Trapped in the hierarchy: the craft of Dutch city managers

ABSTRACT

The position of top public managers implies management in three directions: up (political office holders), down (creating organizational capacity), and out (boundary spanning). We know however, little about what these managers do. I present a close-up analysis of city managers using diary analysis, shadowing, and interviews with stakeholders. The analysis interprets their craft as managing up, down, and out. It finds that despite the contemporary emphasis on collaborative public management, the prevalence of boundary spanning activities in top managers’ activity patterns is easily overestimated. Working in the governmental hierarchy consumes most of the managers’ attention.

KEYWORDS:

• Managerial work
• city managers
• network governance
• elite ethnography

Introduction

How do public sector CEOs, in this instance city managers (CMs) in The Netherlands, give meaning to their craft in their everyday behaviour? CMs are the pivots connecting the city’s executive politicians and its public service. As public sector CEOs they navigate between political and administrative realities, serving and leading, advising and deciding, boundary spanning and getting things done. Yet, we hardly know what their everyday work entails. How do they operate in a role for which no script exists, in a strategic environment that is unfailingly political, ambiguous, and fluid?

Public organizations provide an ambiguous context for public managers (March and Olsen 1979; Noordegraaf 2000). On top of that, public organizations are repeatedly subject to reform (Pollitt and Bouckaert 2011). The popular scholarly refrain about the rise of the network society (governance), that is suggested to replace former hierarchical structures (government), necessitates a reinterpretation of the craft of public managers, as public managers enact public sector reform when new rules of the game are introduced (Agranoff 2006; Ansell and Gash 2008; Cloutier et al. 2016; Pollitt and Bouckaert 2011). After all, they lead organizations that design and/or implement such reforms. Different roles, contexts, and epochs require different skills and expertise (Frederickson and Matkin 2007). At the same time, the craft of public managers by its nature implies responsiveness to democratic political office holder (POH) and a degree of astuteness in reading and operating within the broader political context of public sector organizations (Hartley et al. 2015). This presents a puzzle: the burst of literature on governance suggests that public managers are most likely involved in (meta-)governance work, but is this the case, and to what extent?

Management within the public sector has received abundant attention in the past decades (Pollitt and Bouckaert 2011; Rainey 2014). Less attention has been paid to the work of managers – what it is they actually do. What they do on a daily basis, what activities they prioritize over others is largely unknown. How they act, or for what – explicit or tacit – reasons they do so, remains a mystery. Public managers received even less scholarly attention, compared to either politicians or managers in the private sector, despite the evident influence of these managers and their organizations on public service delivery (Korica, Nicolini, and Johnson 2015). This article studies CMs – public managers who act as CEOs of municipal organizations. They have been studied before in various cities (e.g. Allan 1981; Sancino and Turrini 2009), but the managerial work of Dutch CMs has not yet been covered in international journals. Their work is, however, interesting for international readers as it provides a contextual picture of how non-elected administrators work at the intersection of political, administrative, and societal contexts.

Several scholars stress the contribution that ethnographic/anthropological studies of managerial work can make in opening up this ‘black box’ (Korica, Nicolini, and Johnson 2015, 19; Noordegraaf 2000, 110–3; Rhodes 2016, 645; Watson 2011, 215). These approaches study managerial work up-close, ‘as it happens.’ In this vein, the present article addresses the following research question: How do CMs, in their everyday behaviour, give meaning to their craft?

The next section conceptualizes the managerial work in a changing public sector. The following section describes the design and methods of the study. In the remainder of the article, I present the Dutch case: an analysis of the craft of four CMs, after which I end with concluding remarks.

Capturing the craft: managing up, down, and out

‘Craft’ is a well-established lens to interpret the work of public managers that has recently been revived in the literature (Korica, Nicolini, and Johnson 2015; Rhodes 2016; Tiernan 2015). By ‘craft’ I mean a carefully developed set of skills and knowledge – beliefs and practices – that has been acquired through a process of experiential training. A large part of this is tacit, some is secret. Craft is distinct from a ‘science,’ ‘profession,’ or an ‘art’ by its focus on practical wisdom and activity (Lynn 1996; Raadschelders 2004). It does not suggest a one-best-way, like science does, but unlike the arts it has utility (Rhodes 2016). Originally, crafts were sustained in formal communities of practice known as ‘guilds’ that set standards for practice. In these communities, apprentices were matured in a process of Bildung as journeymen and finally masters of the craft.

Today, public managers also apply a set of developed beliefs and practices in their everyday work. Their craft entails many practices, including serving POHs, leading an organization, active partaking in networks, as well as providing comfort and relief to POHs and tackling budding problems before they become political (cf: Demir and Reddick 2012; Rhodes 2016). In his account of administrative leadership, ’t Hart (2014) distinguishes three key sets of activities for administrative leaders. These activities are aimed in three respective directions.

