Introduction, Moral Agency and Corporate Philanthropy

Moral agency requires us both to act for moral reasons and to provide an account of why those reasons are moral. On Garvin’s (1948, p. 191) account:

It must be granted that the term “moral agent” is commonly used to mean an agent capable of, if not actually engaged in, moral reflection.

This research provides the first systematic empirical enquiry to examine a hitherto theoretically based dispute about moral agency. From a practical perspective, the moral commitment of employees is an increasing concern (Valentine et al. 2014). For instance, millennials believe that their opportunity to be impactful towards moral issues are greater within organisations than anywhere else (Deloitte 2017). Researchers and pressure groups extol organisations to exercise moral agency and to develop virtue (Arjoon 2000; Sison and Fontrodona 2012; Blueprint for Better Business 2014). This enquiry sheds light on the conditions that might enable such agency to be exercised.

A number of scholars have claimed that structural changes could foster moral agency at work. Virtuous structures (Vriens et al. 2018), participatory governance (Bernacchio and Couch 2015), employee control over production (Breen 2012), employee involvement (Sinnicks 2016) and meaningful work (Moore 2017), have all been promoted on these grounds. Drawing together much of this literature, Vriens et al. (2018) argue that virtuous structures are characterised by three mutually supportive elements—they enable employees to deliberate and make decisions, they promote teleological purpose and they involve accountabilities within managerial hierarchies. Such structures would not only enable the exercise of moral agency but may also help address cynicism, distrust and passivity amongst workers. Our enquiry required both an organisation that exhibits these three features and a specific activity that offers a context for moral agency (Vriens et al. 2018). These were provided by an employee-owned business, The John Lewis Partnership through its central involvement of employees, known as partners, in its extensive corporate philanthropic activities. Corporate philanthropy is evidenced through for instance cash donations and in-kind time and as Saiia (2001, p. 58) asserts “when an organisation engages in philanthropy it is confronted with the task of allocating resources for activities not directly related to its immediate business objectives and thus must consider what is important to support”.

The John Lewis Partnership is the largest employee-owned business in the UK employing 83,000 Partners (John Lewis Partnership 2018a) operating two retail brands, John Lewis and Partners, an omni-channel retailer with 51 ‘high-end’ high street department stores and Waitrose and Partners offering quality food, responsibly sourced with 349 shops. It was voted the Best UK Retailer in 2018 (Hosie 2018) for the fifth year running. In 2018, The John Lewis Partnership had gross sales of £11.6bn, profit before tax of £103.9 m and £74 m was distributed as Partner bonus (John Lewis Partnership 2018a). In 2016/17, 3.1% of pre-tax profits (£11.5 million of cash and in-kind support) were donated to charitable causes (John Lewis Partnership 2018b), historically ahead of other ‘responsible’ retailers. The company is owned by a trust on behalf of all its employees who have a say in the running of the business through the Partnership Council and Partnership Board, two of the three governing bodies of the Partnership alongside the Chairman. Through this unique constitutional settlement (Cox 2010), The Partnership has embedded employee decision-making, teleological purpose (centred on the partners) and managerial accountabilities.

A notable feature of the Partnership’s philanthropic activities and decision-making is that employees in stores select beneficiaries themselves. If the argument for a positive relationship between employee involvement in decision-making and moral agency is sound, then we would expect to find evidence of it at the John Lewis Partnership. It thus provided an appropriate context for our research question: How does employee participation in and deliberation about corporate philanthropy evidence moral agency?

Corporate Philanthropy has long occasioned dispute because whilst some regard philanthropy as epiphenomenal of moral agency (Grant et al. 2007; Grant and Hoffmann 2011; Grant 2012; Sulck 2010) others express scepticism (Porter and Kramer 2002). In particular, the concept of strategic philanthropy, with clear business objectives, has met with sustained criticism (Brønn and Vidaver-Cohen 2009). Friedman (1970) argued from an economic perspective that philanthropy could only be justified, as a use of shareholder money, if it led to a direct benefit for the organisation. Following this line of reasoning, strategic philanthropy enables “the synergistic use of a firm’s resources to achieve both organisation and social benefits” (Thorne et al. 2003, p. 360) or more bluntly as criticised by Haley (1991) as a means to “influence corporate stakeholders and to advance managerial interests” (p. 486).

According to its critics (and see the debate in Godfrey 2005), the self-regarding intent of strategic philanthropy precludes it as an example of moral agency regardless of any beneficent outcomes. This understanding of the orientation of corporate decision-making aligns with that of the most cited contemporary virtue theorist in the business ethics literature (Ferrero and Sison 2014), Alasdair MacIntyre. On MacIntyre’s Aristotelian account, moral agency is a potentiality and virtue nothing other than its actualisation as ‘excellence in human agency’ (Lutz 2012, p. 150). Our development towards becoming such agents is highly dependent on the social structures we inhabit and the forms of practical rationality they entail (MacIntyre 1979, 1987, 1999b, 2000, 2016). For MacIntyre, it is not the self-regarding orientation of corporate life but rather the modes of thought and action required of those who pursue such corporate aims at work, the common goods of family life at home and public goods in their thinking as citizens, that corrodes their moral agency by dis-integrating it. MacIntyre’s critique of such compartmentalisation drew from empirical enquiries undertaken in the late 1970s with power company executives (MacIntyre 1979):

The framework of executive reasoning is socially defined so that certain limits are placed upon what questions may or may not be raised about it. These limitations take the form of a definition of responsibility. In his capacity of corporate executive, the manager not only has no need to, but must not take account of certain types of considerations which he might feel obliged to recognise were he acting as parent, as consumer, as citizen (MacIntyre 1979, p. 126)

His compartmentalisation thesis is one example of MacIntyre’s wider account of the intimacy of the relationship between social structures and moral agency. According to McMylor’s (1994) intellectual biography, his insistence that moral philosophy cannot be properly undertaken without an account of social practice distinguishes MacIntyre most clearly from his contemporaries. It is also one reason why business ethics researchers employing MacIntyre’s theses and arguments have so often undertaken empirical enquiries (see Beadle (2017) for summary). This paper extends that project.

