Adopting management philosophies: management gurus, public organizations, and the Economies of Worth

ABSTRACT Management gurus are increasingly prominent in public organizations. However, research generally evaluates guru ideas as primarily pragmatic suggestions, neglecting gurus’ value-laden ‘philosophies’. Particularly in pluralistic public organizations, we suggest that this normative content may significantly shape whether, why, and how such ideas are used. To interrogate this, we investigate a guru idea’s adoption by an English hospital using the Economies of Worth framework, which highlights values and pluralism. We find that guru ideas may be adopted because their normative content can help otherwise divided organizations forge compromises. However, these compromises bring normative constraints and promote risk-aversion in turbulent times.


Introduction
The ideas shaping public organizations have changed radically over the last 40 years. Where traditional public administration knowledge may once have dominated, public organizations today draw heavily on private-sector management texts (Ferlie et al. 2016). This knowledge is increasingly transferred through management consultants (McGivern et al. 2018;Kirkpatrick, Lonsdale, and Neogy 2016) and certified by business school MBAs (Reichard 2017). Particularly in countries heavily influenced by New Public Management (NPM; 2013), public organizations inhabit a new 'political economy of management knowledge' (Ferlie et al. 2016).
Management gurus are pivotal to this transformation. They write influential management texts and provide material for consultants and management education (Engwall 2012;Wright and Kitay 2004). Having once been primarily a private-sector phenomenon, gurus increasingly influence public bodies (Kantola and Seeck 2011). The public sector has sparked its own gurus (Osborne and Gaebler 1992), and privatesector gurus have written for public-sector audiences (Porter 2007).
Given gurus' significance, an extensive literature has developed around them, albeit generally based on private-sector experience (e.g. Carlone 2006). This often highly critical literature is motivated by puzzlement at guru ideas' popularity. Scholars individuals, usually business school academics or business leaders (ii) who spread fashionable management ideas with clear, comprehensible core concepts to (iii) international audiences through (iv) personal presentations and seminal texts (Clark, Bhatanacharoen, and Greatbatch 2012;Clark and Salaman 1998;Groß, Heusinkveld, and Clark 2015;Huczynski 1996;Wright and Kitay 2004). We define management gurus through these four features.
Much of the literature on management gurus doubts the value of their ideas. This is epitomized by Clark and Salaman (1998, 138), key writers in the field: The appeal of guru ideas is curious since . . . there are major doubts about the efficacy of the core ideas [which, some argue] only appear to be novel.
Clark and Salaman find guru ideas' 'appeal . . . curious' because of their practical suggestions' questionable novelty and utility. Consequently, they attempt to find explanations for their popularity that do not rely on the pragmatic usefulness of the management techniques they suggest. Their paper goes on to emphasize the 'narratives' gurus offer their clients.
Clark and Salaman are in good company. With gurus' core ideas commonly doubted, gurus' 'curious' popularity has become the central problem of the field (Alvesson 1993;Kieser 1997;Micklethwait and Wooldridge 1997). Finding that guru ideas lack useful pragmatic suggestions, scholars conclude that these ideas' content cannot explain their popularity. Consequently, they turn to presentational explanations. Some highlight 'supply-side' factors like gurus' rhetorical, discursive, or aesthetic skills (Carlone 2006;Chiapello and Fairclough 2002;Clark and Greatbatch 2004;Lischinsky 2008). Others suggest 'demand-side' explanations like managers facing unmanageably complex worlds and yearning for gurus' crisp explanations of organizational life (Clark and Salaman 1996;Huczynski 1996;Jackson 1996;Kieser 1997). Traditional writing attributes gurus' popularity to their persuasiveness and managers' vulnerability to persuasion.
Yet this critical perspective is not the only view on guru ideas. Others point to these ideas helping to streamline processes and improve performance (e.g. Kannan and Tan 2005;Piercy and Rich 2009). For these scholars, managers might adopt these ideas because they offer exactly the pragmatic utility that critics find lacking (MacIntosh 2003).
We do not seek to adjudicate whether guru ideas contain useful, pragmatic suggestions. Instead, we question the framing that critics and supporters share: both treat guru ideas primarily as pragmatic suggestions, and thus evaluate their novelty, utility, and applicability. Praise of guru ideas tends to focus on their practical usefulness; accounts of their appeal as fundamentally presentational are often motivated by negative evaluations of their practical value.
However, gurus offer more than simply practical suggestions. Their philosophies suggest not only how managers might achieve their goals but also what those goals should be (Pettersen 2009;Theobald 1997). For instance, this paper's case centres on an application of Harvard Business School (HBS) professor Michael Porter and colleagues' 'value-based healthcare' (VBHC; Porter and Lee 2013;Porter and Teisberg 2006): this guru idea explicitly advocates making 'value' (outcomes per dollar spent) one's only goal. Yet despite scholars' awareness that guru ideas are normatively charged (Davies 2014;Salaman 2002), research into their adoption treats them as wellpresented practical suggestions rather than normative interventions.

