Boundary-spanning strategies for aligning institutional logics: a typology

ABSTRACT

This article critically examines strategies used by boundary spanners to align the institutional logics of bureaucracy, management and networks in citizen-state interactions. In-depth interviews conducted within the Dutch municipality of Rotterdam reveal that boundary spanners use entrepreneurial, mediation, and hierarchical strategies to align institutional logics. By providing insight into the strategic toolbox of boundary spanners and the perceived effectiveness of these tools, this article enhances empirical understanding of how the interplay between older and newer institutional logics within public organisations takes shape and how boundary spanners make strategic use of hierarchy to overcome institutional barriers.

KEYWORDS:

• Institutional logics
• boundary spanning
• networks
• new public governance

Introduction

The complex nature of today’s problems has led governments to rethink their approach to the design and implementation of policies and services. An impressive growth of scholarly attention for post-NPM literature, such as network management and collaborative governance, support the idea that governance networks and collaborative governance have become the next ‘big thing’ in public sector management (Agranoff 2006; Ansell and Gash 2008; Klijn and Koppenjan 2015). The New Public Governance (NPG) paradigm is put forward as a viable alternative—a new and dominant logic—to the dominant Traditional Public Administration (TPA) and New Public Management (NPM) paradigms (Osborne 2007; Pestoff 2012; Torfing et al. 2012). Rooted in the literature on networks and co-production, the NPG paradigm’s central assumption is that citizens are no longer to be treated as passive voters or consumers, but as active co-producers that participate in policy-making networks and contribute to the delivery of public services (Osborne 2007; Fledderus, Brandsen, and Honingh 2014). Although we see a sharp increase of scholarly and political attention on networks and co-production, empirical research falls short in exploring how the interplay between the logic of hierarchy (TPA), market (NPM) and networks (NPG) within public organisations takes shape (see Olsen 2010; Laegreid 2016). While many politicians and policy officials certainly ‘talk the walk’ along the lines of the popular scholarly refrain about the importance of collaboration and the rise of the network society (Buser 2013; Nederhand and Van Meerkerk 2018), the question is whether policy officials also ‘walk the talk’ in day-to-day encounters with citizens (see Van Dorp 2018).

This article: the art of boundary spanning in aligning institutional logics

To see how the addition of the newer NPG logic combines and aligns with the older institutional logics of TPA and NPM, it is important to take the work of boundary spanners into account. Due to their strategic positioning in-between policy officials and local communities (Van Meerkerk 2014), boundary spanners are able to play a key role in enabling a productive interplay between ‘’incompatible prescriptions’’ of different institutional logics (Williams 2002; Meyer and Hammerschmid 2006). By linking organisations with their environments, boundary spanners are engaged in building and maintaining sustainable networks (Tushman and Scanlan 1981). As building networks of sustainable relations is very important to the collaborative nature of the NPG paradigm (Osborne 2007, 2010), boundary spanning is the key focus of managerial action within the NPG paradigm (Van Meerkerk and Edelenbos 2018; Osborne 2010). By using their agency, boundary spanners actively foster and shape citizen-government encounters. Although the need to mobilise and activate one’s own organisation to make these encounters work has been discussed frequently (see Streeck and Thelen 2005; Klijn and Koppenjan 2015; Nederhand, Bekkers, and Voorberg 2016), the current literature on boundary spanning lacks a clear typology of boundary-spanning strategies that are used to align clashing institutional logics within public organisations. This article therefore answers the following research question: what strategies do boundary spanners deploy to organisationally align the different institutional logics (hierarchy, market, networks) in citizen-state interactions and do they succeed in their efforts? By providing insight into the strategic toolbox of boundary spanners and the perceived effectiveness of these tools, this article contributes not only to increasing our empirical understanding of how the interplay between older and newer institutional logics within public organisations takes shape (do they peacefully coexist or clash), but also to the effectiveness of boundary spanners in aligning clashing logics by preventing or overcoming barriers. To achieve this, we build on different bodies of the literature. Section 2 describes the literature on institutional logics and boundary spanners. Section 3 and 4 describe our case study approach that focuses on the exemplary case of the municipality of Rotterdam. The results of our analysis are presented in Sections 5 and 6. In the final section, we address important conclusions and limitations and consider avenues for future research.

Institutional logics: TPA, NPM and NPG

The notion of institutional logics is used to provide a bridge between the macro-structural paradigms within organisations and the study of individual behaviour of boundary spanners and policy officials on the micro level (Thornton and Ocasio 2008). Institutional logics are a set of intra-organisational rules, routines and sanctions that individuals in particular contexts create and recreate in such a way that their behaviour is to some extent regularised and predictable (Jackall 1988). This implies the behaviour of policy officials to be structured along the lines of historically grown and accepted rule-based practices. These practices, which are based on dominant organisational paradigms or institutional logics, function as a common meaning system that adds to the stability and predictability of individual interactions (Scott 1995). Institutional logics are thus the sources of legitimacy that provide policy officials with a sense of order (Seo and Creed 2002). This sense of order in modern public organisations has traditionally been provided by the bureaucratic and the managerial logic (Christensen & Lægreid, 2007; Pollitt, 2003). The behavioural repertoire of the more recently introduced NPG is not necessarily congruent with the repertoires represented by the bureaucratic (TPA) and/or the managerial (NPM) logic. We can identify four characteristics of the bureaucratic and managerial logics that can act as institutional barriers that boundary spanners face in their collaborative efforts. These tensions with the NPG network logic will be specified after a short account of what the bureaucratic and managerial logics entails.