The contrast between MacIntyre’s pessimistic account of the prospects for moral agency in the contemporary workplaces and the optimistic account of other virtue-based accounts such as Vriens et al. (2018) could not be starker. Both pessimists and optimists however share the contention that agency and structure are interwoven such that enquiries into agency must also be enquiries into the structures that enable or disable it. Operationalising an enquiry into moral agency, therefore, required both attentiveness to the particularities of structure and a robust analysis of agency.

In respect of the latter, we adopted MacIntyre’s (1999a) four dimensions for the attribution of moral agency for three reasons. First, it closely aligns with contemporary neo-Aristotelian accounts such as that of Alzola (2015). Second, much of the literature on moral agency at work (especially Bernacchio and Couch 2015; Breen 2012 and Moore 2017) has sought to counter MacIntyre’s pessimism regarding its prospects in contemporary organisations (see especially MacIntyre 1999b). Third, MacIntyre’s widely cited work (Ferrero and Sison 2014) has been pivotal to the renaissance of virtue ethics in business (Beadle et al. 2015). By purposively applying MacIntyre’s own schema, this enquiry provides the first empirical enquiry to examine a hitherto theoretical dispute between virtue theorists about the prospects for moral agency in organisations.

Eight partner focus groups were undertaken across two flagship John Lewis stores in Newcastle upon Tyne and Edinburgh (both UK). In total, 52 partners engaged in the focus groups, reflecting all levels of hierarchy within the stores. The transcribed data provided rich evidence of debate around their exercise of moral agency. This primary qualitative data was supported by analysis of archival data through access to the company’s local and national journals.

The paper proceeds as follows. “Literature review: disputes about the potential for moral agency in organisations” introduces the literature on moral agency in organisations and the claims made for virtuous structures. “Background and organisational context” outlines key features of the Partnership in their historical context. “Method” provides an overview of the research methods employed in this enquiry. “Findings and discussion” reports findings using MacIntyre’s (1999a) schema to evidence moral agency. In “Conclusion” we summarise the contributions of this study for theory and empirical research in moral agency.

Literature Review: Disputes About the Potential for Moral Agency in Organisations

The literature on ‘virtuous structures’ presupposes a series of agreements as to the importance of organisational context for the development of employees’ moral agency (Vriens et al. 2018; Sinnicks 2014). In virtue-based accounts, habituation and repetition are critical to the development of the virtues and this includes agents’ own conceptualisation of the moral dimensions of action. A context in which agents are mutually accountable, deliberate together about and towards worthwhile goals and act on the results are believed to enhance capacities for moral reasoning (MacIntyre 2000; Weaver 2006; McPherson 2013; Alzola 2015). On MacIntyre’s account, organisations and wider social structures are thereby understood as either generative of, or threatening to such moral communities but never neutral (MacIntyre 1999b; Brecher 2015). Shoemaker (2007, pp. 70–71) usefully summarises the communal presupposition of such accounts of moral agency:

our practices in voicing the praise and blame expressive of holding someone morally responsible, in the paradigm case, consist of an interplay between at least two agents, one who addresses a moral demand to the other via the praise and blame and the other who ostensibly hears, understands and either accepts or rejects the demand, and such an exchange is possible only for those who have the capacity to enter into a certain kind of relationship with one another. Call those who share this capacity, then, moral agents, and call the collection of agents a moral community.

Virtue based accounts of moral agency are non-reductive (Alzola 2015) and multi-dimensional (Sundararajan 2005; Martin 2007; Miller 2008; Nussbaum 2008; Banicki 2014; Bright et al. 2014; Beadle et al. 2015). MacIntyre (1990) provides a precisive account of the attribution of moral agency, consistent with other virtue ethicists (Alzola 2015) and potentially directive for empirical enquiries in particular contexts. This argues that moral agency may only be ascribed when four conditions are met. First, any claim must be able to identify counterfactuals (through which actions may be regarded as unjust, intemperate and so on). If we are to assert an agent’s virtue, we must be able to recognise the types of circumstances which would negate that attribution. Without an awareness of such conditions, our claims are spurious. Second, we must know enough to be able to establish the intent of the action so that we can distinguish actions taken in pursuit of genuine goods from others. This requires ongoing evidence of correspondence between action and intent. It is when we have both witnessed an act and have listened to the agent’s reasoning that we are able to attribute behaviour to moral commitment. Once again, a context for mutual accountability and shared deliberation is a pre-condition for the quality of relationship that such judgment requires. Third, the moral agent experiences pleasure and pain (pathos) in proportion to their objects. It is only if we can distinguish, for example, relief at the avoidance of another’s harm from a feigned response, that we can make such attributions. This requires us to look for indications of emotional response beyond language. Finally, we need a sufficient range of actions, objects and occasions through which to establish integrity and constancy. Although a single instance of infidelity, evasion or theft may convince us that an agent lacks moral agency, no single occasion provides grounds for attributing moral agency. It is only when all four conditions are met—counterfactuals, intent, pathos and range of occasions, that moral agency may be attributed. (MacIntyre 1999a; Alzola 2015).

