Sandwiched: Exploring role and identity of middle managers in the genuine middle

Introduction

Middle managers play an important part in organizational hierarchies from their position placed ‘between the operating core and the apex’ (Mintzberg, 1989: 98). Since Burns’ (1957) first studies of middle managers’ work in the mid-1950s, a burgeoning stream of research has taught us what middle managers should do, what they actually do and what skills they need for this managerial role (Harding et al., 2014).

Nevertheless, despite an extensive literature on the ‘paradoxical, complex and ambiguous world of middle management’ (Tengblad and Vie, 2012: 35), surprisingly few studies have taken the middle-levelness of this role particularly seriously. When the term ‘middle management’ is employed, the point is often simply to separate this category from top managers and pure subordinates (non-managers). Thus, we find a large variety of middle managers from first line supervisors (Hales, 2005) or small team leaders to very senior managers (Mintzberg, 2009; Watson and Harris, 1999), just below the top of large and sometimes small organizations adhering to this category.

In most studies, middle managers are addressed as a distinct entity rather than in terms of relationality. When the relational aspect is considered, either the superior or the subordinate aspects of middle-levelness tend to be downplayed or neglected, leaving only one type of relationality and positioning upwards or downwards left standing. Furthermore, the managers themselves are often more than willing to talk about themselves as leaders, at the same time as a contradictory organizational situation creates problems for them in their doings of leadership (Alvesson and Sveningsson, 2003a), and they tend to leave out that they, with few exceptions in large organizations, are subordinate managers (Laurent, 1978). Rarely, researchers study middle managers as subordinates, although we do find exceptions such as Sims (2003) that emphasizes how managers are eager to be promoted and therefore sensitive towards stories about themselves told to senior people. As a consequence, and perhaps because researchers follow the interviewees’ preferred storyline of being a senior to junior – leader over followers – we do not really grasp the middle-level experience of the middle managers being studied.

This study addresses the experienced middle-levelness missing in the middle-management literature, and explores what it is like to be genuinely middle in terms of identity, that is, who am I, and role, that is, what should I do, in this middle? Insights on how managers make sense of and navigate this middle-levelness may help advance theory on the middle-management role and identity lacking in in-depth studies, and inform practising managers struggling in a ‘crossfire of pressures’ (Gallos, 2002).

In this study, we focus on the academic heads – a particular representative of middle managers in the professions. These heads are in charge of fairly large groups, often 50–100 people or more, and normally with directors and other workers such as junior managers that typically find themselves at least two hierarchical levels from the top and so need to face both upwards and downwards. Head and Dean of department and school are terms that sometimes refer to the same, and ‘Deans are, in essence, classic middle managers’ (Gallos, 2002: 174).

Traditional conceptions of management and leadership are problematic in professional contexts such as higher education (HE), law, finance and health owing to a characteristic requirement for autonomy among its organizational members, or at least expectancy of such (Empson and Langley, 2015), and the way knowledge workers respond less to control, supervision and direction, which are ‘the most common words used to describe managerial work’ (Mintzberg, 1998: 143). HE is a professional context in which a continued managerial career, according to Sims (2003) a key concern for most middle managers, is not necessarily a priority for senior academics in administrative jobs. Issues around identification, loyalties, drivers and priorities may therefore be particularly interesting to consider. Thus, HE may serve as an extreme case (Pettigrew, 1990) to help highlight issues of relevance, not only for HE as a particular context, but also for other professional contexts in which these issues of role and identity may be more difficult to spot.

The article has two main purposes, the first being to add to our understanding of mid-level management and leadership in HE, more precisely in UK business schools. How do these heads define their work and identity in relation to their role? The second is to illuminate some aspects of middle-managerial work in the professions more generally, and to address how managers deal with their position in between expectations and pressures from senior and junior levels, in addition to their profession marked by innate values and ideals. As universities, not the least in the UK, become more and more similar to other large organizations with marketization, pronounced hierarchy and a high level of bureaucracy (Bolden and Petrov, 2014; Jump, 2015; Keenoy, 2005; Parker and Jary, 1995), research of this particular context should allow for broader insights (more than empirical generalizations) about key aspects of management and leadership in organizations where professionals and knowledge workers’ expectancies of work autonomy and growing bureaucracy tied to performance measures and observations are prevalent.

The article builds upon two overlapping literatures: academic leadership and middle-managerial work. These two streams of literature serve as a backdrop to our analysis as we interpret one particular metaphor that emerged from our interviews, and which aptly portrays this experience in the middle: to be an ‘umbrella carrier’.

The umbrella metaphor suggests an enactment of this role that goes counter to the widespread understanding of the middle manager as ‘responsible for implementing senior management plans by ensuring junior staff fulfil their roles’ (Harding et al., 2014: 1213). It describes how many middle managers attempt to protect subordinates, by shielding them from what is often presented as excessive demands from top management above. The study also reveals two alternative subject positions (broadly in line with the dominant view of the middle manager as implementing demands from above): ‘performance driver’, that is, where the role is understood as a disseminator of pressures from above to ‘wake up sleepy subordinates’; and ‘impotent’, where the manager feels stuck between levels without agency to influence either upwards or downwards.