Bureaucratic logic and potential barriers to network logic

The bureaucratic logic is grounded in classical public administration theories in which government organisations are characterised as impersonal rational systems that prescribe neutral behaviour for policy officials (Weber 1978). From this logic, the function of policy officials should be standardised and executed along the lines of predictable processes and rules. The explicit standardisation of functions makes interaction with the bureaucratic organisation perfectly predictable. This predictability is also safeguarded by the presence of impersonal and stable rules (Dror 1968; Wilson 1989). These rules shield citizens from arbitrariness, power abuse and personal whims of policy officials (Bartels 2013). In this respect, it is also important that the allocation of resources should take place along the lines of clear regulations. Policy officials rely on rigid administrative guidelines, which perfectly fit public values, such as impartiality and equality. As a consequence, policy officials are intentionally internally oriented. Political goals of officeholders are favoured, and so political decisions guide what policy officials should do.

The primary characteristic of this logic that can act as a barrier to the NPG logic of networks is standardisation. Whereas NPG entails the development and negotiation of goals and policies during interaction with citizens and other stakeholders (Osborne 2010; Klijn and Koppenjan 2012), the bureaucratic logic would like policy officials to reason from existing regulations and politically authorised policy programmes. This benefits the predictability and equality of public service provisions; however, tailor-made solutions that are more appropriate to the context at hand become more difficult. The second major characteristic that potentially impedes network-working is the internal orientation of policy officials. The emphasis on both administrative procedures and serving the political officeholders makes policy officials internally oriented. Hence, political decisions guide the actions of policy officials. This internal orientation on policy programmes and rules also enables policy officials to treat each citizen alike. In contrast, the logic of networks requires an openness and external focus from policy officials (Torfing et al. 2012; Klijn and Koppenjan 2015).

Managerial logic and potential barriers to network logic

The managerial logic is grounded in the neoliberal approach of NPM. While it is difficult to provide a definitive image of NPM (Pollitt, van Thiel, and Homburg 2007; see Hood 1991; Lane 2000), most scholars agree on the main features. These main features are the focus on improving the efficiency and effectiveness of public service delivery through management of processes and systems. The use of business instruments (strategic and performance management techniques, performance indicators) is crucial to the conceptualisation of NPM (Hood 1991). After politicians have defined and set the main policy goals for the bureaucratic organisation, public managers are expected to manage the delivery of these policy goals within this budget (Du Gay 2008). Consequently, problems are translated into managerial targets that reflect the internal organisation. Financial resources are subsequently disaggregated into specific organisational units that should realise these targets and results. Results measured in terms of outputs and outcomes are important for purposes of accountability and efficiency (Hood 1991; Haque 2007). Therefore, setting specific performance indicators for each department enables managers and politicians to critically monitor and evaluate their performance.

The primary characteristic of this logic that potentially serves as a barrier to the network logic is functional specialisation. Due to the focus on performance information and monitoring, policy ambitions are broken down into a large set of measurable smaller tasks that are allocated among departments and responsible policy officials. As a result, decision-making power and financial resources are distributed within the organisation. This potentially impedes NPG as local needs of citizens usually cut across the responsibilities of individual policy officials and departments (Bartels 2016). The second major characteristic and barrier is the result-orientation of policy officials. Achieving managerial targets within budget is key for policy officials as that is what they are held accountable for. This potentially leaves little room and time for policy officials to take-up extra tasks that come up during interactions with citizens, and therefore fall outside their performance indicators (see Bartels 2016; Michels and De Graaf 2010).

