (2014). Conflicting norms highlight the need for action. Environment and Behavior, 46(2), 139-162. doi:10.1177/0013916512453992 McGregor-Lowndes, M., & Crittall, M. (2014). An examination of tax deductible donations made by individual Australian taxpayers in 2011-12. Retrieved from Brisbane, Australia: http://eprints.qut.edu.au/72596/1/Tax_stats_2011-12_Working_Paper_ACPNS_63.pdf McKeever, B. S. (2015). The Nonprofit Sector in Brief 2015: Public Charities, Giving, and Volunteering. Retrieved from https://www.urban.org/research/publication/nonprofit-sector- brief-2015-public-charities-giving-and-volunteering McLaughlin, R. P. (2014). Christian theology and the status of animals : the dominant tradition and its alternatives. New York: Palgrave Macmillan. Meer, J. (2011). Brother, can you spare a dime? Peer pressure in charitable solicitation. Journal of Public Economics, 95(7), 926-941. doi:10.1016/j.jpubeco.2010.11.026 Mesch, D., Brown, M. S., Moore, Z. I., & Hayat, A. D. (2011). Gender differences in charitable giving. International Journal of Nonprofit and Voluntary Sector Marketing, 16(4), 342-355. doi:10.1002/nvsm.432 Mesch, D., Osili, U., Ackerman, J., & Dale, E. (2015a). Do Women Give More? Findings from three unique data sets on charitable giving. Retrieved from Online: https://scholarworks.iupui.edu/handle/1805/6984 Mesch, D., Osili, U., Ackerman, J., & Dale, E. (2015b). Where Do Men and Women Give? Gender differences in the motivations and purposes of giving. Retrieved from Online: https://scholarworks.iupui.edu/handle/1805/6985 Mesch, D., Rooney, P. M., Steinberg, K. S., & Denton, B. (2006). The Effects of Race, Gender, and Marital Status on Giving and Volunteering in Indiana. Nonprofit and Voluntary Sector Quarterly, 35(4), 565-587. doi:10.1177/0899764006288288 Michel, G., & Rieunier, S. (2012). Nonprofit brand image and typicality influences on charitable giving. Journal of business research, 65(5), 701-707. doi:10.1016/j.jbusres.2011.04.002 Midlarsky, E., & Hannah, M. E. (1989). The generous elderly: Naturalistic studies of donations across the life span. Psychology and Aging, 4(3), 346-351. doi:10.1037/0882-7974.4.3.346 Miller, B. (2009). Community fundraising 2.0—the future of fundraising in a networked society? International Journal of Nonprofit and Voluntary Sector Marketing, 14(4), 365-370. doi:10.1002/nvsm.373

177

NAB. (2014). Charitable Giving Index [Report]. Retrieved from Online: http://business.nab.com.au/wp-content/uploads/2014/11/nab-charitable-giving-index- indepth-report-august-2014.pdf NAB. (2015). Charitable Giving Index [Report]. Retrieved from Online: https://business.nab.com.au/wp-content/uploads/2015/10/NAB-Charitable-Giving-index- Indepth-report-–-12-months-to-August-2015.pdf Nadler, A. (2002). Inter–group helping relations as power relations: Maintaining or challenging social dominance between groups through helping. Journal of Social Issues, 58(3), 487-502. doi:10.1111/1540-4560.00272 Nadler, A. (2016). Intergroup helping relations. Current Opinion in Psychology, 11, 64-68. doi:https://doi.org/10.1016/j.copsyc.2016.05.016 Nadler, A., & Chernyak-Hai, L. (2014). Helping them stay where they are: Status effects on dependency/autonomy-oriented helping. Journal of Personality and Social Psychology, 106(1), 58-72. doi:10.1037/a0034152 Nadler, A., & Fisher, J. D. (1986). The role of threat to self-esteem and perceived control in recipient reaction to help: Theory development and empirical validation. Advances in Experimental Social Psychology, 19, 81-122. doi:10.1016/S0065-2601(08)60213-0 Nadler, A., Fisher, J. D., & Streufert, S. (1976). When helping hurts: Effects of donor-recipient similarity and recipient self-esteem on reactions to aid. Journal of personality, 44(3), 392- 409. doi:10.1111/j.1467-6494.1976.tb00129.x Nadler, A., & Halabi, S. (2006). Intergroup helping as status relations: Effects of status stability, identification, and type of help on receptivity to high-status group's help. Journal of Personality and Social Psychology, 91(1), 97-110. doi:10.1037/0022-3514.91.1.97 Nadler, A., Harpaz-Gorodeisky, G., & Ben-David, Y. (2009). Defensive helping: Threat to group identity, ingroup identification, status stability, and common group identity as determinants of intergroup help-giving. Journal of Personality and Social Psychology, 97(5), 823-834. doi:10.1037/a0015968 Naskrent, J., & Siebelt, P. (2011). The influence of commitment, trust, satisfaction, and involvement on donor retention. Voluntas, 22(4), 757-778. doi:10.1007/s11266-010-9177-x Neto, A. B. F. (2018). Charity and public libraries: Does government funding crowd out donations? Journal of Cultural Economics, 1-18. doi:10.1007/s10824-018-9318-4 Neumayr, M., & Handy, F. (2017). Charitable giving: What influences donors’ choice among different causes? Voluntas, 0(0), 1-29. doi:10.1007/s11266-017-9843-3

