University of Birmingham Individual responses to competing accountability pressures in hybrid organisations

Purpose – The purpose of this paper is to examine conflicting institutional demands on individual frontline employees in hybrid public sector organisations. Specifically, it examines the competing accountability pressures professional and commercial logics exerted on academics at a business school, how individual lecturers responded to such pressures, and what drove these responses. Design/methodology/approach – The paper draws on a case study of an English business school and is informed by the literatures on institutional logics and hybrid organisations. Findings – The papershowsthatthe co-existenceofprofessional and commerciallogicsatthecaseorganisation exerted competing accountability pressures on lecturers. It moreover shows that sometimes deliberately and purposefully, sometimes ad hoc or even coincidentally, lecturers drew on a wide range of responses to these conflicting pressures, including compliance, defiance, combination and compartmentalisation. Originality/value – The paper sheds light on individual level responses to competing institutional logics and associated accountability pressures, as well as on their drivers. It also highlights the drawbacks of user, customer or citizen accountability mechanisms, showing that a strong emphasis on them in knowledge- intensive public organisations can have severe dysfunctional effects.


Introduction
Over the last 40 years, the rise of neo-liberalism and associated New Public Management (NPM) reforms led to the widespread adoption of business rhetoric and practices in public sector organisations (e.g. Hood, 1995;Humphrey and Miller, 2012;Lapsley, 2008). This has transformed many public service providers into hybrid organisations which need to enact commercial logics as well as historically dominant professional and public service logics (e.g. Caperchione et al., 2017;Denis et al., 2015;Miller et al., 2008;Polzer et al., 2016).
Numerous studies have adopted an institutional logics perspective to examine the challenges such institutional complexity poses. In particular, they have focused on the responses organisations adopt when faced with competing institutional logics (e.g. Greenwood et al., 2011;Oliver, 1991;Santos, 2010, 2013a) as well as on the roles that performance measurement and management control systems play in this context (e.g. Busco et al., 2017;Carlsson-Wall et al., 2016;Dai et al., 2017;Ezzamel et al., 2012;Rautiainen et al., 2017). How individual, frontline employees of hybrid organisations manage competing institutional demands has, on the other hand, received much less attention in the institutional logics literature.
We seek to explore this issue with reference to the competing accountability pressures university lecturers face in the context of English higher education reforms. Historically, English universities, like other knowledge-intensive public sector organisations, were run according to professional logics (Deem, 2004;Deem et al., 2007) and relied primarily on professional notions of accountability (Mulgan, 2000;Romzek and Dubnick, 1987;Sinclair, 1995). Users trusted public sector professionals, by virtue of their education, their ethical codes and their sense of duty, to deliver the service in the users' best interest. More recently, public sector organisations have increasingly adopted commercial logics. NPM reforms have entailed, inter alia, repositioning users as customers, encouraging them to rate professionals and, in some cases such as English higher education, charging them high fees. These developments have given rise to customer accountability or consumer accountability pressures on public sector professionals (e.g. Agyemang, 2009;Bovens, 2005;Meijer and Schillemans, 2009).
We draw on a case study of an English business school to examine the competing accountability pressures professional and commercial logics exerted on academics, how individual lecturers responded to such pressures and what determined these responses. We show that these two logics promoted competing notions of students as learners and customers, respectively, highlight the different accountability pressures associated with these two notions and discuss the conflicts between them. Building on the work of Pache and Santos (2013b), we subsequently show that individual lecturers developed four principal responses to the competing accountability pressures associated with professional and commercial logics (i.e. compliance, defiance, combination and compartmentalisation) and suggest that their choice of response depended on the extent to which they recognise and appreciate the logics in question as well as on their standing within and beyond the organisation.
The paper offers two principal conclusions and contributions. First, we highlight the competing accountability pressures individual, frontline professionals can face in hybrid public sector organisations, and provide insights into how they respond to these pressures. Specifically, we draw on our case study to highlight the rich variety and nuances individual responses to competing professional and commercial logics can take, ranging from internalised compliance to categorical defiance, and show that responses are not always as deliberate or purposeful as suggested by Pache and Santos (2013b) but often improvised, coincidental or inconsistent.
Second, we highlight that user, customer or citizen accountability mechanisms, which have widely been portrayed as almost unambiguously good or desirable things (e.g. Goetz and Jenkins, 2001;Handley and Howell-Moroney, 2010;Meijer and Schillemans, 2009), can have significant drawbacks. We show that a strong focus on one such mechanism, namely, student satisfaction surveys, had potentially detrimental effects on learning and teaching at our case organisation. More generally, we argue that in knowledge-intensive public services like universities and hospitals, where users often cannot immediately understand the quality of the services they receive, a strong emphasis on customer, user or citizen accountability mechanisms can have severe dysfunctional effects.
The remainder of the paper is structured as follows. In the next two sections, we introduce the literature on public accountability and the theoretical concepts which inform our study. We subsequently discuss our case organisation and methodology in Section 4.
In Sections 5, 6 and 7, we respectively examine the competing accountability pressures lecturers faced at our case organisation, how they responded to them, and what drove these responses. The final section discusses the conclusions and implications of our study.