The study presents a theorizing of these three subject positions and attends to the umbrella–protector position in particular (being the most prevalent of the three), including its potential pitfalls. Exploring role and identity from this professional, pressured middle adds to our understanding of managerial work and leadership in the professions, but also helps us gain a better grasp of the middle-levelness that so many managers are faced with also outside the professions.

A brief theoretical overview of academic leadership and middle management

The two literatures of academic leadership and middle management partly overlap. Whereas the academic leadership literature is fairly sparse (Bryman and Lilley, 2009), the middle-management literature is huge, and most of it is not relevant for this article. Since we see a vacuum-cleaning of literatures with superficial and artificial summaries of themes as not necessarily helpful (Golden-Biddle and Locke, 1993), we will only make some dips into this wide-ranging literature up-front, and refer to relevant studies throughout the article.

Methodology

The study uses a combination of three methodological elements: inductively generated meanings, further explorations of these with the use of hermeneutical principles, and crosschecking of reliability of interview statements through interviews with people situated so that they can comment on the actions of the interviewed heads and through follow-up interviews.

The study builds upon material from 29 interviews in total. We purposefully selected 15 ‘heads’ and five ‘non-heads’ known through their publications for a scepticism towards established truths on management and leadership and/or a particular interest in paradox and tensions in management/leadership, from four research-intensive business schools in the UK, to share their ideas on and experience with middle management and managers. We use the label ‘head’ as an overall category to refer to these people who are, or have been, in charge of a department or a school, and refer to them as H1, H2, H3, and so forth. We use the term ‘non-heads’ (N8, N11, N14 etc.) to refer to senior academics without head of school/department management experience. The numbers refer to what number in the line of interviews they were.

The five non-heads were invited from across the four schools to ensure additional observations and interpretations of how the heads enacted their role, and as current heads had previously been non-heads and vice versa, we were able to crosscheck interview accounts to substantiate the interview statements and gain several observations on each head. This is important, as interviewees, in particular in senior positions, may give a selective, flattering or in other ways biased view of themselves and their work, wanting to give off a positive impression and/or because of a common self-serving bias (Schaefer and Alvesson, 2018; Tourish, 2013). Five of our heads were interviewed twice, and two were interviewed three times over approximately 2 years, which helped us follow their interpretations of the middle-management roles and identities over some time.

The interviews were loosely structured around the middle-management role as head, how they themselves and/or their colleagues had enacted this role, and whom they were in relation to this role. By comparing self-report with other people’s observations, we were able to gain a rich understanding of what heads of business schools do in their roles and how their acts were experienced and interpreted by their subordinates and by themselves. The first interviews lasted between 1.5 and 2 hours, and the follow-up interviews lasted about 45 minutes. The follow-up interviews aimed for more concrete examples and elaborations of the emerging themes from the first set of interviews. They also permitted us to go back and check for consistency and reliability in responses once we had gained additional input from other interviewees, and so helped us nuance our findings and first interpretations.

Modes of interpretation are vital in qualitative research. At the same time, it is not possible to work exclusively inductively. Data seldom point in a clear direction. Data-processing approaches such as Grounded Theory privilege surface data and give off an impression of rationality through an emphasis on procedures, rules and a clear route from empirical reality to theory (Corbin and Strauss, 2008; Glaser and Strauss, 1967), although there is some acknowledgement of the constructed nature of the process (Charmaz, 2005). There is an increasing tendency for qualitative research to systematically code data. However, as several authors say, such approaches are not appropriate for all types of qualitative data (Bansal and Corley, 2011: 236) or for all qualitative traditions (Alvesson and Sköldberg, 2009), where coding is seen as a reductionist and mechanistic process that detaches pieces of data from a broader understanding, downplaying the context of, for example, interview statements (Potter and Wetherell, 1987).

This article draws upon hermeneutical principles, which means it engages with the data material in a circular movement between part and whole, and the pre-understanding that the researcher brings to the study is actively used, qualified, challenged and developed in the research process (Alvesson and Sköldberg, 2009). Rather than simply codifying empirical material, this material is looked upon as text, and one tries to go beyond the ‘surface’ to look for something less obvious, or less easily revealed in a coding process (which aims at standardization and a de-contextualization of data). At the same time, the text as a totality is borne carefully in mind, which means that variation and contradiction are taken seriously. Meanings expressed in different parts (passages in interviews) are interpreted in depth, but also related to the interview as a whole as well as to the broader context of the setting and individuals (Alvesson and Sköldberg, 2009). For example, we explored middle managers’ statements that senior levels lacked overview and consideration and had excessive demands, and that junior people were stressed, nervous and vulnerable. We asked ourselves the underlying meanings, and came to an interpretation that by plotting and dramatizing one group against the other, this opened up a space for the middle managers to present themselves as mature, responsible and acting in a parental–protective capacity to balance out the situation. Our interpretation is based on the combination of a set of cues rather than a codified material presented in straightforward patterns.

Thus, meanings in our data material are grasped through a combination of pre-understanding, based on a broad literature review (academic cultures, work situation of business school academics, identity, managerial work, leadership and middle management), general cultural knowledge and personal experiences (including familiarity with the institutions and category of people studied), which are gradually revised as the research process and learning progress. When one analyses material hermeneutically, it is important to get a good feel for what is to be studied and try to spot something that is not obvious in order to produce new insights, while, of course, doing justice to the material. Interpretation, insight and intuition are vital here, more than rational processing of data, for example, through coding.