178

Newman, G. E., & Jeremy Shen, Y. (2012). The counterintuitive effects of thank-you gifts on charitable giving. Journal of Economic Psychology, 33(5), 973-983. doi:https://doi.org/10.1016/j.joep.2012.05.002 Nilsson, A., Erlandsson, A., & Vastfjall, D. (2016). The congruency between moral foundations and intentions to donate, self-reported donations, and actual donations to charity. Journal of Research in Personality, 65, 22-29. doi:10.1016/j.jrp.2016.07.001 Nolan, J. M., Schultz, P. W., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2008). Normative social influence is underdetected. Personality and Social Psychology Bulletin, 34(7), 913- 923. doi:10.1177/0146167208316691 Nook, E. C., Ong, D. C., Morelli, S. A., Mitchell, J. P., & Zaki, J. (2016). Prosocial conformity: prosocial norms generalize across behavior and empathy. Personality & Social Psychology Bulletin, 42(8), 1045-1062. doi:10.1177/0146167216649932 OECD. (2016). Compare your country: Official Development Assistance 2015. Retrieved from http://www2.compareyourcountry.org/oda Oelrichs, D. M. (2018). Promoting efficacy as a means to increase pro-environmental behaviour. (Honours thesis), The University of Queensland. Okten, C., & Weisbrod, B. A. (2000). Determinants of donations in private nonprofit markets. The Journal of Public Economics, 75(2), 255. Osili, U. O., Hirt, D. E., & Raghavan, S. (2011). Charitable giving inside and outside the workplace: the role of individual and firm characteristics. International Journal of Nonprofit and Voluntary Sector Marketing, 16(4), 393-408. doi:10.1002/nvsm.435 Osili, U. O., & Zarins, S. (2018). Fewer Americans are giving money to charity but total donations are at record levels anyway. The Conversation. Retrieved from https://theconversation.com/fewer-americans-are-giving-money-to-charity-but-total- donations-are-at-record-levels-anyway-98291 Ottoni-Wilhelm, M., Vesterlund, L., & Xie, H. (2017). Why Do People Give? Testing Pure and Impure Altruism. American Economic Review, 107(11), 3617-3633. doi:doi: 10.1257/aer.20141222 Oyserman, D. (2009). Identity-based motivation: Implications for action-readiness, procedural- readiness, and consumer behavior. Journal of Consumer Psychology, 19(3), 250-260. doi:10.1016/j.jcps.2009.05.008 Packer, D. J. (2008). On being both with us and against us: A normative conflict model of dissent in social groups. Personality & Social Psychology Review, 12(1), 50-72. doi:10.1177/1088868307309606

179

Park, K., & Lee, S. S. (2015). The role of beneficiaries’ group identity in determining successful appeal strategies for charitable giving. Psychology & marketing, 32(12), 1117-1132. doi:10.1002/mar.20852 Parsons, L. M. (2007). The impact of financial information and voluntary disclosures on contributions to not-for-profit organizations. Behavioral Research in Accounting, 19, 179- 196. Payne, A., Scharf, K., & Smith, S. (2014). Online fundraising - the perfect ask? Retrieved from https://warwick.ac.uk/fac/soc/economics/staff/kascharf/online-fundraising-payne-scharf- smith.pdf Payne, A. A. (1998). Does the government crowd-out private donations? New evidence from a sample of non-profit firms. Journal of Public Economics, 69(3), 323-345. doi:10.1016/S0047-2727(98)00005-X Penner, L. A., Dovidio, J. F., Piliavin, J. A., & Schroeder, D. A. (2005). Prosocial behavior: Multilevel perspectives. Annual Review of Psychology, 56(1), 365-392. doi:10.1146/annurev.psych.56.091103.070141 Perry, R., Paradies, Y., & Pedersen, A. (2015). Religious Ambivalence: Suppression of Pro-Social Attitudes Toward Asylum Seekers by Right-Wing Authoritarianism. The International Journal for the Psychology of Religion, 25(3), 230-233. doi:10.1080/10508619.2014.921473 Piff, P. K., Kraus, M. W., Côté, S., Cheng, B. H., & Keltner, D. (2010). Having Less, Giving More: The Influence of Social Class on Prosocial Behavior. Journal of Personality and Social Psychology, 99(5), 771-784. doi:10.1037/a0020092 Piper, G., & Schnepf, S. (2008). Gender differences in charitable giving in Great Britain. Official journal of the International Society for Third-Sector Research, 19(2), 103-124. doi:10.1007/s11266-008-9057-9 Platow, M. J., Durante, M., Williams, N., Garrett, M., Walshe, J., Cincotta, S., . . . Barutchu, A. (1999). The Contribution of Sport Fan Social Identity to the Production of Prosocial Behavior. Group Dynamics: Theory, Research, and Practice, 3(2), 161-169. doi:10.1037/1089-2699.3.2.161 Plows, S. E., Smith, F. D., Smith, J. R., Chapman, C. M., La Macchia, S. T., & Louis, W. R. (2017). Healthy eating: A beneficial role for perceived norm conflict? Journal of Applied Social Psychology, 47(6), 295-304. doi:10.1111/jasp.12430 Polonsky, M. J., Shelley, L., & Voola, R. (2002). An Examination of Helping Behavior—Some Evidence from Australia. Journal of Nonprofit & Public Sector Marketing, 10(2), 67-82. doi:10.1300/J054v10n02_04

180

Postmes, T., Spears, R., Sakhel, K., & de Groot, D. (2001). Social influence in computer-mediated communication: The effects of anonymity on group behavior. Personality and Social Psychology Bulletin, 27(10), 1243-1254. doi:10.1177/01461672012710001 Pratto, F., Çidam, A., Stewart, A. L., Zeineddine, F. B., Aranda, M., Aiello, A., . . . Henkel, K. E. (2013). Social Dominance in Context and in Individuals: Contextual Moderation of Robust Effects of Social Dominance Orientation in 15 Languages and 20 Countries. Social Psychological and Personality Science, 4(5), 587-599. doi:10.1177/1948550612473663 Pratto, F., Sidanius, J., Stallworth, L. M., & Malle, B. F. (1994). Social Dominance Orientation: A personality variable predicting social and political attitudes. Journal of Personality and Social Psychology, 67(4), 741. doi:10.1037/0022-3514.67.4.741 Prendergast, G. P., & Maggie, C. H. W. (2013). Donors' experience of sustained charitable giving: a phenomenological study. JOURNAL OF CONSUMER MARKETING, 30(2), 130-139. doi:doi:10.1108/07363761311304942 Radley, A., & Kennedy, M. (1992). Reflections upon charitable giving: A comparison of individuals from business, ‘manual’ and professional backgrounds. Journal of Community & Applied Social Psychology, 2(2), 113-129. doi:10.1002/casp.2450020208 Rajan, S. S., Pink, G. H., & Dow, W. H. (2009). Sociodemographic and personality characteristics of Canadian donors contributing to international charity. Nonprofit and Voluntary Sector Quarterly, 38(3), 413-440. doi:10.1177/0899764008316056 Ratner, R. K., Zhao, M., & Clarke, J. A. (2011). The Norm of Self-Interest: Implications for Charitable Giving. In D. M. Oppenheimer & C. Y. Olivola (Eds.), The Science of Giving: Experimental Approaches to the Study of Charity (pp. 113-132). New York: Taylor & Francis. Ratner, R. K., Zhao, M., & Miller, D. (2009). How Can You Say "No"? Deference Granted to Advocates Who Are Victims. Advances in Consumer Research, 36, 202. Reed, A. (2002). Social identity as a useful perspective for self‐concept–based consumer research. Psychology and Marketing, 19(3), 235-266. doi:10.1002/mar.10011 Reese, G., Proch, J., & Cohrs, J. C. (2014). Individual Differences in Responses to Global Inequality. Analyses of Social Issues and Public Policy, 14(1), 217-238. doi:10.1111/asap.12032 Reicher, S., Cassidy, C., Wolpert, I., Hopkins, N., & Levine, M. (2006). Saving Bulgaria's Jews: an analysis of social identity and the mobilisation of social solidarity. European Journal of Social Psychology, 36(1), 49-72. doi:doi:10.1002/ejsp.291