Public accountability
Historically, public accountability in the UK and many other western democracies was hierarchical in nature (e.g. Bovens, 2005;Romzek and Dubnick, 1987). Junior public servants were accountable to senior public servants, senior public servants to government ministers, who in turn were accountable to the legislative. Public accountability was historically also fairly narrow in its focus, emphasising above all the implementation of state functions in accordance with the law as well as government rules and regulations.
This strong focus on hierarchies, rules and regulations was to a certain extent relaxed in the context of knowledge-intensive public services like hospitals, schools and universities. Instead, western welfare states have historically placed great emphasis on professional accountability in such services in recognition of the central role expert knowledge played in their delivery (e.g. Mulgan, 2000;Romzek and Dubnick, 1987). According to Romzek and Dubnick (1987), professional accountability is characterised by the "placement of control over organizational activities in the hands of the employee with the expertise or special skills to get the job done" (p. 229). Under labels like "academic freedom" (Altbach, 2001;Kallio and Kallio, 2014;Kallio et al., 2015) or "clinical freedom" (Armstrong, 1977;Gebreiter, 2016), public sector professionals historically enjoyed great autonomy in determining the nature of their work. Due to the long socialisation periods they had undergone as part of their training, professionals developed a deepseated ethos as well as strong intrinsic motivation to comply with professional norms and values (cf. Kallio and Kallio, 2014;Kallio et al., 2015). As a result, they were trusted to run knowledge-intensive public services in the best interest of their users, and thus, to determine both the needs of users as well as the best way of addressing them (e.g. Clarke et al., 2007).
With the emergence of neo-liberalism and NPM ideas from the late 1970s onwards, the historical reliance on narrow, hierarchical notions of public accountability came under attack. Public services increasingly came to be criticised for their perceived inefficiency and unresponsiveness to the needs of their users (e.g. Denhardt and Denhardt, 2000). The domination of knowledge-intensive public services by professionals and professional notions of accountability similarly came under attack. Fuelled by the ideas of public choice theorists (e.g. Buchanan and Tullock, 1962), suggestions emerged that public sector professionals were rent-seeking monopolists who showed little interest in the needs of users or the wider public. Instead, it was suggested that they abused their privileged position to maximise their budgets, build "empires", pursue pet projects and run public services primarily for their own convenience (e.g. Box, 1999;Clarke et al., 2007).
Consistent with NPM's emphases on markets and competition, many governments have attempted to address the perceived deficiencies of the public sector by gradually withdrawing from the direct delivery of public services. Instead, the role of government has increasingly become restricted to appointing "quangos"[1], which buy or commission public services from a range of private or semi-public providers. This "hollowing-out" of the state (Frederickson and Frederickson, 2007) implied that the historically dominant hierarchical model of public accountability could no longer be operated. Instead, the reformed public sector adopted commercially inspired, horizontal accountability mechanisms (Bovens, 2005;Grossi and Thomasson, 2015). At the heart of these mechanisms were contracts between government-appointed quangos and various private and semi-public service providers which stipulated and assessed the performance of the latter with reference to a range of targets, indicators and measures (Bovens, 2005). These performance measures have been the subject of extensive public-sector accounting research, much of which have been critical of their effects on public services (e.g. Chang, 2006).
Whilst performance-based contracts were arguably the principal commercially inspired horizontal accountability mechanism promoted by NPM reforms, they were not the only one. Governments also started to put greater emphasis on direct accountability mechanisms between public service providers and their users. Bovens (2005) highlighted the numerous user satisfaction surveys and citizen panels that have been introduced into British public services since the 1990s as examples for this development. Clarke et al. (2007) noted that the emergence of such user accountability measures had significant implications for the relationship between the users and providers of knowledge-intensive public services. They suggested that users are increasingly positioned as the consumers or customers of public services and that they are encouraged to question the hitherto sacrosanct decisions of public sector professionals. Thus, the emergence of user, customer or citizen accountability mechanisms[2] could pose a significant challenge to the freedom and autonomy of public sector professionals as well as to related notions of professional accountability. The academic literature has welcomed the advent of user, customer or citizen accountability mechanisms in the public services, highlighting that they can increase user participation, reduce corruption and confer greater legitimacy onto the decisions of public bodies (e.g. Goetz and Jenkins, 2001;Haikio, 2012;Handley and Howell-Moroney, 2010). Their potential downsides, on the other hand, have attracted little attention by accounting or public management scholars.
Public accountability and higher education reforms Higher education was not exempt from the movement towards commercial notions of accountability as discussed in the previous sub-section. Performance measures in general, and research performance measures in particular, emerged as the most visible and debated accountability mechanism in reformed higher education systems across many different countries. In England, which is widely regarded as a pioneer in this area (e.g. Czarniawska and Genell, 2002;Kallio et al., 2015), the government introduced periodic research assessment exercises (RAEs) from 1986 onwards, which aimed to measure the scientific output of English universities. Rebranded as research excellence frameworks (REFs) from 2014 onwards, these exercises were used as a basis for allocating research funding to individual universities [3]. Many other countries, including Finland, Poland, Sweden and the Netherlands (Czarniawska and Genell, 2002;Kallio et al., 2015;ter Bogt and Scapens, 2012), similarly started to measure and manage the research performance of universities, even though the precise measures adopted varied across different jurisdictions.
The emergence of such performance measures, as well as their implications for research and researchers, have been the subject of extensive and largely critical academic debate. Early English studies suggested that performance measures contributed to the commodification of academic work, the separation of research and teaching, enabled panoptical surveillance and encouraged academics to engage in excessive self-citation (Parker and Jary, 1995;Willmott, 1995). More recent studies have argued that performance measures tended to be employed in a judgemental rather than developmental manner (ter Bogt and Scapens, 2012), that they led to stress and insecurity amongst academics (Chandler et al., 2002;Knights and Clarke, 2014), and that they resulted in the adoption of an instrumental career focus by some researchers (Clarke and Knights, 2015). A series of studies conducted in Finland moreover suggested that performance measures promoted the adoption of a competitive rather than collegial ethos at universities, that quality measures tended to be highly quantitative in nature, and that such measures had negative effects on work motivation as they undermined academic freedom and the intrinsic motivation of researchers (Kallio and Kallio, 2014;Kallio et al., 2015Kallio et al., , 2017. Finally, a range of studies argued that research performance measures together with the related emergence of journal rankings could inhibit creativity and lead to more "safe" and conformist approaches to research (e.g. Hopwood, 2007Hopwood, , 2008Neumann and Guthrie, 2002).
Whilst most of the academic literature has expressed significant concerns about the effects of performance measurement on research and researchers, a number of studies have highlighted that these effects are not all-encompassing. Henkel (2000) for example, suggested that academic values and academic identities could overcome managerialist university reforms, Kolsaker (2008) noted that despite such reforms academics could still carve out niches of autonomy, and Clegg (2008) argued that there remained some scope in academia to "create spaces for the exercise of principled personal autonomy" (p. 329).
User, customer or citizen accountability measures have been somewhat slower to emerge in higher education than research performance measures and have attracted much less attention from critical and interpretive accounting researchers. Whilst many universities across the globe operate local student satisfaction surveys, only a few countries have institutionalised such user accountability measures at the national level. England is one such country. The government has run the National Student Survey (NSS) on an annual basis since it was introduced in 2005. Every January and February, all final-year students at English universities are invited to participate in an online survey which asks them 27 questions relating to their satisfaction with various aspects of their university experience (e.g. teaching, assessment and feedback, academic support, overall satisfaction, etc.). On the basis of their responses, the government compiles student satisfaction scores for universities as well as for individual courses they teach, which are published on websites like unistats.ac.uk and form a weighty factor in university rankings [4]. Via these rankings in particular, students' assessments of their learning experience have a significant impact on universities' reputation and their ability to recruit school leavers onto their degree programmes. Thus, the student satisfaction scores collected by the NSS have become a powerful user accountability measure and given students a voice that universities need to listen to.
The emergence of the NSS occurred against the background of profound changes in the relationship between English universities and their students. Historically, the government funded teaching directly and placed strict caps on the number of students universities could recruit. Domestic students did not pay any tuition fees, they were recipients of a government funded social service. In 1999, the government decided that students should make a financial contribution to their education and introduced tuition fees of £1,000 per year for domestic undergraduates, which were increased to £3,000 in 2006. Following the publication of the Browne (2010) report, the government removed direct funding from all but the most expensive degrees (e.g. medicine) in 2012 and, to make up the financial shortfall, tuition fees were increased to £9,000 per year. The government also removed the caps on the number of undergraduates universities could recruit, which gave rise to intense competition between universities for students and the associated tuition fees income.
Together, high tuition fees, the removal of student recruitment caps and the NSS increasingly positioned students as the customers of universities. This development was reinforced by the recent Higher Education White Paper (Department for Business Innovation and Skills, 2016), which highlighted the government's determination to create a "competitive market" in higher education and set out its plans to introduce a "Teaching Excellence Framework" to support students in their choice of universities. The White Paper moreover suggested that NSS ratings, together with employability data, would be directly linked to university funding, which suggests that user, customer or citizen accountability measures could play an even more important role in the future of English higher education.