Although it should be clear from the foregoing that stable social relationships, marked by mutual accountabilities and shared deliberation, are pre-requisite to the attribution of moral agency, the prospects of the latter within organisations has been a matter of dispute between virtue ethicists. MacIntyre (2007 [1981], pp. 27–30) argues that workers’ moral agency is critically undermined by decision-making structures of the workplace. On his account, moral agency could only be fostered where workers exercise deliberative and participative control over production, becoming accountable to one another in light of the standards of the relevant practice (MacIntyre 2016, pp. 130, 170–173; Knight 2015; Beadle and Knight 2012; Breen 2012). As early as 1967, MacIntyre argued that capitalism critically undermines moral agency and adopted Hegel’s concept of ‘the spiritual zoo’ to identify a social order in which:

the agent does his duty in his immediate sphere without asking about the context within which he acts or the wider effects of his actions. He accepts deliberately a limited vision of both his actions and of his responsibilities. His is not to reason why. (He lives in the spiritual zoo; the animals are all in separate cages) (MacIntyre 1967, p. 207).

Others, though sympathetic to MacIntyre’s ethics, argue that varieties of capitalism are differentially conducive to the development of virtues (Keat 2008; Beabout 2012). Bernacchio and Couch (2015, p. 140) argue that employee involvement in governance may provide “experiences that call for the exercise of virtue” even when work routines are Taylorised. Sinnicks (2014, 2016) and Moore (2017) argue for governance and leadership processes that fall short of workers’ control. Bernacchio (2018) argues that virtuous action in organisations is more likely in the context of naturally occurring networks of giving and receiving that develop through organisational processes. Consistent with such reasoning, Vriens et al. (2018, p. 672) integrate research from a variety of disciplines to argue that organisational structures within appropriate teleological, deliberative and social contexts “increase the possibility of exercising and developing moral character” (emphasis retained).

Background and Organisational Context

The John Lewis Partnership on which this research is based has been praised for its constitutional settlement, embedding accountability and employee involvement in decision-making (Flanders et al. 1968; Erdal 2011; Paranque and Willmott 2014; Salaman and Storey 2016). This settlement resonates with its historical roots, and importantly provides a context in which to understand its philanthropic philosophy and tradition. Spedan Lewis, (son of the eponymous John), founded the Partnership through transferring ownership of the retailer (John Lewis) to its employees. According to Cox (2010), he was likely to have been influenced by the 1891 Encyclical ‘Rerum Novarum’, which argued for a rebalancing of the returns to capital and labourFootnote 1. The transfer to employee ownership began once Spedan took control of the firm in 1929 (Cox 2010). Spedan’s reasoning had developed in 1910, centred upon his discovery that the cumulative earnings of the workforce (then 300 employees) was at the same level as that earned by his family as owners. Such a position was justified by his father’s adherence to a form of Victorian rational egoism. By contrast Cox quotes Spedan Lewis as writing:

In fact, as a whole the staff were getting just a bare living, with very little margin beyond absolute necessities and correspondingly little chance to get much fun out of life … In all such ways the management could hardly have been more ruthlessly close-fisted. Obviously such a state of affairs could not have existed unless the general conditions at the time had been more or less similar. To me all this seemed shocking (Cox 2010, p. 32).

Over the subsequent decades, the Partnership transferred ownership to a Trust holding shares on behalf of employees; profit sharing and pensions were introduced alongside an elaborate constitution requiring managers at all levels to be accountable to employees through a variety of democratic and communications mechanisms (Salaman and Storey 2016). The constitutional model for a partnership in which—to use Spedan’s own phrase; ‘Knowledge, Gain and Power’ (Cox 2010, p. 53) were shared, has changed little from its original design.

Spedan’s influence was not limited to the transfer of ownership to employees and the embedding of accountability. He was also committed to philanthropic activities benefitting employees and others but insisted that such ‘random acts of kindness’ must be undertaken ‘quietly’ (Cox 2010, p. 152). His acts of kindness were legion and the Partnership’s accounts manifest this intent at the corporate level with donations reported since 1939 (Cox 2010, p. 112). The Partnership has subsequently maintained both its philanthropic tradition and its reticence about publicising this work. A notable element of its philanthropic work involves volunteering in which partners are supported to give time to work on philanthropic activities (Grant 2012). Since 2000, the Partnership’s Golden Jubilee Trust has sponsored employees to undertake voluntary work for up to six months (whilst being paid by the Partnership); over 330,000 hours were donated to 800 charities in the first 17 years of the scheme (John Lewis Partnership 2018a, b). Partners are willing to take up such opportunities, consistent with its directors’ sentiments regarding the Partnership’s employment selection criteria:

Frankly, we don’t want the kind of people who are only going to join us if the pay rate is so huge. We don’t just want good people, we want good people, those who really do share our values (Cox 2010, p. 274, emphasis retained).