One way to work with hermeneutical principles is through symbolism and metaphors that indicate some ‘core meanings’ or themes. Ortner (1973) refers to key symbols as a rich set of meanings often with a combined cognitive and affective content. In our case, many interviewees talked about ‘shit’. Rather than dismiss this as a mere expression of irritation about trivialities or a way to put some spice into interview talk, we see this as a symbol pointing to a key theme for the studied group. Several of the interviewees explicitly used the ‘umbrella carrier’ metaphor, and others talked of their protector–shielding function in a more general way. We interpreted the metaphors as a way to enact the role and to define the identity as heads. Thus, we decided that the meaning of shit and umbrella–protection called for further exploration. As with all qualitative research, analysis is the ‘result of reality construction processes and subjectively construed by us, the research team’ (Clarke et al., 2009: 331).

Having found or interpreted these key themes in the first part of the study, the study gradually devoted more time and energy to exploring these, not through systematic comparison but through (a) listening to if/when the themes were spontaneously brought up or not and (b) asking questions about how the interviewees saw the themes, explicitly in follow-up interviews and implicitly when interpreting the interviews. This allowed for deeper exploration of the phenomena.

We do recognize the limited and selective empirical material as well as the specifics of the setting (heads of UK business schools), but see the qualified experiences of interviewees, the crosschecking of accounts, and double effect of illuminating both a managerial experience and a subordinate perspective on the heads as compensating for this. We do not aim for statistical generalizations, and acknowledge variation, but seek to present empirically grounded ideas that may progress thinking about the subject matter. We now present the three subject positions revealed through our study, and start off with the most prevalent of the three, namely the ‘umbrella carrier’ or ‘protector’.

Findings

Umbrella protection from what?

The ‘umbrella heads’ were consistent in their accounts of what they tried to protect their subordinates from:

One of the things I find amazing is this one-on-one conversation with other deans. It takes 30 seconds for you and another dean to be commiserating. It’s almost immediate. And I think it is because you’re holding it all in, and suddenly you realize that oh they’re holding the same umbrella. You know the shit might look a little different at University X than it does at University Y, but it’s the same issues, just different flavour. (H6)

New regulations and surveillance routines were seen as obstructive and recurring, eating away at subordinates’ space to do their job. Some surveillance was deemed necessary, but hardly panoptic reporting:

There are a bunch of forms of regulations that seem to be increasingly intrusive, but this isn’t, I just need to be clear this isn’t simply some kind of knee-jerk reaction against surveillance. We’re paid a certain amount of money to do a certain job and I think that it is legitimate that employers should know about what it is that we do. But the problem is when that tilts over to become panoptic and it prevents the autonomous workers. Again, it’s a version of the professional saying, ‘let me get on with my job because the costs of me telling you this stuff all the time is actually very substantial and you don’t seem to realize this and this is not free’. (H3)

Surveillance takes time and leads to the experience of being watched, which disturbs concentration.

‘Financial issues’ was another common form of distraction they would attempt to shield their subordinates from:

The financial crap is probably the most stressful to me and can be to others. I think the financial issues are a big burden on a leader’s head and people need to be responsible for their piece of the pie, and they need to be responsible for all the costs and the revenues, but not in such a way that they think the sky is going to fall. I like the umbrella metaphor because I use the expression, ‘The sky is falling!’ Once upon a time, me as a young administrator, I would go ‘The sky is falling!’ Right? As soon as other people felt my panic it was over. So, I’ve realized I cannot panic. It’s no good for me, and I can’t let anybody else feel my pain. (H6)

Here, H6 bears the pain and protects others from it. We might say there is an element of martyrdom in this meddling-middle position, also implicit in other accounts. The manager is sacrificing her- or himself for others. This also alludes to a form of parental leadership in their protection – to which we will return.

Information from top management regarding periodic crises and urgencies that usually turn out to be less critical or urgent with time was perceived as another form of unnecessary disturbance that needed shielding from. Experience had taught these heads that crises would go away, and so it did not make sense to pass on the full message and worry others about them:

There are periodic crises. There’s a bunch of ways in which we are told that budgets have to be tightened. Depending on who’s talking, this university is 7, 20 or 40 million pounds behind its targets. I see these things come and go, so I do feel a bit more sanguine about them than I used to. And they seem to disappear. It’s an odd thing. You see in many organizations the management will say we’re in a real crisis, this is a real problem, ba, ba, ba, pa, pa, and then suddenly they find 10 million pounds behind the back of a sofa. (H3)

Here, we see the irrationality of university management, which more or less unnecessarily creates crises and panic among staff, and forces the head to not just follow imperatives and communications from above, but also to take an independent stance.

Performance measures were also experienced as particularly offensive towards their subordinates as they would trigger anxieties among many and emphasize individualism and self-interest at the expense of collegiality and meaningful research. Messages regarding the ‘looming REF’ and the ‘ABS-list-bullocks’ (H1, H3, H5), ‘minute changes in school rankings’ (H13), the ‘unexpected shifts in performance expectations’ (H8, N11) and ‘unrealizable targets that everyone knows are unrealizable’ (H10, H7) were, according to our interviewees, constantly repeated by top management and did more harm than good.