Senior public managers manage up when they engage with political office holders – their authorizing environment. The relations between top administrators and their democratically elected masters are likely to be complex and interdependent. Many scholars have argued that these relations are more differentiated than the classic Weberian/Wilsonian notion of strict separation suggests (’t Hart and Wille 2006; Nieuwenkamp 2001). Others refer to these relations as a ‘bargain’ (Hood and Lodge 2006) or use a fivefold typology (Peters 2001). In turn, Svara (2001) claims that the dichotomy of politics and administration is a myth and stresses the complementarity of politicians and administrators. Management of public organizations includes responsiveness to the government of the day – the ‘owners’ of government – as well as keeping long term well-being attended to (Wilson 2016, 66–71). Managing up means advising politicians and negotiating democratic legitimacy for the organization’s output and fostering a productive collaboration with political office holders.

Managing down is about building and preserving organizational capacity in order to deliver public services (’t Hart 2014, 31; Rainey 2014). CEOs of government organizations are responsible for managing large bureaucracies that advise political office holders and execute policy and legislation. Organizational leadership includes transforming, reimagining, and developing this organization. Public managers attend to both the strategy of the organizations and its long-term goals, as well as to ad hoc events that affect the ‘going concerns.’

Managing out is about interactions with stakeholders, societal partners, and competitors outside the leader’s organization (’t Hart 2014, 33). Public organizations ascribe increasing importance to governance in networks (Provan and Kenis 2008). Using networks of valuable partners is an important lever for administrative leaders to create public value beyond the walls of their own organizations (Ansell and Gash 2008). Government managers have a role as meta-governors of these governance networks (Torfing et al. 2012). Managers are expected to be boundary spanners who connect their organization to societal actors and vice versa (Guarneros-Meza and Martin 2016).

I will use the distinction between managing up, down, and out as a structuring device for exploring the craft of CMs (summarized in Table 1). I examine how CMs allocate their time across these three domains and relationships, as well as how they give meaning to these three components of their craft.

Table 1. Summarizing management practices.

Direction	Practices
Up	– Advising political office holders – Negotiating democratic legitimacy for the organization’s output – Fostering a productive collaboration with political office holders.
Down	– Building organizational capacity – Transforming the organization – Managing strategy vs. going concerns
Out	– Interact with societal partners outside the administration – Meta-governance

Managers at work: studying elites

A particular strand of the literature has focused on the question of what managers do (Mintzberg 1971, 1973; Martinko and Gardner 1985; Noordegraaf 2000). Using observational methodology, scholars studied public managers up-close (Kaufman 1981; Rhodes, 't Hart, and Noordegraaf 2007; Rhodes 2011). A landmark study on managerial behaviour is Mintzberg’s (1973) The Nature of Managerial Work. He studied five managers of both public and private organizations up-close and observed what they did. Contrary to classic managerial insights such as POSDCORB¹ activities, Mintzberg associated managerial work with ‘brevity, variety, and fragmentation’ (1990). Managers were not depicted as rational planners, but as people who emphasized action over reflection (Rainey 2014, 348). Their working days were filled with impromptu activities and ad hoc meetings. They would interact with others, rather than lock their doors and think great thoughts (Mintzberg 1973, 37).

Notwithstanding the influence of Mintzberg’s study on the literature on managerial work, his study was neither the first nor the last. Multiple literature reviews indicate that scholars have since followed in Mintzberg’s footsteps and replicated (parts of) his study (Bartelings et al. 2017; Martinko and Gardner 1985; Tengblad 2006), mimicking his observations and shadowing methodology (see also Czarniawska 2014; Noordegraaf 2000), and conducting diary studies (Fleming 2008).

Korica, Nicolini, and Johnson (2015) present ninety-six studies of managerial work in public and private managers on low, middle, and executive level in European, Anglo-Saxon, and Asian contexts, covering classic studies from the 1950s and 1960s (e.g. Carlson 1951) up to recent studies by Cloutier et al. (2016). They discern four research approaches to the study of managerial work. The first studies of managerial work tried to grasp how management affected organizations. Scholars aimed to develop normative ideal types for management in various settings. The second approach, epitomized by Mintzberg’s classic, focused on categorization of managers’ roles. The third approach gained currency in the 1980s and 1990s when scholars noted that managerial work was not a neutral set of skills and behaviour, but a set of political skills and situated moralities that reinforced societal structures and power (im)balances (e.g. Knights and Willmott 1986). Finally, in the first decade of the 2000s, managerial work was being studied as ‘ordinary meaningful activity’ (Korica, Nicolini, and Johnson 2015). The analytical focus is on how the everyday lives of public managers ‘really’ unfold (Rhodes 2011). Managerial work is conceptualized as a ‘practiced craft.’ This craft is increasingly analysed using qualitative research methods.

In contrast to the abundance of literature on what management should be about and what it should entail, relatively few scholars discuss what management actually is – or what managers really do. Managerial work, especially in the public sector, has received relatively little scholarly attention, despite the evident importance of knowing what managers do. Korica, Nicolini, and Johnson (2015, 4) covered only three studies that focussed on managers in government, three studies on healthcare managers, and five on managers in education. Also, of the studies that analyse multiple sectors, few include public sector managers, and even fewer discussed managers at the executive level. The present article mitigates this lacuna by presenting an analysis of the work of pivotal executive managers in local government.