181

Reicher, S. D., Spears, R., & Postmes, T. (1995). A social identity model of deindividuation phenomena. European Review of Social Psychology, 6(1), 161-198. doi:10.1080/14792779443000049 Reinstein, D., & Riener, G. (2012). Reputation and influence in charitable giving: an experiment. Theory and Decision, 72(2), 221-243. doi:10.1007/s11238-011-9245-8 Reno, R. R., Cialdini, R. B., & Kallgren, C. A. (1993). The transsituational influence of social norms. Journal of Personality and Social Psychology, 64(1), 104-112. doi:10.1037/0022- 3514.64.1.104 Roberts, R. D. (1984). A positive model of private charity and public transfers. Journal of Political Economy, 92(1), 136-148. doi:10.1086/261212 Rooney, P. M., Mesch, D. J., Chin, W., & Steinberg, K. S. (2005). The effects of race, gender, and survey methodologies on giving in the US. Economics Letters, 86(2), 173-180. doi:https://doi.org/10.1016/j.econlet.2004.06.015 Ross, L., Greene, D., & House, P. (1977). The "false consensus effect": An egocentric bias in social perception and attribution processes. Journal of Experimental Social Psychology, 13(3), 279-301. Rosseel, Y. (2012). Lavaan: An R package for structural equation modeling. Journal of Statistical Software, 48(2), 1-36. Rudolph, U., Roesch, S. C., Greitemeyer, T., & Weiner, B. (2004). A meta-analytic review of help giving and aggression from an attributional perspective: Contributions to a general theory of motivation. Cognition and Emotion, 18(6), 815-848. doi:10.1080/02699930341000248 Ryazanov, A. A., & Christenfeld, N. J. S. (2018). On the Limited Role of Efficiency in Charitable Giving. Nonprofit and Voluntary Sector Quarterly, 0899764018773899. doi:10.1177/0899764018773899 Sargeant, A. (1999). Charitable giving: towards a model of donor behaviour. Journal of Marketing Management, 15(4), 215-238. Sargeant, A., & Hudson, J. (2008). Donor retention: an exploratory study of door-to-door recruits. International Journal of Nonprofit and Voluntary Sector Marketing, 13(1), 89-101. doi:doi:10.1002/nvsm.301 Sargeant, A., & Jay, E. (2004). Reasons for lapse: the case of face‐to‐face donors. International Journal of Nonprofit and Voluntary Sector Marketing, 9(2), 171-182. doi:10.1002/nvsm.245 Sargeant, A., & Kähler, J. (1999). Returns on Fundraising Expenditures in the Voluntary Sector. Nonprofit Management and Leadership, 10(1), 5-19. doi:10.1002/nml.10102 Sargeant, A., & Shang, J. (2011). Bequest giving: revisiting donor motivation with dimensional qualitative research. Psychology & marketing, 28(10), 980-997. doi:10.1002/mar.20424

182

Sargeant, A., & Woodliffe, L. (2007a). Building donor loyalty: The antecedents and role of commitment in the context of charity giving. Journal of Nonprofit & Public Sector Marketing, 18(2), 47-68. doi:10.1300/J054v18n02_03 Sargeant, A., & Woodliffe, L. (2007b). Gift giving: an interdisciplinary review. International Journal of Nonprofit and Voluntary Sector Marketing, 12(4), 275-307. doi:10.1002/nvsm.308 Satow, K. L. (1975). Social approval and helping. Journal of Experimental Social Psychology, 11(6), 501-509. Saucier, D. A., Miller, C. T., & Doucet, N. (2005). Differences in Helping Whites and Blacks: A Meta-Analysis. Personality and Social Psychology Review, 9(1), 2-16. doi:10.1207/s15327957pspr0901_1 Saxton, G. D., & Wang, L. (2014). The social network effect: The determinants of giving through social media. Nonprofit and Voluntary Sector Quarterly, 43(5), 850-868. doi:10.1177/0899764013485159 Scharf, K., & Smith, S. (2016). Relational altruism and giving in social groups. Journal of Public Economics, 141, 1-10. doi:10.1016/j.jpubeco.2016.06.001 Schlegelmilch, B. B., Love, A., & Diamantopoulos, A. (1997). Responses to different charity appeals: the impact of donor characteristics on the amount of donations. European journal of marketing, 31(8), 548-560. doi:doi:10.1108/03090569710176574 Schmidt, S. (2009). Shall we really do it again? The powerful concept of replication is neglected in the social sciences. Review of General Psychology, 13(2), 90-100. doi:10.1037/a0015108 Schnable, A. (2015). Religion and giving for international Aid: Evidence from a survey of U.S. church members. Sociology of Religion, 76(1), 72-94. doi:10.1093/socrel/sru037 Schroeder, D. A., Dovidio, J. F., Sibicky, M. E., Matthews, L. L., & Allen, J. L. (1988). Empathic concern and helping behavior: Egoism or altruism? Journal of Experimental Social Psychology, 24(4), 333-353. doi:https://doi.org/10.1016/0022-1031(88)90024-8 Schultz, P. W., Nolan, J. M., Cialdini, R. B., Goldstein, N. J., & Griskevicius, V. (2007). The constructive, destructive, and reconstructive power of social norms. Psychological Science, 18(5), 429-434. doi:10.1111/j.1467-9280.2007.01917.x Schwartz, S. H. (1994). Are There Universal Aspects in the Structure and Contents of Human Values? Journal of Social Issues, 50(4), 19-45. doi:10.1111/j.1540-4560.1994.tb01196.x Shang, J., Reed, A., & Croson, R. (2008). Identity Congruency Effects on Donations. Journal of Marketing Research, 45(3), 351-361. doi:10.1509/jmkr.45.3.351