Developing this reasoning and beyond its commitment to philanthropy and the involvement of partners, other actions and awards resonate with a virtuous organisational context. For instance, The Partnership average hourly rate of base pay for non-management Partners is £9.16, which is 17%, above the National Living Wage (2019). The Partnership operates an above market reward model where benefits are higher than competitors (British Retail Consortium, 2019) including pensions, long leave and discount on goods and services. An indicator of its regard as an employer is highlighted by the Indeed.co.uk award as the Top Retailer to work for in 2018 and evidence on that recruitment forum in which one employee (amongst many) specifically remarks, “John Lewis is a great place to work. Well managed and run and staff treated well. You are a Partner in the business, therefore it is your business” (Indeed 2019).

In respect of participation in ownership and decision-making alongside the inheritance of local philanthropic traditions, the Partnership appeared to be an ideal site for an investigation of moral agency. If the arguments of Bernacchio and Couch (2015), Sinnicks (2014), Vriens et al. (2018) and Moore (2017) about the relationship between moral agency and organisational practices were sound, then we could expect the Partnership to evidence moral agency. How would this be investigated?

Method

MacIntyre’s (1999a) four-fold schema of counterfactuals, knowledge of intent, pathos and learning across a range of contexts, was employed as a set of evaluative criteria that would enable us to discern moral agency. To explore intentions, emotions and evidence required a phenomenological method (Reynolds 2013), that could enable morally agentic thinking and action to be identified through analysis of narrative exchanges about the Partnership’s corporate philanthropy (see also Alzola 2015) consistent with Garvin’s (1948) moral reflection.

To achieve the required depth and multi-dimensionality, a focus group approach was adopted as the primary data collection instrument. All participants were guaranteed anonymity to encourage them to speak freely. As part of the research process, and to take forward the focus groups, organisational and individual ethical consent were obtained. This resulted from direct discussions between the research team and the organisation and subsequently with partners in the focus groups.

MacDougall and Fudge (2001) argue that a focus group is a singular data collection method and not an accumulation of interviews (see also Rothwell et al. 2016). The structure and content may vary considerably from one focus group to the next because of the unique dynamics of group interaction. Similarly, Asbury (1995), Murphy et al. (1992) and Morgan and Spanish (1984) claim that group interaction can yield richer information than individual interviews with the same participants. Further, Cowton and Downs (2015) argue that focus groups have potential for engaging participant debate specifically with regard to research in business ethics.

Whilst participants’ behaviour may change with data collection methods (Desmond et al. 1995) and the potential of group and power dynamics leading to false consensus (Sim 1998) is a potential constraint; participant selection, planning and the researcher’s skill set can be used to obviate these dynamics. Gibbs (1997) posits that organising focus group interviews usually requires more planning than other types of interviewing and Barbour (2005) suggests time spent developing and piloting a topic guide should pay dividends in terms of its capacity to encourage discussion. This advice was heeded through the employment of a pilot in this enquiry.

Broom (2005), Mellon (1990) and Fern (1982) advise that the researcher should introduce themselves, taking up to 15 minutes to engage in a ‘warm-up’ conversation to generate rapport, establish a comfortable environment and ensure that the participants fully understand the project. The intention is to encourage the respondents to talk freely and to be honest. Kitzinger (1995) claims that the researcher should consider how he/she is dressed as formal clothes can create a barrier that may impede the interaction between researcher and participants. A number of observation visits were made to the Partnership’s stores at various times and days to enable the first-named author to select appropriate dress for the conduct of the focus groups. Following Dearnley’s (2005) advice on venue, familiar meeting rooms were chosen and refreshments made available.

Focus group discussions were framed by a set of semi-structured questions informed by relevant literature (Kitzinger 1996a, b; Race et al. 1994; Hannabuss 1996 and “Appendix1”). We conducted small-scale pilot studies using both interview and focus group methods. The results of these studies confirmed the advantages of focus groups for generating richer narratives and surfacing both agreement and dispute (Hannabuss 1996; Nicholson 2015). Participants concurred with those reported by Bristol and Fern (1996) in finding the experience more stimulating than participating in either self-administered open-ended surveys or structured group interviews with less spontaneous interaction. The focus group method was able to surface debates between partners who expressed different positions.

There is little guidance in the literature as to the number of focus groups appropriate to different types of question and population, Millward (1995) suggests saturation is likely after ten; Krueger (1994) suggests that the minimum may be three and the maximum twelve whilst Nyamathi and Shuler (1990) argue for no more than four. To gauge the extent to which participants’ understanding was shared between those participating in Corporate Philanthropy across sites, eight focus groups were undertaken across two department stores: five in Newcastle upon Tyne and three in Edinburgh. In total, 52 partners took part in the focus groups following an open invitation to store employees. To gauge patterns at different hierarchical levels, all but one of the focus groups comprised partners at discrete levels (partner, section manager partner and departmental manager partner).

Recordings yielded 67,466 usable words following transcription (some words were excluded due to interruptions, more than one participant speaking at the same time, accents rendering words unintelligible and the deletion of proper names). All participants were offered the relevant transcribed focus group discussion and were free to redact any statements if they so wished. Our analysis was a multi-stage “iterative process in which ideas were used to make sense of data” (Hammersley and Atkinson 2007, p. 158). A series of ‘interpretive repertoires’ (Potter and Wetherell 1987) reflect recurrent themes. To preserve participant anonymity, names been changed.