The few existing studies present managerial work in separate categories, such as desk work, travel, meetings, phone calls, etc. (see e.g. Tengblad 2006). Fewer scholars pay attention to who public managers interact with a relational perspective. This article fills this gap by deploying the logic of ‘up, down, and out’ to study how public managers allocate their time and attention over the vertical axis of internal matters relating to political office holders and the organization, and the horizontal axis of connecting to actors outside to the organization.

In sum, the recent literature on managerial work has turned to a focus on managerial work as ‘ordinary meaningful activity.’ Authors argue that public managers practise their craft through everyday behaviour and the tacit and explicit knowledge they exert. I contend in concert with others, that despite decades of interests in managerial work, the body of knowledge is remarkably small (Cloutier et al. 2016; Dargie 1998). The subset literature that has empirically studied managers in the public sector is even more limited.

Design and methods: being there

This study employs a nested multi-case study design (Stewart 2012). Each case is a CM nested in the context of Dutch local governance. Four CMs have been studied in-depth in order to compare and contrast the findings in each case, borrowing from previous studies of managerial work (Mintzberg 1973), senior public managers (Noordegraaf 2000), local government executives (Dargie 1998), and Whitehall senior public servants (Rhodes 2011). Table 2 profiles the four cities/CMs in middle-sized to large cities in The Netherlands who agreed to be observed and have their diaries analysed. Access to Cases 1–3 was negotiated with assistance from the Dutch Association for City Managers; CM 4 was contacted independently by the author.

Table 2. Case selection.

City manager	Case 1	Case 2	Case 3	Case 4
City size	150,000+	150,000+	80,000+	600,000+
Age category	50–59	60–69	40–49	60–69
Education	Law, public administration	Law, public administration	Law	Law
Years as city manager	5–9	10–19	0–4	10–19
In studied city	0–4	10–19	0–4	0–4
Tenure in studied city	0–4	10–19	10–19	0–4
Organization full-time employee	1,500–2,000	1,000–1,500	500–1,000	10,000–15,000

Data collection: diary analysis, shadowing, and interviews

Three methods have been used to collect data: diary analysis, shadowing, and (elite) interviews. These serve to include the experience of ‘being there’ – to capture the ‘sensation’ of the context in which CMs work. Participant observations are regarded as a well-suited method for answering questions about the practiced craft of public managers (Rhodes 2016), while interviews are suitable for learning about the beliefs of the public managers. While the diary analysis presents helicopter view data, the observations produced up-close and personal data. The combination of these methods together allows for triangulation, by analysing the patterns of practice, speech and written words in parallel (Davies 2001; Gains 2015; Lilleker 2003; Oakeshott 1996). Data were collected in March–July 2016.

Diary analysis

The CMs’ 2015 diaries were content-analysed using a coding scheme inspired by similar previous studies (e.g. Dargie 1998; Fleming 2008). The diary of CM 4 was not made available and was thus omitted from the analysis.² All diary records were coded using the (pretested) coding scheme in Table 3. It uses the logic of ‘up’ (political superordinates), ‘down’ (subordinates), and ‘out’ – activities in networks outside the municipal organization (’t Hart 2014, 26–33). In addition to these three relational categories (who managers interact with),³ the latter three codes (social, travel, and other) are content based. Given that the codes are not fully mutually exclusive, reflexive judgement, grounded in observations, was required.

Table 3. Coding scheme agenda analysis.

Code	Subcode	Explanation
Up	Collective	Meetings with two or more political office holders, i.e. mayor, aldermen
	Individual	Meetings with an individual political office holder
Down	Direct Report	Meetings with employees who directly report to the city manager, i.e. directors, unit heads
	Executive Assistant	Meetings with executive assistants or entourage
	Other	Meetings with other employees, tours, and desk work
Out	Local	Activities in/for local networks
	Regional	Activities in/for regional networks
	National	Activities in/for national networks
Social		Attending social events such as diners or drinks, in a professional capacity
Travel		Time reserved for travelling
Other		Miscellaneous activities (e.g. giving a guest lecture at a university)

Shadowing

Four CMs were shadowed for five days each.⁴ The observed time totals 182 hours. I have been in and out of the field multiple times, reflecting yo-yo fieldwork (Rhodes 2011). With city halls, medieval chambers, meeting rooms, backseats of chauffeured limousines, cross-city bike rides, offices of city agencies, and permeated by information and communication technology, the field was multisited (cf: Huby, Harries, and Grant 2011). I have observed executive board meetings, city council meetings, many bilateral conversations, and CMs answering emails at their iPads. Elaborate note-taking resulted in a corpus of notes (40,000+ words) in multiple fieldwork notebooks⁵ (Emerson, Fretz, and Shaw 2011).