Gurus and normative ideas
Wider literature suggests that the normativity of guru ideas may be relevant to their adoption by public bodies. Observers of guru ideas' political uses highlight their normativity (Davies 2014). Kantola and Seeck (2011) describe Finnish politicians' evolving use of Porter's 'national competitiveness' idea: having initially deployed this idea practically to help move out of recession, they subsequently used it to normatively legitimate their agenda among an elite 'clan' while also delegitimizing civil society opposition. Admittedly, this is a relatively brief, high-level illustration, and one in which guru ideas are initially adopted for practical reasons. However, it highlights that gurus' normative propositions may have their uses.
Normative uses may be particularly pertinent within public organizations, which often have multiple goals that staff may prioritize differently (Pandey 2010). For instance, public bodies frequently contain professionals whose normative outlooks differ from managers' (Freidson 2001). Moreover, researchers note that such organizations frequently have multipolar power structures in which both these groups hold sway, alongside political and civil society actors (Denis, Langley, and Rouleau 2007;Freidson 2001;Smith and Besharov 2019). In such morally fraught contexts, simply getting people to work together despite their normative differences is a key challenge.
In these environments, gurus' normative contentions may be vital to whether -and how -their ideas are adopted. Unlike Kantola and Seeck's (2011) political elites, who used Porter's ideas to sideline civil society, public managers in multipolar organizations are likely to need at least the acquiescence of key constituencies. Consequently, while the guru literature focuses on how gurus persuade individual managers, multipolar organizations contain various other groups who must be convinced (Clark, Bhatanacharoen, and Greatbatch 2012;Clark andSalaman 1996, 1998;Groß, Heusinkveld, and Clark 2015;Huczynski 1996;Lischinsky 2008). If an idea is normatively divisive, even the most persuaded manager may struggle to implement it (Røvik 2011).
Conversely, guru ideas may be adopted because they offer useful points of consensus to normatively divided groups. Persuasive and calibrated for wide appeal (Clark and Greatbatch 2004), guru ideas and their philosophies may appeal to multiple groups with different normative outlooks. Consider a stereotypical dilemma in a hospital department: a manager wishes to focus on cost savings, whereas a nurse-leader wishes to pursue quality improvement; with collaboration needed to get anything done, little is achieved. However, 'Lean Management' compellingly advocates the goal of eliminating waste while focusing on what the customer (patient) values. Handled correctly, Lean may become a banner under which managers and nurses could collaborate. Organizations may value guru ideas whose normative content has this effect, regardless of their pragmatic suggestions' novelty or usefulness. Circumventing normative divides that inhibit collaboration can be invaluable.
This section has highlighted that while analysts treat guru ideas as fundamentally pragmatic suggestions, guru ideas' normative dimensions may be vital to their adoption in public organizations. Divisive ideas may be avoided, but those that align people behind a shared goal may be enthusiastically adopted. This latter possibility has strong face validity for those familiar with public managers' thinking around strategic initiatives, in which fostering collaboration can be key (Harris, Dopson, and Fitzpatrick 2009). Consequently, we chose to investigate how guru ideas' normative content influences their adoption and use. For this purpose, we sought a framework suited to theorizing normative ideas' use to facilitate coordination in pluralistic organizations. Boltanski and Thévenot's (2006) EW suits this purpose well. While focusing on moral values, EW resists leaping to critique normative positions. Instead, it investigates the social role of those values and the practices surrounding them. Accordingly, EW theorizes 'public justification': the practices whereby actors make someone or something seem normatively acceptable or worthy. This is achieved by persuasively associating the object of justification with socially acceptable (normative) principles. EW is distinctive among approaches to organizational normativity in seeing normative pluralism as ubiquitous and justification as a means of coordination: whether through successful persuasion or compromise, justification can transform disagreement into joint action towards shared goals (Thévenot 2001). As Denis, Langley, and Rouleau (2007) observe, this makes EW distinctively suited to theorizing how normative ideas are used to foster coordination in pluralistic organizations.

Economies of Worth
EW is well established in organization studies. It has been used to analyse environmental controversies (Patriotta, Gond, and Schultz 2011), care services (Oldenhof, Postma, and Putters 2014), and wage determination (Pohler 2019). Pertinently, scholars have deployed it to interrogate the construction of management knowledge (Mailhot and Langley 2017) and guru ideas (Chiapello and Fairclough 2002). While demonstrating EW's promise in researching these related ideas, these studies did not interrogate how and why organizations adopted guru ideas. To help us do so, we now outline the key EW concepts we deploy: coordination, worths, compromises, and qualified objects. Boltanski and Thévenot (2006) cast people as facing perennial uncertainty. Coordination is difficult because we cannot know how others will respond to our actions. However, as we repeatedly experience similar situations, we gradually build expectations about others' likely behaviour therein. These progressively solidify into shared 'conventions': expectations about contextually appropriate behaviour in such situations.

Coordination
Workplace conventions may be broad, formal, and far-reaching, like an organization's strategic plan and its required behaviours; or quotidian, like an office's expectation that those making tea should offer others a cup. Both are shared understandings of a situation and how one should behave. They help us anticipate one another's behaviour and coordinate.

Worths
EW's public justification involves the broadest, farthest-reaching conventions: those describing highly abstract normative goods, or 'worths'. Rather than being limited to any one community of shared norms, they represent socially accepted goods. This makes them particularly useful when coordinating with people with whom we are less acquainted.
For instance, if two departments are merged, managers may encounter one another's contrasting approaches. Such differences might be overcome by agreeing to pursue 'efficiency' (the 'industrial' worth) and choose the more efficient option when practices differ. Associating one's ideas with a socially acceptable (normative) principle broadens the audience to which one's argument can appeal.
EW observes that multiple principles constitute socially acceptable worths and can, therefore, be invoked to justify something. When making a strategic proposal, one might appeal to the industrial worth, but one might equally cite the 'domestic' worthwhich values tradition and status -by noting that the proposal represents a return to the organization's roots. While invoking a socially acceptable worth may broaden an argument's appeal, one must also navigate the plural worths with which a situation can be evaluated. Table 1 outlines the six worths that Boltanski and Thévenot (2006) identify.
The co-existence of plural worths means that people can hold differing views about which should be used. Where people encounter mutually familiar situations, they often agree on which worth is situationally appropriate. Conversely, in unfamiliar situations or groups, disputes about which worth suits the situation are more likely. As such disputes can impede coordination, managers may be keen to limit them.

