Introduction

An elite donor whom I had long-hoped to interview was listed as a fellow panellist at a philanthropy conference, so I emailed to suggest we travel together on the train from London and chat en route. ‘And don’t worry’, I wrote, eager to show my familiarity with the conventions of his more luxurious world, ‘I can get a first-class ticket so we can sit together’. My cunning plan failed when he responded that he would be flying himself up in his own private plane.

This paper explores the challenges and benefits of pursuing research designs that include interviewing elite donors. As the opening (true) story suggests, the first problem is gaining access to philanthropic elites, but there are complicating factors after successfully navigating that first hurdle: developing rapport in order to get good quality data on a topic that involves both money and morals and is therefore considered a doubly private matter in many cultures, and analysing the resulting data without succumbing to the ‘ethnographic dazzle’, a term coined by the anthropologist Robin Fox to describe how easy it is for outsiders to overlook similarities because we are dazzled by striking surface differences (Fox, 1969).

Despite the triple challenges of gaining access, establishing rapport and sense-making, there are good reasons to persevere with interviews in our efforts to ‘study up’, to use Laura Nader’s phrase (1972), in order to meet both public and practitioner demands for better understanding of both the accumulation and distribution of wealth. Public concerns about the existence and impact of wealth elites, whose number and wealth levels are expanding in every region of the world (Capgemini, 2020, 9), are joined by academic analysis of wealth and its societal consequences (notably Piketty, 2014) and by ongoing media and popular interest in the lives of the wealthy. Wealthy elites are known to have a higher propensity to give and to make larger lifetime charitable gifts, than the general population (Indiana University Lilly Family School of Philanthropy, 2016), and to be far more likely to leave charitable bequests (James, 2020, 282). This prompts practitioners to seek a better understanding of the attitudes, motivations and giving practices of elite donors in order to realise their untapped philanthropic capacity (Breeze & Scaife, 2015, 586–587).

Contemporary interest in studying elite donors builds on a long-standing sociological tradition, initiated by C. Wright Mills’ study of The Power Elite in 1956, which has been revived and renewed in response to the recent dramatic rise of the super-rich (Savage, 2014, Glucksberg, 2018). Difficulties in undertaking research interviews with all types of elites, including in politics and business, are widely acknowledged (Bakkalbasioglu, 2020; Goldstein, 2002; Ma et al., 2021). Thus, when elite donors are studied, three alternative research methods dominate: surveys, secondary analysis and interviews with proxies. An example of the former is the biennial survey of High Net-Worth Philanthropy using sampling frames of wealthy US postcodes (Osili et al., 2019). Secondary analysis underlies studies such as the US Million Dollar List (milliondollarlist.org) and content analysis of letters written by signatories to the Giving Pledge (Cossu-Beaumont, 2016; Coupe & Monteiro, 2016). The third type of common approach, studying people such as foundation staff, wealth managers and philanthropy advisers, assumes that proxies can faithfully represent the views of the elites they work for, and is further complicated when ‘philanthropoids’ and philanthropists are combined within the same sample (for example, Jansons, 2015; Maclean et al., 2015; Shaw-Hardy & Taylor, 2010).

This preponderance of arms-length methods that study elite donors from a distance results in our understanding of elite philanthropy being largely impersonal, based on secondary sources that were generated for ends other than research, or mediated via proxies, resulting in very limited primary data about donors’ direct experiences (Dale & O’Connor, 2021). The need for greater implementation of interview methods is clear but, as noted above, challenges face those seeking to pursue this type of primary data collection.

In order to assist those considering using this methodology, this paper offers an overview of the interview method as it relates to elites, and a review of notable examples that utilise this research design, followed by reflections from my own experiences to demonstrate the feasibility and benefits of this method. The paper concludes with advice on three stages of designing and implementing elite donor interviews.