Institutional logics, hybrids and individual responses
According to Thornton (2004), institutional logics are a "set of assumptions and values, usually implicit, about how to interpret organizational reality, what constitutes appropriate behavior, and how to succeed" (p. 70). They represent an array of historically contingent, taken-for-granted beliefs and practices that help organisations and individuals make sense of the social world and to act in a socially legitimated manner (e.g. Friedland and Alford, 1991;Thornton and Ocasio, 1999;Thornton et al., 2012).
Against the background of the rise of neo-liberalism (e.g. Rose and Miller, 1992), governments, think tanks, consultants and other actors have promoted the adoption of commercial, private sector ideas and practices in many areas of social life. As a result, the boundaries between historically distinctive commercial, public sector and not-for-profit logics became increasingly blurred and hybrid organisations emerged which needed to accommodate two or more of these different logics (Battilana and Lee, 2014). Specifically, in the public sector, a succession of private sector inspired NPM reforms (e.g. Hood, 1995;Humphrey and Miller, 2012;Lapsley, 2008) have resulted in a need for public sector organisations to accommodate and enact commercial logics in addition to their traditional public service logics. How public service providers managed and responded to the often conflicting demands of these two logics became a significant research focus (e.g. Caperchione et al., 2017;Denis et al., 2015;Polzer et al., 2016).
Arguably the most influential contribution to our understanding of how organisations manage competing institutional pressures was the "typology of strategic responses" developed by Oliver (1991), which suggested that organisations can adopt acquiescence, compromise, avoidance, defiance and manipulation strategies when faced with conflicting institutional demands. Oliver's (1991) typology has been widely adopted to explore organisational responses to competing logics, including by accounting researchers (e.g. Chang, 2006;Guerreiro et al., 2012;Modell, 2001). It has also formed the starting point for more recent models such as the one developed by Pache and Santos (2010), which considered intraorganisational processes and helped us develop a more nuanced understanding of how organisations manage conflicting pressures.
Whilst these studies, alongside many others, have shed much light on organisations' or senior managers' responses to competing institutional pressures, the question as to how individual, rank-and-file organisational members deal with conflicting institutional pressures has attracted much less attention in the institutional logics literature. In order to make a step towards addressing this perceived shortcoming, Pache and Santos (2013b) have built on Oliver (1991) and Pache and Santos (2010) to develop a typology of individual level responses to competing institutional logics. This typology comprises the following five elements: (1) Ignorance, which Pache and Santos (2013b) defined as "an individual's lack of reaction vis-à-vis institutional demands" (p. 12). This lack of reaction, they suggested, is not a deliberate act of resistance against a particular institutional logic but instead simply reflects the individual's lack of awareness of its existence or significance.
(2) Compliance, which Pache and Santos (2013b) defined as "an individual's full adoption of the values, norms, and practices prescribed by a given logic" (p. 12). This response requires individuals to fully go along with one logic, even if this runs counter to the prescriptions of the other logic.
(3) Defiance, which Pache and Santos (2013b) defined as "an individual's explicit rejection of the values, norms, and practices prescribed by a given logic" (p. 13). They suggested that, unlike ignorance, defiance is a conscious response which can entail both the refusal to adhere to a particular logic but also attempts to attack and remove it.
(4) Combination, which Pache and Santos (2013b) defined as "an individual's attempt at blending some of the values, norms, and practices prescribed by a given logic" (p. 14). This response requires individuals to bring together elements of the two competing logics in some shape or form.
This response requires individuals to be conversant with both of the competing logics, and to selectively comply with or reject one of the logics depending on their context or audience.
In addition to articulating this typology of individual responses, Pache and Santos (2013b) suggested that, when faced with similar circumstances, different people may adopt different responses to competing institutional pressures. They moreover suggested that the principal driver behind the responses they adopt is their degree of adherence with each of the competing logics, where adherence denotes the extent to which individuals recognise and appreciate a particular logic. In other words, Pache and Santos (2013b) suggested that individuals are more likely to comply with logics they are familiar with and committed to, and more likely to defy those which they are unfamiliar with or reject.
In the subsequent section, we introduce the methodology adopted by our study. Next, we first highlight the competing institutional logics and accountability pressures lecturers encountered at our case organisation and then draw on Pache and Santos's (2013b) model to discuss lecturers' responses to these pressures. In the subsequent section, we offer some reflections on the drivers of different individual responses to competing institutional demands.

Methodology
This paper draws on a single case study of ABC, a large business school which forms part of a research-intensive public university in England. The school has a strong research tradition and is accredited by AACSB, AMBA and EQUIS. Like many other English business schools, ABC made some steps towards corporatisation from the 1980s onwards (cf. Humphrey and Gendron, 2015;Parker, 2011). Specifically, ABC aimed to increase its revenues by recruiting high fee paying international students and placed considerable emphasis on publications in leading international journals in response to successive RAEs and REFs. Despite such developments, ABC had in many respects retained a strong professional bureaucracy ethos in the early 2010s which entailed an emphasis on individual autonomy, consultation and collegial governance (cf. Deem et al., 2007).
Against the background of increases in tuition fees and the removal of caps on the number of students universities could recruit from 2012 onwards, ABC appointed a dean with a background in the financial services industry who emphasised that, to him, the school was a business and that he intended to run it as such. Under his leadership, ABC started to adopt commercial logics to a much greater extent. The dean positioned the student as the customer of the business school and identified an excellent performance in the NSS as crucial to the school's future ability to attract students in the emerging market for undergraduates. To this end, the school started to put a very strong emphasis on the internal satisfaction surveys students were asked to fill in at the end of each module [5]. It also introduced explicit and binding rules regarding, for example, the availability of lecture notes, the recording of lectures and the nature and number of assessments on each module in response to perceived student demands. Yet, during our case study we got the strong impression that the vast majority of ABC's lecturing staff remained committed to the historically dominant professional logic. As a result, we suggest that business school was an organisation with "high hybridity", in which both professional and commercial logics played important roles (Pache and Santos, 2013b, p. 15).
For the purposes of our study, we interviewed 14 accounting and finance lecturers of various levels of seniority at ABC. Eight of our interviewees held teaching and research appointments, whereas the remainder were employed on a teaching-only basis. The former group were predominantly career academics who had completed PhDs and had little or no recent work experience beyond the higher education sector. The teaching-only staff were more diverse in their backgrounds. Some had PhDs and/or significant experience of teaching at universities, whilst others had only recently entered higher education after careers in practice and at private accountancy tuition providers. The interviews were semi-structured (Horton et al., 2004), lasted between 48 and 99 minutes and aimed to develop a rich understanding of our interviewees experiences as well as the context in which they occurred (Ahrens and Dent, 1998). They focused inter alia on lecturers' views on higher education and their students, on the strong focus on student satisfaction adopted by ABC, and how they were affected by and responded to this emphasis on student satisfaction. The interviews were electronically recorded and transcribed.
We subjected the data thus collected to several stages of thematic analysis (Boyatzis, 1998;Braun and Clarke, 2006;Creswell, 2009). The first stage was "inductive" or "bottom-up" in nature, in the sense that we sought to explore themes that emerged from the data without, as far as possible, imposing pre-existing theoretical or analytical schemes onto it (Boyatzis, 1998;Braun and Clarke, 2006). We started the analysis by repeatedly reading the transcripts in order to get an in-depth feel for our data. Then, we manually coded the data into codes and subsequently grouped relevant codes into themes (Braun and Clarke, 2006). This process was neither linear nor straightforward but a lengthy and iterative endeavour which involved repeatedly going back and forth between transcripts, codes and themes. After much reflection and refinement, we settled on a small number of principal themes which included professional notions of teaching, students and associated accountability pressures; commercial notions of teaching, students and associated accountability pressures; the tensions and contradictions between the first two themes; and lecturers' responses to the conflicting accountability pressures they faced.
The second stage of the thematic analysis focussed specifically on how individual lecturers managed commercial and professional accountability pressures. This stage was "deductive" or theory driven (Boyatzis, 1998;Braun and Clarke, 2006) as we adopted Pache and Santos' (2013b) model of individual responses to competing institutional demands to inform our analysis. We found that this conceptually derived model provided a helpful tool for understanding and organising our data. We however also found that it could not account for the significant amounts of variation and nuance we observed within the different categories of responses identified by Pache and Santos (2013b).
We therefore conducted a third stage of thematic analysis, which focussed on getting a more detailed understanding of the various responses our interviewees adopted in response to competing institutional demands. This stage of the analysis allowed us to propose eight sub-categories of the responses identified by Pache and Santos (2013b). For example, we identified four distinctive types of compliance from our interview data (i.e. enforced compliance, internalised compliance, compliance through exhaustion and instrumental compliance). We believe that these sub-categories, all of which will be defined and explained in Section 6, allow us to develop a fuller appreciation of the diversity and richness of individual responses to conflicting institutional pressures.

