Accountability and inter-organizational collaboration within the state

ABSTRACT

This article discusses collaborative arrangements within the state and looks at their consequences for democratic accountability in networked governance involving public actors at different levels and in different policy areas. Applying an institutional perspective, it examines two such collaborations in Norway: work-related organized crime and vulnerable youths. It finds that horizontal collaborationsincorporate local differentiation and expertise resulting in a ‘collaborative culture’ facilitating broader involvement and more deliberative processes. However, top-down mechanisms, organizational commitment, resource allocation, and monitoring are also important. The cases demonstrate a shift towards more informal, voluntary, and horizontal accountability existing in the shadow of hierarchy, constraining the collaborative arrangements’ effectiveness and making them more complex and multi-layered.

KEYWORDS:

• Accountability
• collaborative arrangements
• institutional perspective
• wicked problems
• state-centric
• whole-of-government

View correction statement:

Correction

Introduction

Twenty-first century problems are increasingly seen as ‘wicked’, and the solutions to them are not simple. ‘Wicked’ problems are ones that entail complexity, uncertainty, and diverging viewpoints about what precisely the problem and solution is (Head and Alford 2013). But ‘wicked’ may also apply to political-administrative arrangements for dealing with these problems (Christensen, Lægreid, and Røvik 2020; Lægreid et al. 2015) as well as to the associated governance (Pierre and Peters 2020). Increased collaboration and coordination, both within the state and between actors and organizations in the public and private sectors, is frequently seen as the answer to such problems (Lægreid et al. 2014; Lægreid and Rykkja 2015), but the accountability issues that these raise are often consigned to the back burner and not thoroughly addressed or consciously thought through.

This article discusses accountability issues related to the shift from Weberian bureaucracies to collaborative arrangements. It examines the merits and limitations of the novel practice of inter-organizational collaboration within the state, and in particular the related issue of democratic accountability within the context of contemporary administrative reforms. We address the potential and pitfalls of accountability related to collaborative arrangements in public administration. Generally, the main focus in research and practice has been on network and collaborative arrangements involving public organizations and actors from the private sector and civil society. A systematic literature review of collaboration among public organizations reveals that accountability issues have not been up front (Costumato 2021), although the issue has been addressed linked to the concept of horizontal accountability (Schillemans 2008) and also to joined-up and whole-of-government reforms (Wilkins 2002; Hallligan, Buick, and O’Flynn 2011). The novelty of our article is its explicit focus on collaboration inside government and the consequences thereof for democratic accountability in networked governance involving public actors at different decision-making levels and in different policy areas (Papadopoulos 2007). To illustrate how collaboration plays out between different parts of the government we draw on experiences from two innovative collaborative arrangements in the Norwegian public administration: a policy programme for more holistic services for vulnerable children and youths; and a collaborative arrangement set up to fight organized crime in the employment sector.

As more and more societal problems cross ministerial lines, inter-ministerial and inter-agency collaborative frameworks are becoming increasingly essential. Indeed, they have become a core component of post-NPM reforms, such as joined-up government, whole-of-government, and the new public governance reform trajectories (Pollitt and Bouckaert 2017). In this article, we examine how collaborative arrangements play out between different parts of the government, traditionally characterized by a strong degree of specialization, vertical silos, and hierarchy (Lægreid and Rykkja 2015). Our main interest is in how innovative collaborative arrangements within the state can evolve as well as how and to what extent accountability issues are considered in the design and operation of such collaborative arrangements. This resonates with a broader conception covering different collaborative governance regimes, which may range from policy- or programme-based intergovernmental cooperation to geographically-based (regional) collaboration with nongovernmental stakeholders and public-private partnerships (Emerson, Nabatchi, and Balogh 2012).

The article surveys attempts to ‘bring the state back in’ (Bell and Hindmore 2009) and addresses the inner workings of the state and how different levels and policy areas cooperate to strengthen its coordinating role (Jacobsson, Pierre, and Sundstrøm 2015). Horizontal policy programmes that go beyond single organizations and single policy sectors are at the heart of whole-of-government approaches (Christensen and Lægreid 2007; Molenveld, Verhoest, and Wynen 2020) and also form the focus of this article. In this context, we understand collaborative governance as consisting of ‘processes and structures of public policy decision-making and management that engage people constructively across boundaries of public agencies and different levels of government, aiming to carry out public purposes that could not otherwise be accomplished’ (Emerson, Nabatchi, and Balogh 2012, 2).

The potential frequently attributed to such collaborative frameworks is a main reason for this interest. Collaborative frameworks cutting across organizational boundaries are seen as challenging the traditional hierarchical ‘silo’ structures and closed top-down processes that characterize public organizations (Scott 2020) and as therefore resulting in more innovation (Bommert 2010). As society and governments face more complex cross-cutting policy challenges, calls for better coordination of government policies and more transboundary collaboration – including within government and public administration – have become stronger (Lægreid et al. 2014). Through the instrumental use of networks and collaborative governance arrangements, powerful government actors are seen as having increased their capacity to shape and deliver public policy in a complex world (Skelcher, Mathur, and Smith 2005; Klijn and Skelcher 2007; Fimreite and Lægreid 2009).

An essential question is how democratic accountability is affected by this increasing prevalence of collaborative arrangements in government. Here accountability is understood as the relationship between an actor and an engaged forum in which the actor voluntary commits him- or herself to offering information about and explanation of his or her conduct, in a situation where the actor may also face consequences (Bovens 2007). Different ways of assessing various types of accountability, relating to accountability to whom (legal, political, administrative, managerial, professional or social accountability), accountability for what (processes, procedures, performance) and the nature of accountability (vertical, horizontal) can be examined. Collaborative innovation processes draw on mixed rather than fixed accountability standards. The positions of accountability holders and accountability holdees tend to shift during the governance process and result in shared rather than divided responsibility between involved actors. Thus, accountability relations become multiple (Christensen and Lægeid 2015), involving both the hierarchical and authorative logic of the state and notions of social accountability typical of more loosely coupled network arrangements. These multiple forms of accountability mix actors, values, and mechanisms from different accountability regimes (Benish and Mattei 2019).