Compromise
Disputes may conclude with one side's victory but frequently precipitate a 'compromise'. Compromises are means of normative evaluation combining multiple worths. Although difficult to achieve (Huault and Rainelli-Weiss 2011), they can enable coordination despite normative differences (Thévenot 2002). However, they remain fragile as contradictions between conflicting worths may emerge and undermine the compromise.
Preserving compromises may depend on not examining them too closely. Because compromises embed conflicting worths, attempting to define them too precisely can make them collapse. Consider the organization dedicated to environmentally sustainable development which tries to specify how it balances environmental and industrial commitments: such efforts may reignite tensions the compromise initially suppressed; left vague, the compromise may persist.

Qualified objects
Objects can solidify compromises (Mailhot and Langley 2017) because they are central to justification. Successfully justifying something in terms of a given worth requires evidence. If advocating a proposal's industrial efficiency, one might provide time-andmotion study data highlighting inefficiencies to be corrected. Conversely, one might use archival photographs to demonstrate a proposal's domestic worth by showing it represents a return to the organization's roots. Socially accepted measurement devices are known as 'qualified objects'.
With repeated use, qualified objects become associated with the worth they measure, and serve as both sources of evidence and signals of the nature of the situation. They thus shape our sense of which worth is situationally appropriate and likely to persuade others (Boltanski 2011). For instance, a factory filled with time-and-motion study reports feels like a setting befitting the industrial worth. These associative clusters of objects and worths are known as 'worlds' (see Table 1).
Appropriately designed objects help solidify compromises (Daigle and Rouleau 2010;Demers and Gond 2019). Objects like 'balanced scorecards' measure two or more worths, facilitating compromises between them. Used repeatedly, they make the situation feel like one which befits neither worth individually but both together, making compromises more enduring.

Applying the Economies of Worth
EW's emphasis on coordination and normative worths helps to theorize guru ideas' use in pluralistic, multipolar public organizations. In particular, it helps to conceptualize these ideas' normative dimensions and link them to organizations' coordination challenges and the compromises which may resolve them. Like compromises, guru ideas are frequently deliberately vague (Alvesson 1993;Kieser 1997) and come with measurement tools to assess compliance with their principles (Kantola and Seeck 2011). EW thus seems well suited to conceptualizing how guru ideas may be used in pluralistic organizations.
This use of EW is novel. EW analysts who have discussed guru ideas focus on their discourse, not their adoption and use (Chiapello and Fairclough 2002). Yet gurus often translate into organizations the social values that EW investigates (Salaman 2002), making such analysis overdue. Accordingly, this study draws on EW to ask, 'What role do justification and compromise play in guru ideas' adoption and use?'

Case and methods
To address this question, we focus on one English hospital's implementation of VBHC. We select this case as it typifies management gurus' influence in public organizations. The UK's public sector is among those most affected by management gurus, especially in healthcare (Ferlie et al. 2016;Micklethwait and Wooldridge 1997, 336-341). As Denis, Langley, and Rouleau (2007) observe, healthcare organizations are generally pluralistic and multipolar. We study a UK hospital because it represents the type of organization underrepresented in the literature and a key sector for guru ideas.
VBHC is a highly influential idea generated by 'the guru who has most obviously involved himself in the public sector' (Micklethwait and Wooldridge 1997, 321). Originally developed for US audiences, VBHC has rapidly internationalized (e.g. Porter 2007). Porter exemplifies gurus as defined above: highly personally prominent, he has a long record of internationally spreading fashionable ideas through seminal texts (Porter 1980(Porter , 1998 and a dizzying array of personal presentations (see Porter's online HBS profile). Consequently, scholars categorize him as an exemplary 'academic guru' (Clark 2004;Huczynski 1996;Kantola and Seeck 2011). We study VBHC because it is a prominent idea, by a key public-sector guru.

Value-based health care
VBHC's central idea is that healthcare organizations should make their goal maximizing 'value' for each condition they treat. Value is defined in the 'value equation': where 'costs' include all expenditure associated with treating patients with a given condition, and 'outcomes' are those which matter to those patients. For Porter and Teisberg (2006), value is a worthy goal because it reflects patients' priorities and encompasses both quality and financial sustainability considerations. They see all other goals as unhelpful distractions.
To maximize value, VBHC encourages organizations to adopt it as their sole goal, and to measure and publicly report it for each condition they treat. Measuring outcomes and costs is intended to help organizations track their own improvements, while public reporting is predicted to spark interorganizational competition on value, prompting a 'virtuous cycle' of improvement.
Vitally, VBHC insists that value is measured not individually for specific interventions or departments but holistically across care processes, from initial consultations through to aftercare. This aims to (i) avoid encouraging supposed improvements to one step in a process that cause harmful downstream consequences and (ii) foster radical, whole-process improvements to care.
Our case study hospital was awarded £400,000 to pilot and evaluate a version of VBHC through a four-step process: (1) Identify three conditions in which to pilot VBHC, and work with staff to define 'value' for those conditions to make it measurable; (2) Use existing and new data to measure value accordingly; (3) Use these measurements to identify, implement, and evaluate improvements; and (4) Institutionalize this 'value-based management' approach for these conditions and across the hospital.
We were fascinated by this hospital adopting VBHC. England's publicly funded National Health Service (NHS) is steeped in EW's civic world (associated with rights, solidarity, and social welfare). Seen as a proud civic achievement, the NHS even has a written constitution (Department of Health 2009) which enshrines good healthcare as an inalienable right. While the value equation accommodates civic concerns for healthcare quality under 'outcomes', it also includes cost (market world) as a substantive goal. More strikingly still, the equation embeds both civic and market goals within an overall efficiency calculation characteristic of the industrial world. This made an NHS hospital adopting VBHC intriguing.