Brief Overview of the Interview Method

An interview can be simply defined as ‘a conversation with a purpose’ (Robson, 1993, 228). In the case of interviewing elite donors the purpose is to discern their understanding of the nature, meaning and goals of philanthropy in general and of their own philanthropic actions in particular.

Qualitative interviewing is a broad term that covers a range of interviewing styles. The two styles most commonly employed in studies of elite donors are semi-structured interviews, in which the interviewer has a set of questions from which she deviates as necessary to maintain flow and to follow up interesting lines of enquiry, and life history interviews in which the interviewer is guided by a list of topics and issues followed in the order that best elicits the relevant biography of the interviewee. In both cases the interview process is flexible because accessing the respondent’s worldview matters more than sticking slavishly to a script (Bryman, 2016, 377 & 468). To use metaphors offered by Kvale (1996) the interviewer is not ‘mining’ to unearth buried truths, rather we are ‘traveling’ with the interviewee to understand their journey and how they interpret what they see and experience. The metaphor of a journey usefully highlights that the interview method is generative: it is a process that results in the creation of new knowledge or thoughts by the interviewee as they think through their responses to questions and debates that they may not have previously considered or explored (Legard et al., 2003, 142; Silverman, 2007).

This approach to interviewing, and the set of skills and techniques required to collect good quality data using the interview method are, of course, not unique to the elite interview situation. But there are known distinct challenges of implementing this method with elites, beginning with gaining access to people who are usually time-constrained, and may live in ‘golden ghettos’ with multiple addresses and paid gatekeepers. As Goldstein notes, ‘none of these [research] skills matter if you don’t get the interview… [everything] depends on getting in the door, getting access to your subject’ (Goldstein, 2002, 669).

Assessment of the Use of the Interview Method in the Field

To highlight the issues and strategies deployed when undertaking interviews with elite donors, this section describes and reflects on four extant interview-based studies.

Between 1985 and 1988 Paul Schervish and colleagues conducted 130 ‘intensive interviews’ with elite donors in eleven major US cities (Schervish, 2005, 61–62). The eligibility criteria for inclusion was net worth of $1 million or gross annual income exceeding $100,000, but this sampling frame is described as ‘flexible’ due to ‘the perennial problem of gaining access to the wealthy’ (Schervish & Hermans, 1988, p.12); hence, ten interviewees were included despite not meeting the set criteria, largely because they were young adult children of wealthy parents due to inherit substantial sums. To recruit participants an Advisory Board comprised of well-connected people was created to provide leads, plus introductions from the study’s funder (the T. B. Murphy Charitable Trust). Personal letters of introduction requested interviews, and where successful these initial interviewees were asked for referrals to wealthy individual within their networks. This enabled a snowball or branching sampling method (see Tansey, 2007 for a discussion on the merits of non-probability sampling for elite interviews), with extra efforts to recruit under-represented interviewees including female and non-white philanthropists. Concerns that interviewees would be reluctant to discuss ‘taboo’ subjects were largely unfounded:

much to our surprise and benefit… they were quite willing to disclose the financial information we desired and generally offered more detailed and elaborate answers to our questions than we anticipated… Moreover, they tended to be disarmingly honest in recounting to us intimate details of their family and personal lives (Schervish & Hermans, 1988, 20).

Schervish and colleagues attribute the relative ease of establishing rapport as being due to the participants’ shared interest in the topic and ‘the simple therapeutic value of being closely questioned and listened to by someone who was sincerely and unjudgmentally interested in what was being said’ (ibid). The analysis involved multiple stages of coding to discern narratives about how the wealthy use money in the world, and this data is presented in numerous papers (e.g. Schervish, 1992, 2000, 2007).