Transcripts were thematically analysed for the presence of the four elements required for the attribution of moral agency posited by MacIntyre (1999a): counterfactuals, knowledge of intent, pathos and learning across contexts. We moved beyond an impressionistic view of the data by specifically noting each occurrence of the four ascriptions. In doing so, we have endeavoured to show how our interpretations are “embedded in empirical material” (Reinecke et al. 2016, p. xvii).

It is important to acknowledge that the results of small-scale phenomenological enquiries cannot meet the conditions that positivists would require to assert the validity, representativeness or generalisability of findings. Our study deliberately sought to engage a purposive sample of partners involved in corporate philanthropy; participants were self-selecting and thus may exclude those who were unavailable or preferred not to discuss their involvement. None of this however is material to the method’s appropriateness to the central research question, because to elicit whether employee engagement in corporate philanthropy could occasion moral agency, the self-selection of participants was advantageous.

Findings and Discussion

To explore how employee deliberation about corporate philanthropy provides evidence for the attribution of moral agency, we first sought to understand employee involvement in philanthropic activities and related decision-making. Unexpectedly, this initial area of enquiry met with animated responses. Recent public relations’ initiatives to involve customers in determining the recipients of philanthropic donations were roundly criticised by participants across all of the focus groups. They were typified by this response:

As we’ve already said we have our four charities we can choose from. If customers get involved are they going to say ‘well actually no we want that this’ and stuff like that so I think because it’s a partner and it’s a partner giving their time then I think it should just be coming from us rather than involving customers (Lizzy).

Customer involvement was viewed as both a breach of the Partnership’s established democratic processes and as an injustice towards those who would be conducting philanthropic work. Partners valued their involvement in decision-making through both nominating and decision–making rights about the philanthropic causes to be supported. To further understand this from a partners’ perspective, participants were asked about their involvement in the nomination of projects. This is facilitated by the employment of a full-time community liaison co-ordinator (CLC) in each store to garner partner opinion. The CLC manages the process of selecting beneficiaries, creating a short-list circulated to partners before an open meeting and partner vote.

Such an explicit employee interface is in stark contrast to commonly centralised decision-making about corporate philanthropy by a member of the senior management group or within marketing or human relations departments (Brammer and Millington 2004). One-off support for particular events and causes is also routed through the CLC and results of charitable campaigns are reported through the branch newspaper, The Chronicle.

Beyond their decision-making input, evidence of partners’ involvement in philanthropic activities was extensive. Volunteering projects recalled by participants included removing Japanese knotweed from bird habitats, a school cooking project, a fashion show for a school for children with special needs, the creation of a scented garden for the blind, donations of display stock to local charities, fund-raising events and painting a care home and a flat for a newly released prisoner. At the same time, particular stores positively sought opportunities to enable their space to be used by local charities in addition to the offer of a dedicated ‘community’ room for local groups to use at no cost. Events included the sale of children’s hand-made Christmas cards and a choir of military wives and girlfriends raising funds for wounded troops. Perhaps most surprisingly one participant reminded colleagues that employees had helped clean up after a competitor’s department store experienced a flood. Other activities employed participants’ core competencies. These included assisting a local hospital shop and a homeless charity shop with merchandising, training a chef working in a soup kitchen for the homeless and teaching charity shop workers about window-display. One participant, Stuart, recalled specific training in customer relations:

One of our managers on the shop floor has been involved with [name of charity – ed.] as well as kind of speaking to all their volunteers about what our attitude is to customers, and how they then take that to their customers.

Whilst, such types of philanthropy might well be undertaken by any commercial organisation, this initial evidence suggests a strong social contracting (Slack et al. 2015) of the partners with philanthropic acts and the underpinning decision-making processes. The focus group method additionally encouraged participants to engage in the kind of substantive moral deliberation (Garvin 1948) that develops moral agency (MacIntyre 2000).

One debate recurred across all eight focus groups; this concerned the intent condition and specifically whether the Partnership should publicise philanthropic activity. Employee support for the visibility of corporate philanthropy (Cowton 1987; Robertson and Nicholson 1996; St Clair and Tschirhart 2002) and scepticism towards its intent (Porter and Kramer 2002; Brønn and Vidaver-Cohen 2009) have both been reported in the literature. This research goes beyond previous studies in reporting animated moral argument of the kind noted by Shoemaker (2007), as indicative of moral community. Three distinct positions about the appropriateness of corporate self-regarding intent recurred across focus groups.

A first position was taken by those who argued that the Partnership should promote its philanthropic activity in support of, even if not animated by, the reputational advantages this brings. Support for such visibility involved three related themes, arguments from justice, arguments about commercial consequences and arguments about potential beneficiaries. The first theme was the perceived injustice that the Partnership experiences in failing to be recognised for good work. The second theme highlighted customers’ enhanced expectations of businesses’ philanthropy and argued that the increasing visibility of competitors’ philanthropic activities damaged the Partnership’s reputation relative to competitors. The third theme argued that potential beneficiaries may not approach the Partnership for support as they may be unaware of the extent of the Partnership’s giving. All of these themes were predicated on the importance of reputation. Whilst none argued a rational egoist position that philanthropy should be strategic, many argued that impact on the organisation should operate as a legitimate consideration.

For instance, Hannah, drawing on reputation, with particular regard to customers, argued as follows:

I think we could do more in terms of shouting about what we do and getting more back from press and PR….I think customers and other agencies and organisations would actually benefit from knowing about what we do…. I think it’s really important to help us upkeep our trading policy and it helps customers believe in us more that we are a trader of distinction…you know above and beyond the norm. I’m passionate about it and I’m pleased that we do it.