183

Shariff, A. F., Willard, A. K., Andersen, T., & Norenzayan, A. (2015). Religious Priming: A Meta- Analysis With a Focus on Prosociality. Personality and Social Psychology Review, 20(1), 27-48. doi:10.1177/1088868314568811 Shier, M. L., & Handy, F. (2012). Understanding online donor behavior: the role of donor characteristics, perceptions of the internet, website and program, and influence from social networks. International Journal of Nonprofit and Voluntary Sector Marketing, 17(3), 219- 230. doi:10.1002/nvsm.1425 Showers, V. E., Showers, L. S., Beggs, J. M., & Cox, J. E. (2011). Charitable giving expenditures and the faith factor. American Journal of Economics and Sociology, 70(1), 152-186. Sibley, C. G., & Bulbulia, J. (2015). Charity explains differences in life satisfaction between religious and secular New Zealanders. Religion, Brain & Behavior, 5(2), 91-100. doi:10.1080/2153599X.2014.899509 Siegel, J., Tan, C., Rosenberg, B., Navarro, M., Thomson, A., Lyrintzis, E., . . . Jones, N. (2016). Anger, frustration, boredom and the Department of Motor Vehicles: Can negative emotions impede organ donor registration? Social Science & Medicine, 153, 174. Silvergleid, J. E. (2003). Effects of watchdog organizations on the social capital market. New Directions for Philanthropic Fundraising, 2003(41), 7-26. Simmons, J. P., Nelson, L. D., & Simonsohn, U. (2012). A 21 word solution. SSRN. doi:10.2139/ssrn.2160588 Simmons, W. O., & Emanuele, R. (2004). Does government spending crowd out donations of time and money? Public Finance Review, 32(5), 498-511. doi:10.1177/1091142104264364 Simon, A. F. (1997). Television news and international earthquake relief. Journal of Communication, 47(3), 82-93. doi:doi:10.1111/j.1460-2466.1997.tb02718.x Singer, P. (2009). The Life You Can Save. New York: Random House. Sleesman, D. J., & Conlon, D. E. (2016). Encouraging Prosocial Decisions: The Role of Fairness Salience and Uncertainty. Journal of Behavioral Decision Making, 30(2), 502-515. doi:10.1002/bdm.1970 Slovic, P. (2007). "If I look at the mass I will never act": Psychic numbing and genocide. Judgment and Decision Making, 2(2), 79-95. Slovic, P., Västfjäll, D., Erlandsson, A., & Gregory, R. (2017). Iconic photographs and the ebb and flow of empathic response to humanitarian disasters. Proceedings of the National Academy of Sciences, 114(4), 640. Small, D. A., & Loewenstein, G. (2003). Helping a victim or helping the victim: Altruism and identifiability. Journal of Risk and Uncertainty, 26(1), 5-16. doi:10.1023/A:1022299422219

184

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(2), 143-153. doi:https://doi.org/10.1016/j.obhdp.2006.01.005 Small, D. A., & Simonsohn, U. (2008). Friends of Victims: Personal Experience and Prosocial Behavior. Journal of Consumer Research, 35(3), 532-542. doi:10.1086/527268 Small, D. A., & Verrochi, N. M. (2009). The Face of Need: Facial Emotion Expression on Charity Advertisements. Journal of Marketing Research, 46(6), 777-787. Smith, J. R., & Louis, W. R. (2008). Do as we say and as we do: The interplay of descriptive and injunctive group norms in the attitude–behaviour relationship. British Journal of Social Psychology, 47(4), 647-666. doi:10.1348/014466607X269748 Smith, J. R., & McSweeney, A. (2007). Charitable giving: The effectiveness of a revised theory of planned behaviour model in predicting donating intentions and behaviour. Journal of Community & Applied Social Psychology, 17(5), 363-386. doi:10.1002/casp.906 Smith, L. G. E., Gavin, J., & Sharp, E. (2015). Social identity formation during the emergence of the Occupy movement. European Journal of Social Psychology, 45(7), 818-832. doi:10.1002/ejsp.2150 Smith, L. G. E., McGarty, C., & Thomas, E. F. (2018). After Aylan Kurdi: How Tweeting About Death, Threat, and Harm Predict Increased Expressions of Solidarity With Refugees Over Time. Psychological Science, 29(4), 623-634. doi:10.1177/0956797617741107 Smith, L. G. E., Thomas, E. F., & McGarty, C. (2015). “We must be the change we want to see in the world”: Integrating norms and identities through social interaction. Political Psychology, 36(5), 543-557. doi:10.1111/pops.12180 Smith, R. W., Faro, D., & Burson, K. A. (2013). More for the Many: The Influence of Entitativity on Charitable Giving. Journal of Consumer Research, 39(5), 961-976. doi:10.1086/666470 Smith, S., Windmeijer, F., & Wright, E. (2015). Peer effects in charitable giving: Evidence from the (running) field. The Economic Journal, 125(585), 1053-1071. doi:10.1111/ecoj.12114 Soetevent, A. R. (2005). Anonymity in giving in a natural context—a field experiment in 30 churches. Journal of Public Economics, 89(11-12), 2301-2323. Sparks, A., & Barclay, P. (2013). Eye images increase generosity, but not for long: the limited effect of a false cue. Evolution and Human Behavior, 34(5), 317-322. doi:https://doi.org/10.1016/j.evolhumbehav.2013.05.001 Srnka, K. J., Grohs, R., & Eckler, I. (2003). Increasing fundraising efficiency by segmenting donors. Australasian Marketing Journal, 11(1), 70-86. doi:10.1016/S1441-3582(03)70119-0