A second major study involving interviews with US elite donors was also conducted in the 1980s, when Theresa Odendahl interviewed 140 millionaires in three different research projects which were combined to provide the empirical basis for her book ‘Charity Begins at Home’ (1990). Whilst the sampling frame and strategy are broadly similar to that used by Schervish—‘verifiable millionaires’ from a variety of geographic locations in the USA, predominantly white, Protestant and middle-aged or older, reached through personal contacts made during her time working at a grant-making foundation—there are some notable differences in reflections on the ease with which rapport was established and analysis was conducted. Odendahl describes wearing an ‘interview outfit’ of a subdued brown business suit, a silk blouse and a leather attaché case to fit in with the ambience of interviewees’ homes and offices where interviews took place (p.5). Unlike Schervish and colleagues, Odendahl did encounter a taboo in talking about money:

asking direct questions [about the size of their wealth] proved fruitless… it became apparent that such questions jeopardised rapport with potential informants. (Odendahl, 1990, 249).

Odendahl relied instead on ‘scant’ information in the public domain on both wealth and philanthropic gifts and added case studies of named families based on material in the public domain.

The third study in relatively quick succession of elite US donors is presented in Francie Ostrower’s book ‘Why the Wealthy Give’ (1995), based on interviews with 99 wealthy donors who live or work in the New York City area. Unlike the previously described reliance on convenience samples, Ostrower created a sampling frame from the lists of major donors to 48 of New York’s largest nonprofit organisations. Whilst this more rigorous research strategy enables replication, limitations include a narrower geographic focus, the absence of wealthy people who do not give and those who support ‘offbeat’ (non-mainstream) nonprofits, and the inability to plug gaps in under-represented donor groups through purposeful sampling. The study achieved an extraordinarily high 80% acceptance rate from those approached for interviews, which Ostrower attributes to respondents’ shared belief in the importance of philanthropy as a ‘good thing’ that is worthy of serious study, and her interviewee’s frequent involvement in fundraising (as well as being donors) and who were therefore keen to understand why other rich people give (Ostrower, 1995, 24). The promise of confidentiality and assurance that unwelcome questions (such as relating to specific sums of money) could be skipped was also deemed helpful in securing interviews.

A fourth study of elite donors took place in the following decade on another continent and generated useful insights into the elite interview process. Michael Gilding interviewed 43 of the 200 names on the Australian ‘Rich List’ of 1999 and identifies three important success factors: organisation, protocols and image management. These factors are illustrated with advice to take extra care with letters of introduction to emphasise research credentials including relevant international links to help gain assent and establish trust, and being willing to dress and behave in a way that fits the norms expected by elites and their gatekeepers (Gilding, 2010 757–758). Gilding shares an illuminating error to help others seeking to access elite donors: he explains that naming the Australian Rich List as his sampling frame was not helpful because many of those listed are unhappy about having their personal details and estimated wealth in the public domain:

One man who ‘vetted’ me by phone changed his tone when I stated that my starting point was the Rich 200. ‘I enjoy being on the rich list as much as I’d enjoy having my head pushed down the toilet!’ he declared and ended the conversation.” (Gilding, 2010, 762).

Despite this drawback of his sampling strategy, Gilding recruited over a fifth of the 200 names on the Australian rich lists and attributes his success in securing participants to the existence of two goals held by interviewees: ‘to promote their concerns to a wider audience and… to reflect upon their predicaments in confidence’ (Gilding, 2010, p.756). The suggestion that elite interviewees have both an outward-looking media sensibility and an inward-looking therapeutic need is consistent with my own experience, as discussed in the next section.

Reflections from a Decade of Studying Elite Donors in the UK

Before becoming a philanthropy scholar, I worked for a decade in fundraising and charity management roles. For most of that time (1997–2006), there was no empirical data available on the motivations and experiences of elite donors in the UK, until 2004 when Theresa Lloyd, a fundraising expert, published her book Why Rich People Give, modelled on Ostrower’s study described above. Lloyd interviewed 76 people living in the UK with income over £200,000 or high net worth (largely between £5 million and £100 million). In 2011 Theresa asked me to co-author an updated version of her study in which we revisited some of her original sample and added interviews with elite donors who had emerged in the intervening years, resulting in my involvement in 22 in-depth interviews, published in Richer Lives: Why rich people give (Breeze & Lloyd, 2013). Additionally, since 2008 I had been conducting in-depth interviews with UK-based donors who had donated at least £1 million (and often much larger sums, up to £100 million), for inclusion as case studies in the Coutts Million Pound Donor Report, which was published annually until 2017 (Coutts, 2017). I conducted a total of 27 of these interviews, three of which involved elite donors who had also been interviewed for the Richer Lives study, resulting in a total of 46 distinct elite interviews.