The strands of justice and reputational standing were commonly voiced in this group. For instance, Brenda highlighted the injustice of this lack of recognition:

I think that when you see other companies; I shan’t mention names but you know quite near companies to us that do have big signs up, it does make us think well our customers will think -ooh they’re good but John Lewis aren’t.

These arguments are not so much advocating traditional strategic philanthropy as legitimising a defensive response to the strategic philanthropy of others. But even this step in the direction of visibility was too much for others. A second position that moral agency is betrayed by such reputational intent (Mescon and Tilson 1987). They pursued a single argument—namely philanthropy should be exclusively other-regarding. Once any form of self-regarding consideration was entered into, this undermined the intent proper to philanthropy and the founding principle of ‘giving quietly’. For partners arguing this position, such as Betty (see also Table 2); the only question at issue was whether reputational effects should be the subject of any consideration at all. On this account, philanthropy precluded the legitimacy of any strategic interest; Chloe succinctly argued: “I think that’d be the wrong reason for doing it as well.”

The following exchange from one of the focus groups highlights this tension between partners and their respective positioning regarding the external visibility of philanthropy:

Susan: “I sometimes think that we should shout about it because when you go into the likes of [name of retailer] it’s really in your face erm and I also I kind of think why aren’t we doing that you know do people really know what we do and I don’t think;”

Arthur: “But then you’ve got to question why - why do they- are [name of retailer] doing it for the right reason or are they just doing it for something to shout about there’s an awful lot of businesses do it because they want to look good from a CSR point of view;”

Susan: “Yeah there’s plenty recognition;”

Arthur: “When actually we do it because we want to do it and actually if no-one else knows it doesn’t matter.”

Arthur (and others who argued similarly) held to a strict distinction between self-regarding and other-regarding action. Moral agents must act ‘for the right reasons’ and this condition is critically undermined by those who ‘boast’ about philanthropic achievement. On this account, strategic philanthropy is oxymoronic as only other-regarding action counts as philanthropic. By contrast, those who supported publicising the Partnership’s philanthropic work did not regard the object of intention as the only arbiter of moral agency. They held that extending the range of beneficiaries and securing just recognition were proper objects of moral action.

A third position comprised those who believed there to be a happy co-incidence. They held that commercial interests were legitimate and served by the company’s traditional reticence. Such an argument had, incidentally, informed the company’s refusal to advertise at all from the 1920s to the 1990s (Cox 2010). Their arguments against enhancing the visibility of the Partnership’s philanthropy were subtle in acknowledging the reputational benefits and commercial advantages that arose. These included Hazel:

But there’s no getting away from that commercial benefit either and that’s what I was suggesting earlier on getting involved with the younger generation we’re introducing ourselves as a company because we don’t advertise as much. Everyone loves the Christmas ads but they’re new, whereas being involved in the community is something we’ve being doing for a very long time

Jenny, though committed to philanthropic work for ‘the right reasons’, also sought a balance with the reputational arguments:

Especially because we don’t advertise it so much it does kind of make you feel that you’re doing it for the giving reasons not because ‘look at us’ and I think sometimes to be fair I think people think that we don’t advertise it quite enough because people don’t realise necessarily that we do it. So it’s kind of like getting that right balance we’re not doing it for the advertising but I think we do need to let people know that we do do something.

Jenny’s discussion acknowledges the views of both those committed to acting ‘for the giving reasons’ and those who think that visibility should be enhanced before concluding with her own preference for the ‘right balance’. She thereby incorporates both affective (‘make you feel’) and intentional (‘not doing it for the advertising’) dimensions and inasmuch as her moral reasoning presupposes community (‘us’ an ‘we’ recur six times in this short extract), Jenny illustrates the type of reasoning within a moral community that aligns closely with Shoemaker’s (2007) understanding of moral agency. Tensions between a notion of moral agency focused on philanthropic action and a commercial environment which formed the setting for organisational action was manifested in each of these three positions. Partners contested accounts that provided evaluations of themselves, their organisation and others as acting towards or departing from what they regarded as moral purposes. As we have seen, contestation, in the form of the establishment of counterfactuals, is essential to MacIntyre’s account of moral agency; for we cannot claim moral agency without an ability to discriminate the ‘moral’. The animation of debates between those who held diverging opinions of the legitimacy of self and other-regarding action in a context of strong organisational identification itself provides evidence for the fostering of moral agency and the opportunities provided for its exercise within the Partnership.

Regardless of their disagreements, participants shared an overwhelmingly positive appraisal of the company’s philanthropic work, one that elicited pride and provoked commitment. Participants displayed a high degree of identification between partners, the Partnership and with the local community. Such identifications provided both presuppositions for the debate and rhetorical resources for argumentation. Whilst corporate moral agency remains contested in the literature (Moore 2015), participants in this enquiry regarded the Partnership as a corporate actor with reasons, interests, actions and a history (Cox 2010). From this, philanthropic activities could be understood as part of an ongoing narrative identity. For example, Chloe drew on the purposes with which the Partnership’s founder, Spedan Lewis (1885–1963) had attempted to imbue the co-owned company:

And you’re major major employer of course you’re a major retailer you’re a major business in the UK you’ve got a responsibility actually as part of the wider society …and there’s little bit of that just sits well with what our founder was trying to do really. It just feels right.