In a discussion of the relationship between collaborative arrangements and different types of accountability, we specifically address accountability challenges in unsettled situations characterized by wicked and transboundary problems, and propose different ways of assessing various forms of accountability in this context. Transboundary issues are typically characterized by uncertainty regarding means-end relations and ambiguity regarding goals and values. Accountability relations will tend to differ across different tasks, situations and problem areas, making it important to pay attention to the context in which account-giving and account-taking takes place. An institutional perspective as outlined by James G. March and Johan P. Olsen (March and Olsen 1989; Olsen 2017) underlines the organizational basis for democratic accountability and takes into account such different contexts, considering that different rules, norms, routines, values, identities, and historical path dependencies influence social behaviour and therefore also accountability relations in different settings.

In the next part of the article, we introduce our theoretical approach. We then go on to describe our two empirical cases. Lastly, we discuss the findings and draw some conclusions and lessons related to institutional design, leadership, and systemic changes.

A transformative theoretical approach

Traditionally, the accountability literature has been anchored in the principal-agent model. However, this perspective on accountability has been challenged and developed further in recent years (Schillemans and Busuioc 2014, Christensen, 2017; Lægreid 2014, Olsen 2017). A principal-agent model refers to the relationship between a principal who authorizes another person or organization to act on their behalf as an agent. The accountability literature has further mostly addressed routine and settled situations, characterized by ‘business as usual’, clear goals, strong means-end knowledge, and predetermined superior and subordinate actors in a hierarchy, where it is rather clear who is accountable to whom and for what. Here, a logic of consequence and dyadic, monocentric relationships between leaders and subordinates are at the forefront (Christensen and Lægreid 2016).

In unsettled situations characterized by ‘wicked’ problems that cross organizational boundaries, there is a need to address the complexity and dynamics of accountability relations. This calls for a supplementary institutional approach. According to Olsen (2017), there is a need to examine how accountability relations play out in relation to the way authority is organized, exercised, and controlled, rather than assuming that the preferences and actions of principals and agents are predetermined. An institutional approach holds that institutions affect actors’ identities and roles through socialization, internalization, and habitualization as well as via external incentives (Olsen 2013). This means that organizational and institutionalized values matter and that behaviour within an organization is influenced by normative pressure arising from within as well as outside that organization. Democratic accountability plays out in a mixed order of government consisting of hierarchical authorities with strong sectoral specializaiton and horizontal, inter-organizational collaborative arrrangements. In collaborative arrangements with transboundary and hybrid characteristics, accountability channels are normally multiple (Bovens, Schillemans, and t Hart 2008), goals and values are ambiguous, means-end knowledge is uncertain, participation is more fluid, authority is more contested, and experiences are less transparent (Olsen 2017). Collaborative arrangements, especially in unsettled or wicked policy areas, may be less institutionalized, weaker and more contested; and accountability relations and processes may therefore be more controversial, politicized, and dynamic. The question of who is accountable to whom and for what is unstable and evolves and changes. Who is responsible depends on ambiguous and contested interpretations by different actors and forums in different contexts and is characterized by institutional constraints rather than by rational instrumental considerations. Motivation and control are based on traditions, culture, and socialization into a professional ethic and public ethos, and actors and forums are expected to behave in a more altruistic and interactive way (Selznick 1957; March and Olsen 1989). If dealt with consciously, this may result in accountability relations that are more interpretive, interactive, and reflexive.

To understand better how collaboration works in practice, and how it relates to accountability, we apply a transformative perspective (Christensen, Lægreid, and Røvik 2020; Christensen and Lægreid 2016b). This approach implies that it is necessary to study not only formal organizational structures, but also cultural features (such as existing informal behavioural norms and values) to understand how public institutions work and how the actors within these structures act. The perspective takes into account that formal and informal institutions and accountability relations coexist. They might be complementary or competing, but they can also accommodate or substitute for one other (Helmke and Levitsky 2004). There might also be different forms of formal accountability, such as traditional bureaucratic or managerial accountability related to performance management (Askim, Christensen, and Lægreid 2015). A main idea is that collaboration and the exercise of accountability can be initiated via formal structural interventions, e.g. formal rules and regulations, as well as through cultural changes, e.g. via the development of new, but informal ways of acting or thinking with respect to what are considered appropriate problems and solutions.

In a study of practitioner experiences and preferences related to horizontal policy programmes, Molenveld et al. (2019) identify three ‘images of coordination’ combining structural and cultural elements. First, central frame setting, where participants support collaboration by design from a structural perspective, favouring a framework introduced by the centre but with broadly defined objectives giving room for local autonomy. Second, networking via boundary spanners, reflecting a cultural perspective where collaboration is seen as having evolved culturally; and third, collaboration beyond window dressing, where participants consider collaboration as merely symbolic and less useful. Together, these images can enhance multiple accountability relations including formal vertical and more informal horizontal accountability relations.

Olsen (2017) makes a distinction between accountability within an established regime with stable power relations and role expectations, and accountability in (re)structuring processes during transformation periods. Thus, accountability can be both about maintaining routines within an existing order and about transforming this order. Collaborative arrangements are often about transforming the established order, implying that accountability relations can become more demanding. Indeed, wicked problems transcend administrative levels and policy areas and are characterized by complexity, ambiguity, and uncertainty (Head and Alford 2013). The introduction of new collaborative arrangements to deal with those that do not follow the existing organizational structure raises new accountability challenges. New accountability relations are added to traditional ones, making them more complex and multi-dimensional.