During the observations, I adopted a fly on the wall approach (cf. Czarniawska 2014): I shadowed the CM, was present but overwhelmingly silent at all meetings and conversations I attended. Being a fly on the wall implied maintaining a fragile balance between unobtrusively observing the CM and his various colleagues and associates (so as to minimize any observer effects), and actively managing my presence among them (so as to legitimize my presence and establish rapport in view of future access for interviews). In practice, this boiled down to acting the part of a professional stranger – who does not interrupt the ongoing activities (Agar 1996 [1980]). When asked, the CMs indicated that they had not been affected by my presence.

Interviews

The observations and diary analysis were complemented by thirty-three semi-structured interviews with twenty-five CMs and other stakeholders who operate close to CMs (some were interviewed more than once). The goal of these interviews was to enrich the observations with the CMs’ own sense-making and to include additional reflections from their counterparts on site. The interviewees were selected using the same logic of ‘up, down, and out.’ The selection includes – besides (deputy) CMs – mayors, aldermen, directors, unit heads, and executive assistants. Most of these interviews would qualify as elite interviews (Lilleker 2003; Littig 2009). Interviews usually took place in the offices of the interviewees. All interviews were recorded and transcribed verbatim. All interviewees were promised anonymity. The interviews were semi-structured, partially based on a general topic list, but also on case-specific information or triggered by the observed events and conversations.

Data analysis

Field research resulted in four notebooks with field notes, written documents (policy papers, meeting agendas, local newspapers), and photographs. Not all data are reported in the present article. Interpreting actions and implicit ‘theories-in-use’ is an elaborate process (Mintzberg 1973). Understanding ‘what was done by what was said, and what was said by what was done’ is not an easy task for ethnographers, and requires social skills and reflexivity (Van Hulst, de Graaf, and van den Brink 2012, 440). ‘How is such an unruly experience transformed into an authoritative written account?’ (cited in Van Maanen 2011, 1). Data were analysed by thematic coding, complemented by extensive reading and lexical searches (Braun and Clarke 2006). A first phase of open and axial coding resulted in a list of potential themes for analysis. A second phase of selective coding resulted in the interweaving of empirical results and the existing literature.

Empirical generalization was not the aim of the exercise. Rather, the aim was to present complex specificity in context (Rhodes 2011, 298–302). The resultant account of the four managers’ working lives can best be interpreted as provisional unfolding stories about possible realities in the world of top public managers (see also Rhodes, Hart, and Noordegraaf 2007, 225).

Findings

CMs in Dutch local government

Pollitt and Bouckaert (2011) describe The Netherlands as a consensual polity with a history of consociational democracy. Institutionally, the Dutch state can be characterized as a ‘decentralized unitary’ state (Toonen 1990). Dutch municipal government evolves around three institutions: the ‘city council,’ the ‘Board of Mayor and Aldermen,’ and the ‘mayor.’ The directly elected city council is regarded as the highest ranking body, because it decides on the appointment and dissolution of the board of mayor and aldermen, and holds the board to account (Andeweg and Galen 2002). The board of mayor and aldermen drives the executive of the municipality. The mayor is appointed by the crown, the aldermen by the city council – neither has to have a local electoral mandate. The mayor chairs both the board of mayor and aldermen and the city council. Depending on the size of the municipality, two to eight aldermen have a seat in the board.

Although ‘CM’ is most common in the international public administration literature, in Dutch local government practice, another vocabulary is used. The Dutch term for CM is ‘gemeentesecretaris,’ which literally translates as ‘municipal secretary.’ In common parlance, they are referred to as: ‘secretary general’; ‘the secretary’; ‘the sec’; or ‘madame/mister secretary.’ Most CMs have ‘general director’ (chief executive officer) as part of their formal job title. In The Netherlands, the CM is appointed and fired by the board of mayor and aldermen. S/he acts as the secretary of the board. S/he attends the board’s weekly meetings and along with the mayor signs the decision papers of the board. Unlike the other board members, the CM does not have voting powers. Every municipality is legally obliged to have a CM. The CM traditionally also acts as the civil servant-in-chief. S/he is the CEO of the municipal organization which advises the board of mayor and aldermen and executes the administration’s policies.

The CMs and their working week

The observed CMs were mid- to end-of career. They are rooted in legal university training and have spent most of their career within public administration organizations. All witnessed CMs were Caucasian. They wore a two-piece suit and tie (male) or a blazer (female). Their operating gear included an access pass, a thin folder with papers, and invariably an iPad.

The analysed aggregated diaries reflect a mean of 30.7 hours of planned activities per week. Adding up daily ad hoc meetings, an occasional crisis situation and mundane activities such as waiting for people that do not show up on time, walking through the building and fine-tuning diaries, the result is a full schedule. These recorded hours preclude a lot of the work done by CMs; reading papers, writing proposals (if they write any), making phone calls and reading emails and replying to them. A lot of these activities are postponed until the evenings or weekends (cf: Dargie 1998).