Conducting the study
Three of the four authors followed the project's first two years, reviewing 139 documents, conducting 46 observations, and interviewing 22 clinical, service management, and corporate staff. The fourth author was originally a participant but then switched to an academic career and joined the project at the theorization stage. Interviews focused on participants' opinions and experience of VBHC. Research offices designated the project as service evaluation, so formal ethical approval was unnecessary.
First-round coding focused on understanding the change processes involved in using VBHC. Drawing on Porter and Teisberg (2006) and on Pettigrew, Ferlie, and McKee's (1992) writing on strategic change, we deductively developed initial codes. During coding, we recognized the importance of the project's and funding stream's objectives, and of issues like clinical-managerial relationships, and accordingly expanded the coding scheme. The resulting overview of our material highlighted actors' use of VBHC's normative ideas.
We therefore engaged in second-round coding focused on extracts relevant to people's normative positions, reasons for adopting or opposing VBHC, and ways of measuring success. For instance, extracts describing debates about how to define value for a given condition were coded under 'editing value'. We grouped developing codes into a hierarchy, using key EW terms as sensitizing concepts. Thus, 'editing value' and other codes related to shaping agreed versions of VBHC were grouped as 'locally editing compromise'. The resulting coding hierarchy described the key processes associated with VBHC's normative ideas.
In a third, theory-generating step, we scrutinized the processes that coding had identified. Initially, we explored how well EW explained them. For instance, EW expects compromises to be supported by qualified objects. We examined data coded as related to compromises and qualified objects (e.g. 'locally editing compromise', 'using compromise objects') to look for such relations. We then turned to those processes that EW left unexplained. We scrutinized their contents, contexts, and timing. For instance, 'ossification' described the project contracting; we asked when this happened, and what work was most affected.
Throughout this third step we tested explanations of processes across the whole project and within its three clinical sub-groups. Treating the latter as sub-cases allowed us to test emergent explanations through a logic of cross-case replication. By the end of this third step, we could robustly explain in EW's language the processes we had identified.

Results
We identified five processes central to VBHC's adoption and use: (i) identifying justificatory resources; (ii) deploying compromise; (iii) locally editing compromise; (iv) materializing compromise; and (v) ossification. This section reports each in turn, noting their interrelations.

Identifying justificatory resources
The hospital's interest in Porter began in 2011, when senior managers attended a seminar he delivered. Yet contrary to some critical scholars' expectations, they did not leave fervently supporting VBHC and only revisited the idea when a funding body invited bids for projects to improve clinical-managerial cooperation. Even then, they were no acolytes: VBHC was 'a great concept, but it's just a concept', one commented. However, concluding that 'we're the kind of organization that can operationalize it', they selected VBHC to frame their bid. The decision to adopt VBHC came not from commitment to its policy guidance but from belief it was a 'great concept' to persuade potential funders.
Even at this early stage, staff supported VBHC for a range of reasons. Some felt it could protect the NHS by making it financially 'sustainable', i.e. by improving its market worth. For others, the aim was emphasizing healthcare outcomes: So value, as you know, is outcomes that matter to patients divided by the cost of producing those outcomes over the full cycle of care. The outcomes are, therefore, an important component of value. So we've been measuring outcomes. (Interview, senior clinician) This senior clinician transforms value writ large into the narrower, civic-world concern of healthcare quality, part of his longstanding agenda. This does not happen by chance: it requires a carefully constructed, well-presented argument. This clinician deliberately mobilizes those elements of VBHC which help advocate his priorities.
It gradually became clear that the proposed pilot would focus on three different medical conditions: endocarditis, hepatitis, and stroke. Accordingly, clinicians and managers from these departments were engaged in the project. These groups' reasons for supporting VBHC varied. Endocarditis staff hoped the data-rich VBHC project could support their long-held industrial-world ambition of developing a specialty database, while hepatitis staff hoped to further their market-world aim of winning funding for a new service.
Despite their diverse objectives, these groups were united by their belief that VBHC could help advocate their goals. As one hepatitis interviewee explained: This interviewee highlights VBHC's ability to confer 'a jump ahead of the competition' when 'bid[ding] for a service'. It can help 'demonstrate' the value of one's existing service and provides a 'metric' for supporting one's case. While this interviewee may also believe that VBHC is a useful guide to action, here they highlight its use in 'demonstrat[ing]' the worth of one's service and proposals. In Boltanski and Thévenot's (2006) terms, they emphasize VBHC's use in crafting justifications.
Finding justification was unsurprising: it is a normal part of organizational life. However, we were struck by the sheer range of goals which elements of VBHC were considered useful in justifying. We refer to individuals identifying components of VBHC which could help justify their goals as 'identifying justificatory resources'.