Students as learners or customers? Professional and commercial logics at ABC business school
In this section, we show that both professional and commercial logics emphasised lecturers' accountability to students. We however also show that the notion of the student differed dramatically under these two logics, and that this resulted in conflicting accountability pressures on lecturers.
According to the professional logic, students were perceived as learners and the task of the lecturer was to help them develop into good citizens, critical thinkers as well as into employable and successful professionals. The overwhelming majority of the lecturers we interviewed seemed to feel a strong sense of professional accountability to the notion of the student as a learner: Well, my view on teaching is that when I am finished with a student, the student should be transformed. I am not just passing on technical knowledge. I am hoping that the students would be able to be inquisitive, to explore. That way they'll want to engage in lifelong learning, to be critical in their thinking, don't take anything at face value. To always look at the various sides, for want of a better word, to any problem or any situation. And for them to come up with their own conclusions, so they never accept whatever is put in front of them. (L1) I want a student who can be given any sort of information, any task in their future job, whatever that job is and feel confident that they can work with that information.
[…] And I look at my students and I want them to have that confidence that they can look at something and work it out rather than just push it straight back. Because I think in the world of work we don't want people who can't at least show some initiative. So I want someone who can come through, not be fazed by whatever you give them, have some good technical skills, be able to communicate findings in an intelligent and professional way and ideally have the confidence to debate and present those findings. And be professional, so they know the importance of turning up and meeting deadlines and those sorts of elements. (L12) Our interviewees suggested that these objectives could be achieved by adopting a case study based teaching approach which exposed students to complex, real-life problems and encouraged them to address these problems in a proactive and innovative manner: I like to give them tasks for which there is no yes or no answer. So I give them something where there are grey areas and then they can argue their point. I love to make up little cases, cases that relate to real-life situations, so it cannot be a yes or it cannot be a no or it can be a yes or it can be a no, depending on how it is argued. So I like to give them little situations. Plus I like to give them situations say in accounting, I will give them an accounting problem and ask them to imagine that they take on certain roles and they are to use the information there to make a decision. So I like to give them those things to do. (L1) According to the commercial logic, on the other hand, students were perceived as customers who needed to be satisfied. The notion of the student as the customer was supported by the high tuition fees students were charged and by the strong emphasis ABC placed on the results of student satisfaction surveys. The interviews we conducted suggest that students at ABC readily embraced their customer status, indicating that the teacher/student relationship had become more transactional. Many students felt that, by virtue of the high tuition fees they were paying, they were entitled to good marks, which they saw as the key to securing good graduate jobs. Our interviewees moreover highlighted that many students were very passive and expected lecturers to spoon-feed them the contents of their modules. Students tended to be most satisfied if lecturers focused on the technical aspects of accounting and finance, offered simple black and white solutions and provided mechanical step-by-step guides or recipes for reaching these solutions[6][7]: The students are so passive, they see this whole learning experience as a passive thing, they've got no proactivity whatsoever. And that to me is really disappointing because it's become a bit like a transaction. They pay their £9,000 a year and expect to just sit back and be given all this stuff and it'll come in like by osmosis. And they don't see that they have to put an effort in, it's almost like they're paying the money and they shouldn't have to do much more to come out as this perfect employable person with an accounting degree. (L4) The students want something which is, they want to be taught something or given some sort of exercise where they know exactly which are the steps so they can solve it. It is the opposite of critical thinking, they are more looking for sort of recipes if you will, every bit to get the mark or pass the class. These are the exercises students are feeling comfortable with and the best strategy to keep students happy as well because they won't complain as long as you tell them that this is how assessment works. (L14) Some teaching-only lecturers, especially those with a background in the private sector, showed qualified sympathy for the notion of the student as a customer: I think you know when I was at university the tuition fees were £1,500 a year and at the time I felt like that was a lot of money to be paying out when you're a student with no income and your debt is ramping up. So for the guys now that are faced with £9,000 a year, I think it's a tremendous amount of money when you think their starting salaries could be say £20,000. It's going to take them years, if not decades, to pay this sort of money back. So I do feel that they need to be treated like customers in a way. And you know, I mean we can still tell them off if we don't feel they've done enough work and put enough effort in you know, we're not going to spoon-feed them but at the same time I think we should be really trying to give them the best experience and the best learning facilities possible. (L2) The majority of interviewees however rejected the notion of the student as a customer as they felt that the "wants" of the customer were incompatible with the professionally determined "needs" of the student as a learner. They highlighted the youth and inexperience of the students and questioned their ability to make informed judgements about teaching quality. Our interviewees argued that fulfilling the wishes of the students would impede rather than further students' education and suggested that lecturers, by virtue of their professional knowledge and experience, were best placed to determine the teaching approach which was in the long-term best interest of the student[8]: What gives students satisfaction is not what makes them get the most value from their degree. Unlike buying a sandwich, education is something where you understand the quality say 5 or 10 years down the line.
[…] Student satisfaction and learning are pretty much uncorrelated. […] Students would be happy if you give them an easy exam or it was very clear how they are going to be assessed, or you know, you just make them happy. It's like me going to the gym and saying you are really strong, you should be an Olympic athlete. Well I'm not, I understand the praise, I will feel good about it but it wouldn't make me fitter in the long-term. I don't think students can actually evaluate what they are getting the moment they are getting it. The whole university idea is that you have someone who is an expert in the field, he can see several years down the line, so he sees what should be important in the future, so he should be determining what students should be taught and how, and what should be expected of them as a minimum. (L14) You see the pressure that we are now put under because of the students' perception of themselves as the customer rather than the learner. And that's an issue which leads to this: "You're supposed to spoon-feed me, you're here to deliver my education, I'm paying for this. Yeah?" attitude. So there is a conflict between the student as the customer and the student as the learner because the student as the customer is entitled tonot entitled but expects you to provide -X, Y and Z. But the student as a learner only develops if you don't provide X, Y and Z. You just give them the means by which they can do X, Y and Z themselves, if you see what I mean? And that's the main dilemma that we're in now. And it's a difficult one because we don't survive without the student, we don't survive without the student satisfaction but we're doing the student a disservice if we don't give them this resilience and these other skills that mean that they become independent learners. (L10) Thus, lecturers were caught between two seemingly irreconcilable logics and their associated accountability pressures. On one hand, they were personally committed to the professional logic and felt a strong professional accountability towards the students as learners, who needed to be transformed into good citizens, critical thinkers and employable professionals. Lecturers argued that higher education was different from other services and that the customer status was not applicable to students for two reasons. First, they suggested that students could not assess the quality of higher education at the time they were consuming it. Unlike most other goods and services, whose quality becomes apparent at the time of consumption, they suggested that the quality of university teaching could not be appreciated by students until years after graduation. Second, lecturers argued that a customer mentality, which values comfort, simplicity and convenience for the service user, was detrimental to the effectiveness of higher education. They believed that students needed to experience complexity and ambiguity as well as put in large amounts of effort in order for them to derive the greatest long-term benefits from their university experience. In light of these considerations, and consistent with traditional notions of professional accountability (Mulgan, 2000;Romzek and Dubnick, 1987) and academic freedom (Altbach, 2001;Kallio and Kallio, 2014;Kallio et al., 2015;), lecturers believed that they should determine what and how students were taught. They believed that, because of their professional knowledge and experience, they had a better understanding of what was in the long-term interest of the students than the students themselves.
On the other hand, amid an increasing emphasis on commercial logics at ABC, lecturers also faced strong customer accountability pressures from student surveys and the students themselves, who readily embraced the customer role. Student-customers demanded that lecturers taught predominantly technical content and adopted a spoon-feeding approach to teaching. They moreover expected to receive good grades in return for the high tuition fees they were paying rather than as a reward for hard work and diligent study. Such demands not only challenged the historic right of academics to determine what and how they were teaching (e.g. Altbach, 2001;Clarke et al., 2007), but were also incompatible with the professional accountability lecturers felt towards their students as learners. In the next section, we will explore how they responded to these conflicting professional and customer accountability pressures.