The heads of school believed that these perpetual reminders of performance targets would present a ‘source of incredible anxiety’ (H13) if not softened and/or re-written. Thus, our participant heads saw it as their task to protect their subordinates from these messages or at least cushion their impact. The studied heads (with a few exceptions) placed themselves clearly in the middle between senior and junior levels. From here they operated as an antidote to all the anxieties that otherwise may take over, or as a ‘toxic handler’ that would absorb and soften emotional pain and this way help save organizations from self-destructing (Frost and Robinson, 1999). There is a double looking upwards and downwards among our middle managers, where a careful eye and critical scrutiny of the demands from above is informed by a consideration for the pressure on subordinates (in particular, juniors) below. The umbrella metaphor illuminates the in-between experience of this middle-level role.

Protecting what and how?

So, what were the heads trying to protect underneath the umbrella, and how did they go about this protector role?

Organize and manipulate time

Time is a scarce resource for academics faced with the juxtaposed demands of teaching, administration and research. We learned that the heads attempted to carve out time for their subordinates, for example by attending to subordinates’ timetables, and to create pockets of air in an otherwise cramped calendar. Time could also be created or bought by running the department well financially: ‘it allows you to help people with their travel expenses so they can attend conferences, and to buy in more teaching and more administrators’ (H4).

Yet another way was to simply make time up. Manipulation is, of course, a management evergreen, although not so often voluntarily mentioned by managers in research interviews (an exception is Jackall, 1988). Our interviewees were not so cautious. Some of them described how previous heads had engaged in a creative weighting of teaching hours that would create time, on paper, which was then used for research:

The first one didn’t create any space. S/he just didn’t get it at all. S/he couldn’t see beyond her/himself. I think I had one or two more heads who didn’t. They just expected you to find it yourself. Work longer hours. But the last head of department did. S/he managed to create the space by rethinking the weighting of modules. And overnight our teachings increased on paper, although we had the same amount of teaching. (H4)

This last head appeared to be an effective organizer of time, and superior to the previous one, who ‘could not see beyond [her/him]self’ but who expected people to work more and more. Still, this head could of course also be viewed as a manipulator of fake performances. Nevertheless, s/he is presented as a good exemplar.

A similar way of conjuring time out of thin air was done in relation to the hours that academics spent performing administrative tasks. An interviewee talked admiringly about how one head had introduced:

. . . a scheme that weights administration very heavily, making administrative jobs that on paper took up more hours than they actually did. But then with time administrative jobs did take up more time. And so, reality caught up with the imaginary system. But it did create a system that allowed academics to get on with research and pursue publications. (H4)

The interviewees expressed some concerns regarding who these time-conjuring systems were stealing from, which indicates downsides to time protection and creation.

Protect from and clean away anxiety

To stop negative messages from clogging up the subordinates’ minds and creating unnecessary anxiety, the heads would filter information from their place in the middle, rather than simply pass it on. This meant they would omit certain negative messages and place more focus on positive messages:

As department head, there is a whole bunch of stuff that you know, that quite often you don’t tell people because you don’t want to worry them, you don’t want to frighten them. So, if you were explicit about the mess that the university was in, because you know this stuff, you heard something in these meetings, you would scare a lot of people and lose a lot of people too. So, I think there is a kind of temptation, and not a bad one really, to kind of focus on the positive stuff and hope that the negative stuff disappears. (H3)

We can here notice how anxious junior people are believed to be. H3 is eager not to worry or scare them. We learned that some strategies aimed to stop ‘shit’ from reaching subordinates, while others aimed to take away the filthiness that got through:

There are two main ways to protect space: one is a process of filtering, which allows a certain amount of messages to get through if you think it is necessary for them to get through, and the rest is in a sense deflected somewhere else, you absorb them or you bounce them back to whichever part of the university where you think they should be properly dealt with. The other one, it’s difficult not to be fairly blunt about this, but I think it involves a good deceit. (H3)

Another form of protection was offered as the heads took on a parent-like role and guided particularly the juniors through tainted organizational life. This parental theme resurfaced in several of the interviews, and indicated a view of subordinates similar to children, good but vulnerable, to be protected by the parent who bears the full burden and keeps others ignorant for their own good:

I mean, this, I have to be careful maybe about this analogy, but this is a classic parent analogy. I remember this from my psychology books when I had young children and it just said you never talk to your kids about something that they have no control over and will just worry about. (H6)

The interviewee here appears reflexive and careful. However, without discarding some real need for parental protection, it may also be, as is common in human life, an element of self-serving bias and identity work in the story-telling. It may be tempting for heads to rationalize their work and importance by portraying others as in need of their parental support, facilitating the head as mature, powerful and good, against the Others in the form of child-like junior colleagues and negative (shit-spreading) forces. The citation of N8 (re ‘paternalism’) above hints at this. Still, irrespective of this parent–child type of relationship, our interviewees confess that protection is not always possible, and some ‘systemic shit’ (H5) will always ‘seep through the umbrella’ and reach the subordinates, with the risk of causing unnecessary worries. Thus, several of the heads would attempt to ‘clean away’ this dirt through talk that provided an alternative meaning of research and the proper academic. One previous dean in particular was renowned for this alternative picture:

Sam made a great talk once when s/he said to us all, maybe in the first staff meeting I was at, s/he said, ‘We need to read more and write less’. And I was thinking finally somebody has said it. Everywhere else you were told how to write more. So, Sam was of that kind of understanding that it takes time, and that actually reading itself, for itself, is a valuable activity. (H1)

This talk on ‘reading more and writing less’ was referred to by several of our interviewees and had made a long, lasting impression, the way it emphasized scholarly ideals over obsessions with publishing and career.