Friday is the moment when I look out to the next week and decide what I ought to do the upcoming weekend. I spend about 4 to 8 hours a weekend on preparations for the week and reading. A lot of reading happens during the weekend. I have less time for that on weekdays. I also answer emails, when I’m lagging behind. (Interview 8 March 2016)

A sample of a CM’s evening is given below (FWNB, Case 1).

A city manager’s Monday evening

18:00–19:00 Groceries at super market, cooking, and having dinner at his pied-à-terre

19:00–20:00 Emails and phone calls

20:00–21:00 Working on a proposal about organizational leadership

21:00–22:00 Private phone call with family member

22:00–23:00 Writing letter to works council

23:00–23:30 Reading papers for executive board meeting

The aggregate diaries show that the week of a CM is structured around multiple cyclical events and practices that establish a weekly rhythm, much like the rituals that Rhodes (2011) observed among Westminster elites, e.g. the minister’s red box. CMs tended to get in at 08:45 and remained at the office until 18:00 (FWNB). Mondays were dominated by bilateral meetings, starting with the mayor and direct reports. The weekly meeting of the board of mayor and aldermen and possible follow-ups or joint site visits are regular fixtures on Tuesdays. This board meeting sets the pace of the municipal organization. It was perceived by civil servants as the pivotal locus of political decision-making affecting the organization; this is where policy proposals submitted by the administration survive or get killed. Wednesdays started with the city’s management team meeting of all top public servants, another key reference point in the week. The remainder of the Wednesday is often spent on visits to city agencies or in regional networks. Thursdays are mostly devoted to managerial work inside the organization, often with time reserved for the works council, bilateral meetings with staff and internal socials, often followed by a meeting of the city council or a council committee. Fridays are days for contemplation, and various activities. Friday is often cut short till no later than 14:30, after which the weekend comes with ample readings that feed into yet another week.

Interpreting the CM’s craft

Figure 1 pictures how three CMs allocated their time during the year immediately preceding the observation. How did they distribute their attention between managing up, down and out? I discuss the observed practices and the influences they employ to get things done. This is summarized in Table 4.

Table 4. Craft of city managers.

Direction	Core of craft	Influences	Allies	Competencies
Up	Adviser-in-chief	– Experience – Standing	– Executive board advisers	– Political antennae (’t Hart and Wille 2006) – Political astuteness (Hartley et al. 2015) – Empathy (FWNB) – Dienstwissen (Lassman and Speirs [original 1918] 1994)
Down	Organizational stewardship	– Position	– Directors – Entourage	– Analytical skills (FWNB) – Public performance (FWNB) – Discipline (FWNB)
Out	Boundary spanning	– Membership – Social capital	– Colleague city managers	– Diplomacy (Rhodes 2016) – Orchestration (Bartelings et al. 2017)

Figure 1. Allocation of time in city manager’s diaries.

Managing up: adviser-in-chief to political office holders

As Figure 1 pictures, all three managers clearly prioritized their role as adviser-in-chief by their attendance in settings with political office holders (POHs). The majority of this time was passed in collective settings, with multiple POHs present such as at the weekly executive board meeting, or a (bi)weekly meeting with the mayor and the council clerk. Despite the fact that CMs spent relatively many hours in the company of POHs, these hours were largely represented by meetings that take a long time: executive board meetings and city council meetings. The remainder of the week was spent apart from POHs.

The CM is adviser-in-chief to the mayor and aldermen. This position however is not a given and must be earned:

It is not the case that when someone is a city manager and gives his opinion, that all board members will take a bow and say: ‘’Thank you, mister Secretary. That’s the way forward‘’. It is a position that needs to be earned and, to some extent, needs to be fought for. (Mayor, 3 June 2016)

CMs need to acquire ‘standing’ among POHs to fully fulfil this role, in a field also populated by political advisers or executive directors who are content experts on a certain dossier. The CMs I observed all gave both content-based and procedural advice. Much of the latter related to tactical matters of political feasibility and reputational concerns. Some would ask questions such as ‘is alderman [name] on board with this?’ or give them guidance, such as ‘communicate solely about what you are doing, not about what you intend to do.’ In doing so they draw on their experience and past performance of the CM (FWNB, Case 2), rooted in what Weber (Lassman and Speirs [original 1918] 1994, 178) coined Dienstwissen – a deep understanding of procedural and legal know-how – as well as in the political astuteness built up through years of direct exposure to the politics of executive government. Advice is delivered in real-time during both formal meetings, little asides and impromptu interactions, and is more often than not verbal (hence not retrievable through Freedom of Information Act) or via private digital media such as WhatsApp.

Giving counsel to POHs includes the element of speaking truth to power. For instance, after a series of budget cuts imposed by the board, CM3 believed that new efficiency cuts were no longer acceptable – stating ‘I don’t intend to return with new budget cuts’ to his/her colleagues before an executive board meeting (FWNB). This implies that CMs are involved in politics – the art of the possible – but not necessarily party politics. The CMs had their private political views, but showed little interest in local party statements or positions. They were more involved in ‘small-p’ politics – managing the interests of their organizations vis-à-vis political office holders. Their involvement in politics effectively makes them ‘political administrators’ (Rhodes 2016, 639).