Individual responses to competing accountability pressures
We discuss how lecturers responded to the competing accountability pressures they faced at ABC with reference to Pache and Santos' (2013b) categories (i.e. compliance, defiance, combination and compartmentalisation)[9] as well as the eight more nuanced sub-categories we developed in the course of our data analysis (i.e. enforced compliance, internalised compliance, compliance through exhaustion, instrumental compliance, categorical defiance, defiance with justification, blending and translation).

Compliance
Pache and Santos (2013b) suggested that one common response is compliance, which entailed "an individual's full adoption of the values, norms and practices prescribed by a given logic" (p. 12). We found many instances of lecturers complying with commercial logics and associated customer accountability pressures, and that this type of response took on a range of different shapes and nuances, including enforced compliance, internalised compliance, compliance through exhaustion and instrumental compliance. The vast majority of the instances of compliance we encountered focused solely on the adoption of practices associated with the commercial logic. Our interviewees remained largely committed to the values and norms of the professional logic and often emphasised that they adopted practices associated with the commercial logic with significant reluctance and regret.
Enforced compliance. Many of our interviewees suggested that they gave in to accountability pressures from their "customers" when these were reinforced by official rules set by the management team of ABC. The business school had articulated rules relating to a wide range of issues including the number of assessments on each module and the recording of lectures, which reflected student demands but were often inconsistent with the professional aims of the faculty. Faced by such rules, and the largely inflexible nature in which they were enforced, lecturers were more or less coerced into complying with the commercial logic even when the practices in question were fundamentally at odds with their professional judgement and the accountability they felt towards their students as learners. For example, one lecturer reluctantly removed an assignment from one of his modules in response to the introduction of a policy which banned coursework in the final year: I like assessment, I think the more students are assessed, the better they become and the better marks they get overall. So I wouldn't have been happy when the university or the undergraduate office abolished the final year coursework completely, I was not happy, I remember. Probably that came from the student feedback overall or we thought because they were over-assessed. So I wasn't happy about that but I had to do it, I had no other option. So we had to stop assessing the students during term time and they had 100% exams instead. (L5) Internalised compliance. Compliance with the commercial logic however not only occurred when lecturers were effectively forced to do so by managerial rules. The strong emphasis on student surveys and the consumerist attitudes and behaviours students often displayed in their interactions with lecturers resulted in the emergence of a form of internalised compliance with commercial logics amongst many of our interviewees. Although they remained committed to their professional values and norms, some lecturers suggested that they now routinely complied with the commercial logic and the associated accountability pressures in their interactions with their students. One lecturer, for example, considered rerecording a lecture after he had made a small mistake, not because students had complained but because he "felt the aura of pressure that they would be unhappy": Sometimes it is not something they [the students] ask, sometimes you do it indirectly because you think they would complain about it. So, for example, last week when I was teaching the last class, the exam paper preparation class, and we were supposed to record the class, I recorded the class but I forgot to press the box that records the screen. So I was talking about the exam paper but they couldn't see to which question I was referring unless they had the exam paper in front of them. […] After I realised that, I actually spent 5 minutes thinking about whether I should do the recording again. That wasn't something the students had complained about, but I felt the aura of pressure that perhaps they would be unhappy because, when they will do their exam preparation, they will say I'd prefer to see what is actually on the screen to just hearing the discussion of the solutions. (L14) Compliance through exhaustion. A number of our interviewees highlighted that in the context of the student satisfaction focus adopted at ABC, running challenging modules required huge amounts of engagement with and support for the students. Given the large student numbers on many accounting and finance modules, some lecturers suggested that they had dropped educationally valuable aspects from their teaching because they were exhausted by the demands these aspects had placed on them. One lecturer, for example, explained that she had moved away from a particularly open-ended approach to assessment for final-year students because she "just ran out of steam": I just ran out of steam […] The problem was that it was the first time ever that they [the students] were given a range of resources that they could study on and a range of assessments that they could choose from, and that choice made them even more unsettled. My take is that final year students should have choice, they should engage with the things that they are more interested in but when, in other modules, they are given a set pack of activities and a set of assessments that look the same from year to year, when they are then presented with a lot of choice in the final year in one module, many students find that [confusing] -"What should I choose, why should I choose this, how can I be assessed the same way as others?" And that takes a lot of engagement from the lecturer, and I think that's why I ran out of steam. (L13) Instrumental compliance. A number of our interviewees suggested that lecturers may respond to the customer accountability pressures they faced by watering down the difficulty of modules, teaching to the exam[10] or effectively giving students the exam questions in advance. Unlike the other types of compliance discussed above, which restricted themselves to the practical level, this type of compliance arguably entails the "full adoption of the values, norms and practices" prescribed by the commercial logic. For if we adopt the transactional mindset that underpins the commercial logic and we fully subscribe to the notion of students as customers who like to achieve the highest possible mark whilst putting in minimal levels of effort, responses such as reducing the difficulty of modules and teaching to the exam appear to be perfectly consistent with the values and norms promoted by commercial logics. None of our interviewees suggested that they adopted such a response. Instead, several of them suggested that they came across other lecturers who had done so: I know certain colleagues at this institution, certain modules at this institution, where the average mark is 97%. So, clearly, I would say this colleague has decided: "Right, the students, it's very difficult to get them to engage with this material and so on, I know that if I challenge them they're gonna fail, it's gonna cause issues, so I'll make it very simple, very straightforward, very easy." I don't know, either by the way they assess it or by spoon-feeding them all the way. I'd say that's a completely instrumental approach. I don't think the students from what I understand complain about these modules at all, I think they are very happy. (L9)