Deploying compromise
VBHC's wide normative appeal was important to the project. Both the project and its funding scheme targeted building collaboration between managers and professionals, whose perceived conflicting agendas fuelled longstanding mistrust. Senior managers had adopted VBHC partly because they intuited that it might help get these groups 'on the same page' (internal project evaluation).
This argument proved compelling to not only senior managers but also their prospective funders, and the hospital's bid ultimately succeeded in 2012. Subsequently, a small project team was gradually established within the hospital's change management function. However, while this team could support coordination and analysis, successful implementation would rely on widespread collaboration. With much of the awarded funding consumed by this central team, departments had little financial incentive to participate, necessitating other ways to motivate participation.
Accordingly, staff mobilized the broad normative appeal of value. When soliciting others' participation, they cast value variously as focusing on patients ('we're all interested in improving patient care') and as promising to 'save the NHS' from rising costs. These simultaneous appeals mobilized value to appeal to audiences with very different normative outlooks.
While these efforts predominantly highlighted broadly appealing elements of VBHC, they also omitted potentially divisive ideas. Competition was systematically omitted from the hospital's version of VBHC. One senior manager explained: We are [publicly] very explicit about that -here's the three things that Porter says and [competition] is the bit that we're not really interested in at the moment. And that's very much subscribed to and supported, I think, both by [a senior internal forum], [senior manager], and indeed by [an external VBHC forum]. And it's, you know, I've definitely always had the idea that there's a really, for me, a really interesting nub of possibilities in Porter, but swallowing the whole Porter pill is not something I was particularly keen on because it felt that it was culturally, in terms of the health service as I understand it, kind of slightly different in a way that I wasn't personally committed to. (Interview, senior manager) This interviewee explicitly rejects adopting VBHC wholesale. They characterize VBHC as a 'nub of possibilities': a resource to draw from as appropriate, not a cohesive outlook one takes or leaves. Competition is omitted partly to ensure VBHC 'culturally' fits the organization, something this manager has reflected on with colleagues ('And that's very much subscribed to and supported I think by [list of people]'). Indeed, during one early meeting, a senior manager warned that overemphasizing competition could alienate potential participants.
This highlighted how managers prioritized VBHC being widely normatively acceptable. As explained above, competition is important to VBHC's theory of change: it should drive the eventual virtuous cycle of improvement. Senior managers sacrificed this supposed route to practical improvement to ensure VBHC's broad normative acceptability. This was a key early hint that VBHC may have been valued more for its normative than its pragmatic utility.
By selectively emphasizing and omitting elements of VBHC, central staff accentuated its normative appeal. Staff in the pilot departments seemed interested in and broadly receptive to the hospital's version of value. Interviewees reported seeking out further Porter materials, such as texts and online lectures. One clinician reported being 'taken very much by the idea of VBHC; it was a great idea, and I was very enthusiastic'. Managers noted that framing discussions around value helped them work 'on a level' with clinicians and fostered more fruitful conversations. Around late 2012, clinical teams were set up for each condition and became the project's core delivery mechanism.
This section has described how staff deployed VBHC's broad normative appeal to win support from those with various normative outlooks. We term this 'deploying compromise'.

Locally editing compromise
With collaboration established, early 2013 saw the project begin in earnest. Its first task was agreeing how to measure value for each condition.

Editing value
Agreeing how to measure value meant going beyond broad shared positivity about the concept to more specific consensus on what it should mean in practice. VBHC only partly specifies how this should happen. As different outcomes are important to different conditions, VBHC leaves particularly open the question of which outcomes to include. Therefore, clinical teams needed to agree internally, and with central staff, how value should be measured for their focal conditions: that is, they had to decide what was value-able.
There followed lengthy, often fractious discussions, generally focused on which outcomes to include. Even those who agreed that outcomes were pivotal differed about which to prioritize. While healthcare quality writ large is associated with the civic world, participants appealed to diverse worlds to advocate the specific outcomes they prioritized. Consequently, they mobilized various forms of evidence, including formal scientific validations of metrics (industrial world) and reports of patients' subjective experiences (inspired world). Meanwhile, central staff were keen to stay true to Porter and so disseminated his guidance on selecting outcomes. For up to 11 months, people mobilized diverse worlds, worths, and associated evidence -reflecting their varied normative priorities -to craft and challenge justifications about what should be valued.
However, each clinical team gradually agreed a local definition of value and an attendant suite of measures. Measures which some initially criticized were included, and measures which others initially considered central were excluded. Some had changed their views about which measures were important: as one clinician explained, 'my eyes were opened . . . it's not just M[orbidity] and M[ortality] . . . other things mattered'. Others had not changed their views but were convinced it was worth compromising to progress the project.
Consequently, rather than passionately espousing each measure, participants reconciled themselves to the agreed combinations of measures. Resultant ways of measuring value were not perfect realizations of any one normative standpoint, but combinations of multiple worths which together constituted mutually acceptable ways of measuring and evaluating participants' work. That is, they became compromises (Boltanski and Thévenot 2006). As one manager reflected about the whole project: There was a mutual understanding about 'we need this, we need this, and we've got a shared area of concern here, and we can each agree that we're going to help out with the bits that are maybe not a direct aim but [which] we both recognize are good things, so that we can get a genuine cooperation going'. (Interview, manager) For this interviewee, 'genuine cooperation' emerged despite differing 'direct aim[s]'. Useful in justifying various objectives, value became a 'shared area of concern' without any illusion of complete consensus. Rather, VBHC incorporated enough of what participants prioritized to make it a satisfactory compromise worth pursuing.
Our observations tracked a change in the concept of value: initially a placeholder term for diverse, contradictory goals, it evolved through lengthy, conflictual deliberations into a more specific set of aims which participants agreed to treat as shared.
While people wanted different things from the project, they each relied on it to attain them. Consequently, VBHC gradually became not only a mutually appealing if somewhat vague concept but also a substantive compromise and a shared initiative. Because it involved further specifying the meaning of value for each condition, we call this process 'locally editing compromise'.

Non-editing
However, these modifications were not spread evenly across all elements of VBHC. The value equation specifies three things: (i) outcomes, (ii) costs, and (iii) how the two should be weighted. Most deliberation focused on outcomes. One senior manager cast costs as the easy part of the equation because less debate about them was needed. Equally, while staff held a workshop to determine how to weight outcomes relative to one another, no equivalent session pondered how to weight outcomes and costs.
This was not wholly surprising. Porter and Teisberg (2006) leave room for debate about which outcomes to measure, acknowledging they should vary by condition. Conversely, they tightly specify that one should measure all costs in a care process, and the value equation makes explicit how to relate outcomes and costs. 'Editing' focused on which outcomes to measure.