Aligning institutional logics: the art of boundary spanning

As institutional logics—with contradictory rules and routines—are confirmed or changed during interactions (Edelenbos 2005), we will take a closer look at the role and agency of boundary spanners in aligning the institutional logics. The boundary-spanning literature depicts boundary spanners as connectors of people and processes: they facilitate contacts between internal and external parties. Different ideas exist in the literature about when a person qualifies as being a boundary spanner. Some scholars focus on holding a structural position within the organisational structure (Fernandez and Gould 1994; Tushman and Scanlan 1981), while others define boundary spanning as an activity that is not bound to a particular organisational position (Levina and Vaast 2005; Quick and Feldman 2014). In this article, we focus on boundary spanners that hold strategic brokerage positions. As a result of the central positioning of the boundary spanners in-between their organisation and its environment, boundary spanners are able to shape perceptions by controlling information and access to various parts of the network (Williams 2002; Levina and Vaast 2005). Hence, boundary spanning is commonly thought of as the management of the interface between organisations and their environments (Sullivan and Skelcher 2002; Williams 2002). To accomplish a better fit between organisation and environment, boundary spanners (1) connect different people and processes across organisational boundaries, (2) select relevant information on both sides of the boundary and (3) translate this information to the other side of the boundary (Van Meerkerk 2014). They are thus involved in a two-step informational flow, collecting and transferring information from one side of the boundary to the other. However, boundary spanning is more than a simple matter of translation between different ‘worlds’ (e.g., internal and external, professional and amateur). Key to managing and coordinating the interface is not only the ability to connect, but also the ability to align organisations and actors of different backgrounds (Lodge and Wegrich 2014). The central positioning of boundary spanners enables them to strategically shape perceptions through controlling information and to access various parts of the network. The rapidly expanding scholarship on boundary spanning has recently put more emphasis on the innovative component of boundary spanning: to transform particular institutional arrangements (Van Meerkerk and Edelenbos 2018; Baker, 2008). Boundary spanners are considered to be entrepreneurs and innovators in the sense that they try to link different policy issues and policy streams across boundaries. They highlight contradictions in institutional rule-sets to make a case for changing existing routines, or they attempt to recognise and exploit policy windows to create turns in the ‘paths’ of internal routines (Kingdon 1984). These kinds of strategies have received far less attention within the public administration literature on boundary spanning. Consequently, insight into the strategies that boundary spanners employ to prevent or overcome institutional barriers within public organisations is limited. Therefore, in this article, we will identify and construct a typology of boundary-spanning strategies that play a role within public organisations.

Research site

This research is conducted within the Dutch municipality of Rotterdam. This municipality is internationally known for its experience with partnership working due to its involvement with the Rotterdam Harbour. Moreover, Rotterdam was following a political programme highlighting the importance of NPG by stressing citizen activation and participation (Municipality of Rotterdam 2014). In order to diminish the legitimacy gap between local government and citizens, the municipality of Rotterdam introduced the so-called GGW-approach in 2006 (Gebiedsgericht Werken). The GGW approach gives primacy to local experiences as opposed to the institutionalised understandings and routines of policy officials (see Bartels 2016). In 2010, a large reorganisation completed the shift of the administrative organisation to fit this new governance philosophy of putting the local needs of districts and citizens first. Hence, the municipality has moved from a ‘policy-centred’ organisation towards a ‘citizen-centred’ organisation (Rekenkamer Rotterdam 2011). Policy officials of the municipality or Rotterdam thus build on years of experience with collaborative working (Van Steenbergen et al. 2017). This is not only mirrored in the current political programme of the municipality, but also in the institutional structure built around the GGW-approach. Therefore, the municipality of Rotterdam is exemplary in studying how different institutional logics (hierarchy, market, networks) in citizen-state interactions take shape.

This study focuses on public officials working in the municipality of Rotterdam as district managers. District managers are explicitly employed to function as the link between the perspectives and interests of a specific district (environment) and the perspectives and interests of the policy-departments of the city (organisation). They are responsible for collecting information on the local needs of citizens and transferring this information to policy officials and vice versa. The work of district managers is thus characterised not only by representing the local needs of citizens to policy officials, but also by representing the policies of these officials to citizens living in the relevant districts. As such, they connect two different worlds. District managers thus not only hold a structural brokerage position within the organisational structure (Fleming, Mingo, and Chen 2007), their job description also implies a process of constantly interacting back and forth in-between various actors: the very activity of boundary spanning (Quick and Feldman 2014). District managers play a key role in the GGW approach. This approach divides the urban area of Rotterdam into 14 districts. Each of the 14 districts is represented within the municipal organisation by a separate district-department in which district managers are employed. Besides spanning boundaries between policy officials and communities, district managers collaborate with a district-committee that consists of elected citizens living in the specific district and who communicate the local needs that they want the policy officials to take up. These needs are in turn taken up by district managers who are responsible to find a solution for these local problems in their district through informing and calling into action policy officials and, if necessary, the political head of the municipality. The district managers have no formal power and budget; thus, they rely solely on the commitment of policy officials of the policy-departments. Consequently, for the GGW-approach to work, the district managers have to align the interests of policy officials and districts.

Methods

In this study, we interviewed 16 district managers. Respondents were selected on theoretical grounds: respondents are all working as boundary spanners who have encounters with both citizens from their district and with policy officials working in the policy-departments. This selection made it possible to study the barriers that boundary spanners experience in their encounters with policy officials and to examine how they strategised upon these barriers. Within this selection frame, this study aimed for a sample consisting of district managers who are spread over the different city-districts to be able to grasp a variety of experiences. We succeeded in interviewing district managers from 12 of the 14 district-departments in the municipality of Rotterdam. Although Hoek van Holland and Charlois were not included, they were indirectly represented by the 12 district managers, who were also familiar with these districts. In four interviews the district manager brought along a colleague (district IJsselmonde, Prins Alexander, Feijenoord and Overschie). The respondents were introduced to the study by either the director of their own district or the director of another district and were then requested to participate as part of a broader evaluation of the GGW-approach within the municipality.