185

Stack, L. (2017). Donations to A.C.L.U. and other organizations surge after Trump’s order. The New York Times. Retrieved from https://www.nytimes.com/2017/01/30/us/aclu-fund- raising-trump-travel-ban.html Stavrova, O., & Siegers, P. (2014). Religious Prosociality and Morality Across Cultures: How Social Enforcement of Religion Shapes the Effects of Personal Religiosity on Prosocial and Moral Attitudes and Behaviors. Personality and Social Psychology Bulletin, 40(3), 315-333. doi:10.1177/0146167213510951 Steinberg, L., & Monahan, K. C. (2007). Age Differences in Resistance to Peer Influence. Developmental Psychology, 43(6), 1531-1543. Steinberg, M. A., Crow, R., Cain, L., & Milford, C. (2005). Living and giving: Older Australians and charitable donations. Australasian Journal on Ageing, 24(1), 41-43. doi:10.1111/j.1741- 6612.2005.00062.x Stephenson, A. L., & Bell, N. (2014). Motivation for alumni donations: a social identity perspective on the role of branding in higher education. International Journal of Nonprofit and Voluntary Sector Marketing, 19(3), 176-186. doi:10.1002/nvsm.1495 Stevenson, C., & Manning, R. (2010). National identity and international giving: Irish adults' accounts of charitable behaviour. Journal of Community and Applied Social Psychology, 20(4), 249-261. doi:10.1002/casp.1029 Stuart, A., Thomas, E. F., & Donaghue, N. (2018). “I Don't Really Want to Be Associated With the Self-Righteous Left Extreme”: Disincentives to Participation in Collective Action. Journal of Social and Political Psychology, 6(1), 242-270. doi:10.5964/jspp.v6i1.567 Stürmer, S., & Siem, B. (2017). A Group-Level Theory of Helping and Altruism Within and Across Group Boundaries. In E. van Leeuwen & H. Zagefka (Eds.), Intergroup Helping (pp. 103- 127): Springer International. Stürmer, S., Snyder, M., Kropp, A., & Siem, B. (2006). Empathy-Motivated Helping: The Moderating Role of Group Membership. Personality and Social Psychology Bulletin, 32(7), 943-956. doi:10.1177/0146167206287363 Stürmer, S., Snyder, M., & Omoto, A. M. (2005). Prosocial Emotions and Helping: The Moderating Role of Group Membership. Journal of Personality and Social Psychology, 88(3), 532-546. doi:10.1037/0022-3514.88.3.532 Subašić, E., Reynolds, K. J., & Turner, J. C. (2008). The Political Solidarity Model of Social Change: Dynamics of Self-Categorization in Intergroup Power Relations. Personality and Social Psychology Review, 12(4), 330-352. doi:10.1177/1088868308323223

186

Sun, S., Zagefka, H., & Goodwin, R. (2013). Predictors of intergroup concern for disaster victims of the Japan earthquake. Asian Journal of Social Psychology, 16(2), 152-157. doi:10.1111/ajsp.12023 Svanberg, A. S., Lampic, C., Geijerwall, A.-L., Gudmundsson, J., Karlström, P.-O., Solensten, N.- G., & Sydsjö, G. (2012). Gamete donors’ motivation in a Swedish national sample: is there any ambivalence? A descriptive study. Acta Obstetricia et Gynecologica Scandinavica, 91(8), 944-951. doi:doi:10.1111/j.1600-0412.2012.01430.x Szper, R., & Prakash, A. (2011). Charity Watchdogs and the Limits of Information-Based Regulation. Voluntas: International Journal of Voluntary and Nonprofit Organizations, 22(1), 112-141. doi:10.1007/s11266-010-9156-2 Tajfel, H. (1981). Human Groups and Social Categories. Cambridge: Cambridge University Press. Tausch, N., Becker, J. C., Spears, R., Christ, O., Saab, R., Singh, P., & Siddiqui, R. N. (2011). Explaining Radical Group Behavior: Developing Emotion and Efficacy Routes to Normative and Nonnormative Collective Action. Journal of Personality and Social Psychology, 101(1), 129-148. doi:10.1037/a0022728 Taylor, L., & Hanna, J. (2018). Altruism born of suffering among emerging adults in Northern Ireland. Journal of Aggression, Conflict and Peace Research, 10(3), 157-169. doi:10.1108/JACPR-01-2017-0271 Tempel, E. R., Seiler, T. L., & Burlingame, D. F. (2016). Achieving Excellence in Fundraising (4th ed.). New Jersey: John Wiley & Sons. Terry, D. J., & Hogg, M. A. (1996). Group norms and the attitude-behavior relationship: A role for group identification. Personality and Social Psychology Bulletin, 22(8), 776-793. doi:10.1177/0146167296228002 Teson, K. (2017). Rage Donations: Is Your Nonprofit Ready for this New Trend? Wired Impact. Retrieved from https://wiredimpact.com/blog/rage-donations-nonprofit-trend/ Thomas, E. F., Amiot, C. E., Louis, W. R., & Goddard, A. (2017). Collective self-determination: How the agent of help promotes pride, well-being, and support for intergroup helping. Personality and Social Psychology Bulletin, 43(5), 662-677. doi:10.1177/0146167217695553 Thomas, E. F., Cary, N., Smith, L. G. E., Spears, R., & McGarty, C. (2018). The role of social media in shaping solidarity and compassion fade: How the death of a child turned apathy into action but distress took it away. New Media & Society, 1461444818760819. doi:10.1177/1461444818760819

187

Thomas, E. F., Mavor, K. I., & McGarty, C. (2012). Social identities facilitate and encapsulate action-relevant constructs: A test of the social identity model of collective action. Group Processes & Intergroup Relations, 15(1), 75-88. doi:10.1177/1368430211413619 Thomas, E. F., & McGarty, C. (2009). The role of efficacy and moral outrage norms in creating the potential for international development activism through group-based interaction. British Journal of Social Psychology, 48(1), 115-134. doi:10.1348/014466608X313774 Thomas, E. F., & McGarty, C. (2018). Giving versus acting: Using latent profile analysis to distinguish between benevolent and activist support for global poverty reduction. British Journal of Social Psychology, 57(1), 189-209. doi:10.1111/bjso.12228 Thomas, E. F., McGarty, C., Lala, G., Stuart, A., Hall, L. J., & Goddard, A. (2015). Whatever happened to Kony2012? Understanding a global Internet phenomenon as an emergent social identity. European Journal of Social Psychology, 45(3), 356-367. doi:10.1002/ejsp.2094 Thomas, E. F., McGarty, C., & Mavor, K. I. (2009a). Aligning Identities, Emotions, and Beliefs to Create Commitment to Sustainable Social and Political Action. Personality and Social Psychology Review, 13(3), 194-218. doi:10.1177/1088868309341563 Thomas, E. F., McGarty, C., & Mavor, K. I. (2009b). Transforming “Apathy Into Movement”:The Role of Prosocial Emotions in Motivating Action for Social Change. Personality and Social Psychology Review, 13(4), 310-333. doi:10.1177/1088868309343290 Thomas, E. F., McGarty, C., & Mavor, K. I. (2010). Social psychology of Making Poverty History: Motivating anti-poverty action in Australia. Australian Psychologist, 45(1), 4-15. doi:10.1080/00050060903447095 Thomas, E. F., McGarty, C., Reese, G., Berndsen, M., & Bliuc, A.-M. (2016). Where there is a (collective) will, there are (effective) ways: Integrating individual- and group-level factors in explaining humanitarian collective action. Personality & Social Psychology Bulletin, 42(12), 1678-1692. Thomas, E. F., Smith, L. G. E., McGarty, C., Reese, G., Kende, A., Bliuc, A.-M., . . . Spears, R. (2018). When and how social movements mobilize action within and across nations to promote solidarity with refugees. European Journal of Social Psychology, 0(0). doi:10.1002/ejsp.2380 Thomsson, K. M., & Vostroknutov, A. (2017). Small-world conservatives and rigid liberals: Attitudes towards sharing in self-proclaimed left and right. Journal of Economic Behavior and Organization, 135, 181-192. doi:10.1016/j.jebo.2017.01.023 Thornton, J. P., & Helms, S. (2013). Afterlife incentives in charitable giving. Applied Economics, 45(19), 2779-2791. doi:10.1080/00036846.2012.678984