Recruit for and create (academic) culture

Finally, the heads would attempt to protect and preserve an institutional culture around threatened, marginalized core (academic) ideals, which they believe may be threatened by initiatives from top management. Recruitment was seen as a particularly powerful way to nurture a specific culture: ‘The collegial ethos is maintained through recruitment of team players. Everyone fits. It’s not only about four-star publications. There’s a vibe in the corridor. The norm is to be open and engaged (N1).’

Of course, there is here a balance between meritocracy and nepotism, where those who fit may not necessarily be the best but add to the right tribe-orientation. The heads would also try to facilitate the protection and emergence of a distinct culture through role modelling:

I try to create an environment in which everyone can thrive. And I try to be collegial in my leadership style. I listen to people, treat them respectfully. I consult people, and talk to people, and I try to be available for people if they want to talk. (H18)

The heads repeated certain messages, for example, ‘this is a place where it is fun to work’ (N2) or ‘where management theories are up for debate’ (H7), and then follow suit by attempting to live and demonstrate the values they preached, for example, ‘collegiality’:

S/he facilitates collegiality. S/he develops people and enables them to do their best. Management and leadership have a pejorative connotation to it. S/he is not manipulative, or calculated, to get more out of us. S/he is committed to a collegial ethos. (N1)

To describe an experienced, or a hoped for, culture in the way that these heads did is a classic example of leadership understood as ‘management of meaning’ (Smircich and Morgan, 1982), and yet few of our interviewees referred to these acts as ‘leadership’. Most were sceptical of the leadership term, even in the form of management of meaning. To some, leadership was simply too ‘foul, besmirched and tainted’ a word (H5). The ‘above-ness’ signalled by the leadership term triggered a distancing among many of our heads, who emphasized collegiality rather than superiority. Leadership signalled elitism, hierarchy and non-collegiality, and was therefore a negative symbol for several of our subjects. Umbrella work represented a positive symbol as a representative and protector of colleagues. The middle-level position was for them not one of superiority indicated by the term leadership, which implicitly turns colleagues into followers, but about being a senior person who mediates between levels with the use of an umbrella.

It was interesting to note how (most of) the heads seemed to operate in a context between several negative forces: the bad things that come from above (and sometimes below) and the leadership discourse that threatens to make them appear elitist and distanced from ‘followers’. So, the umbrella also works as a metaphor that protects the heads from being besmirched by discourses that may undermine their sense of self and view of the role. The umbrella-holder identity allows them to uphold a consistent self-view as a collegial academic doing management without becoming a manager or a leader wannabee. Thus, there is a primary meaning of umbrella as protector of juniors against the senior levels, and a secondary, supplementary meaning where the umbrella metaphor shields the manager from elitist or grandiose (non-collegial) leadership ideas, without reducing the manager to just an administrator. As a key symbol, ‘umbrella’ then has more than one meaning that facilitates not only meaning-making of the role, but also ‘identity work’ (i.e. ‘forming, repairing, maintaining, strengthening or revising the constructions that are productive of a sense of coherence and distinctiveness’ [Sveningsson and Alvesson, 2003: 1165]) for managers caught in a difficult in-between position (Sims, 2003). The protection metaphor then does multiple protections: that of junior people from seniors, of the head from juniors, and of the head’s working identity from leadership discourse. We illustrate the multiple functions of the umbrella in Figure 1. The big umbrella directed at the senior level is the most significant one and the one explicitly mentioned by the studied heads. The other two are less salient, but we can see how there are more implicit references to ‘umbrella thinking’ in relationship to junior levels and the leadership idea.

OPEN IN VIEWER

Figure 1

Model of umbrella management.

Variation in middle-levelness: Three subject positions

Most interviewed heads thought that shielding was necessary, but we did find some variation. As previously mentioned, one head deviated completely from the protector role and saw it as her/his job to wake up sleepy academics. This stance was confirmed by several subordinates. S/he had implemented a performance management system that reviewed and monitored performance and ‘gave control back’ to the level above. We found another head who believed that there was not much protection s/he could offer: ‘Yes, people are overstretched, but that is beyond my control’ (H15). S/he saw the effects of the REF, and messages from above as more or less unstoppable: ‘The shit comes down whether you like it or not’ (H15). This lack of protection was also confirmed by a non-head at the same business school: ‘The pressure is unbelievable. I don’t think you’ll ever get an umbrella that’s big enough …’ (N17).

We can here see three distinct interpretations of the middle-levelness of this role: we find the middle manager as superior, who identifies upwards and with being a leader, who wants to improve performance and uphold pressure, and who experiences the agency to do so. But we also find a head who simply feels s/he is one of the middle people in the hierarchy, caught in the chain of command, unable to do much protection, and forced or inclined to follow the pressure from above and from within the profession. And third, we find heads who see themselves as umbrella protectors, taking the side of subordinates, in particular junior people, and who identify as ‘middle down-up’ people with agency to influence and engage in protection.