Materializing compromise
As clinical teams gradually agreed what value should mean for their respective conditions, the project's next phase began: using these definitions to measure and report on value. This section describes how participants created and used objects to measure value ('materializing compromise').

Reformulating measurement devices
Measuring value consumed most of the project's effort. VBHC demands measurement of value across whole care processes. This requires extensive outcome and cost data, structured around those care processes. Hospital data systems are generally structured around organizations and departments and contain limited outcomes data. Creating the desired measures entailed gathering new data, seeking datasharing agreements with other care providers, and extensively restructuring existing data. We labelled this technically complex, time-consuming effort 'reformulating measurement devices'. This work had begun by early 2013, while clinical teams were still negotiating which outcomes to measure. Consequently, staff initially focused on costs, with analysts and even clinicians -stereotypically suspicious of financial motives -devoting considerable time to refining financial data.

Using compromise objects
The data thus derived were repeatedly reviewed by clinical teams throughout the project, prompting further debate and changes ('locally editing compromise'). From October 2013, the data began to be presented on so-called 'value dashboards', which displayed the clinical teams' outcome metrics alongside costs, composite quality, and overall value scores. While managers hoped to implement electronic interfaces, these dashboards remained colourful, large-format paper reports when our observations finished.
To us, the eventual production of the dashboards seemed anticlimactic. Months of work had ostensibly produced little more than the documentation of largely preexisting data. Nor did the dashboards seem especially useful: more serviceimprovement ideas came from clinicians' earlier insights, from patient discussion groups used to identify key outcomes, or from ad hoc analyses that clinicians requested. We expected the dashboards to be neither welcomed nor sustained.
We were wrong. While project managers recognized that early dashboards were 'crude', they were surprisingly well received, even by clinicians not normally enamoured of the project. One senior clinician seemed to surprise managers with their positive response on first seeing the dashboards: 'you've done yourself an injustice in the [pessimistic] progress chart. This is very good'.
To understand these positive reactions, we had to go beyond our observation that the dashboards' data lacked novelty and consider how they were used in practice. Physical copies of the dashboards were handed to clinical teams as the culmination of months spent negotiating what value meant. Although the collated data were largely pre-existing, they documented and operationalized clinical teams' eventual agreements. As one clinician observed, 'this is verified . . . data we're collecting on the outcomes about what we're doing'. The dashboards became both symbols of those shared definitions and ways to measure value in line with them. Boltanski and Thévenot (2006) describe compromise objects as enabling simultaneous measurement of multiple worths. This can reinforce tentative compromises between worths by making their combination enduring and avoiding choices about which to measure. Accordingly, the dashboards concretized delicate agreements on the outcomes and costs to measure and enabled their simultaneous measurement. As Boltanski and Thévenot predict of compromise objects, dashboards' physical concreteness seemed welcome: they made value data 'more apparent and palpable' than before. In turn, this physicality helped 'deploy compromise' and win round those not yet participating in the project: 'when you show people . . . the dashboard, it does show people what can be achieved'.
In line with this status as compromise objects, dashboards were mobilized to firm up agreements and coordinate work. For instance, at one stage the stroke team had agreed a set of metrics to use in their definition of value but were reluctant to finalize them, while the central team were eager to sign off those measures and move on; only after the unveiling of a draft stroke dashboard were the measures signed off.
Dashboards remained important during early steps towards service improvement. Although rarely the eventual source of improvement ideas, they were used as mutually acceptable starting points in a way that would have been rare for individual metrics like cost. These dashboards supported coordination because they materialized and solidified teams' tentative compromises.

Ossification
With dashboards in place, the project seemed poised to begin its service-improvement phase in earnest, and subsequently to institutionalize 'value-based management'. Yet progress stalled. By 2014, facing a rapidly destabilizing environment, the project had turned inwards and remained focused on earlier phases of work. This section describes this 'ossification'.

Sticking to compromise groups
The project's immediate and organizational environments destabilized. Several members of the small central team departed around the same time. Meanwhile, following the organization's merger with a nearby failing hospital, intense financial pressures diverted management attention. The project seemed decreasingly likely to receive the intra-and interorganizational support needed to make major service improvements and institutionalize value-based management.
Consequently, previously widespread engagement and enthusiasm became more restricted. In mid-2014, one manager explained the reduction in efforts to promote the project internally: Lovely, a project which has got money from outside and which is doing a long-term improvement and has got academic[s] involved and has got a slow burn, it's not something that gets people's attention, is not well received, and if I'm honest I've been a bit worried that if I opened it up too much, people would say, 'you know what, they just don't care, I just don't care' and 'get rid of it, stop it and don't [do] it or just, you know, just don't do any more work on it'. So to be honest, I've kept it under the radar in order to keep it going. (Interview, manager) This manager feared that the developmental project jarred with senior managers' more immediate financial priorities, so exposing the project to their evaluation could lead to its demise. Similar pressures affected managers and senior clinicians throughout the organization. Gaining support and attention grew harder and riskier at all levels.
Consequently, participants kept the project to themselves. While remaining aligned behind the 'slow burn' pursuit of value, urgent financial priorities reduced the perceived likelihood of others sharing their enthusiasm. Intraorganizational engagement efforts thus grew patchy, while external engagement became restricted to groups explicitly interested in VBHC. Keeping the project 'under the radar' was safer but also curtailed its potential. Limiting engagement hampered service improvement and prevented widely institutionalizing value-based reporting.