Access: How I Recruited Participants

I had no pre-existing personal connections in the world of wealth and philanthropy so my access to interviewees came primarily through collaboration with a better-connected colleague (Theresa Lloyd) and through introductions facilitated by an intermediary organisation, Coutts bank, to wealthy clients and contacts. Latterly, those I had interviewed also made introductions, following the ‘interviewee-as-fixer’ method described by Bakkalbasioglu (2020). I sent potential interviewees a letter of introduction printed on university headed notepaper, or an email from my university account, depending on my best guess as to which mode of communication they would prefer. This first approach began with a personalised hook (e.g. ‘I enjoyed the Financial Times article in which you were quoted’ or ‘I believe we both know X’) and included a brief overview of the purpose of the study, the topics I hoped to discuss, and an offer to meet whenever and wherever was convenient for them. As my career developed, I also met potential interviewees when attending and speaking at philanthropy sector events. And as my body of work grew, I was able to include samples of my writing on elite donors with my letter of introduction, with examples carefully chosen to include, where possible, elite donors in the networks of the person being approached. I offered to share the transcript for amends or augmentation to correct inaccuracies or add information not recalled during the interview. The combination of my credentials, interaction with people they know and trust, flexibility in terms of timing and location and my offer to let them check the data afterwards, resulted in only a handful of unsuccessful requests, always attributed to logistical incompatibility (although that is potentially a very British way of saying ‘no’). My rationale for this approach was not only driven by a desire to secure interviews, but also by a belief that sharing as much information as possible about the purpose of the study, and welcoming checks on the accuracy of transcripts, would build trust and involve elite donors as partners in the research.

Rapport: How I Built Trust with Interviewees

Securing an interview is only the first step because the free flow of information is dependent on mutual trust (May, 1997, 117), without which interviewees are unlikely to open up and share personal insights on private issues. Developing rapport was easier when this grew organically from the means of access, for example if we had met at a charity reception, or been co-panellists at a discussion on philanthropy. Establishing rapport was hardest with those who had agreed to be interviewed because of their relationship with an intermediary, a situation reminiscent of my days as a fundraiser pitching to potential major donors who sat frostily throughout having promised a friend they would see me. At a mundane level, I attempted to quickly establish rapport by ‘looking the part’ in terms of dress and presentation of self (I rarely wear jewellery or make up outside of research encounters), and ‘playing the part’, for example by not expressing outward surprise at the cost of coffee in places we met, or being too gushing about original artwork on office and home walls. I also tried to develop rapport during introductory small talk by emphasising relevant parts of my autobiography, for example when interviewing people from the north of England where I grew up. This had the added benefit of highlighting similarities in the face of status and wealth differences, pre-empting the ‘dazzle’ that can distort encounters with elites.

Rapport was also strengthened by our evident shared interest in philanthropy. I concur with Gilding that interviewees did not have any particular enthusiasm for social science or a concern to advance knowledge (Gilding, 2010, 765). Rather I noted the same motivations highlighted by Gilding—to promote their charitable concerns and enjoy a ‘therapeutic’ conversation about being philanthropic—and I also identify a third concern: a belief that participation in my research could encourage more and better philanthropy through role modelling and encouraging social norms related to charitable giving, and thus leverage greater value from their philanthropic investments. This motivation reflects the contemporary paradigm of ‘impact’ or ‘strategic’ philanthropy, in which donors seek to maximise the utility and outcomes of their donations.