One of the possible remedies to increasing fragmentation and complexity is to introduce more horizontal forms of accountability, supplementing traditional, hierarchical accountability (Schillemans 2008). Formal and vertical accountability relationships are often supplemented by informal and horizontal ones, resulting in more complexity. A multi-dimensional accountability approach is hence a response to the growth of more collaborative arrangements in the public sector. Several accountability dimensions are activated simultaneously, involving various actors with different opinions on appropriate problems and solutions being called to account by a variety of account-holders within various forums. Hence, collaborative arrangements foster change in basic government arrangements and create new forms of accountability relations. This changes vertical specialization and coordination, where balancing a need for central political control with institutional and professional autonomy is important. It can also alter horizontal specialization and coordination, where balancing a need for unambiguous roles with a need for a more holistic approach is central. These issues are typical for post-NPM reforms, adding to and partly modifying earlier reform trajectories.

The empirical evidence for a positive relationship between accountability and reforms in unsettled situations is inconclusive. Lack of motivation, time, knowledge, and capability may result in both agency drift (agents who shun accountability) and forum drift (principals who choose not to hold agents accountable) (Schillemans and Busuioc 2014). In collaborative arrangements aimed at joint service delivery, there are normally multiple principals who have to balance competing interests and conflicting expectations of numerous stakeholders and find multiple accountability relations (Voorn, van Genugten, and van Thiel 2019; Schillemans and Bovens 2011). In the following case studies, we explore some of the challenges, but also solutions, that were found in two cross-sectoral, innovative collaborative arrangements in Norway.

Accountability and state-centric collaborative arrangements in Norway

Several innovative collaborative arrangements have been introduced in Norway to enhance the quality of coordination in wicked policy areas (Nesheim et al. 2019), but their specific accountability design has not been a prominent feature. In the following, we illustrate potential challenges related to accountability in such collaborative arrangements within the state, by showcasing two high-profile and innovative arrangements in two policy areas dealing with cross-cutting, wicked problems. The two cases illustrate typical state-centric collaborative arrangements and are included to strengthen their representativeness).

The ‘0-24 collaboration ’ for more holistic services for vulnerable children and youths

The 0–24 collaboration was launched by the Norwegian government in 2015 as a five-year cross-sectoral programme. The overall goal of the programme was to strengthen the follow-up on vulnerable children and youth and their families through increased collaboration in service delivery and more customized, earlier, and more holistic services. The principals behind the programme were the Ministry of Education and Research, the Ministry of Health and Care Services, the Ministry of Labour and Social Affairs and the Ministry of Children and Families. These ministries gave a joint letter of assignment to the following central agencies, called the programme owners: the Directorate for Education and Training, the Labour and Welfare Organization, the Directorate for Children, Youth and Family Affairs, the Directorate of Integration and Diversity, and the Directorate of Health.¹

The directors of the five central agencies constitute a steering group chaired by the head of the Directorate for Education and Training. The steering group has overall responsibility for the programme. It is assisted by a programme secretariat with one member from each of the five participating central agencies. The programme secretariat is responsible for the daily operations of the programme and is chaired by the Directorate of Education and Training. The education agency has an internal steering group and the health agency has a programme board while the other agencies have less formalized arrangements. Most of the many projects are organized as cross-sectoral working groups subordinate to the steering group. They are chaired by one of the central agencies and include professionals from the line departments in the different agencies.

On behalf of the Directorate of Education and Training, a consultancy firm (Deloitte) conducted an ‘underway’ evaluation of the programme. The evaluation followed, documented, and assessed the implementation of the programme and its goal attainment and was set up to facilitate and contribute to learning, development, and adjustment of the programme during the period that it ran. A special focus was on the challenges linked to horizontal collaboration and how the programme worked as a tool for coordinating the handling of wicked problems. Two underway reports have been published so far (Deloitte 2018, Deloitte 2019). The reports are based on extensive document studies, interviews, surveys, observations, workshops, and meetings.² A final report will be published after the programme finishes at the end of 2021. The data presented below draw heavily on the first two underway reports.

Deloitte assessed the organization of the programme, its activities and goal attainments from the start of the programme in 2015 up to June 2019 and concluded that the programme had developed significantly since it started. The reports identified positive developments mainly brought about by a willingness and ability to learn and change. Collaboration between the actors in the programme and in the general progress of the joint work was strengthened. In 2019, the programme received the ‘Better State’ award, demonstrating its contribution to better governance and efficient use of resources within central government.

That said, the programme has also faced significant challenges, primarily linked to its organization, anchoring, and ownership in all the central agencies. The programme was said to lack a clear direction and commitment from the involved directorates. In addition, accountability relations were ambigious and the different measures were rather fragmented during the early years (see also Helsedirektoratet 2015). These problems were addressed by the formulation of a strategy for the programme in 2017, which broadened its scope from improving completion rates of high school to enhancing the living conditions of vulnerable children and youths. Leadership commitment and the role of central organizational arrangements in the programme was rather ambiguous, partly due to its transboundary and inter-sectoral challenges. In the first years, there was also a lack of communication and information within the programme.

Cross-sectoral working groups were crucial in formulating means and measures. However, these were rather loosely linked to the overall goals of the programme and to the core activities of the central agencies. A main challenge for the central agencies was to sustain emphasis on the importance of cross-sectoral coordination. Improving collaboration between the central agencies as well as collaboration across administrative levels, sectors, and different services, was challenging. There were significant differences in cultures and professional perspectives regarding how to coordinate the joint work. Establishing a joint understanding about challenges and the most important goals of the programme was also difficult, and the commitment from parent ministries as well as from line departments in the central agencies was rather low. There was a rather big gap between the steering group and the programme owners, and it was difficult to communicate responsibility and accountability, engagement and ownership to the central agencies. Up to 2018, the efforts and results of the programme were judged to be modest.

The 2019 evaluation report revealed a more positive trend, but it was still not clear for all partners what the aim of coordination was, who the target groups were, what measures were appropriate, and what was needed to produce significant results. It was unclear who was accountable for what and to whom. Initially, the stated aim of the programme was vertical coordination and accountability between central and local government for the benefit of vulnerable children and youths and their families. The involved actors, however, perceived the aim more as the horizontal coordination of means and measures across governmental agencies. The importance of stronger commitments from line departments, of clarifying roles and accountability was underlined by the evaluators. The working groups’ efforts were seen to be only loosely anchored within the core activities of the participating agencies.