All except CM4 resided in offices right next to the mayor and aldermen – providing them with an ‘open door’ to the powers that be at all times (FWNB, Cases 1–3). They work with and for the political executive; the legislature is kept at arm’s length – its needs are looked after by the City Clerk. CMs do attend council meetings, even though they play no part in them. Their reasons for doing so are opportunistic: ‘knowing first-hand what is going on in the council’; ‘earning bonus credits for being present’; and ‘to quickly prompt aldermen on issues’ (Interview 8 March 2016; FWNB, Case 3). This allows CMs to understand POHs more astutely because they share ‘big-P’ political experiences.

Managing down: organizational stewardship

Figure 1 shows that the observed CMs spent most time on the management of their organizations. The majority of this time was spent attending to direct reports, other managers and staff. Explicit dealings with executive assistants reflect but a mere fraction of the time, though they may be heavily under-represented in the official diary. CM3 spent almost half of his/her time (42.7 per cent) on management, especially dealing with direct reports. The diary of CM2 too reveals a priority of management. CM1 is an exception, as s/he spents most time in dealings with political office holders (31.2 per cent), although closely followed by management (29.0 per cent). All CMs spent most time with direct reports, reflected in less time spent with other people in the organization. The absence of desk work (e.g. writing, reading) in the diary is confirmed by the observations – it has little place in the diary nor in everyday life. CMs 1 and 4 did not even have a desk. The absence of desk work is consistent with findings from managers in other sectors (Tengblad 2006).

CMs actively reform the organizations in which they and their co-workers work. They do not intend to leave the organization as they found it. It is part of their job-perception to monitor and steward the organization’s current performance as well as (re)imagine its future in light of evolving public and political demands as well as financial, technological, and other contextual changes. This reflects an attitude of ‘stewardship’ vis-à-vis POHs rather than classic principal–agent relations (Schillemans 2013). ‘I think it is very important that there is the will to do better in the organisation’ (Interview 12 April 2016).

All four CMs were involved in some form of a programme aiming to improve the organization and reshape its behaviour. CM1, for example, employed a mantra of four core principles guiding the changes that s/he imagines: ‘one organisation’; ‘the challenges are central’; ‘efficient operations’; and ‘entrepreneurial employees’ (Interview 8 March 2016). S/he used these four principles regularly as beacons to guide his/her way in managing the organization, constantly articulating the principles during his meetings. S/he believed that disciplined repetition of the same story was essential, as a new managerial narrative trickles down only slowly. And s/he often told stories about his/her interventions:

This story is about a certain policy unit. An additional two million Euros had already been spent on improving their processes. This unit was not located in the municipal office, but off site. I hadn’t been there before, because every time we spoke, we met in my office. After a while I thought to visit them, because I had trouble explaining their budget exceeding without any tangible results – I was worried. So I went there. When I entered the premises, there was an office space with a terrarium and carpets on the floor. It looked like a living room. I immediately understood the problem; this was not a professional outfit; it had become a proxy family unit. Within the month we had cancelled the lease, and they were in my sight at the municipal offices. (Interview 17 March 2016)

The most important lever to create organizational capacity for CMs is the board of directors, the ‘management team’ of the city administration (cf: Steyvers, Reynaert, and Block 2010). CM2 spoke for all four when s/he observed that ‘The quality of the organisation starts with the quality of the board of directors’ (Interview 22 April 2016). CMs were keen on having the right people in this board: CM1 let go of two directors on his board and hired a new director in his/her first months in office.

In board meetings, CMs lead discussions, seek consensus, and, if need be, make unilateral decisions on organizational matters. To be effective in this role requires effective chairmanship: bringing people with diverse outlooks and interests on board. One director commented on how his fellow director developed after he was appointed CM.

Director:

I have always perceived him as very critical, almost as an outsider

van Dorp:

As a devil’s advocate, you mean?

Director:

Yes. That’s one of the reasons why he grew so fast. Also the reason why I stimulated him to go in senior management. [….] By now, he has let go of that role. He currently is a connector. I think it’s impressive when someone develops like that. (Interview 2 June 2016)

In all cities, the management team monitored the state of the organization based on a dashboard of parameters and structured information streams. They analyse organizational performance by both ‘hard’ and ‘soft’ indicators. ‘Hard’ indicator parameters include: absenteeism (absolute and compared to national average); annual employees’ appreciation survey; exhaustion of budgets; external hiring; and full-time employee formation (FWNB). ‘Soft’ indicators are gossip and hearsay about how well pivotal actors or units perform. This information reaches the CM through (informal) talk with civil servants, POHs, and members of his/her entourage. When certain units underperform, or certain dossiers are regarded as politically explosive, a CM may declare them Chefsache and place them under his/her direct supervision. At the same time, CMs acknowledge that control of their large organizations is an illusion. They ‘steer’ and ‘adjust’ the meta strategy of the organization, but the execution of these directions is in the hands of others (FWNB).