This study focused on the experience of boundary spanners. Thus, to study these experiences, we focused on detailed examples and stories about encounters with policy officials. Stories present highly textured depictions of practices in which the norms, beliefs, and decision rules that guide actions and choices become clear (Maynard-Moody and Musheno 2003). This method allows respondents to illustrate what particular situations call for certain routines and how the specifics of a case fit or do not fit standard practices (see Bartels 2013; Raaphorst 2018). Within an interview setting, respondents were asked to tell stories about struggles they had experienced. Additionally, we asked respondents to come up with examples and stories on how they strategised facing these struggles. To make valid and replicable inferences, we analysed the transcribed interviews by making use of the step-by-step approach of the constant comparative method to identify boundary-spanning strategies (Boeije 2002). We first segmented our data into relevant categories by making use of an open coding process. Open coding is the process of breaking down, examining, comparing, conceptualising and categorising data (Strauss and Corbin 1990). The fragments were then compared among each other, grouped into categories dealing with the same subject, and labelled with a code. The list of codes was then grouped in categories by means of axial coding and reassembled into the findings that are presented in this article. We used ATLAS.ti software for qualitative data analysis. The full coding scheme is presented in Appendix A. We now turn to the discussion of the main patterns and most exemplary stories. See Appendix B for additional supporting quotes.

Clashing institutional logics

The public management practice is characterised by the coexistence of multiple institutional logics. How does the logic of the more recently introduced NPG paradigm combine with the older institutional logics of the TPA and NPM paradigm? This section explores the four barriers that we theoretically expect boundary spanners to encounter as the result of clashing logics. We first turn to the standardisation barrier.

Standardisation barrier (experienced by 12 respondents)

It often happens that boundary spanners clash with policy officials who are ‘guarding’ current standards in terms of rules and policies. In the experience of boundary spanners, policy officials do not reason from the perspective of the problem but from the perspective of existing rules and policies. Consequently, many solutions do not quite fit the needs of citizens from the districts. This is a two-fold problem: a strict interpretation of standards, and also a controlling approach of policy officials ‘guarding’ these standards. According to boundary spanners, guarding these standards seems more important for policy officials then deviating from them because of an external need or initiative. As long as requests and ideas of citizens can be arranged within existing rules and standards, they are realised easily. However, when a request or initiative falls outside this framework, “it gets very tough’’. The following story is exemplary for this case.

‘’When citizen initiatives deviate from existing standards and rules, it gets very tough. For example, a group of citizens had the idea of installing new bicycle parking facilities in their neighborhood. This is of course really nice. They say to me: we will hire an artist to design these parking facilities. However, this is incredible difficult for policy officials, as they want to adhere to the standard way of designing bicycle facilities in Rotterdam. […] The same goes for the design of paving stones. Policy officials don’t want these to include wood for a whole range of reasons… It doesn’t fit the pre-defined standard. Well, and so on.’’ (Respondent 14)

The respondents who experienced this barrier indicated, like the above story aptly shows, that it often seems easier for policy officials to point at reasons why something is not possible, than to adapt the standard solution. Hence, citizens are often urged to adapt their idea or initiative so that it fits policy standards.

Internal-orientation barrier (experienced by 11 respondents)

The following story shows that nearly three quarters of the boundary spanners also feel hindered because the primary point of reference for public servants from the policy departments is the inner bureaucracy and political process. Boundary spanners argue that policy officials are so busy with making sure the internal machinery runs well that they often lose sight of the more personal story of initiatives and requests from citizens.

‘’The alderman promised the district committee that he would send them a letter. Well, that letter was important, also for other citizens from the district. The letter would enable the district committee to show what they had arranged with the alderman. It took a very long time for the letter to arrive. Each time, there was someone who felt that something else had to be included in the letter. While I was thinking like, this letter has to get through otherwise there will be a quarrel in the district, and with … So that is… This is where you really notice that the logic of the community and the municipality are very different. Policy officials say, ‘’yes, yes, but if this is not part of the letter, then the alderman… [gets in trouble]. That political party will definitely ask questions.’’ All the time I am waiting and saying like: hurry up, this letter needs to become public. We have to get started. So that… And eventually the alderman would profit. But OK, this is my experience, and I also get that the ‘other world’ looks different.’’ (Respondent 9)

Boundary spanners describe policy officials as being mainly oriented towards their organisation and the proper answering of internal questions. It comes more natural for policy officials to take a formal perspective on matters than to take a more personal approach and look at initiatives proposed by citizens as actions taken by ‘real people’. However, this barrier is hardly experienced as an issue of principle regarding equal treatment of citizens, but more as a pragmatic working-attitude.