The evaluation identified differing sector interests as a main obstacle for the programme. The different meeting places and organizational arrangements strengthened cross-sectoral collaboration, however, and a stronger awareness and understanding of each other’s tasks, work methods, culture, and priorities developed over time. Nevertheless, strong sectoral interests were continually perceived to be one of the strongest barriers to producing results. Even though the agencies had developed a greater joint conceptual framework and understanding of the goals of the collaboration, sectoral interests remained strong when it came to deciding about specific projects. A main precondition for achieving results identified by the report was that all participants needed to be willing to give up something. This was not always the case. Everyone wanted better collaboration and coordination, but few were willing to be coordinated. Owing to the sector principle, each agency was more interested in delivering on their own assignments than in contributing to others’ assignments. The agencies were more interested in bringing their own assignments into the programme than to contributing to cross-sectoral collaboration. Barriers in the sector-related laws and regulations contributed to this, and the government has now proposed changes in the sector laws to enhance transboundary collaboration and coordination. They were also constrained by the performance management system with its sector-specific performance indicators Thus, accountability tended to be upwards within the formal hierarchal line to the parent ministry while the more informal types of horizontal accountability were less developed.

Another barrier concerned the links to the participants’ core activities. There was a positive development in attitudes towards the 0–24 programme in the agencies. However, getting engagement and commitment from the leadership of the involved agencies was still challenging. One reason for this was that the participants had trouble seeing the collaborative arrangement as related to the agencies’ core activities. It was difficult to develop shared accountability relations, and the agencies were not sufficiently accountable for cross-boundary tasks, partly due to strong vertical accountability relations within their own ministerial area, such as the performance management system. There were conflicting expectations and frictions between the agencies, and a loose coupling between activities within the programme and core activities in the agencies. A central success factor for cross-sectoral collaboration thus seems to be linkage with the core activities of the participating agencies.

A third barrier was lack of commitment from the parent ministries. The agencies requested stronger involvement and engagement from their parent ministries and underlined that it was challenging to keep up the pressure when political engagement was weak due to forum drift (Schillemans and Busuioc 2014). It was difficult to incorporate horizontal cross-sectoral goals into the performance management systems between the ministries and central agencies because they mainly focused on vertical accountability within each ministerial area. The different agencies directed attention towards their own sectoral targets rather than towards prioritizing areas in which the results depended on collaboration with authorities outside their own ministerial area, leading to agency drift.

The 0–24 collaboration is briefly mentioned in the annual letter of assignment from the ministries to their subordinate agencies, but this mention is largely limited to a call to participate in the collaboration. The lack of priority allocated to the programme in the letter of assignment is likely to result in less attention to the collaboration than to sector-specific objectives. Core activity and legally mandated tasks tend to be more prominent than cross-sectoral assignments. In some participating agencies the 0–24 collaboration is more closely linked to the agencies’ core activities, which makes accountability relations less blurred. The education agency has a coordinating role in the programme, but the responsibility and role of the coordinator with respect to accountability is somewhat ambiguous. One challenge is that it is difficult to improve transboundary government collaboration and accountability relations without changing the legal arrangements, the state budget system, and the performance management system.

A fourth barrier is the organization and leadership of the programme. The programme has some coordination forums but no specific programme leader. There is no ‘backbone’ support from an independent coordinator. The horizontal project and programme organization is established across the sectoral hierarchies but has in itself no specific internal hierarchy. There are several dilemmas with this type of arrangement. The collegial structure encourages consensus across different organizations. This can make it difficult to deliver according to expectations. Assigning ownership to one lead agency can weaken the ownership in the other agencies, however. On the one hand, there is a need for clear and predictable governance signals from the steering group. On the other hand, there are limitations on how detailed the steering can be, owing to lack of professional expertise, time, and attention. The working group model has brought together professionals from different agencies with similar backgrounds, which has simplified collaboration but also produced tensions between professional accountability and vertical accountability. However, the projects are loosely anchored in the line organizations within the agencies, making the accountability of the programme difficult.

The evaluation concludes that the programme has developed in a positive way towards improved collaboration, resulting in better knowledge and understanding of each other’s tasks, work processes, cultures, and priorities across sectors. At the same time, accountability relations are fuzzy and blurred. Sector interests are strong and there is a need to strengthen the links between the programme and the different core tasks of the agencies, as well as the links upwards to the parent ministries. Traditional formal accountability relations within each silo need to be supplemented by stronger accountability to the joint transboundary collaborative arrangements by the participating actors, both horizontally and vertically.

One lesson that the evaluators emphasize is that collaboration should be seen as an added value for the participating agencies. It is also important to limit this to projects where the cross-sectoral effort is essential to reach the objectives of the programme. There is a need to clarify whether the participants in the programme are also supposed to collaborate to handle tasks at the strategic level, and whether tasks more directly related to service provision for users and clients at the local level should also be addressed. Another lesson is the need for stronger attention to accountability relations. It is especially important to strengthen efforts to develop shared and joint accountability relations across the involved agencies and to reduce barriers arising from strong vertical accountability towards the parent ministry.

Collaborative arrangements to fight work-related crime

Since 2014, Norwegian politicians and government have shown increased concern about ‘work-related crime’ or crimes taking place in the labour market, defined as ‘activities that breach laws concerning pay and working conditions, social security and taxation’, and ‘organized activities that exploit employees, distort competition, and undermine the social structure’ (PMO 2017a, 2017b).³ The overall goal of the initiative is to safeguard a good and sound working life, prevent and combat work-related crime, and ensure compliance with rules governing working life through a joint effort by the police, the tax administration, the labour inspection authority, the labour and welfare service, and other related public agencies operating at the national, regional, and local levels. The crimes include, but are not limited to, traditional types of ‘white collar’ and economic crime such as tax evasion, fraud, document falsification, and corruption; breaches of labour market legislation related to wages, health, and safety; but also migrant labour force regulations including exploitation in the form of forced labour and human trafficking. Heightened attention to these issues has resulted in novel collaboration arrangements between government agencies at the local, regional, and national level, and also in European-level initiatives designed to inspire more collaboration at the supranational level (Jahnsen and Rykkja 2020).