Thus, to sum up, we suggest that middle managers take one of three subject positions that influence how they interpret their actions, craft their stories of self-in-role and are experienced by others in their attempts at organizing from the sandwiched middle: ‘performance driver’, ‘impotent’ and ‘umbrella–protector’ (see Figure 2 for an illustration of the three).

OPEN IN VIEWER

Figure 2.

Three subject positions.

Performance driver (1) implies a superior position where performance is emphasized and the manager is energized by the power to improve results. This positioning is illustrated by a megaphone in the conceptual model, as performance pressure from above is not simply passed on, but accentuated: ‘Work harder, produce more!’ The direction of identification is first and foremost with the level above and with the manager/leader role, and agency is experienced as high. The impotent manager (2) is a low-energy position and without (experienced) agency. This position is influenced by an unclear direction of identification neither upwards nor downwards, and an experienced inability to influence either way. Umbrella–protector (3) is characterized by faculty-identifying, subordinate-energized managers. Their direction of identification is clearly downwards and with the profession which fuels a strive, and some capacity, for protection.

These three positionings of self as performance driver, impotent and protector represent very different managerial identifications, that is, social group affiliations, and identities, that is, self-in-role understanding (Burke and Reitzes, 1981), and correspond to differing views of subordinates as pressured and suffering for (2) and (3), respectively, and as sometimes sleepy and non-performing for (1). As most of our interviewees placed themselves firmly in category 3, we cannot say much about the other two, but believe a good task for future research is to explore these further. Given the widespread enactment of and identification with the protector role among our studied heads, we will dwell some more on this particular interpretation of life in the middle, and touch upon its potential downsides, before we move on to the discussion of our findings.

Discussion

So, how do middle managers in the professions, and higher education in particular, experience their role and identity in the ‘sandwiched middle’ (Bryman and Lilley, 2009)? What is it like to deal with conflicting interests and different ‘zones of rationality’ (Kallenberg, 2015)? We find some variation among our studied heads, and suggest they experience this middle-levelness through one of three main subject positions: (1) performance driver, (2) impotent manager and (3) umbrella–protector. We argue that each position is informed by the direction of their identification (up, down or with neither), how they experience their degree of agency in relation to institutionalized performance demands and pressures, and their worldview of superiors as ‘good’ or ‘bad’ and subordinates as either under- or over-performing/-pressured. We focus on the major tendency in our material, which is the umbrella–protector position.

Our findings and theorizing on the umbrella–protector role and identity suggests a new perspective of middle-management identity and role in the professions – or at least new in terms of nuances developed. The middle managers in our study see themselves as protectors of time, mental states and a culture needed to conduct professional (academic) work. The heads’ acts aim to stop and reduce, rather than implement, pressures from above, which is a more commonly assumed objective of this role (e.g. Kallenberg, 2015; Tengblad and Vie, 2012). The way the middle managers attempt to protect both subordinates and a professional culture adds to how previous studies have addressed the typical dilemma of organizational control versus individual autonomy in academia (Raelin, 1995) and how academic leaders go to great measures to protect cherished values such as academic freedom (Levinson, 2010) and collegiality (Bryman, 2007; Middlehurst et al., 2009). Still, we find that the way they address these dilemmas tilts them in a slightly opposite direction to these previous studies, and in a way that may seem counterintuitive to the typical managerial/leadership stance.

Many authors suggest that leadership involves interpretation of values and the representation of collective purposes and interests (e.g. Middlehurst, 1993). Others point at the therapeutic, supportive role of managers (Western, 2008). Often these acts are interpreted as ways to facilitate productivity – although they are sometimes presented as an exaggerated preoccupation with making people feel good at work – that takes time and energy away from core issues (e.g. Zaleznik, 1997). The umbrella position taken by our studied heads represents a very different role interpretation in comparison with this productivity facilitator function. It also stands in stark contrast to how, for example, Mintzberg once described the middle manager as a ‘disseminator’ of information, receiving and transmitting information to subordinates (Mintzberg, 1973: 57), and the departmental head as someone to offer ‘free flow of information’ on a regular basis (Middlehurst, 1993: 141).

The experienced and enacted middle-levelness is central for this protector positioning and is framed from a subordinate rather than superior interpretation. Again, this is counter to how middle managers tend to emphasize their superiority and distance to subordinates (doing leadership, e.g. Alvesson and Sveningsson, 2003a), and define themselves in terms of pro-top management, loyal and career-oriented towards senior management positions (Sims, 2003). Our managers take on burdens, but they do not emphasize their heroism or fighter capacity. They engage in discussions, manipulations and filtering rather than great battles. They hold up an umbrella, not a sword, and the underlying image is more of a martyr than a heroic warrior type.