Functional specification barrier (experienced by 9 respondents)

Functional specification is experienced as being a barrier for more than half of the boundary spanners. Because functions and tasks of policy officials are clearly demarcated, power and resources are spread over different units. Hence, a lot of interdependencies exist within the organisation. The following story shows that it is hard to reconnect the resources and responsibilities in line with priorities of citizens within the districts. Such issues can linger very long simply because it is difficult to combine the various resources from different units, and, moreover, there is no real mechanism for doing that.

‘’It is difficult to get a hold on the cash flow that enables you to say: Ok, I can promise on this table that this issue will be handled next week. You can only promise that to citizens if you possess the money. These days, there are a lot of dependencies and a major distance between the policy officials ‘on the tables’ and policy officials guarding the cash flow. (…) [These] policy officials work in one of the big towers, you know, somewhere in the new building. Their systems show that somewhere in another district a street is in a much worse condition, so they won’t provide a budget for our street. (…) Often my contact person within the policy department doesn’t even know who this asset manager is. How are we supposed to contact this person? How can we arrange that not the street in district A, but the street in district B… [is fixed]? Well, that is all at a distance.’’ (Respondent 1)

Interestingly, more than one third of the boundary spanners didn’t perceive the functional specification barrier to be problematic at all. They perceived the spread of functions and resources as a structural feature of organisations that not inherently impedes collaborative working. In fact, this feature could be made productive by public officials—if it wasn’t for their, for-instance, result-oriented or inward-oriented behaviour or attitude. These boundary spanners thus perceived other more behavioural barriers to cause problems for collaborative working.

Result-orientation barrier (experienced by 12 respondents)

Policy officials and managers are judged on realising their programme; however, seeking cooperation with citizens and entrepreneurs is something ‘extra’ and, therefore, not always part of their performance agreements. Because it is not ‘in the programme’, three quarters of the boundary spanners feel that when a collaborative activity does not add to a performance target of policy officials, few internal managers feel responsible for it; for them, it literally does not count. The following story shows that processes of co-production could easily go wrong when they are not anchored in the performance agreements of policy officials.

‘’Take for example the redesign of a street. We did that in co-production with citizens. (…) Let’s say you work for the policy department Urban Development, and you have drafted this plan together with citizens while also taking into account safety and social development arguments. This could imply that, for instance, the parking spots are not entirely optimal placed, and the spots are 10 cm shorter than the norm. Then a senior [policy officials who works as] urban planner sees the plan and dismisses it immediately, with all its citizen ideas, regardless of whether the citizens thought it was the best plan possible, given all their other interests. This person is not held accountable for integral working [in co-production with citizens], this was not specified in his performance indicators. You should not give this one person the power to destroy the integral plan just with one point-of-view.’’ (Respondent 12)

Furthermore, part of the orientation on results is the allocation of budgets to specific managerial targets. This makes the municipal budget quite inflexible; there is little space for re-allocating money to purposes outside a pre-defined programme. Boundary spanners indicate that because of the labelling of money, there is not always space for new issues that pop-up. This is also part of the result-orientation barrier that they experience.

Four institutional barriers to NPG

Our findings show that traditional logics produce significant barriers to the work of boundary spanners. We found little evidence of major objections to the requests of boundary spanners. Most initiatives were not far-fetched and fitted within existing policy goals. That shows how traditional institutional logics provide an ever-present but also implicit barrier for aligning actions of policy officials with demands of local citizens. Of course, not all encounters with policy officials are cumbersome. Some boundary spanners explicitly stress the good-intentions of many of the policy officials. Nevertheless, that does not take away from the tensions experienced almost on a daily basis. We will now examine the strategies that boundary spanners use to align the clashing institutional logics by preventing and/or overcoming the barriers that they experience.

Boundary-spanning strategies

Within our case, we can identify three types of strategies that boundary spanners use to avoid or overcome the barriers they experience: an entrepreneurial strategy, a mediating strategy, and a hierarchical strategy. In the next section, these strategies are described, including their effectiveness as perceived by the boundary spanners.¹

Entrepreneurial strategy (used by 7 respondents)

This strategy involves taking a creative approach to rules and routines, but also to contacts. Boundary spanners who employ this strategy strategically think about the best starting point for the request or initiative from citizens. They think carefully about whom to go to and what battles they would pick. When a policy official fails to properly react to the request, boundary spanners try another way into the system by approaching officials at different positions. The following story shows an example of this entrepreneurial approach:

‘’Take for example the MOE-landers, that formed a problem in our district. (…) It takes a lot of time before the policy officials working to get their policies moving. Rules and all sorts of frameworks should be followed. That is quite difficult sometimes. Simultaneously, like I said before, I try to work at the boundaries of what is appropriate, I am inventive and creative. You are going to look at… take the consultation hour that we set-up for the MOE-landers. They [policy officials] then say like, the aldermen don’t want to facilitate special target groups. I think, OK, so be it. Policy makers will not deviate. I think OK, but it has also something to do with integration and there may be room within the policy framework of integration for something like this. So you have to be very creative and open-minded, but also show courage.’’ (Respondent 16)