Work-related crime is seen by the Norwegian government as a threat to the economic stability and the very foundations of the Norwegian welfare state (MLSA 2019, 5). Its work to combat this type of organized crime involves different ministries, agencies, and government units at different administrative levels and in different sectors. At the national and strategic level, the issue involves the Ministry of Labour and Social Affairs (supervising the Labour Inspection Authority and the Labour and Welfare Administration), the Ministry of Finance (supervising the Tax Administration), and the Ministry of Justice (supervising the Police Directorate and the immigration authorities) most closely. The cross-sectoral nature of the issue is illustrated in three national government strategies for combating work-related crime (from 2015 to 2019), jointly signed by the Prime Minister and as many as ten different ministers (MLSA 2019). At the local and operational level, too, efforts to combat work-related crime involve a broad spectrum of national agencies, government authorities, and municipal actors, including the police, the local and regional branches of the labour authorities, the Norwegian Customs Authorities, the Food Safety Authority, the Norwegian Public Roads Administration, the municipal fire and safety departments, and the municipal treasurers.

National government strategies to combat work-related crime were introduced by the Norwegian government in 2015 and updated in 2017 and 2019. These strategies emphasize a need for more coordinated efforts to handle work-related crime across sectoral and organizational boundaries, favouring more and improved collaboration between government authorities and the enforcement of harsher penalties for this type of crime (Jahnsen and Rykkja 2020). Prior to the establishment of a national strategy, there were several examples of local operational initiatives seeking to pool local resources to achieve a more concerted effort to prevent the deterioration of conditions on the labour market and the use of an illegal workforce.

In 2015 a group of researchers evaluated the cooperation between state agencies in the field of work-related crime at the local/regional level (Neby et al. 2016; Nesheim and Neby 2017). Their report concluded that collaboration across agencies and sectors was by and large promising and that some important things had been achieved. At the same time, they emphasized that further adjustments and adaptations were likely to occur owing to the internal and external complexity of this ‘crossagency’ collaboration. Collaboration was challenging in particular due to existing divisions between agencies and a prevailing sectoral and hierarchical organization of the field. The evaluators pointed to information and communication barriers across agencies and identified challenges related to finding and agreeing on joint performance measures. The participants in the collaboration also perceived tensions between their primary and secondary organizational affiliations and sometimes found it difficult to prioritize collaboration over commitments to their line organizations. Furthermore, introducing and strengthening cooperation initiatives on the one hand and ensuring cross-cutting standardization on the other were identified as a challenge. The evaluators pointed to a delicate balance between a need for standardization and formalization and for more flexible and dynamic collaboration arrangements across agencies. They emphasized the importance of information-sharing and communication, in particular regarding technical and operative aspects, and recommended gearing performance measurement more directly to cross-agency cooperation, emphasizing collaboration initiatives more strongly in national strategic considerations, and further developing intelligence tools and analytical capacity at the operational level. These barriers between actors in different sectors and with different primary organizational affiliations also complicate accountability relations.

In 2017, the government instructed the four main authorities working in the field – the Labour Inspection Authority, the Tax Authority, the Police, and the Labour and Welfare Administration – to strengthen their collaboration and develop a joint governance model to combat work-related crime, specifying the roles and main framework for the involved authorities. This governance model was presented in a joint report by the main agencies the following year (Joint Report 2018a). Here, the main coordinating arrangements and bodies established at the national level, supplementing the existing ‘line’ authorities with separate responsibilities within the field, were presented. Concurrently, the involved authorities were asked to develop joint performance indicators for combatting work-related crime. A joint working group was established to develop these, and shared, cross-cutting performance indicators were integrated in each of the agencies from 2019.

Adding to the central governance structure, there are currently seven co-located inter-agency teams organized as work-crime centres at the regional level. At these centres, dedicated personnel seconded from the main bodies with responsibility in the field work together in joint and co-located operational teams. A joint report from 2018 concluded that the regional centres worked well and that the collaboration largely had been successful, not least because of the co-location and establishment of dedicated teams (Joint Report 2018b). According to the report, the teamwork at local level resulted in better operational collaboration, valuable knowledge-sharing across organizations, and a more coordinated use of sanctions towards criminals. However, interviews with centrally placed people within the collaboration indicate that many of the participants are frustrated over non-compatible regulations, a lack of a common ICT system as well as differing priorities within the collaborating agencies (Jahnsen and Rykkja 2020).

Concurrently, important challenges related to collaboration and accountability are still evident. The partners succeeded in overcoming geographical distance through co-location, but still exhibit a certain ‘structural distance’ (Nesheim et al. 2019). This distance is observable, for example, between the Labour Inspection Authority and the Police on one side, who have a more operational focus and investigative/punitive approach, and the Tax Authorities on the other, who tend to work more in-house, utilizing existing (digital) information. There are also challenges related to power asymmetries between the authorities (e.g. differences in discretion, resources, and sanctions) and conflicting interests related to sector-specific aims, mandates, and resources (Jahnsen and Rykkja 2020). Although they introduce a clear governance structure, these factors impact on accountability relations within the field.

Secondary and supportive collaborative arrangements have been established at both the regional and central levels, but like in the 0–24 collaboration, there is no central coordinating (lead) agency. At the regional level there are regional steering groups consisting of the regional leaders of the involved authorities. At the national level, there is a central coordinating forum⁴ and a joint steering group with representatives from the involved authorities working together to develop joint goals and performance targets. The steering and accountability relations are recognized as complicated, whereas the different involved authorities also have their own, separate priorities, budgets, personnel resources, and performance targets that translate into different and sometimes incompatible signals vis-à-vis the involved subordinate authorities at the regional and local levels (Joint Report 2018b). Hence, it is still sometimes unclear who is accountable for what and to which exact forum one is accountable.