188

Tillerson, R. W. (2018). FY2018 Congressional Budget Justification - Secretary's Letter [Press release]. Retrieved from https://www.state.gov/documents/organization/271282.pdf Tinkelman, D., & Mankaney, K. (2007). When is Administrative Efficiency Associated With Charitable Donations? Nonprofit and Voluntary Sector Quarterly, 36(1), 41-64. doi:10.1177/0899764006293176 Tremblay-Boire, J., & Prakash, A. (2017). Will You Trust Me?: How Individual American Donors Respond to Informational Signals Regarding Local and Global Humanitarian Charities. Voluntas: International Journal of Voluntary and Nonprofit Organizations, 28(2), 621-647. doi:10.1007/s11266-016-9782-4 Trump, D. J. (2017). Executive order protecting the nation from foreign terrorist entry into the United States. Retrieved from https://www.whitehouse.gov/presidential-actions/executive- order-protecting-nation-foreign-terrorist-entry-united-states-2/ Turner, J. C., Hogg, M. A., Oakes, P. J., Reicher, S. D., & Wetherell, M. S. (1987). Rediscovering the Social Group: A Self-Categorization Theory. New York: Basil Blackwell. United Nations Statistics Division. (2003). Handbook on Non-Profit Institutions in the System of National Accounts. Retrieved from New York: http://unstats.un.org/unsd/publication/SeriesF/SeriesF_91E.pdf USCB, U. C. B. (2015). Statistical Abstract of the United States: 2012. Retrieved from https://www.census.gov/library/publications/2011/compendia/statab/131ed/population.html Vaidyanathan, B., Hill, J. P., & Smith, C. (2011). Religion and charitable financial giving to religious and secular causes: Does political ideology matter? Journal for the Scientific Study of Religion, 50(3), 450-469. doi:10.1111/j.1468-5906.2011.01584.x Vaish, A., Kelsey, C. M., Tripathi, A., & Grossmann, T. (2017). Attentiveness to eyes predicts generosity in a reputation-relevant context. Evolution and Human Behavior, 38(6), 729-733. doi:https://doi.org/10.1016/j.evolhumbehav.2017.07.004 Van de Vyver, J., & Abrams, D. (2015). Testing the prosocial effectiveness of the prototypical moral emotions: Elevation increases benevolent behaviors and outrage increases justice behaviors. Journal of Experimental Social Psychology, 58, 23-33. doi:10.1016/j.jesp.2014.12.005 van Der Toorn, J., Nail, P. R., Liviatan, I., & Jost, J. T. (2014). My country, right or wrong: Does activating system justification motivation eliminate the liberal-conservative gap in patriotism? Journal of Experimental Social Psychology, 54, 50-60. doi:10.1016/j.jesp.2014.04.003

189

Van Doesum, N. J., Tybur, J. M., & Van Lange, P. A. M. (2017). Class impressions: Higher social class elicits lower prosociality. Journal of Experimental Social Psychology, 68, 11-20. doi:10.1016/j.jesp.2016.06.001 van Doorn, J., Zeelenberg, M., & Breugelmans, S. M. (2017). The impact of anger on donations to victims. International Review of Victimology, 23(3), 303-312. doi:10.1177/0269758017710819 Van Lange, P. A. M., Bekkers, R., Chirumbolo, A., & Leone, L. (2012). Are conservatives less likely to be prosocial than liberals? From games to ideology, political preferences and voting. European Journal of Personality, 26(5), 461-473. doi:10.1002/per.845 van Leeuwen, E. (2007). Restoring identity through outgroup helping: beliefs about international aid in response to the December 2004 tsunami. European Journal of Social Psychology, 37(4), 661-671. doi:10.1002/ejsp.389 van Leeuwen, E. (2017). The SOUTH model: On the pros and cons of strategic outgroup helping. In E. van Leeuwen & H. Zagefka (Eds.), Intergroup Helping (pp. 131-158): Springer International. van Leeuwen, E., & Harinck, F. (2016). Increasing intergroup distinctiveness: The benefits of third party helping. Personality and Social Psychology Bulletin, 42(10), 1402. doi:10.1177/0146167216662867 van Leeuwen, E., & Mashuri, A. (2012). When Common Identities Reduce Between-Group Helping. Social Psychological and Personality Science, 3(3), 259-265. doi:10.1177/1948550611417315 van Zomeren, M., Kutlaca, M., & Turner-Zwinkels, F. (2018). Integrating who "we" are with what "we" (will not) stand for: A further extension of the Social Identity Model of Collective Action. European Review of Social Psychology, 29(1), 122-160. doi:10.1080/10463283.2018.1479347 van Zomeren, M., Postmes, T., & Spears, R. (2008). Toward an integrative social identity model of collective action: A quantitative research synthesis of three socio-psychological perspectives. Psychological Bulletin, 134(4), 504-535. doi:10.1037/0033-2909.134.4.504 van Zomeren, M., Postmes, T., Spears, R., & Bettache, K. (2011). Can moral convictions motivate the advantaged to challenge social inequality? Group Processes & Intergroup Relations, 14(5), 735-753. doi:10.1177/1368430210395637 van Zomeren, M., Spears, R., Fischer, A. H., & Leach, C. W. (2004). Put your money where your mouth is! Explaining collective action tendencies through group-based anger and group efficacy. Journal of Personality and Social Psychology, 87(5), 649-664. doi:10.1037/0022- 3514.87.5.649