This story shows that this particular boundary spanner detected a policy window and acted upon that opportunity. Boundary spanners thus attempt to recognise and exploit policy windows to act upon resource opportunities and couple solutions to existing problems. This strategy is used before actual barriers are experienced (e.g., approaching the right policy officials) as well as after barriers have occurred (e.g., approaching other policy officials). Hence, experience about previous barriers is used to inform future behaviour, for example, not to approach certain policy officials again. By seeking another way in (as the above story shows), boundary spanners try to avoid all four types of existing barriers. While all boundary spanners who use this strategy agree with its effectiveness to avoid and get around barriers, two boundary spanners were more pessimistic than the others by stressing that it takes considerable time to get in contact with the right person. The other—more positive—respondents indicated that if this strategy is to work, the formulation of the request is very important. Issues should be framed and split up into smaller parts to match the interests of the receiver.

Mediating strategy (used by 14 respondents)

The mediation strategy is used by boundary spanners to find common ground for the development of a solution. The mediating strategy can be applied in two ways. The first way involves the usage of an argumentative approach to persuade officials. Boundary spanners use terms like engaging in the battle or starting the fight to describe these encounters. They try to make the policy officials see the importance of bending existing rules and policies. The second way involves trust-building and facilitating compromises by listening and showing respect to the interests of policy officials. The following story shows the importance of keeping in mind the interests of their colleagues. On the basis of this knowledge, boundary spanners try to facilitate a shared understanding on which they themselves and the policy officials can build. Facilitating a shared understanding becomes easier if the relationship between the policy officials and boundary spanners is good. Therefore, investing in relationships beforehand is part of the strategy boundary spanners use to avoid future barriers (see Appendix for supporting quotes). Boundary spanners then have a better basis to start the negotiation and mediation process. This strategy is also used to overcome barriers by exploring common ground and entering in a negotiation process as the following story shows.

‘’The art is to get as many things done as possible. In all fairness, aye. Knowing each other’s world on the basis of arguments is very important. Contacts and relations are key. They [policy officials] have to get that there is a problem in the neighbourhood, and simultaneously I have to get that there are certain policy assignments at stake. In the organization, there should also be scored. (…) You have to get that policy officials take a lot of [internal] interests into consideration. If you get that, you have created room for a good conversation. In this case, you don’t work against the currents, but you can adjust the main current a little.’’ (Respondent 3)

This strategy is used in combination with all barriers, but mostly with the standardisation, and result-orientation barrier. However, the boundary spanners disagree on whether this strategy is effective or not. Of the 14 boundary spanners that use the strategy, 3 boundary spanners find mediating effective (although some of them do complain about the effort it costs and the delay accompanying it), and 2 boundary spanners find the mediating strategy effective only when they combine it with the entrepreneurial strategy of picking the right colleagues. Seven boundary spanners stress that the effectiveness of the mediating strategy depends, for instance, on the competences and willingness of colleagues to see (and act upon) the added value. Lastly, 2 boundary spanners are entirely negative about the potentials of the mediating strategy and indicate that the traditional logics are too dominant to be aligned with the logic of networks. They thus plea for institutional reform.

Hierarchical strategy (used by 9 respondents)

Sometimes policy officials need a little help in breaking through red tape and engaging in processes of collaboration and co-production. Boundary spanners indicate that when requests and communication get stuck (often for a long time), it may be necessary to involve managers. This more confrontational approach to overcome barriers is sometimes unavoidable to getting the collaboration moving. In order for this strategy to be successful, it is important for boundary spanners to clearly formulate what they expect from these managers and why it is of crucial importance to change existing rules or policies and/or mobilise extra manpower within the policy-departments. The preparation needs to be excellent and to-the-point. This strategy is usually applied after experiencing some kind of barrier (‘’as a last resort’’) rather than preventing barriers from happening in the first place, as is the case with the other two strategies. The following fragment underlines this strategy.

‘’Yes, but something goes wrong here. Then we need to escalate, like ok, you consign the choice or dilemma, what you see as unfair, to [the managers on] the table who possesses the power to solve the issue. They then might say: ‘’Yes you are right, this is indeed undesirable.’’ Subsequently, you hope, of course, that the manager [of the policy department] who seems most appropriate to take the lead, then says: You are right, I will instruct several policy officials to help you.’’ (Respondent 1)

This strategy is used in combination with all barriers, but mostly with the standardisation and internal orientation barrier. All the boundary spanners, but one, agree on the effectiveness of this strategy. It is important though to be very specific to the management what exactly is needed so that they know what to do. While threatening with the hierarchical strategy may sometimes be enough to get policy officials moving, at other times, the strategy is developed together with policy officials as managerial or political involvement may help the policy officials to resolve the issue by getting more resources or leeway. Although effective, the boundary spanners indicate that they use this strategy not very often, and almost only when the other two strategies have failed. One respondent is negative about the effectiveness. He indicates: ‘’If my mangers talks to their manager and arrange that someone gets assigned to this task, you and I both know that that this person lacks the intrinsic motivation to give his or her best.’’