In 2016, a National Inter-Agency Analysis and Intelligence Center (NTAES) was established with a mandate to work within the field of work-related crime (Government Instruction 2016). This centre added another layer to the collaborative arrangements, now at the national level. The centre was mandated to provide intelligence to the national agencies as well as to the teams at the regional level and was introduced to encourage more coordination between the national authorities within the field as part of a more overarching national crime prevention strategy (NTAES 2019). Overall responsibility for the centre is held jointly by the Ministry of Justice and Public Security and the Ministry of Finance (ASD 2015), and it includes personnel from the same agencies as the regional inter-agency teams. Hauge (2019, 94) concludes that the centre got off to a somewhat abrupt start and to some extent still lacks well-defined working routines and a hierarchical/lead authority, although the collaboration within the centre is deemed to work fairly well because of its pronounced horizontal structure. Marked differences in analytical capacity (knowledge, available information, and competence) between the different authorities involved were, however, an important barrier to collaboration.

From 2015 to 2018, three joint reports evaluating the results of the overall collaborative efforts in combatting work-related crime were published.⁵ The reports were based on the annual reports from the seven regional centres, but also included joint efforts implemented at the central national level (Joint Report 2018b, 4). Summarized, the reports described a successful operational collaboration at all levels, resulting in a perception of increased detection of criminal activity and a more efficient use of instruments and sanctions. At the same time, barriers to collaboration were also identified, mainly related to information-sharing across organizations.

The two cases compared

The two cases portray some important similarities. They represent interesting examples of innovative, cross-sector collaborative arrangements within the state designed to tackle wicked problems and solve situations where the problem structure does not fit the existing organizational structure. In both cases, the problems were located within strong sectoral ‘silos’ and were dealt with by many separate (semi)autonomous agencies. The emphasis had been on accountability for organization-specific objectives and targets within these silos rather than on shared, joint, and collaborative objectives, targets, and performance measures. In both cases, the new collaboration initiatives were supported by the establishment of new joint governance structures, mutual strategic thinking, and organizational backing for the collaboration. Motivating the single government organizations and those working in them to invest and engage in the collaborative arrangement was a challenge for both sides. This is typically the case when actors, as in these two examples, compete for resources and have their own separate organizational and sector-specific priorities relating to their primary organization.

They also exhibit similar barriers to collaboration as well as limitations related to accountability. Strong sectoral interests linked to organizational goals and sector-specific performance indicators as well as legal and financial obligations linked to the state budget system impede actors’ willingness to commit and deliver to the collaborations. When the commitment to sector-based core activities is stronger, or when the main goal of the collaboration is not seen as part of the core activity of the organization in question, the collaborative goals are not prioritized within the various sector organizations. Furthermore, accountability becomes blurred if goals and performance indicators for collaboration are not specified, and if clear requests for participation, commitment, and priority are not made. The sector-specific performance management system was a clear barrier to development of joint accountability relations across organizations in different ministerial areas. A more collegial, cross-sector organization allows room for the participation of more relevant actors, facilitates knowledge transfer, and might result in more attention to the issue and policy area. But without clear leadership and commitment, accountability relations remain blurred. This might run counter to prioritization of the problem area.

In unsettled problem areas, institutions, accountability relations, and processes are less developed and hence less stable. Both our cases showcase these features. Accountability relations can be – and are to a certain degree – clarified by introducing formal governance arrangements (e.g. through the alignment and formalization of goals, common performance indicators, budget allocations, and staffing or by introducing transboundary targets and performance indicators), as seen in both cases, but need to be followed by institutionalized norms and values that guide organizational behaviour. In the case of work-related crime, these were developed within the regional collaborative centres, but to a lesser extent at the central agency level.

Revisiting the theoretical approach

The two cases illustrate that collaborative arrangements across sectors and government agencies exhibit important tensions and challenges, and accountability relations are not clear. In the discussion below, we return to the question of accountability in collaborative arrangements by using the two cases as examples.

In both cases the three images of collaboration found by Molenveld et al. (2019), – central frame setting, networking via boundary spanners, and collaboration beyond window dressing – seem to be relevant. Collaboration is introduced via structural interventions as well as through cultural ones. The two cases reveal a hybrid picture where hierarchical as well as horizontal accountability relations exist side by side. In both collaborations, there was a certain cultural resistance to hierarchical central control because the individual agencies had their own priorities. There was a need for local differentiation, but also evidence of the evolution of a more collaboration-oriented culture over time. As a result, top-down mechanisms to achieve collaboration and accountability were supplemented by adaptive arrangements, involvement, and deliberate processes enhancing more informal horizontal accountability relations.

Our cases illustrate that collaborative governance arrangements are often mixed orders and compound and composite systems. They represent typical collaborative cross-boundary arrangements in unsettled situations and exhibit complex and multiple accountability relations (Olsen 2017). Although certain joint steering mechanisms, structures, performance goals, and results measures are introduced, thereby creating clearer lines of responsibility, it is not always obvious who is accountable to whom, and for what. This illustrates that accountability in collaborative governance arrangements is often blurred and fuzzy. For example, measures and efforts designed to help vulnerable children and youths that straddle several sectors were often introduced without involving the collaborative arena that the ‘0–24 collaboration’ was supposed to be (Difi 2016; Hildrum 2020). This indicates that short-term and sector-oriented considerations might still displace the comprehensive and long-term policy development and accountability relations related to cross-boundary collaborative networks.