190

Veldhuizen, I., Kort, W. d., Ferguson, E., & Atsma, F. (2012). Exploring the pattern of blood donor beliefs in first-time, novice, and experienced donors: differentiating reluctant altruism, pure altruism, impure altruism, and warm glow. Transfusion, 52(2), 343-355. doi:10.1111/j.1537- 2995.2011.03279.x Walker, P. (2016). Charities to get formal warning about 'chuggers'. The Guardian. Retrieved from https://www.theguardian.com/society/2016/feb/28/charities-formal-warning-chuggers- fundraising-methods Wang, L., & Graddy, E. (2008). Social capital, volunteering, and charitable giving. Voluntas, 19(1), 23-42. doi:10.1007/s11266-008-9055-y Warr, P. G. (1982). Pareto optimal redistribution and private charity. Journal of Public Economics, 19(1), 131-138. doi:10.1016/0047-2727(82)90056-1 Webster, R. J., Burns, M. D., Pickering, M., & Saucier, D. A. (2014). The Suppression and Justification of Prejudice as a Function of Political Orientation. European Journal of Personality, 28(1), 44-59. doi:10.1002/per.1896 West, S. G., & Jan Brown, T. (1975). Physical attractiveness, the severity of the emergency and helping: A field experiment and interpersonal simulation. Journal of Experimental Social Psychology, 11(6), 531-538. doi:https://doi.org/10.1016/0022-1031(75)90004-9 White, K., & Argo, J. J. (2009). Social identity threat and consumer preferences. Journal of Consumer Psychology, 19(3), 313-325. doi:10.1016/j.jcps.2009.03.007 White, K. M., Poulsen, B. E., & Hyde, M. K. (2017). Identity and personality influences on donating money, time, and blood. Nonprofit and Voluntary Sector Quarterly, 46(2), 372- 394. doi:10.1177/0899764016654280 Wiepking, P. (2007). The Philanthropic Poor: In Search of Explanations for the Relative Generosity of Lower Income Households. Voluntas, 18(4), 339-358. doi:10.1007/s11266-007-9049-1 Wiepking, P. (2010). Democrats support international relief and the upper class donates to art? How opportunity, incentives and confidence affect donations to different types of charitable organizations. Social science research, 39(6), 1073-1087. doi:10.1016/j.ssresearch.2010.06.005 Wiepking, P., & Bekkers, R. (2012). Who gives? A literature review of predictors of charitable giving. Part two: Gender, marital status, income, and wealth. Voluntary Sector Review, 3(2), 217-245. doi:10.1332/204080512X649379 Wiepking, P., Bekkers, R., & Osili, U. O. (2014). Examining the association of religious context with giving to non-profit organizations. European Sociological Review, 30(5), 640-654.

191

Wiepking, P., & James, R. (2013). Why are the oldest old less generous? Explanations for the unexpected age-related drop in charitable giving. Ageing and Society, 33(3), 486-510. doi:10.1017/S0144686X12000062 Wiepking, P., & Maas, I. (2009). Resources That Make You Generous: Effects of Social and Human Resources on Charitable Giving. Social Forces, 87(4), 1973-1995. Wiepking, P., Scaife, W., & McDonald, K. (2012). Motives and barriers to bequest giving. Journal of Consumer Behaviour, 11(1), 56-66. doi:10.1002/cb.369 Wilson, M. S., & Sibley, C. G. (2013). Social Dominance Orientation and Right‐Wing Authoritarianism: Additive and Interactive Effects on Political Conservatism. Political Psychology, 34(2), 277-284. doi:10.1111/j.1467-9221.2012.00929.x Winterich, K. P., Mittal, V., & Ross Jr, W. T. (2009). Donation behavior toward in- groups and out- groups: The role of gender and moral identity. Journal of Consumer Research, 36(2), 199- 214. doi:10.1086/596720 Winterich, K. P., Zhang, Y., & Mittal, V. (2012). How political identity and charity positioning increase donations: Insights from Moral Foundations Theory. International Journal of Research in Marketing. doi:10.1016/j.ijresmar.2012.05.002 Yen, S. T., & Zampelli, E. M. (2014). What drives charitable donations of time and money? The roles of political ideology, religiosity, and involvement. Journal of Behavioral and Experimental Economics. doi:10.1016/j.socec.2014.01.002 Zagefka, H. (2018a). It could have been me: Proximity motivates disaster giving. International Journal of Nonprofit and Voluntary Sector Marketing, 23(1), n/a-n/a. doi:10.1002/nvsm.1587 Zagefka, H. (2018b). Triadic intergroup relations: studying situations with an observer, an actor, and a recipient of behaviour. Journal of Theoretical Social Psychology, 00, 1-13. doi:10.1002/jts5.26 Zagefka, H., & James, T. (2015). The psychology of charitable donations to disaster victims and beyond. Social Issues and Policy Review, 9(1), 155-192. doi:10.1111/sipr.12013 Zagefka, H., Noor, M., & Brown, R. (2013). Familiarity breeds compassion: Knowledge of disaster areas and willingness to donate money to disaster victims. Applied Psychology, 62(4), 640- 654. doi:10.1111/j.1464-0597.2012.00501.x Zagefka, H., Noor, M., Brown, R., de Moura, G. R., & Hopthrow, T. (2011). Donating to disaster victims: Responses to natural and humanly caused events. European Journal of Social Psychology, 41(3), 353-363. doi:10.1002/ejsp.781

192

Zagefka, H., Noor, M., Brown, R., Hopthrow, T., & Moura, G. R. (2012). Eliciting donations to disaster victims: Psychological considerations. Asian Journal of Social Psychology, 15(4), 221-230. doi:10.1111/j.1467-839X.2012.01378.x Zhou, X., Wildschut, T., Sedikides, C., Shi, K., & Feng, C. (2012). Nostalgia: The Gift That Keeps on Giving. Journal of Consumer Research, 39(1), 39-50. doi:10.1086/662199

193

Appendix A: Evidence of Ethics Clearance

Ethics Clearance: 15-PSYCH-PHD-29-JH

194

Ethics Clearance: 15-PSYCH-PHD-73-JH

195

Ethics Clearance: 17-PSYCH-PHD-4-AH

196

Ethics Clearance: 17-PSYCH-PHD-79-JS

197

Appendix B: Supplementary Materials for Chapter 2

Charity Categories and Coding Procedure Nominated charities in both studies were coded by volunteer research assistants into categories using publically searchable data on the Australian Charities and Not-for-profits Commission website (www.acnc.gov.au). During the process of annual registration with the commission, charities select at least one charity sub-types from 14 potential options (see Table 20) and also specify the communities their organization serves. Each named charity was coded for all mentioned sub-types and whether or not they served communities overseas (each coded 0 = is not, or 1 = is in this category). Two sub-types (“Advancing health” and “Health promotion”) explicitly related to health and were combined to make one category (“Health”) for analyses.