We suggest that the umbrella interpretation of role and identity is informed by the heads’ identification with and perception of subordinates (in particular junior colleagues) who they experience as overly occupied by productivity, partly triggered by performance management systems. Their stance also needs to be understood in the light of how performative demands have crept into universities and business schools with new public management (Keenoy, 2005; Parker and Jary, 1995; Prichard and Willmott, 1997; Willmott, 1995), which has expanded bureaucracy (Jump, 2015) and created insecure, pressured academics (Bristow et al., 2017; Clarke and Knights, 2015; Knights and Clarke, 2014), and anxiety about publications. The academic middle-manager context knows these worries, struggles and insecurities from within. The strong identification with subordinates as junior colleagues seems to make heads particularly prone to want to protect them. Against this anxiety-enhancing, horrific world, the heads take on the role as parental actors who offer care, advice and protection.

This parental theme is present also in other studies on academic leadership where, for example, a department head is described as a ‘supportive parental figure’ who influences ‘a supportive environment in which faculty feel safe’ (Gomes and Knowles, 1999: 170), and where heads should ‘buffer’ staff against externally imposed changes that may demoralize and de-motivate staff (Evans, 1999). However, it is not clear if (or rather how much of) this need for parental protection comes from real threats to professional work and culture, a need from subordinates, or from the managers’ need to construct an identity-supportive story of themselves.

It may very well be that our studied heads take on the protector role in response to explicit calls similar to the expectations revealed in a study where academic leaders were expected to provide ‘a holding space/enabling environment to facilitate academic work’ (Ambrose et al., 2005, cited in Bolden et al., 2012: 31). It may also be that what used to be academics’ attempts to resist managerialism by fiddling ‘time and space from themselves in order to fulfil their obligations’ (Anderson, 2006: 578) has now been passed on to the manager, who is expected to take on an opposite approach to how we tend to understand regular management work.

Still, we find it more likely that the umbrella–protector positioning is a way for middle managers to handle the complicated middle-levelness in which they are in crossfire (Alajoutsijärvi and Kettunen, 2016; Gallos, 2002; Sims, 2003). To navigate this complicated middle, we suggest they need a combination of stories and acts to organize meaning and direction for their interpretations and enactments of identity and role. We suggest the heads’ own constructions of the umbrella–protector, performance driver and impotent manager help them navigate dilemmas such as having two stories to tell different audiences, as Sims (2003) emphasized is a challenge, and deal with superior and subordinate levels while they bridge or modify too sharp positions.

Although we have sympathy with the story-telling approach taken by Sims (2003), and acknowledge that the narration and dramatization of organizations, people and events as well as self are crucial (Fraher and Gabriel, 2014; Sims, 2003), we are somewhat worried about studying only words and assuming that stories capture major parts of organizational reality or people’s experiences (Alvesson and Kärreman, 2011). There may be a difference between the narrative-discursive and the behavioural-dramaturgical (Down and Reveley, 2009), as managers, like most people, tell stories but also interact in other ways. Talk, decisions and action often diverge (Brunsson, 2003). Thus, in this study we do not stress issues like plot, narrative structures and the situatedness of storytelling, but explore their behaviours and the way they are interpreted by their subordinates, peers and themselves.

We suggest a theorizing of middle-level managers in the HE sector with a possible bearing also to other professions. The three main positionings and their adhering discourses (that trigger and are produced by these) offer an action space that is not uncomplicated, but which helps them navigate in potentially troubled waters. The heads can position and re-position themselves in various ways and handle the multiple demands and inconsistencies of their work situation shared with most middle managers, according to our initial literature review (e.g. Clarke et al., 2009; Thomas and Linstead, 2002). And so, this framework of self and significant-other positioning helps them deal with a demanding and difficult work situation in the genuine middle, pressured from several significant sources both above and below.

Previous studies have described how managers constantly move between subject positions of controller and controlled, resister and resisted (Harding et al., 2014), whereas in our study the three subject positions seem to be enacted in a more stable manner. The middle managers identify themselves either as senior people (performance driver) with responsibility for and some capacity to protect their good subordinates, as subordinates (umbrella–protector) in relationship to senior levels with many problematic and destructive demands and interventions, or as neither senior nor subordinate (impotent manager).

The umbrella protector, the most thoroughly explored position in our study, places itself strongly in the middle, with reference to both senior and junior levels and with a clear identification with and sympathy for their subordinates. The heads protect the level below from (perceived) ruthlessness and stupidity of senior levels above, a behaviour confirmed not only by other heads but also by non-heads and during our informal observations. Faculty and juniors in particular were said to be subjected to an ‘incredible source of anxiety’ in the face of the full horror of some of the stupidities of top management. Here, we find some strong story-telling, which legitimatizes certain actions and creates a scene for strong identification with faculty in need of support.

The use of the loaded vocabulary, for example, ‘shit’, ‘excrements’, ‘horror’, indicates a fairly strong emotional response and pejorative attribution towards senior level. This dramatization of a horrible context full of irrationality allows the heads to uphold a ‘positive’ working identity as a balancing force that softens all the badness from above. We have no reason to say that the heads are not acting in line with these stories and interactions in terms of actual ‘shit-prevention’, as their accounts are often being backed up by others. But irrespective of actual substantive behaviour and accomplishment, it is key to manage impressions through stories and symbolic acting in order to attain legitimacy. Impression management reduces problematic pressure and opposition, not only from above, but also from below. And so, we argue that the umbrella carrier position is not just the result of an objective need for protection, but also facilitates intensive identity work (Brown, 2015; Sveningsson and Alvesson, 2003). One should here bear in mind that senior managers often come from academic ranks and perhaps have some reasons for their acts, not necessarily being as bad as our interviewees suggest. Without denying the experiences and assessments of our heads, they may still construct reality based on identity work (the great protector against the badness of the upper echelons) as much as nuanced mirroring of ‘facts’.