Typology of boundary-spanning strategies

Boundary work thus involves continuously avoiding and overcoming various institutional barriers; it can be seen as the core feature of NPG. About half of the boundary spanners use multiple strategies, while the other half has a strong preference for only one strategy (mostly the mediating strategy). Table 1 depicts the three strategies that boundary spanners use to avoid and overcome institutional barriers that are based on traditional institutional logics.

Table 1. Boundary-spanning strategies.

	Entrepreneurial strategy	Mediating strategy	Hierarchical strategy
Goal	Avoid barriers	Avoid and overcome barriers	Overcome barriers
Focus	Working around rules and contacts	Search for shared meanings to facilitate compromises	Using the power of managers to pave the way
Activities	Framing issues to match interests of receiver, splitting up issues in smaller parts, using political knowledge to pick right policy agendas, seeking opportunities, avoiding inflexible colleagues	Talking with colleagues to find common ground and a solution, using charms and/or arguments to change viewpoints	Taking the issue to a higher managerial level by involving public managers and/or politicians.
Competences and skills	Creativity, courage, political sensitivity, proactivity	Listening, openness, negotiation, persuading	Prioritising, perseverance, delegating, result-orientedness
Mechanism	Policy windows	Communication and personal relations	Formal hierarchy
Applied	All barriers	All barriers, but mostly standardisation and result orientation	All barriers, but mostly standardisation and internal orientation
Effectiveness	Yes, if applied well	Mixed results	Yes, but 1 respondent

Conclusion and discussion

This study aims to empirically examine the interplay between traditional hierarchical and more horizontal institutional logics within the Dutch municipality of Rotterdam by connecting the literature of institutional logics (structure) to the literature of boundary spanning (agency). The central aim of this article was to examine what strategies boundary spanners deploy to organisationally align the different institutional logics (hierarchy, market, networks) in citizen-state interactions and examine if boundary spanners succeed in their efforts.

To align clashing institutional logics, boundary spanners make use of three strategies: entrepreneurial, mediating, and hierarchical. The entrepreneurial strategy is used to recognise and exploit policy windows, to act upon resource opportunities, and to couple solutions to existing problems. Boundary spanners not only act entrepreneurial when it comes to policy-opportunities, but also when it comes to approaching ‘the right’ people. Past encounters inform the strategic behaviour, for example, not to approach certain inflexible policy officials again. To make this strategy work, boundary spanners have to do their homework to find the perfect policy-person fit for the specific collaborative issue and frame the issue accordingly. This strategy is, if applied well, effective in aligning logics by avoiding problematic clashes. The mediating strategy is used to facilitate a shared understanding and to negotiate the conditions of a potential solution. While mediating is the most-applied boundary-spanning strategy in this study, boundary spanners have mixed feelings when it comes to assessing its effectiveness. While some boundary spanners find this strategy very helpful in aligning logics (e.g., aligning logics requires good communication), other boundary spanners find this strategy a waste of time (e.g., aligning logics requires institutional change). Most boundary spanners however find this strategy to be occasionally effective, depending on the flexibility and goodwill of the policy officials concerned. The hierarchical strategy is employed mostly as a reaction to barriers that stem from the traditional hierarchical bureaucratic logic, such as standardisation and internal orientation. While hierarchy poses problems for boundary spanners, using hierarchy simultaneously also provides the solution to align clashing institutional processes and make collaboration work. Although this strategy is very effective, boundary spanners are careful in applying it.

This study makes two major contributions to the literature on NPG and boundary spanning. First, this study underlines that the real challenge of working collaboratively does not lie across the borders of the public administration, but lies deeply rooted within it. Although the need to mobilise or activate one’s own organisation has been discussed frequently in the governance literature (see Klijn and Koppenjan 2015; Bartels 2016), this article connects this discussion to the literature on institutional logics. Consequently, this article responds to an important theoretical challenge to the study of hybrid governance structures by identifying important structural institutional incompatibilities in the assumptions and principles that underlie traditional paradigms and the NPG paradigm (Klijn and Koppenjan 2012). Second, this study shows that boundary spanning not only uses strategies based on trust-building, communication and entrepreneurship (see Williams 2002), but also uses strategies that requires the mobilisation of the power of politicians and managers (e.g., hierarchy) to align institutional logics and handle conflicts. This point is an important addition to the current boundary-spanning literature. As boundary spanners extensively handle non-routine tasks, political and managerial support to safeguard their maneuver room enables them to better handle conflicts inherent to their position (Chebat and Kollias 2000; Stamper and Johlke 2003). While boundary spanners often get portrayed as the ultimate ‘network champion’ and ‘postmodern non-hierarchical leader’ (Williams 2002; Guarneros-Meza and Martin 2016), aligning clashing institutional logics require not only ‘bending’ and ‘renegotiating’ dominant traditional routines, but also strategically using these hierarchy-based routines to break through critical institutional barriers.