Sticking to compromise objects
Yet this was not the only reason the project stalled. The unstable environment encouraged staff to not only avoid engaging with those not yet supportive of the project but also stick to the project's most consensual elements.
Even among enthusiasts, the emphasis on service improvement diminished. Instead, a striking amount of project activity remained focused on gathering and analysing data. One central manager bemoaned this situation: And it's remarkable, the tendency within those meetings for people to want to avoid getting to the 'well what do we do about it?' conversation. They just keep wanting to say, 'well that's interesting and if we had more data on that we could find out a couple of more interesting things or we could check that further'. And my interventions are, to summarize and to make them crude, are basically to say we have enough data. (Interview, central manager) Nevertheless, difficulties making service changes persisted. Initially, we wondered whether clinical teams had simply lost interest or were unwilling to make the changes to their practices that service improvement might require. Yet they still invested time in the project and changed their practices to incorporate fresh data collection. Their reluctance was more specific: they seemed unwilling to shift from assembling data dashboards to improving services. There was something about the dashboards that was acceptable in a way that service improvement was not.
This reminded us of our earlier surprise regarding the dashboards, which transcended mere reports to serve as compromise objects holding project participants together. By further developing these dashboards, participants reinforced devices which stabilized their compromises at an otherwise destabilizing time. Conversely, service-improvement efforts could have imperilled those compromises. Most service improvements likely focus on a single component of value, like costs or patient experience. Progressing to service improvement would thus risk surfacing participants' competing objectives by requiring them to choose a priority. In uncertain times, focus on reinforcing, not utilizing, compromise objects helped participants solidify, rather than risk, their compromises.

Modelling guru idea adoption
The previous section is structured around the five processes identified through secondround coding. Its narrative highlights the relationships between the processes, which EW helped us theorize. This section collates these processes and relationships in a model. Figure 1 represents our five processes as solid boxes and causal relations between them as arrows. Figure 1 illustrates that people initially 'identified justificatory resources' by seeking extra-organizational ideas (here, elements of VBHC) which could help justify their goals. This prompted 'deploying compromise', 'locally editing compromise', and 'materializing compromise'. The figure highlights their continuous, mutually reinforcing nature: compromises were continually suggested, documented, and revised.
These processes operated under apparent normative constraints. Whereas locally editing compromise involved extensively debating outcomes, other factors (e.g. cost) seemed considerably less negotiable. Figure 1 illustrates this by the thick dashed line surrounding deploying, materializing, and locally editing compromise.
Compromise development was enabled and constrained by the contexts in which it occurred. People found VBHC's justificatory resources within their intra-and extraorganizational environments. However, as their environment destabilized, people stuck to existing coalitions and compromises (ossification). Figure 1 illustrates these relations with arrows to and from the external environment.

Discussion
This paper began by highlighting a key limitation of traditional management guru literature: its neglect of gurus' normative contentions when explaining their ideas' popularity. In public organizations, where successful strategic initiatives must appeal to powerful groups with differing normative standpoints, this is a vital omission.
We addressed it by using EW to theorize the use of guru ideas. EW emphasizes the frequency of pluralistic environments and, therefore, of compromises (Denis, Ferlie, and van Gestel 2015). This helped us to see our hospital's staff as pursuing different worths and thus struggling to coordinate. VBHC was useful because its widely appealing normative philosophy won initial support, enabling local negotiation of what it should mean in practice. This helped the formation of mutually acceptable compromises, facilitating coordination.
Complementing previous arguments that guru ideas may be used to legitimate single viewpoints (Davies 2014), we highlight how these ideas can be used to craft compromises between several normative positions. Public managers value guru ideas not because they simply support managers' own views but because their broader normative acceptability can facilitate compromise and coordination. Gurus need not spellbind managers like 'witch doctors' (Clark and Salaman 1996) if they can offer normatively charged devices which facilitate valuable organizational compromises. We call such devices 'justificatory resources'.
This reframing should prompt reconsideration of how we characterize (i) managerial agency, (ii) publicness, and (iii) context in relation to guru ideas.

Managerial agency
Managers are maligned in the traditional literature on gurus. Studies are generally motivated by scholars' puzzlement about managers adopting ideas perceived to have little pragmatic merit. Their explanations tend to position such managers as cultural dupes taken in by gurus' supposed ability to serve their psychological needs through charismatic speeches (Groß, Heusinkveld, and Clark 2015) and persuasive texts (Clark and Greatbatch 2004). Managers' agency is thus minimized.
Our study suggests that this view of managers is over-socialized. However, unlike previous appeals for greater recognition of managerial agency, we resist casting managers as pursuing purely 'personal ambitions' (Collins 2019, 224). Instead, we challenge traditional accounts by highlighting that adopting managers deploy a social and moral agency to identify ideas around which to build intraorganizational compromises. Going beyond simply avoiding controversial ideas (Andersen and Røvik 2015), managers deploy detailed knowledge of colleagues' normative orientations to identify ideas with wide, active normative appeal. Where such ideas help facilitate compromise, adopting them can be wholly rational, regardless of whether they offer novel, useful, pragmatic suggestions. Managers' social and moral agency extends to not only selecting but also utilizing guru ideas. These ideas facilitate compromise because, following their initial appeal, their local meaning must be negotiated by organizational actors. This recalls NGO managers' use of strategic plans to promote integration (Harris, Dopson, and Fitzpatrick 2009, 423): they deployed deliberately 'ambitious' initial plans to prompt criticism and debate, allowing 'thorny issues [to be] aired' and working practices agreed. Vitally, these plans do not themselves resolve issues but help managers prompt processes through which that resolution may occur. Similarly, rather than guru ideas autonomously reconciling differences, their adoption and use provide opportunities for managers to foster compromise. However, unlike strategic plans, guru ideas seem to facilitate compromise by prompting not dissensus but initial apparent consensus, making them particularly useful in morally fraught organizations where more antagonistic disputes could be dangerous. Guru ideas help build compromise by enabling managers' social and moral agency.
Yet if guru ideas enable managerial agency, they also limit it. In our focal organization, actors freely negotiated the meaning of VBHC's vaguer tenets but found it harder to deviate from its more precisely defined prescriptions. This recalls two apparently contradictory characterizations attributed to guru ideas: they are said to be vague, to broaden their applicability (Alvesson 1993;Kieser 1997;Thomas 2003), and unambiguous, to increase their clarity (Clark and Greatbatch 2004;Clark and Salaman 1998). We argue that both characterizations may simultaneously be true of different components of a guru idea. This juxtaposition of vagueness and specificity facilitates locally negotiated compromises within normative constraints. Guru ideas both enable and constrain managerial agency and organizational compromises.