Table 20. List of current Australian Charities and Not-for-Profits Commission registration sub-types

Sub-type Examples Preventing or relieving the suffering Animal protection societies, shelters of animals Advancing health Hospitals, medical research, support groups Advancing education Kindergartens, universities, scholarships Advancing social or public welfare International aid, soup kitchens, elder care, disability services Advancing religion Religious congregations, buildings, or education Advancing culture Museums, ballet, theatre Promoting reconciliation, mutual Promoting equality, restorative justice respect and tolerance between groups or individuals Promoting or protecting human Promoting rights and freedoms rights Advancing the security or safety of Neighbourhood watch, volunteer emergency services the Australian public Advancing the natural environment Protecting flora and fauna

Advancing public debate Promoting change that aids one of the other charitable

purposes

Health promotion Community healthcare, medical research, awareness raising

Public benevolent institutions Hospices, disability services, aged care

Other Any other purpose beneficial to the general public International Serves: communities overseas Note. Charities self-identify as having missions under each category. Charities can register under multiple sub-types.

198

Appendix C: Supplementary Materials for Chapter 6

In the paper above, factor scores were used to evaluate the fundraising actions in the principal analyses, allowing greater methodological robustness, parsimony, and theoretical focus. However, the influence of specific best practices is surely of relevance to fundraising practitioners and is therefore reported here. The relative frequency of individual best practices (i.e., the percentage of fundraisers who reported they had done each practice) for both the 2013 and 2014 samples are reported in Table 21. as well as the zero-order correlations between each action. Hierarchical regression analyses for both 2013 and 2014 retaining original best practice items are presented in Table 22.

199

Table 21. Frequency of individual best practice actions with means, standard deviations, and zero-order correlations. Data from 2013 is presented below and 2014 above the diagonal

Frequency Zero-order correlations 2013\2014 2013 2014 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 1 Fundraiser identification 0.67(0.35) 3.92(0.78) .06 .01 .13*** .11** .08* .12*** .04 .13*** .13*** .14*** .15*** .18*** .07* .07* .12*** 2 Emailed everyone 57% 44% .16*** .26*** .17*** .18*** .16*** .09** .03 .10** .12*** .19*** .20*** .14*** .04 .23*** .24*** 3 Email reminders 47% 43% .05 .31*** .20*** .19*** .20*** .19*** .15*** .17*** .19*** .13*** .15*** .14*** .09* .18*** .30*** 4 Asked in person 52% 56% .12*** .14*** .24*** .26*** .15*** .18*** .08* .09** .14*** .17*** .19*** .15*** .09* .14*** .16*** 5 Social media ask 58% 57% .15*** .16*** .22*** .31*** .22*** .44*** .16*** .23*** .30*** .20*** .16*** .18*** .12*** .08* .11*** 6 Thanked 86% 81% .04 .15*** .19*** .18*** .16*** .17*** .15*** .18*** .29*** .17*** .13*** .18*** .06 .09* .25*** 7 Social media post 74% 68% .11** .06 .12*** .19*** .43*** .10** .15*** .23*** .34*** .12*** .17*** .13*** .08* .07* .10** 8 Updated page 39% 39% .04 .00 .11** .09** .20*** .10** .21*** .22*** .29*** .07* .07* .14*** .07* .11*** .15*** 9 Shared reasons 61% 69% .15*** .08* .19*** .10** .23*** .19*** .25*** .29*** .32*** .20*** .22*** .27*** .11*** .16*** .24*** 10 Uploaded photo 66% 68% .08* .01 .07 .06 .23*** .19*** .25*** .32*** .32*** .21*** .25*** .20*** .08* .12*** .27*** 11 Donation impact 40% 39% .20*** .17*** .14*** .15*** .16*** .17*** .07 .13*** .18*** .12*** .53*** .29*** .21*** .14*** .17*** 12 Fundraising impact 23% 30% .21*** .20*** .13*** .15*** .19*** .11** .07 .06 .19*** .14*** .49*** .32*** .17*** .15*** .20*** 13 Shared charity info 30% 36% .15*** .13*** .17*** .19*** .17*** .17*** .10** .17*** .22*** .16*** .36*** .34*** .11*** .15*** .18*** 14 Specific donation 7% 11% .02 .06 .11** .03 .11** .07 .05 .08* .12*** .07 .18*** .19*** .13*** .06 .06 15 High target 0.13(0.84) 0.08(0.79) .14*** .16*** .16*** .13*** .06 .08* .06 .02 .06 -.09* .13*** .16*** .14*** .04 .49*** 16 Raised (log) 981(2381) 1009(1603) .14*** .20*** .24*** .17*** .13*** .27*** .08* .06 .21*** .13*** .19*** .20*** .19*** .05 .49*** N2013 = 768; N2014 = 878 (Listwise) Note. Best practices coded 0/1. Target coded -1 = less than $700; 0 = default value $700, 1 = more than $700. Figures reported in the frequency column for identification, target, and raised are means (and standard deviations). Means and standard deviations reported for Raised are untransformed. *p < .05, ** p < .001, *** p < .001

200

Table 22. Hierarchical regressions with individual best practices for 2013 and 2014 2013 Dollars Raised: Log (ß) 2014 Dollars Raised: Log (ß) Step 1 Step 2 Step 3 Step 4 Step 1 Step 2 Step 3 Step 4 Fundraiser identification .14*** .10** .04 .02 .12*** .10** .05 .05 Solicitation ** *** * * Emailed everyone .10 .06 .05 .16 .07 .06 *** ** ** *** *** *** Email reminders .15 .09 .09 .24 .15 .15 * Asked in person .06 .04 .03 .07 .03 .03 Social media ask .01 .01 .00 .01 -.04 -.04 *** *** *** Thanked .21 .17 .16 - - - Signaled investment *** *** Thanked - - .12 .12 Social media post -.04 -.03 -.05 -.05 Updated page -.04 -.04 .01 .01 ** ** ** * Shared reasons .11 .10 .08 .08 ** ** *** *** Uploaded photo .11 .10 .14 .13 *** *** *** *** High target .44 .44 .40 .40 Signaled efficacy Donation impact .03 .00 Fundraising impact .04 .04 Shared charity info .02 .01 Specific donation -.02 -.01

R2 ch. .02*** .12*** .20*** .00 .01*** .12*** .21*** .00 Model R2 .14*** .34*** .34*** .13*** .35*** .35***

Note. N2013 = 768; N2014 = 878; Items are entered based on the factor they loaded highest on, including cross-loading items. Thanked loaded higher on Solicitation in 2013 and Signaled Investment in 2014. * ** *** p < .05; p < .01; p < .001

201