Managing out

‘Out’ – activities in networks – tended to come third in the attention hierarchy. Van Der Steen et al. (2012) note that regional cooperation is highly contingent on the CM and his/her personal priorities. CMs1 and 2 spent 17.3 per cent and 18.6 per cent of their time in networking activities. CM1 divided his/her time across local, regional and national networks, while CM2 focussed on regional and national networks. CM3 spent comparatively little time on external management (11.5 per cent). That 11.5 per cent was predominantly spent on large national events that virtually all Dutch CMs or local government managers attend, as well as on a monthly regional network meeting with colleagues (FWNB, Case 3). Thus, networking activities thus were less prominent in his diary and were fairly isolated rather than habitual.

As administrator in chief of their respective organizations, boundary spanning is part of their craft. The pivotal position of CMs – linking POHs with the organization – inherently implies connecting different actors and rationalities. Boundary spanners are loosely defined as ‘individuals who work across different organizational cultures and exercise influence through formal and informal channels in order to strengthen the connections between actors’ (Guarneros-Meza and Martin 2016, 240).

CMs connect and lobby with colleagues in neighbouring cities and governance levels in regional cooperation and lobby. They come in various guises, but none of them work in splendid isolation – beautifully expressed by the prominence of the meeting table in their offices (FWNB). Of course they are by and large preoccupied with the POH–organization nexus. Given their pivotal position in the organization and role as secretary of the board of mayor and aldermen, this is hardly surprising. Still, spanning boundaries in networks is part of their craft.

All observed CMs engage in regional networks and intra-city alliances. Jokingly referred to as ‘playing outside’ (FWNB), these networks include in particular fellow CMs and other administrators, and sometimes POHs. These networks range from collaborative governance networks in which municipalities co-produce public services (FWNB, Cases 2 and 3) to informal deliberation fora (FWNB, Cases 1 and 4). This requires CMs to be diplomats on behalf of their administration (cf: Rhodes 2016). The instrumental function of permanent diplomacy is believed to be important to CMs. A CM explained: ‘you first need to have a network in order to get things organised’ (Interview, 18 March 2016). Bartelings et al. (2017) use the language of ‘orchestrational’ work when referring to public managers that manage collaborative networks/chains. In a similar vein, CM4 stressed the importance of ‘being able to make arrangements’ in a public address to fellow CMs (FWNB). Three CMs indicated that they would engage in lobby among administrators and/or politicians in ministerial departments in The Hague (FWNB, Cases 1, 2, 4). One CM explained that s/he would go to lunch with a director general, to ‘put my city on the map’ (FWNB, Case 1). Another CM joined committees of the Dutch Association of Municipalities (VNG) to engage with other municipal administrators.

The boundary spanning work of the CM should not be exaggerated. They all agree that collaborative governance is important, but the demands of managing up and managing down can and often do take over. The ‘rules of the game’ in local government seem to favour internal affairs over external activities. CM1 started his/her position by actively investing in regional involvement, but paused some of these efforts when his/her vertical managerial tasks seemed more pressing. CM3 too started in his/her current position and prioritized vertical managerial work over horizontal networking activities. In contrast, CM2, who had been in office since 2004, made ample time for regional and national networks. This may hint at a ‘life-time-cycle’-effect suggesting that upon taking office, CMs first attend internal matters and try to get their organizations in shape, before they get involved in external networks.

Other activities: social, travel, other

The final categories, ‘social’ (average 8.0 per cent); ‘travel’ (average 6.4 per cent); and ‘other’ (average 6.5 per cent) consumed only a relatively modest part of the diary. Within these categories, there are relatively large differences. For example, CM1 was more prominently engaged in ‘social’ activities than others (10.5 per cent) while CMs 2 and 3 devoted 7.7 per cent and 5.7 per cent of their appointments to social events, respectively. These activities usually include receptions, dinners, and lunches. CM2 spends more time (8.7 per cent vs. 6.0 per cent and 4.6 per cent respectively) on the road than others do, which is largely explained by his/her prioritization of networking activities. Coherently, the diary of CM3, who spends less time on activities in/for networks, reflects 4.6 per cent time spent on travelling. Travelling is of course no idle time. The CMs in the three largest cities (1, 2, and 4) have access to chauffeured cars which permits working while travelling. It is used for reading, phone calls and answering emails (FWNB, Cases 2 and 4).

Table 4 summarizes the craft of Dutch CMs, the influences they employ, the allies they have to gather to do their work, and the competencies this requires.

Discussion: greedy hierarchies

All three CMs’ diaries reflect a dual focus on both managing the political office holders and management of the organization. The activities coded as ‘up’ and ‘down’ and the combination of the two are recognized as the most important. A CM remarks in conclusion: ‘The core of my craft as city manager is to be a bridge between the board of mayor and aldermen and the organisation’ (Interview 31 March 2016). Using Noordegraaf’s (2000, 243–246) typology, all four CMs best fit the public organization manager profile, rather than political or policy managers. Public organization managers are focused on the organizations they lead; involvement in political arenas and processes is markedly less time-consuming. Though ‘managing up’ is expected of such managers, and political nous/astuteness a key part of the skill set, their strategic focus is first and foremost ‘downwards’ – let alone outwards.