Publicness
This paper was motivated by a desire to better conceptualize guru ideas' role in pluralistic, multipolar public organizations. While we lack empirical studies of the phenomenon of public-sector gurus, some scholars have briefly considered whether guru ideas suit the sector. They generally cast guru ideas as too single-minded for such pluralistic, multipolar organizations (Micklethwait and Wooldridge 1997;Radin and Coffee 1993). Some argue that gurus focus excessively on a given goal, like an 'entrepreneurial ethos' (Du Gay 1993, 646). Others suggest that their focus on 'the customer' is inherently too unitary for public bodies serving multiple masters (Jung 2010;Theobald 1997;Radnor and Osborne 2012). Finally, some challenge gurus' preference for unitary priorities writ large, casting this as a vain attempt to 'depoliticize' an inevitably politicized context (Spicer 2004, 358). For such scholars, gurus' unitary ideas are unhelpful in public bodies' pluralistic, multipolar contexts.
Conversely, this study suggests that public managers may value guru ideas precisely because of those contexts. It is public bodies' pluralism and multipolarity that make coordinating between disparate staff groups difficult, and thus make compromises valuable. If guru ideas cannot truly reconcile public bodies' plural goals into a single, coherent aim, temporarily seeming to do so may facilitate otherwise elusive cooperation (Boltanski and Thévenot 2006). Although this cannot speak to whether guru ideas benefit public bodies overall, it suggests one reason for using them which is likely to be particularly compelling in public organizations.
Yet the pluralism and multipolarity that scholars associate with public bodies (Pandey 2010) are not unique to them. Private organizations can equally be pluralistic (e.g. Demers and Gond 2019) and multipolar (e.g. Daigle and Rouleau 2010). Such private organizations may also use guru ideas as justificatory resources. For instance, a professionalized but private healthcare organization may turn to VBHC to forge internal compromise. Our findings are likely relevant to certain private organizations.

Context
Third, we highlight the importance of the contexts facing guru ideas. We noted that, amid turbulence, people stuck ever-closer to established compromises (ossification). This initially surprised us: most scholars cast guru ideas are transient fads (Clark and Greatbatch 2004;Huczynski 1996;Kieser 1997), which primed us to expect radical contextual change to make VBHC seem outdated. Indeed, for the hospital's upper echelons this expectation was fulfilled: VBHC seemed to become 'old news'. Yet, surprisingly, staff closer to the ground seemed to cling to VBHC ever tighter.
Our findings suggest that this occurred because guru ideas produced valuable means of coordination after lengthy, painstaking negotiation. When turbulent times make coordination harder, people may reasonably stick with what has worked. In our case, this manifested in the project reducing external engagement and focusing internally on its most consensual elements. Projects ossify as risk-averse people protect fragile compromises threatened by turbulent times.

Conclusions
This paper's core contribution is to explain guru ideas' adoption and use by public organizations in terms of their utility in crafting intraorganizational normative compromises. This explanation precipitated three further observations. First, we highlight managers' social, moral, and strategic agency in selecting justificatory resources to build local compromises. Second, we note that contrary to traditional expectations, gurus' single-minded ideas might be particularly useful in forging compromises in pluralistic, multipolar public bodies. Third, we observe that, in turbulent times, ossification may occur, protecting compromises but hampering progress. We thus provide novel explanations of guru ideas' use in public organizations.
While our contribution focuses on management gurus, it also advances EW research. For instance, while EW is fascinated by compromise formation (e.g. Demers and Gond 2019;Patriotta, Gond, and Schultz 2011), it rarely considers how people generate the ideas around which compromises are based. While this sometimes involves identifying a pre-packaged idea, EW currently lacks the vocabulary to conceptualize these ideas' organizational roles. In this regard, we contribute the concept of 'justificatory resources'. Similarly, EW largely considers the limitations of compromises in terms of potential causes of their collapse. Conversely, we highlight the possibility of ossification, the protective stagnation of fragile compromises amid contextual turbulence (Boltanski and Thévenot 2006;Lemasson 2017).
Although this paper reports on a single management guru idea operationalized at a single site, Porter is among the most prominent academic gurus of recent decades and covers fields from healthcare to philanthropy and government. Nor has our single site prevented comparative analysis. By treating the three clinical teams as sub-cases, we interrogated the cross-case replication of our findings.
These contributions suggest three future research avenues. First, guru ideas' normative uses within organizations demand further research. To complement our observation of their use to create compromise, studies could investigate whether less conciliatory normative uses are common. Second, we do not know whether the ossification observed here is guru ideas' inevitable fate, or an exigency of the focal organization's unusual turbulence. Studies of guru ideas' use in other cases could investigate whether ossification is a broader risk. Finally, while we identify a guru idea ossifying, others highlight guru ideas' decline as they become unfashionable. Further studies could investigate the potential tension between these two forces.