Findings from my Interviews with 46 Elite UK Donors

As with other research that uses the interview method to study elite donors, my findings reveal a wide range of motivations, experiences and views on the role and purpose of philanthropy. In support of identification theory (Schervish, 2007) I find that elite UK donors support charitable beneficiaries with whom they identify as a result of personal connections, common experiences and shared membership of social networks. My findings also underline the importance of institutional connections as highlighted by Odendahl (1990) and Ostrower (1995). Additionally I find that elite donors seek self-actualisation (the satisfaction of personal development and creating a tangible legacy); are motivated by a sense of duty and responsibility to acknowledge their privilege and ‘put something back’; and hope for enriching experiences and personal fulfilment as a result of being involved with interesting nonprofit organisations and developing enjoyable relationships with charity leaders, fellow donors and beneficiaries. My interviewees also describe philanthropy as the ‘right use’ of surplus money; their preference to give in a way that complements rather than substitutes the role of government; and their understanding of philanthropy as a parenting tool to help children cope with wealth and live meaningful lives (Breeze & Lloyd, 2013, 90–101).

Discussion of Implications When Implementing an Interview Method

This final section recaps the main methodological implications within three identifiable stages involved in using the interview method with elite donors.

Recruiting interviewees

There is no centralised, accessible list of elite donors that can be used as the basis of a sampling strategy, so identifying and gaining access is the core challenge when using the interview method. Samples are typically created by consulting publicly available information on people with wealth (such as those named on rich lists) and on people who are philanthropic (such as media reports of significant donations). Once potential names are identified, the researcher makes creative use of direct or indirect contacts or makes cold calls that are preferably supported by credentials, endorsements and evidence of the importance of the project. Researchers should expect to be ‘checked out’: ‘You get in and get useful data from them if you know others that they know and respect’ (Ostrander, 1993, 12). However it is constructed, the sampling frame is inevitably compromised by triple pre-filtering: we can only interview those we have heard of, those we can reach, and those that are willing to be interviewed.

Conducting the Interview

Advice on conducting interviews benefits from the largest amount of guidance offered in research methods handbooks. Key recommendations that are applicable to elite interviews include: start with an easy, non-threatening ‘warm-up’ question; establish your credibility and knowledgeability without dominating the encounter; be structured but flexible; listen more than you speak; ask open-ended questions in a straightforward and clear way; steer where necessary to keep the conversation relevant; avoid leading questions that prompt ‘correct’ or socially desirable answers; make strategic use of nonverbal cues such as ‘mmhmm’ and expectant silences to draw out all relevant responses; convey neutrality towards the topic whilst displaying interest in the answers; make it a comfortable and ideally enjoyable experience for the interviewee (summary of Bryman, 2016, 473; Fielding, 1993, 141; Kvale, 1996; Legard et al., 2003, 141; Robson, 1993, 232; Silverman, 2007). Rapport is the key in this stage, but it can be challenging to establish ‘insiderness’ with elites. Strategies to develop rapport include being an empathic researcher, asking questions in a way that encourages interviewees to feel relaxed and cooperative, and framing the researcher and interviewee as ‘epistemic partners’ sharing a common concern and reflecting on a shared purpose (Holmes & Marcus, 2008; May, 1997, 117). Elites are used to occupy powerful positions and may seek to dominate the interview encounter and prioritise impression management during responses (Ma et al., 2021, 84). To avoid losing control over the agenda, and thus the data available for analysis, it is important to set boundaries and avoid being overly deferential (Ostrander, 1993, 26). For example, arriving first at the chosen venue and beginning by setting out the planned structure of the conversation enables the interviewer to take the lead and emphasise the purpose and shared goals inherent in the encounter. When data appear questionable due to social desirability bias, such that the interviewee is motivated to present a favourable picture of their charitable actions and motivations (Bekkers & Wiepking, 2011), the interviewer could use other sources, such as charity reports and accounts, to check the veracity of specific cited donations. The process I follow of sharing transcripts swiftly after the interview, also creates an opportunity for unintentional inaccuracies to be excised once viewed in black and white.