Defiance
Another response to the challenges commercial logics and the associated customer accountability pressures posed to professional accountability was defiance which, according to Pache and Santos (2013b), entailed "an individual's explicit rejection of the values, norms, and practices prescribed by a given logic" (p. 12). We identified two principal types of defiant behaviour in our data, which we call categorical defiance and defiance with justification, respectively.
Categorical defiance. Categorical defiance refers to those instances in which lecturers adhered solely to the professional logic and made no attempts to engage with customer accountability pressures. Given the increasingly pervasive nature of the commercial logic at ABC, instances of categorical defiance were relatively rare and often restricted in their scope. Occasionally, categorical defiance was however also adopted on more significant issues. One lecturer, for example, highlighted his refusal to go along with student expectations for high exam pass rates: Failure rates at the exam are very high. I guess mid-20s is the average over the last six years. It has varied from something like 18% to, you know, 34%. And that's because I think there is a certain level at which they should know the material. If they are below that I don't think people should be entitled to, you know, if they don't get a degree they don't get a degree.
[…] Instead of making the material easier year after year, I actually kept it quite difficult. I am keeping the quality, the difficulty of the class generally high. And I probably get the heat in terms of the evaluations because the students prefer if they had something easier but you have to take this cost, I guess. (L14) Defiance with justification. Similar to categorical defiance, defiance with justification refers to instances where lecturers rejected student demands and adhered to the professional logic. It however differs from categorical defiance in the respect that lecturers acknowledged and engaged with customer accountability pressures, even though this engagement is restricted to providing a justification as to why student demands are not met. For example, one lecturer persistently rejected student demands for model solutions to a series of practice tests with the justification that, in his view, this would undermine their pedagogical value: I give them a lot of quizzes to do online but I don't give them the answers, and I keep on telling them: "You figure them out, you talk to each other, come together as a group, try to figure out the questions. I am not gonna give you the answer because what you are going to do at the end is just read the answer, you are not gonna try it, you're gonna fail it, you're just gonna read the answer, you're not gonna practice, you're not gonna try to figure out where you've gone wrong." So I am adamant, I don't give them the answers. (L5)

Combination
On numerous occasions our interviewees suggested that they attempted to combine the conflicting pressures that emerged from their own sense of professional accountability and the demands of the students. We found that lecturers adopted two distinctive types of strategies in this context, which we call blending and translation, respectively.
Blending. Blending entailed bringing together elements from both logics and combining them in a manner that went some way towards satisfying the professional accountability of lecturers on one hand, and keeping students happy on the other hand (Glynn and Lounsbury, 2005;Miller et al., 2008;Rautiainen and Jarvenpaa, 2012). This response involved a trade-off between the requirements of the professional and the commercial logic. For example, one lecturer suggested that when designing assessments she tried to allocate a proportion of the marks to simple, mechanical exercises to keep the student-customers content and ensure that they pass, whilst the remainder was allocated to more ambitious exercises which, consistent with her sense of professional accountability, were designed to encourage students to think independently: We were looking at designing module assessments, you can set like the midpoint mark to be where you've met your learning objectives. And to an extent that is the bit that's been spoon-fed. But if you want to achieve a much higher grade, then you've got to go above and beyond that yourself autonomously to get such a mark. (L2) Translation. Some of our interviewees sought to bring professional and customer accountability pressures together by translating the interests of the student-customers into their own (Callon and Latour, 1981;Rose and Miller, 1992). This strategy involved convincing the students that the narrow, technical and passive approach to education they favoured was contrary to their long-term interests, and that only the broader, open-ended and proactive approach endorsed by the professional logic would prepare them for the careers they aspired to pursue. Translation allowed lecturers to move beyond the simple trade-off between professional and customer accountability pressures, which is implicit in other responses, by seeking to "take students on board" (L2) and creating an overlap between the two otherwise conflicting pressures. For example, a lecturer enrolled the help of "real" case studies and a guest lecture from a professional body to convince students that going beyond the narrow, technical approach to education was in fact good for them and helped them prepare for their future careers: I think the convincing case was telling them: "You will be exposed to more than calculations in your work, you will be asked to make a judgement on the calculations, you will be asked to make proposals for scenario 1, scenario 2, scenario 3, you will be asked to think wider than the organisation and its costs, you will be asked to think of other stakeholders when you think of performance." […] We gave them the chance to read real cases and brought the professional body in to support that this is actually happening [in business]. Bringing in the professional body, bringing in more cases and examples showing that this is real business made the students realise that this [i.e. going beyond the numbers] is relevant. (L13)

Compartmentalisation
Compartmentalisation, according to Pache and Santos (2013b), denotes "an individual's attempt at purposefully segmenting her compliance with competing logics" (p. 13). We found one instance in our data which appears consistent with this definition. A lecturer (L14) who taught one big module which contained many mediocre students and one small module populated by very good students, explained how he deliberately adopted very different teaching approaches for the two modules. On the small module, he adopted an ambitious and challenging teaching approach which was consistent with the professional logic and the accountability he felt to the student as a learner. He felt that, due to the large number and the limited abilities of the students, this teaching approach was not a "realistic possibility" for the other module. Instead, he went with "the will of the students" and adopted a more mechanical, spoon-feeding approach for this module.
We however also found many instances of lecturers switching between different logics which did not appear to be particularly purposeful or aimed at securing legitimacy with different audiences, as suggested by Pache and Santos's (2013b) notion of compartmentalisation. Instead, lecturers' compliance with different accountability pressures at different times and in different contexts often seemed much more improvised, coincidental or even contradictory. One lecturer, for example, highlighted that she put much emphasis on students working things out for themselves in tutorials. Rather than showing the students exactly what they needed to do, she just kept "batting things back, answering questions with questions". She however also described that, for no particular reason, she adopted a very different approach in a recent tutorial in which, as was often the case, none of the students had done the "pre-work": I caved a bit. I could have gone to the tutorial and I could have said "Well, you've done nothing, you were supposed to do this, get on and do it and I'll stand here and watch you". But I didn't […] I helped them to do the entire exercise. (L10)

Drivers of individual responses to competing accountability pressures
The previous section has shown that lecturers adopted a great variety of responses to competing accountability pressures. This wide range raises the question as to why different lecturers, when faced with the same institutional demands, adopt such a big variety of different responses. In this section, we offer some reflections on this subject based on the data we collected at ABC Business School. We suggest that Pache and Santos' (2013b) notion of adherence plays a role in this context, but not in a deterministic manner. We moreover suggest that individuals' standing within and beyond the organisation is also a significant driver of individual responses to competing institutional pressures in our case organisation.