When the heads tell stories and engage in interactions that indicate sound anti- or countermanagement, they gain credibility and reduce pressure from subordinates. Subordinates typically trust managers who are perceived to share the same social identity and are seen as ‘one of us’ (Haslam et al., 2011).

It may also be that some of the protection offered comes from a need for the heads’ self-protection. This is partly because subordinates may protest strongly to what is passed on from top management above and creates problems for the non-protecting head, and partly because all the (perceived) bad things may also affect the head if s/he returns to a faculty position. So, here we also find some protection from subordinates, which motivates some distance and a slight paternalistic orientation. In a sense, the umbrella is held not only over subordinates, but, and through demonstrating this, also over themselves to protect against potential critique and frustration from the subordinates regarding their own managerial work.

Mintzberg (1998: 146) once suggested that professionals in general require little direction and supervision but need ‘protection and support’ to perform their professional roles at their best. We believe that the study’s theorizing on role and identity in higher education as an extreme case for the professions may cast light on managerial life among other professionals such as lawyers, hospital physicians and consultants as well.

However, this experienced need for protection may also go beyond the professional contexts. The great openness of modern identities can be anxiety provoking in itself. The so-called ‘meritocratic’ society in which selves are no longer ascribed by birth and made legitimate through religion and family status, but need to be achievers through practice, triggers anxieties (Collinson, 2003) and instils a need for protection. And the intense visibility that follows from organizations’ constant performance monitoring of their employees may do just the same. The ‘glass cage’ of contemporary organizations that comes from such systems implies that employees are not so dependent upon managerial actions for results to materialize (Gabriel, 2005), but may need an ‘ethic of care’ (Gabriel, 2010) rather than a furthering of managerial pressures from above. Thus, our theorizing on the umbrella-carrying protector, the performance driver and impotent manager in the middle should help advance our understanding of role and identity among middle managers also beyond the professions.

Conclusion

This article highlights three main subject positions taken by academic middle managers in higher education – performance driver, impotent and protector – and explores the latter position in particular. The umbrella–protector position sees the managerial role as protection of subordinates and professional work. Protection is made through a good deceit of affairs by re-writing messages, restructuring work schedules, slight manipulation of audits, filtering of information, repeated messages that provide meaning and affect culture, and the recruitment of people who are believed will help develop and sustain this meaning.

We have no intention of empirically generalizing from this. Qualitative studies aim to give rich understanding, insights and new conceptualization, not broad and thin empirical generalizations. As our small but carefully chosen sample indicates, there is variation in how heads see their work, from waking up sleepy academics (performance driver) to ‘protectors’ of anxiety-free sleep (umbrella carrier) and being stuck between levels without agency (impotent manger). We see these three main positionings as valuable input in our understanding of how many heads may see their role and themselves as they enact it from their sandwiched middle. We encourage development of more rich and nuanced understandings of how middle managers deal with their role as superior, middle and subordinate in the genuine middle, and how they engage with performance pressures or may feel caught between strong forces and without agency.

The presented insights and theorizing supplement conventional understandings of managerial work (Mintzberg, 1973; Tengblad, 2012), and we argue that ‘umbrella protection’ conceptualizes a key aspect of middle-managerial work role and identity, not only in a context such as higher education but in any profession. Protection may be central in professions such as law, health and the management consultant sectors. Still, academic institutions are not the only organizations that produce a surplus of policies, requirements and communications that may create problems for lower levels in their attempts to carry out the essence of their jobs. Information overload may lead many managers to see themselves as protectors rather than mediators of information. And so, one could perhaps say that the umbrella–protector role is as important as, for example, any one of Mintzberg’s (1973) 10 managerial roles.

Our findings may also cast light on leadership studies that address issues of loyalties and identifications. The middle manager who is perceived by subordinates as their (wo)man and who is met with gratitude for their protection will have a much better chance of doing the job without suspicion, disloyalty and conflict. According to social identity theory, a shared social identity with the group and the leader perceived as embodying group characteristics is key for leadership (Hogg et al., 2003). Loyalty downwards may motivate loyalty upwards and result in grateful subordinates and/or follower attitudes (Haslam et al., 2011). Why should people have trust in and follow their manager? Well, if s/he is a great protector this appears a wise thing to do. If subordinates believe that requests from a manager are not just passed on from above but are also carefully scrutinized and filtered, there is a faith that what one is asked to do is reasonable or unavoidable. Responsiveness to (middle) managerial influence may thus be reinforced.

And so, this umbrella metaphor may also add to our understanding of middle-management work with subordinates and the preconditions for doing leadership. Even if leadership is not seen as important by many of our managers who reject the label, this (espoused or storied) anti-top-down and anti-leadership may, particularly in professional contexts that value collegiality, facilitate occasional doings of leadership. We have only touched upon this issue here, but would say this is an interesting aspect to consider in future research, as is the issue of the umbrella-holding metaphor more generally in organizational relations.