Analysis

As with all types of research interview studies, the analysis stage involves sense-making by coding data to impose a hierarchy that highlights the more important themes, creates categories for comparison and provides a structure for the discussion. This is the stage with the highest risk for being ‘dazzled’ by differences exhibited by elite donors, who likely come from a very different milieu to the researcher. We can strive to dodge the dazzle by keeping in mind that:

the millionaires interviewed are probably as different from each other as they are similar to the non-philanthropic rich or the population at large (Odendahl, 1990, 250).

Sustained care is required to avoid over-emphasising superficial differences between elites and non-elites, and consequently identifying unwarranted causal relationships between wealth and philanthropic behaviour. For example, every wealthy donor I have interviewed prefers to support efficient and effective nonprofits, but this desire for prudence and impact is not a quirk of being rich, rather donors of all wealth levels wish their money to be spent well.

Conducting parallel studies of non-elite donors can help avoid drawing misplaced conclusions of ‘distinct’ features in a sample with no control group, a point acknowledged by Schervish and colleagues. Despite the frequency with which their body of work is cited to substantiate claims about the distinctive philanthropic behaviours and attitudes of the wealthy, such as the concept of hyperagency (Schervish, 2000, 20) and identification theory (Schervish, 2007), these researchers offer a clear warning that, in the absence of a representative random sample of elite donors and a rigorous comparison with non-wealthy donors:

Nothing can be said about the relative prominence of certain motivations among the general population of wealthy and whether the same motivations enjoy prominence among the non-wealthy (Schervish & Hermans, 1988, 36).

In the same time period as my elite interview projects, I also conducted sixty interviews with non-wealthy donors (Breeze, 2013). This study demonstrated many similarities between everyday donors and elite donors regarding their giving decisions and motivations, notably the role of personal taste, the influence of connections, networks and autobiographical experiences, and the desire for donations to be spent well and with tangible impact.

As these three stages show, there are many factors that researchers can attend to in advance of, during, and after the interview to increase the likelihood of a successful encounter. They also underline that elite interviews are time-consuming to set up, conduct and analyse; this method relies on considerable skill to develop rapport and generate useful data; and it requires flexibility in implementation and in coping with making sense of the non-standardised data that is produced.

Conclusion

Despite growing academic interest in ‘studying up’, as well as demands from the public and practitioners for better understanding of those who can make significant financial donations, there is to date a rather limited body of research that uses the interview method to study elite donors. There are well-understood challenges in implementing this methodology including: gaining access to interviewees, developing rapport to elicit useful data and avoiding the ethnographic dazzle that results in over-emphasising superficial differences between wealthy and non-wealthy donors. A key goal of this paper is to convince future researchers that it is worth facing these challenges in order to avoid over-reliance on studying elite donors at a distance, through surveys, secondary analysis or via proxies, because remote observation leads to limited understanding of elite cultures and subjectivities (Gilding, 2010, 757).

The advice set out in this paper shows how pragmatic steps can be taken to meet the known challenges and help researchers to ‘create their own luck’ in scheduling and completing elite interviews, as Goldstein advises (2002, 671). Luck has certainly played a role in my research efforts—not long after the elite donor flew himself to that conference, we bumped into each other in London and soon arranged an enlightening interview, without the need to purchase a first-class train ticket.

Confidence in using this methodology can be heightened by awareness of the motivations of elite donors for taking part in scientific studies: their outward-looking media sensibility and inward-looking therapeutic needs (Gilding, 2010) and my additional finding that donors believe participating in research can leverage greater value from their philanthropic investments. Thus research interviews enable elite donors to meet some of their philanthropic goals, whilst simultaneously providing much-needed primary data for social scientists. A better understanding of the suitability and challenges of conducting elite interviews is therefore useful for all engaged in third sector studies.