The core of their craft materializes within the organizations they manage. Managing ‘down’ is therefore what the three CMs spend most of their time on (up to 42.7 per cent). They have an organizational ‘mental agenda’ and prioritize organizational change over external networking activities. CMs 1 and 3 were both relatively new in their positions, and each had reorganization ambitions, aimed at ‘getting the organization into shape,’ either mentally or both mentally and structurally. Implementing their view of a ‘good municipal organization’ formed a raison d’être for the managers. For the longer-serving CM2 this was less of a priority. His/her focus in managing down was on ‘continual strengthening’ and preserving the existing qualities of the organization; his/her prime mental focus was in managing ‘out’ aimed at shoring up and shaping regional cooperation processes. Yet his/her actual pattern of attention and attendance (18.6 per cent in networks) did not differ all that drastically from CMs 1 and 3 (respectively 17.3 per cent and 11.5 per cent).

I have described the work and behaviour of public managers as a shared craft, something all four CMs relate to. As follows from the analysis above, public managers are individuals who each show unique expressions of beliefs and behaviour in their respective social and institutional contexts. Though they are not dissimilar, they are not one and the same. Ultimately, this typology is not about a hierarchy of tasks, roles, and competencies. The point is that senior public manager such as CMs deploy a repertoire of skills, tricks of the trade, and rules of thumb. Mastering the craft means that public managers can in situ judge to apply the right mix of skills and interpretations at a given moment in a local context. My fieldwork however suggests that the central tension Dutch CMs experience in doing so is that between a ‘greedy’ vertical axis of managing up and down, and a strategically important but always somewhat less pressing horizontal axis of managing out.

Conclusion

This article set out to address a less than fully developed aspect of public management research: examining close up how senior public managers – in this instance Dutch local government CEOs – give meaning to their craft in their everyday behaviour. Using diary analysis and ethnographic methodology, four CMs in Dutch middle-large and large cities were analysed. The findings show that these CMs consistently allocate most of their time to management of their organizations (down) and advising political office holders (up). Taking part in networks (out) is believed to be important, but the diary analysis shows that less time is spent on networking activities than is spent on the former levers. The bulk of their work takes place within the hierarchy, despite the deafening chorus of the Network Society (Agranoff 2006; Ansell and Gash 2008; Pollitt and Bouckaert 2011). I do not argue that boundary spanning does not happen, but rather that the observed CMs themselves engaged much less in boundary spanning work than would be expected. The rhetoric of governance seems at odds with the everyday reality of government in the working life of CMs. To date, their craft continues to be mainly organized along the vertical up-down axis – emphasizing traditional elements of the bureaucratic craft (see also Olsen 2006; Rhodes 2016).

At the same time, we know from a great deal of research that network governance and collaborative problem-solving has become common place in many government jurisdictions and policy sectors. This raises the question who gives strategic direction to this work and undertakes the meta-governance work associated with it – if not the CEO of the municipal organization (Torfing et al. 2012). CMs may for example deliberately recruit and empower designated boundary spanners or delegate responsibility for network management down the hierarchy, but in the current study there was not much evidence to suggest that this was indeed the case. In the four cases studied here, CMs in effect were largely ‘missing in action’ when it came to managing out. They did realize its strategic importance and paid lip service to it, but in their day-to-day routines they are largely trapped in the demands of their hierarchy-related roles as chief adviser and CEO of the municipal organization.

Acknowledgements

I would like to thank Paul ’t Hart, Mirko Noordegraaf, Rod Rhodes, René Grotens, Anna Killick, and three anonymous reviewers for their insightful comments on an earlier version of this article.

Disclosure statement

No potential conflict of interest was reported by the author.

Additional information

Notes on contributors

Erik-Jan van Dorp

Erik-Jan van Dorp is a researcher based at the Utrecht School of Governance working on a comparative research project on cabinet government in collaboration with Griffith University. His interests include core executive studies, public governance, and managerial work. He co-authored a professional monograph on the craft of city managers in collaboration with the Netherlands School of Public Administration.

Notes

1. Planning, Organizing, Staffing, Directing, Coordinating, Reporting, and Budgeting.

2. Triangulation in Case 4 was limited accordingly.

3. Whether a given diary entry is perceived as up/down/out by city managers, is not solely determined by relational categories, but is also contingent on, e.g. practices and the content of the recorded diary items. The applied method cannot easily capture practices and the content of the diary entries, resulting in a limited interpretation.

4. City manager 4 was shadowed for four days. On the fifth day, the deputy city manager was shadowed in the same organizational context.

5. When referring to notes in my fieldwork notebooks, I use (FWNB, case X) as a reference.