Without the strong sense of organisational identity (Albert and Whetten 1984; Dutton and Dukerich 1991), the debate as to whether philanthropy should be visible would not have engaged partners as it did. The situated nature of the moral arguments brought to bear here derive from inherited traditions. In particular, the opposition to ‘shouting about’ their achievements resonates strongly with the commitment to other-regarding moral agency that characterised the philanthropic work of the Partnership’s founder, Spedan Lewis. According to the testimony of Spedan Lewis’s Personal Assistant from the 1930s:

Spedan dispensed such kindness quietly; many [letters from beneficiaries – ed.] turn up unexpectedly and it led to a Partnership tradition of random acts of kindness dispensed quietly (Cox 2010, p. 152).

Indeed, in maintaining such an identity through quiet giving, Bill suggested that it was the very absence of external visibility that he appreciated about the philanthropic work of the Partnership nationally and in his branch:

I think the thing I like as well is that we’re not slap bang on papers we keep it quite quiet, we keep it personal.

The auditory metaphor was far more commonly invoked than other semiotic channels; whilst the claim that this evidences the cultural resonance of Spedan Lewis’s ‘quiet’ dispensing of kindness would be difficult to press, the frequency of references to tradition is undeniable. This mediation of relationships between individual and organisational agency in terms of Partnership tradition presupposes a strong organisational identity, one that provides rhetorical resources for argumentation.

By contrast, the motives of competitors’ ‘marketised philanthropy’ (Nickel and Eikenberry 2009) were often derided. The contrast with the other-regarding motives of the Partnership caused argumentative obstacles for those who would support the promotion of its philanthropic activity whilst maintaining loyalty to received tradition. At their most elegiac, participants lauded partners’ conformity to other-regarding norms inimical to their own economic interests. For instance, Susan contended that:

we’ll hear things that happen throughout the year and partners’ first reaction will be ‘well if we didn’t do that would that eat into our bonus?’ you never hear that about charities so you never hear partners going ‘well if we didn’t give that what would that mean to us?’ and I think that’s great.

Teresa echoed this observation in another group:

Profit is our bonus … so really we are giving from ourselves as well, not just the company, and we’re all happy to do that.

Susan’s ‘never’ and Teresa’s ‘all’ represent the strongest possible claims for a “local moral landscape” (Harré et al. 2009, p. 9) in which corporate philanthropy is unquestioned. If these generalizations are to be believed (and neither of these examples were contradicted by other participants), then we have grounds to claim that partners identify with the wider social ambitions that animated the creation of the Partnership. Partners’ praised their own and others’ moral agency and associated it with a notion of historically informed organisational identity.

On the virtue-based account, moral agency requires practical deliberation. The level of dispute between participants, and their own reflections, evidenced in these debates is itself emblematic of agency that would be consistent with Garvin’s (1948) account. But does this evidence exhibit all four dimensions of moral agency (MacIntyre 1999a; Alzola 2015), the framing on which this research is based? To answer this question requires us to isolate elements of text. This is necessary to illustrate the conditions that MacIntyre (2000) and Alzola (2015) require before assertions of moral agency be made. The analysis that follows can neither capture the animation of the conversations, nor estimate the moral agency exhibited by particular participants. Instead, it enables us to exemplify dimensions of moral agency and to provide the analytical rigour required to evaluate whether employee engagement in corporate philanthropy has occasioned moral agency at the John Lewis Partnership. In order to demonstrate the regularity of illustrations, different speakers are quoted in each extracted text.

First, counterfactuals of different kinds were deployed to contrast positive accounts of the Partnership against those of other businesses (both competitors and those in other sectors). The three repertoires, from 31 identified through thematic analysis, illustrate the range of counterfactual conditions identified by partners. Each example is taken from different participants in different focus groups (Table 1).

Table 1 Examples of counterfactual conditions
Full size table

In each of these cases, participants established counterfactual conditions through which to test propositions about moral agency. Competitors fail to meet counterfactual conditions around temporality (‘We’ve always done it’, ‘doing it on a regular basis’), embeddedness (‘it’s cultural you know’) and financial advantage (‘for the tax reasons’). The discriminator here is not that tests were made but rather conditions established and connected to factual claims. In the first case, Walter argues that the publicity generated and the consequent reputational advantage by another retailer demonstrates that in such instances agency was not directed towards good purposes with the philanthropic act itself at its heart. In the second example, Cal argues for constancy as a counterfactual condition and in the third case, Ali argues that tax advantages provide a motive whose absence may allegedly change competitor behaviour; once again implying that it is the act itself as the legitimate purpose. In each case the counterfactual condition is such as to undermine the claim to philanthropic intent and it is this that provides the criterion by which other organisations were judged by participants. Whilst not all participants were critical of self-interest or used counterfactual conditions to test this; those who provided credible counterfactual conditions against which to judge other companies clearly did.

The second of MacIntyre’s (1999a) dimensions of moral agency is the intent condition. The repertoires in Table 2, from 65 identified in thematic analysis, illustrate the proposition that moral agency is to be distinguished by the philanthropic intent of the kind tested by the counterfactuals recorded above. Again, the three participant examples are drawn from different focus groups.