Adherence
Pache and Santos (2013b) suggested that adherence, or the degree to which individuals are committed to the competing logics they are faced with, is the primary driver of individual responses. In the context of our case study, this would suggest that those who are more familiar with and sympathetic to the commercial logic are more likely to comply with it, whereas those who are not are more likely to defy it. We found that by virtue of their backgrounds, those lecturers who had significant private sector experience were almost instinctively sympathetic to the commercial notion of students as customers, and more likely to treat them as such. One lecturer explained: Perhaps reflecting their commercial backgrounds, lecturers with significant private sector experience tended to be more compliant with student demands and customer accountability pressures than those who did not have such experience. However, they had also come to recognise the perceived limitations of the customer notion and associated accountability pressures in the higher education context. Thus, whilst they were more likely to comply with student demands, they did not fully submit to them and they made use of the full range of responses to competing accountability pressures discussed in the previous section, including defiance.

Individuals' standing
The second driver of individual responses was related to the individual's standing within and beyond the organisation. Lecturers who had a good reputation in terms of their teaching track record and research outputs and, as a result, felt that they could easily move to a different university if they wanted to, were more prepared to defy student demands than their less successful colleagues. For example, one lecturer who had won several teaching awards, explained that she had the confidence to ignore student demands if she felt they were inconsistent with her professional mission. Asked how she would respond if the leadership of the business school were to challenge her on the basis of student complaints or low student satisfaction ratings, she replied: Whilst high performers felt that they had a certain amount of leeway in dealing with student expectations, their less successful colleagues tended to be more reluctant to defy student demands in a substantial manner. For example, a lecturer, who was struggling to meet the research and student satisfaction expectations of ABC, explained that he had to sell his "soul to the devil" and comply with student demands because otherwise his life would become "very, very difficult": You have to sort of sell your soul to the devil. You know that ultimately you are not teaching things that you want to teach or think are important for the students to know, you know you are not pushing the students, you are tailoring the way you teach just to put a smile on their face. If they are generally happy you will be seen as doing a good job, and that's the most important thing for you as an individual to get on in life. It's the path of least resistance, if you don't do this your life becomes very, very difficult. Students are going to complain to you, to associate deans, it is a very negative process. (L9)

Discussion and conclusions
In this paper, we have examined responses to competing institutional logics in a hybridised public service organisation. Unlike most of the extant literature on this subject, which has focused on organisational responses and/or the role of accounting therein (e.g. Greenwood et al., 2011;Hyvonen et al., 2009;Oliver, 1991;Santos, 2010, 2013a;Rautiainen et al., 2017), we have explored how individual, frontline professionals experience conflicting institutional logics and how they manage them. In our case organisation, an English business school, we found that, on one hand, the historically dominant professional logic and associated notions of professional accountability required lecturers to adopt broad, challenging and open-ended teaching approaches. On the other hand, the emerging commercial logic and associated customer accountability pressures pushed lecturers towards narrow technical content, black and white solutions, spoon-feeding and watering down the curriculum.
Our study subsequently drew on Pache and Santos (2013b) to discuss lecturers' individual responses to these conflicting logics and accountability pressures. We found that the responses fell into the broad categories of compliance, defiance, combination and compartmentalisation identified by Pache and Santos (2013b). We however also found considerable variation and nuance within these broader categories, leading us to develop eight sub-categories for Pache and Santos' (2013b) model, namely, enforced compliance, internalised compliance, compliance through exhaustion, instrumental compliance, categorical defiance, defiance with justification, blending and translation.
Some of the responses we observed indicated that the strong emphasis on customer accountability mechanisms at ABC Business School resulted in a significant erosion of the professional logic and professional notions of accountability that had historically underpinned the behaviour of lecturers. The most extreme example of this was instrumental compliance, whereby lecturers responded to customer accountability pressures by deliberately turning away from the professional logic and embracing the transactional mindset associated with the commercial logic. This response was associated with behaviours like telling students exam questions in advance or reducing the difficulty of exams.
Other responses we observed, however, indicated that lecturers and their professional sense of accountability showed a significant degree of resilience in the face of the strong shift towards commercial logics at ABC Business School. This was possible because lecturers developed clever combination strategies (Pache and Santos, 2013b) which sought to balance or bring together the competing professional and customer accountability demands they faced. We observed two distinctive sub-categories of such combination strategies. The first, which we called blending (cf. Glynn and Lounsbury, 2005), involved the implementation of careful trade-offs between professional and commercial logics with the aim of offering both constituencies enough to at least partly satisfy them. The second, which we called translation (cf. Callon and Latour, 1981), involved attempts to move beyond such tradeoffs and to overcome the apparent conflicts between the two logics by convincing studentcustomers that their long-term interests were in fact best served by the professional logic.
Our findings indicate that even though the professional ethos of academics is under attack by the ongoing corporatisation and commercialisation of universities (Kallio et al., 2015), academics can still find space for a degree of professional autonomy and individual agency (Clegg, 2002;Henkel, 2000;Kolsaker, 2008). Our study however also indicates that as universities become increasingly corporatized and customer-focussed, realising such space will require careful consideration and lots of hard work, as exemplified by the blending and translation strategies discussed above. Yet, in a higher education environment that places ever greater publication, research funding and student satisfaction demands on individual lecturers, there is a danger that increasing numbers of academics may no longer be able or prepared to invest large amounts of time and effort into developing such strategies. Instead, they may be tempted to adopt instrumental compliance responses to student demands. Pache and Santos' (2013b) conceptually derived model of individual responses to competing institutional pressures tended to emphasise their deliberate or even strategic nature. Whilst some of the responses we observed as part of our empirically grounded study were broadly consistent with this characterisation (e.g. blending, translation and instrumental compliance), others were, however, much less purposeful or calculated. For example, some of our interviewees complied with student demands not as the result of deliberate or strategic decisions but because they were simply exhausted by the strains maintaining the professional logic in an increasingly commercialised environment placed on them. Similarly, whilst Pache and Santos' (2013b) notion of compartmentalisation suggests that individuals switching between responses was strategic or purposeful in nature, we found that lecturers moved back and forth between compliant and defiant responses for much more mundane reasons. Thus, we argue that, in the real world, individual responses to competing institutional demands are often much more improvised, coincidental and inconsistent than suggested by Pache and Santos' (2013b) conceptually derived model.
In addition to providing insights into which responses individual, rank-and-file professionals adopt when faced with competing institutional logics, we have also explored why they at times respond to similar institutional demands in fundamentally different ways. In this context, we have identified two different drivers for these responses in our data. First, we found support for Pache and Santos' (2013b) suggestion that adherence (i.e. the extent to which individuals were familiar with and committed to a particular logic) was an important driver of individual responses. Specifically, we found that lecturers with significant private sector experience were more sympathetic to commercial logics and more likely to comply with customer accountability pressures than their counterparts who lacked such experience.
Second, we have suggested that an individual's standing within and beyond the organisation represents another important driver of individual responses to competing institutional pressures. We found that lecturers with weak research and/or teaching track records were more likely to comply with student demands, whilst those with better records were more inclined to defy them. Our data indicates that the relationship between individuals' standing and their responses to competing institutional pressures was strong at ABC Business School, and we would encourage future research to examine whether this relationship applies to a wider range of organisations.
Beyond insights into individual responses to competing institutional logics and their drivers, our study also furthers our understanding of user, customer and citizen accountability mechanisms in public services. Previous literature has portrayed public service providers as remote and unresponsive to the needs of its users, and called for more such accountability mechanisms, which are often assumed to be unambiguously beneficial (e.g. Goetz and Jenkins, 2001;Handley and Howell-Moroney, 2010;Meijer and Schillemans, 2009). Whilst we recognise the benefits of user, customer and citizen accountability mechanisms, we argue that they can also have significant drawbacks, especially in the context of knowledge-intensive public services like universities or hospitals. In such organisations, users often cannot fully evaluate the quality of services at the time they receive them, and obtaining the long-term benefits of these services often requires users to make short-term sacrifices. As a result, placing strong emphasis on user, customer or citizen accountability measures, which are usually based on information collected at the time the service is delivered, can have severe dysfunctional effects in knowledge-intensive public organisations. More specifically, they may compel public sector professionals to act in a manner that prioritises the short-term satisfaction of users over the long-term effectiveness of the services they provide. Our study provides a case in point. It indicates that a strong emphasis on one such accountability measurestudent satisfactioncan lead to a watered-down curriculum and the adoption of instrumental teaching approaches, which keep students happy and satisfied in the short-term but fail to impart the critical thinking and independent working skills which are key to their long-term development.
We conclude with a call for further research into teaching related accounting and accountability issues at universities. The extant literature on higher education reforms has predominantly focused on research performance measures (e.g. Agyemang and Broadbent, 2015;Hopwood, 2007Hopwood, , 2008Kallio et al., 2015;ter Bogt and Scapens, 2012;Tourish and Willmott, 2015), which have historically been the fulcrum of performance management at universities in the UK and in many other countries. At least in England, the recent introduction of the Teaching Excellence Framework, which aims to establish a direct link between universities' funding and their performance on a number of teaching related measures (Department for Business Innovation and Skills, 2016), indicates that university teaching could soon become subject to similarly powerful accountability and performance management pressures as research. The findings of our study indicate that an increasing emphasis on student satisfaction measures could have profound effects on higher education professionals and users alike. Further research exploring the changing roles of lecturers and students in corporatising universities will be required to develop a fuller understanding of these issues.