We find that collaboration through cross-cutting policy programmes needs organizational commitment, resource allocation, certain structural steering mechanisms, and monitoring. This is crucial for developing clear accountability relations. And, as also noted by Molenveld, Verhoest, and Wynen (2020), such collaborative arrangements need to take both the structural and cultural diversity of the collaborating organizations into account, balancing horizontal and hierarchical (vertical) accountability as well as formal and informal commitments and incentives. Judging from the evaluations of our two cases, such a hybrid collaborative approach, combining soft informal and hard formal features, might be a successful strategy for strengthening both collaboration and accountability relations (see also Voets, Verhoest, and Molenveld 2015). The challenge is to find an appropriate balance between the two features. Hence, both formal structural and informal cultural variables matter for state-centric collaborative arrangements and for how accountability plays out – especially in transboundary settings. This leads us to contend that one needs to go beyond an instrumental approach and supplement it with an institutional approach when trying to understand how collaborative arrangements and accountability relations work.

The hybrid collaborative governance arrangements we see in our two cases have overall moved formal, mandatory, vertical accountability relations towards more informal, voluntary, horizontal accountability. However, these new accountability dimensions seem to exist in the shadow of hierarchy, constraining their effectiveness in reducing the vertical silos in government and also adding to the complexity.

Accountability ‘to whom’ was a main concern in these two collaborative arrangements. The problem of the ‘many eyes’ and multiple accountability forums also played out in the two cases. Vertical and horizontal accountability relations as well as legal and professional accountability mechanisms led to accountability confusion. Also, accountability ‘for what’ was an issue in both cases. In addition, process or procedural accountability (Day and Klein 1987) was important, and different legal arrangements constrained information-sharing and cross-boundary procedural accountability. There is also an apparent tension between the collaborating activities and the accountability systems in the parent ministries (Moynihan et al. 2011). Collaborative actors face a trade-off between their organizational accountability upwards and accountability for their collaborative performance. To develop a ‘both/and’ mindset rather than an ‘either/or’ mindset might be a way to handle this challenge (Waardenburg et al. 2020). Moving in the direction of accountability for outcomes was also challenging. In the case of the joint work designed to combat organized crime, however, the development and use of shared performance indicators and joint annual reports was a way to deal with this accountability problem.

Our cases illustrate the development of a more general pattern of complex accountability relations and interorganizational activities, generating accountability challenges (Mulgan 2014). As the two cases show, when responsibility for a policy area is shared, accountability relations become blurred. The simultaneous vertical and horizontal accountability mechanisms that are present in our two cases seem to be supplementary and complementary, providing a mixed political order that combines and blends different modes of governance, as highlighted by Olsen (2010). This also corroborates with Mulgan (2014), who argues that in unsettled situations, accountable actors typically face ambiguous obligations.

Studying accountability in collaborative governance processes requires us to analyse the complexity and dynamics of such accountability relations and processes (Olsen 2013). By their very nature, collaborative arrangements redefine patterns of influence, and accountability inherent in existing structures and cultures will be redefined, creating new dynamics. Rather than just asking whether government officials are more or less accountable after collaborative governance arrangements have been introduced, one should also focus on what kind of accountability is perceived as appropriate (Romzek 2000).

Collaboration does not necessarily happen through deliberate design and institutional engineering and may be constrained by administrative tradition and culture. This is also the case for accountability relations. Unless there is an awareness of this, the distance between collaborative partners may result in ambiguous and conflict-driven arrangements, producing more accountability challenges. In the case of work-related crime, there has been more deliberate design and institutional engineering over time. At the same time, there was a certain resistance and barriers to collaboration that can be linked to different traditions and cultures within each of the collaborating organizations, alluding to the ‘structural distance’ that Nesheim et al. (2019) find. The same tendency can be seen in the ‘0–24 programme’.

What lessons can be drawn regarding how to promote collaborative governance and accountability within government? To answer this question, we will address the importance of institutional design, leadership, and systemic changes (Torfing, Sørensen, and Røiseland 2016). First, institutional design matters. Smart meta-governance of hybrid network arrangements through legal framing, design, facilitating, and participation are important (Torfing et al. 2012; Voets, Verhoest, and Molenveld 2015). The way political ownership of collaborative governance at ministerial level is organized is crucial (Molenveld and Verhoest 2014), and introducing a governance structure with clear responsibilities and organization at the local level is also important. In our cases, joint accountability towards involved ministries was, at least initially, not organized well, and the leverage of ministries and agencies over other ministries and agencies to make them contribute was rather weak.

Second, the links between ministerial decision-making and the dynamics of collaborative governance arrangements at lower administrative levels are often problematic and might lead to accountability challenges. High intensity of policy integration and administrative coordination reforms seems to be affected by both the policy agenda of governing political parties and by the political role of central agencies (Trein, Maggetti, and Meyer 2020). This might be due to the salience of the topic, different time horizons in ministries and subordinate collaborative arrangements, and a mismatch between ministerial sector policies and transboundary collaborative arrangements.

Third, the way central agencies contribute to collaborative governance arrangements is crucial and is affected by the way ministries organize coordination and accountability. The contribution of implementing organizations to cross-cutting objectives is influenced by structural conditions (tasks, size, organizational boundaries), legal constraints, cultural features (trust, innovativeness, customer orientation), professional features (expertise), and by the amount of vertical control the parent ministries exert.

Conclusion

In this article we have examined how inter-organizational collaborative arrangements within the state to handle transboundary and wicked problems are designed, how they work in practice, and how accountability relations are considered in the design and operation of such arrangements. Our case studies have revealed that collaborative governance arrangements exhibit multiple accountability regimes in which different horizontal and vertical accountability mechanisms do not substitute for each other (Schillemans 2008) but are more overlapping, complementary (Scott 2020), and supplementary. A new accountability regime with more layered accountability forms emerges, combining hierarchical (vertical) and horizontal accountability.