Table 2 Examples of intent condition
Full size table

The intent condition is exemplified in notions about the depth of commitment (‘actually want to do it’), voluntariness (‘not the company pushing you’) and through informing the types of counterfactuals (‘we’re so secretive’) that we have seen fellow participants invoke when evaluating the philanthropic activities of the Partnership and its rivals. They highlight disputes about intent that were evident across the focus groups and that informed conflicting judgments about the merits of visibility and reputation enhancement. The selected repertoires in Table 2 hint at the potential tension between the quiet discharge of giving compared to external reputational benefits alongside some insight into those participants who held most firmly to the view that self-regarding intent undermined claims on behalf of both the Partnership and the individual to acting for ‘the right reasons.’ This position starkly contrasts with strategic philanthropy whose purpose in undertaking philanthropic activity to bestow reputational benefits is fundamentally self-regarding (Haley 1991; Brammer and Millington 2005; Brammer et al. 2006).

The third dimension of moral agency is the articulation of appropriate feelings (pathos), expressed in the virtue literature as feeling pleasure and pain about the right things, to the right extent, in the right way; that is by some measure of proportionality (Gagliardi 1999). The three repertoires in Table 3, from 73 in the thematic analysis, are again those of different speakers in different focus groups (Table 3).

Table 3 Expressions of pathos
Full size table

In each of these cases, participants’ stories elicited affective responses (‘lovely feeling’, ‘tearful’, ‘attached’, ‘brilliant feeling’, ‘proud’, ‘tears were in your eyes’), including 17 occasions in which participants were visibly weeping. These exemplify emotional responses to both the beneficial impact of their agency and the attachment to beneficiaries. It is not only that goods have been created and distributed but that the beneficiaries are understood as properly deserving; on occasion this estimate has involved a transformation in regard occasioned by the activity itself. The effect of pro-social giving in the context of contact with beneficiaries (e.g. Grant et al. 2007; Grant and Hoffman 2011; Grant 2012) hence receives further support in this research.

The fourth dimension of moral agency is the range and consistency of examples required for moral agency to be attributed. The three examples in Table 4, from 91 identified in the thematic analysis, are also drawn from different participants and focus groups and indicate a range and a consistency of philanthropic activity which enabled participants to attribute moral agency to particular colleagues, organisational units and to the Partnership.

Table 4 Expressions of range and consistency condition
Full size table

Participants exchanged examples from across a wide range of activities and schemes (‘catering department’, ‘partners’ time’, ‘reading programme’), some involving individuals choosing from a range of options for volunteering, others involved with colleagues supporting particular causes and others involving partners nominating causes to be supported.

This research, drawing on participant voices in the focus groups, surfaces numerous examples of employee engagement in both decision-making about and voluntary activity in support of the Partnership’s philanthropic work. Participants associated these practices with an understanding of moral agency consistent with the neo-Aristotelian conception.

Conclusion

Against a background of ideological contest and contradictory empirical claims (Raub 2017), the research reported here makes three contributions. First, it introduces a robust and theoretically grounded method for evidencing moral agency of employees within a hitherto disputed organisational context. Specifically, drawing on the multi-dimensional conceptualisation of moral agency that inheres in virtue-informed accounts, our use of MacIntyre’s (1999a) schema demonstrates how empirical studies may generate and analyse narrative material. The findings help to illustrate understandings of morally agentic arguments advanced by the partners against four dimensions.

First, we have provided evidence to demonstrate that the intellectual dimension of moral agency (Alzola 2015) may be indicated by the application of counterfactual conditions (MacIntyre 1999a), by organisational agents. We have further shown that the motivational dimension/intent condition (Alzola 2015; MacIntyre 1999a) can be evidenced by both expression of, and debate about, purpose and thirdly, that the emotional dimension (Alzola 2015) may be elicited when impact on beneficiaries is experienced by agents. In addition, the behavioural dimension (Alzola 2015) is examined through the range (contextual and temporal) of illustrations that agents provide.

Second, this paper provides empirical evidence that develops our understanding of the relationship between organisational features and the potential for moral agency situated within an organisational context. The John Lewis Partnership is a commercial organisation but provides a special case (Siggelkow 2007) that combines employee ownership, the continuing cultural influence of its philanthropic founder and a constitutional settlement that involves employees in decision-making. These features cohere with those who argue that employee involvement may enable moral agency within large-scale commercial organisations (e.g. Bernacchio and Couch 2015; Sinnicks 2014; Moore 2017; Vriens et al. 2018). Our evidence provides support for their arguments as to the positive relationship between moral agency and employee involvement. Future research may explore moral agency in the context of private, state or employee-owned contexts with different types of decision-making structure to further aid our understanding of the relationship between organisational ownership, structure and moral agency.

Our third contribution is to refute MacIntyre’s pessimism about the prospects for moral agency in commercial organisations (MacIntyre 2000, 2007, 2015). Moral agency is only activated in particular and concrete circumstances. The appeals made by contemporary partners at the John Lewis Partnership to the philanthropic ambitions of its founder exemplifies moral agency. They drew inspiration from shared inheritance; and on this occasion found in a commercial organisation, albeit one with exceptional features.

This paper has evidenced moral agency, provided examples that are at one and the same time intellectual, motivational, emotional and behavioural, and corroborated arguments as to the structural features that foster it. A half century ago, Flanders et al. (1968, p. 132) reported that senior partners at John Lewis “have committed themselves to a moral rather than a calculative relationship”; an extension of this argument to partners more widely is consistent with our findings.