Inevitably this study has limitations that we hope will inspire future research. This study was performed in a specific context—boundary spanning in a large-sized municipality in the Netherlands. While we believe our findings to hold in comparable contexts, more empirical research is needed to test the generalisability and to further develop the theory on intra-organisational boundary-spanning strategies. To this end, future research could compare strategies of boundary spanners within different-sized municipalities both within one country and across countries to theoretically advance this field of study. The findings indicate that more attention should be given to the role of hierarchy in boundary spanning.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

José Nederhand

José Nederhand is a PhD researcher in the Department of Public Administration and Sociology at Erasmus University Rotterdam, the Netherlands. Her research focuses on collaborative governance, public-private partnerships and co-production.

Martijn Van Der Steen

Martijn van der Steen is Professor of Public Administration in the Department of Public Administration and Sociology at Erasmus University Rotterdam, the Netherlands. He is also Associate Dean of the Netherlands School of Public Administration. He specialises in network governance and public strategy.

Mark Van Twist

Mark van Twist is Professor of Public Administration in the Department of Public Administration and Sociology at Erasmus University Rotterdam, the Netherlands. He is also Academic Director of the Internal Audit & Advisory Program at Erasmus University and (extraordinary) member of the board of the Netherlands Court of Audit. His research interests focus on strategic management and on accountability issues of network governance in a multilevel context.

Notes

1. See Appendix for inductive coding scheme and additional exemplary quotes.

Appendix A.

Coding Scheme

Table

Open coding	Axial coding	Final code
Frame issues to match interest receiver	Issue formulation	Entrepreneurial strategy
Splitting up issues in small parts
Using political knowledge to pick right policy agendas	Seeking collaborative and resource opportunities
Avoid inflexible colleagues
Collaborate with colleagues with decision-power
Taking advantage of circumstances	Opportunism
Learning about each other’s culture	Empathy	Mediating strategy
Listening
Togetherness and trust
Exploring common ground	Collaboration
Bringing people together
Organising commitment
Influencing choices	Persuasion
Debating choices
Arguing
Change or force decision	Decision power	Hierarchical strategy
Seeking power
Power over resources
Involve supervisors	Higher managerial level
Supervisors will deliberate
Managers use hierarchy
Involve executive politicians	Higher political level
Threatening with politics
Collaborating with politicians

Appendix B.

Supporting Quotes

Table

Category	Quotes
Standardisation	‘’All kinds of objections are raised. Whilst as you explain it the way I do, everyone is like: this is beneficial for all parties. The municipality reduces maintenance costs and citizens have more fun. But the rules say that the area has to adjacent your own property, and that is not the case here. So it is not possible.’’ (Respondent 6)
	‘’It is a hard and sluggish process. General solutions are employed, while you have specifically asked for custom solutions. They then come up with a policy measure that is not quite what we need in our district. This happens often.’’ (Respondent 8)
Internal orientation	‘’This is, for example, making sure the organization is run smoothly and all questions are answered properly.’’ (Respondent 8)
	‘’[Policy officials say] We do not handle complaints in this way. You should tell [the citizens] that they should report it again. For us [district managers], this is a very sluggish, administrative, annoying way of working.’’ (Respondent 11)
Functional specification	‘’Somewhere in the organization an asset manager sees in his/her system that other streets should be handled first. So I hear that there is no money.’’ (Respondent 1)
	‘’I spent 1.5 years (!) negotiating with two governmental clusters, boys, we suffer from it and get reports of angry citizens, and rightly so, you have to renovate this field. Well, cluster 1 said the space was the responsibility of cluster 2, while cluster 2 said: no it’s theirs…’’ (Respondent 5)
Result orientation	‘’For the civil servant working on urban development it makes no difference that we have a district plan, because he is judged on realizing his own program. So if he has to build 30 houses, and he built 30 houses, he did a good job. However, it could well be that the 30 houses were quite unnecessary for the area.’’ (respondent 5)
	‘’Yes, that [achieving the policy program] is of course not something [that they are judged upon], you know, that is not part of the result-oriented way of working, which is required here.’’ (Respondent 7)
Entrepreneurial strategy	‘’At a certain point you know which colleagues to approach for a fruitful dialogue, and which ones to avoid.’’ (Respondent 3)
	‘’You can try different persons.’’ (Respondent 4)
Mediating strategy	‘’They have to get that there is a problem in the neighborhood, and simultaneously I have to get that there are certain policy assignments at stake. In the organization there should also be scored.’’ (Respondent 3)
	‘’I invest in relationships. Whether it is a colleague or inhabitant, I invest in them. I notice that these people… Well, of course I am not always nice, but in general these people are willing to do things for you.’’ (Respondent 11)
Hierarchical strategy	‘’Present it to the managers, like, listen, it doesn’t work. Your employees repeatedly say: these are the rules. I want to have these rules adjusted.’’ (Respondent 6)
	‘’Cooperation is hardly possible. This [co-creation process] is so beyond all conventions. In this process I have direct contact with the mayor. Which of course is very weird if you take into account all the managers and advisors. They are non-existent here.’’ (Respondent 7)