A key challenge is how to handle such multiple accountability relations. It is claimed that different, co-existing conceptions of accountability may undermine organizational effectiveness by paralysing decision-makers, because not all forums can be satisfied at once, by allowing more strategic and unaccountable actions, or by introducing more symbolic accountability (Schillemans and Bovens 2011). However, this might not always be the case. Multiple or hybrid accountability may also present an appropriate solution for an increasingly pluralistic system of governance. In unsettled situations with a fluid, complex, flexible, semi-autonomous, and fragmented multi-level governance polity, it seems that a better solution could be to go beyond the traditional vertical and formal hierarchical models of political accountability to resonate with the new reality of these transitional, cross-cutting governance systems. Collaborative governance arrangements represent dynamic situations. Reform agents often have limited powers and a rather weak knowledge of means and end results. We need to study the dynamic relationships between state-centric collaborative arrangements and the ensuing accountability relations and especially how multiple and hybrid accountability relations interact and change over time.

Our study of the relationship between collaborative arrangements and accountability highlights that informal horizontal accountability mechanisms are needed as a supplement to formal, vertical, and hierarchical accountability relations. Furthermore, the study supports the argument that there is a need to develop collaborative arrangements that counteract blame avoidance tendencies, low trust, and scepticism towards other ministerial silos (Flinders 2014). More attention to the design of relevant accountability mechanisms and the conditions and contexts in which they will work well might be a good approach going forward (Bovens and Schillemans 2014). Accountability design has not been a prominent feature in collaborative governance arrangements so far, although our two cases showcase some developments in this direction. A stronger focus on meaningful accountability design and on what type of accountability to choose and when would be helpful. Accountability needs to be operationalized to fit the specific situation and organization and identify clearly who is accountable to whom, for what, and how.

The main broader lessons to be learned from these two cases are, first, that interorganizational collaborative arrangements within the state to handle wicked transboundary problems are constrained by traditional vertical specialization and strong silo structures in government. Second, accountability relations in such arrangements are complex and tend to be less addressed than collaboration. Shared accountability relations are difficult to develop in a system with strong vertical accountability. However, in collaborative arrangements some form of shared accountability between the different actors involved is necessary and helps ensure commitment to the common goal(s) of the collaboration. Third, both cultural and structural factors affect how collaborative arrangements and accountability relations play out in practice. Fourth, time matters. Collaborative arrangements tend to work better when they become more settled and adjusted, and once those involved have come to work better together following more interaction over time and experiential learning.

One conclusion is that a theory of democratic accountability that aims to capture the role of accountability processes in novel collaborative arrangements within the state, addressing wicked and complex transboundary issues, is necessary. Such a theory needs to relax key assumptions from the mainstream rational choice and principal-agent literature and become more open to an institutional approach that takes into account the development of informal norms, routines, and values through socialization. As illustrated by our two cases, collaborative arrangements exhibit coexisting, overlapping, partly competing, and mutually affecting relationships, especially when they address wicked transboundary problems (Olsen 2010). A recombination and repurposing of existing collaborative governance initiatives in an adaptive and agile way within the constraints of instrumental, formal, and cultural features therefore seems worthwhile (Ansell, Trondal, and Øgård 2017; Greve et al. 2019).

A possible approach to handling accountability challenges in collaborative arrangements could be to adopt a probing strategy whereby feedback obtained via small-scale investigations and experimentation could form a response strategy (Ansell and Bartensberger 2017). A further unpacking and understanding of the mixed order and dynamic relationship between collaborative arrangements and ensuing accountability relations seems to be a promising first step. Hence, more empirical studies of collaboration and collaborative arrangements within the state, emphasizing the presence and development of accountability relations in such arrangements, is necessary.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the European Union’s Horizon 2020 research and innovation programme under grant agreement No [726840].

Notes on contributors

Per Lægreid

Per Lægreid is Professor Emeritus at Department of Administration and Organization Theory, University of Bergen, Norway. He has published extensively on public sector reform, public management policy, institutional change and crisis management from a comparative perspective. His latest books include Societal Security and Crisis Management. Governance Capacity and Legitimacy (with L. H. Rykkja, eds). London: Palgrave Macmillan and Nordic Administrative Reforms. Lessons for Public Management (with C. Greve, and L.H. Rykkja, eds). London: Palgrave Macmillan and Organization Theory and the Public Sector (with T. Christensen and K.A. Røvik) London: Routledge.

Lise H. Rykkja

Lise H. Rykkja is Professor at the Department of Administration and Organization Theory at the University of Bergen. Her research concentrates on public administration in Europe from an institutional and comparative perspective and focuses on administrative reforms, public management, collaboration, crisis management and public security. Her latest publications include articles in Public Administration, Public Administration Review, Public Management Review, Studies in Conflict & Terrorism and Risk, Hazards & Crisis in Public Policy. Rykkja currently leads the Horizon 2020 project: Transforming into Open, Innovative and Collaborative Governments (TROPICO).

Notes

1. In 2019 the Ministry of Justice and its subordinate central agencies of the Police and the Secretariat of the Conflict Council was included in the network.

2. Fifteen interviews with key people in the programme were carried out in 2016. In 2017 an additional 20 core participants in ministries, central agencies, and counties were interviewed, and interviews were conducted in four municipalities. In 2019, interviews were conducted in the steering group, secretariat, working groups, and central agencies. A survey was conducted in 2017 covering 155 persons in ministries, central agencies, county governors’ offices, and counties. The response rate was 62%. Deloitte also observed steering group and secretariat meetings during this period. There has also been a process evaluation of the ‘0–24’ collaboration, focusing mainly on Nordic collaboration on such arrangements, which complements Deloitte’s findings (Skog Hansen, Steen Jensen, and Hansen 2020).

3. While traditional forms of state-centred labour market regulations have focused on issues such as undeclared work and deteriorating/degrading?? working conditions, offering protection to citizens against health, environmental, and safety hazards in the work place, more recent policies have aimed to increase the national capacity to prevent and prosecute violations.

4. Consisting of the leaders/directors of the Public Prosecution Authority, the Police Directorate, the Authority for Investigation and Prosecution of Economic and Environmental Crime, the Tax Agency, the Customs Agency, the Labour Authority (NAV), and the Labour Inspectorate

5. The reports were published jointly by the Labour Inspection Authority, the Tax Authority, the Police, and the Labour and Welfare Administration.