Notes
1. Quangos is short for quasi-autonomous, non-governmental organisations. The term denotes organisations which are tasked by government departments to manage public services at arm's length.
2. Bovens (2005) used the term social accountability to refer to these mechanisms. Given that this term has a different meaning in the accounting literature, we have decided not to employ it in this paper.
3. The specific RAE/REF rules have changed repeatedly over the last 30 years. For the next REF in 2021, each participating business school needs to submit on average 2.5 publications for each research-active member of staff, which will be rated on a five-point scale from "unclassified" to "world-leading". In addition to this, business schools' research environment (which covers issues like doctoral supervision) and the impact of their research on the wider society will also be rated on this scale. The vast majority of government research funds allocated through REF go to schools which produce large numbers of publications and impact case studies that are rated world-leading.
4. For example, the 2018 edition of The Guardian's university ranking allocated 25 per cent of institutions' overall scores on the basis of measures derived from NSS results.
5. Student satisfaction scores were effectively the sole measure ABC used to assess teaching quality as well as the overall teaching performance of lecturers. Lecturers who scored poorly on this measure were asked to explain themselves in meetings with the associate dean responsible for student satisfaction, departmental heads and often also directly with the dean, who took a very keen interest in student satisfaction. Our interviewees suggested that these meetings could be highly acrimonious and that on some occasions lecturers had been explicitly encouraged to seek alternative employment. In addition to the routine satisfaction surveys, students were also encouraged to report any emerging concerns or complaints to the above-mentioned associate dean who, according to our interviewees, tended to take the students' side in such matters. Besides student satisfaction, research-active lecturers were also assessed on their research output and their engagement with the wider society. This occurred in the context of a performance and development review process which included annual meetings with departmental heads and the dean.
6. Many of our interviewees described students in general in these terms. Some qualified such statements and highlighted that a minority of students preferred critical and intellectually stimulating teaching approaches. One lecturer suggested that approximately 20 per cent of the students fell into this category, whilst the remaining 80 per cent preferred recipes, technical content and black and white solutions.
7. The teaching approaches favoured by lecturers and students, respectively have significant commonalities with the notions of deep and surface learning as discussed in the accounting education literature (e.g. Beattie et al., 1997;Boyce et al., 2001;Hall et al., 2004). Interestingly, even though many of our interviewees either held or were studying towards teaching qualifications such as the Postgraduate Certificate in Higher Education, none of our interviews contained any explicit references to learning and teaching theory, or the various different learning and teaching approaches discussed by the accounting education literature.
8. During our field work, we got the strong impression that our interviewees' rejection of the customer notion reflected a genuine concern for the intellectual and professional development of their students. The proponents of neo-liberal higher education reforms might however argue that lecturers were trying to invoke the supposed interests of students to protect their historically high levels of autonomy, which had allowed them to, for example, prioritise their research over their teaching or to take more and longer holidays than other employees. 9. We did not find any instances of Pache and Santos' (2013b) fifth category, ignorance, in our data. Our discussion will focus on the remaining four categories set out by Pache and Santos (2013b). It would appear that, as a result of the enthusiastic adoption of commercial notions of higher education by the student body as well as the managerial hierarchy of ABC, all our interviewees displayed a strong awareness of the commercial logic.
10. We should note that teaching to the exam is not necessarily problematic, especially if exams are designed in a way that encourages critical thinking and reflection. However, at ABC, student demands made it increasingly difficult to set such exams. Lecturers experienced strong pressures to adopt a narrow curriculum which focussed on providing students with recipes for solving technical questions. More open-ended or critical teaching approaches were often met with student complaints and low satisfaction scores. Since exams needed to be consistent with what students were taught in class, lecturers experienced pressure to design exam questions which effectively were slight variations of exercises they had taught during term-time and which could be solved by following such recipes in a mechanical manner. Exam questions which placed a strong emphasis on critical thinking, or which asked students to apply the knowledge and skills learned in class to a less familiar context, were often met with strong student complaints.