Strategies and transitions to public sector co-creation across Europe

This article analyses 14 co-creation cases in six European countries from macro and meso theoretical perspectives on strategic public management. The comparative case study includes focus group sessions and interviews with 107 professionals and 26 managers. Findings show, ﬁ rst, that co-creation is not limited to countries with a strong af ﬁ nity to new public governance, but also occur in more traditional public administration – or new public management contexts. Second, co-creation cases show a large variety in strategies at the meso level of organizations and networks, but strategic planning seems a precondition for process-based strategies to succeed. Third, despite many differences between the cases and countries, frontline actors often encounter similar barriers in co-creation, in particular in public leadership and vertical organizational structures. Our ﬁ ndings contribute to the literature on strategic public management and transitions in public administration.


Introduction
Co-creation is considered a panacea for many complex social and environmental problems that cannot be solved by traditional hierarchy-based policy-making or marketbased principles alone.Like literature on public networks or collaborative governance, the co-creation literature emphasizes that solving complex problems requires coordination between many actors from different subfields (Bryson et al., 2015).Additionally, it stresses the involvement of lay people (citizens, communities) and innovation (new solutions beyond existing policies, creating public value) (Torfing et al., 2021).Co-creation should improve the functioning of the public sector through a better understanding of social challenges (e.g.climate change, welfare issues) and attract new resources to meet such needs (Ansell and Torfing, 2017).It should lead to integrated services and coordinated governance (Mavrot and Hadorn, 2021); build shared ownership of solutions to social issues and increase democratic legitimacy (Sørensen et al., 2021;Torfing et al., 2021).Although co-creation is seen as a promising concept, there are few empirical studies (Voorberg et al., 2015;Ferlie, 2021), let alone from a comparative and multi-level perspective (Baptista et al., 2020;Voorberg et al., 2017).This article's purpose is to theoretically and empirically examine co-creation, taking a comparative view on strategies and transitions across Europe.The theoretical framework is based on the literature of strategic management at multiple levels and with regard to the public sector.We combine literature on public management models at the macro/national level (Osborne, 2006;Pollitt and Bouckaert, 2017) and strategic management models at the meso level of organizations and networks (Ferlie and Ongaro, 2022).Research into the impact of macro contexts and how they are dealt with in organizations points to differences between European countries where legal frameworks, expertise and working arrangements have evolved in nationally specific ways (Pollitt and Bouckaert, 2017).Thus we are aware of different national contexts across Europe, but there is a need for theoretical and empirical comparisons to better understand how public management models at the macro level take shape at the meso level of organizations and networks and the micro level of relationships that arise in the co-creation between public actors, citizens and other parties (Brodkin, 2011;Pandey and Wright, 2006;Zacka, 2017).
Professionals traditionally have a key role in balancing national and organizational public policy with the needs of the citizens and communities they serve (Hendrikx et al., 2022;Zacka, 2017).To allow for in-depth knowledge of the strategies and transitions in co-creation, our empirical focus is on co-creation cases where public professionals and their managers are in direct contact with citizens and other stakeholders.Our research questions are 'How do macro public management models facilitate and/or hinder the transition to co-creation at the meso and micro level?Which meso-strategies do organizations and networks use to develop co-creation in the public sector across Europe?'The empirical research is a comparative case study of 14 co-creation projects across six European countries: Croatia, Denmark, France, Netherlands, Slovenia and the United Kingdom.Data are based upon (focus group) interviews with 107 frontline professionals, and 26 public managers involved in the 14 co-creation cases.

2
Public Policy and Administration 0(0) Contributions of this study to the literature are threefold.First, we show how public management models considered dominant for a country (e.g.New Public Management (NPM) in the UK, Traditional Public Administration (TPA) in France, Croatia or Slovenia, or New Public Governance (NPG) in Denmark or the Netherlands) are recognized in some cases of co-creation, but with notable exceptions, so macro level models offer an insufficient explanation as to how co-creation develops.Second, contrary to expectations, at the meso level of organizations and networks in the six countries, a strategic planning approach seems to be a precondition for co-creation, also in countries with a tradition in networking and participation.Third, despite major differences in the macro and meso contexts of strategic management in the six countries, street-level bureaucrats encounter remarkably similar barriers to co-creation.They point to the support of public leaders and the transformation of the vertical organization of the public sector as important conditions for co-creation.
The article is structured as follows.We present our theoretical framework and expectations based on the literature on strategic (public) management.We then explain our research methodology and present our findings on strategy and transitions in 14 cocreation projects in six European countries; followed by a discussion and conclusion.

Theoretical framework
Co-creation has different meanings and definitions.Here, we adopt a broad view of cocreation which encompasses collaboration between the public sector with citizens, and actors in the private and voluntary sector (Torfing et al., 2021).We do not distinguish between co-creation in service delivery from other policy phases, such as agenda setting and policy design (Brandsen and Honingh, 2018); this means that when making theoretical assumptions we lean on the literature that uses the term 'co-creation' as well as 'coproduction'.
A theoretical framework for studying co-creation should discuss the form and content of strategies, their creation and implementation, and the actors involved in this process (Torfing et al., 2021).Taking these elements as a basis for our theoretical framework, we first discuss strategic management at the macro level of public management models (TPA, NPM and NPG); second, strategy models at the meso level of organizations and networks (strategic planning and process-based models); and third, at the micro level, professionals and public leaders and their relationships with other parties in co-creation.Three broad expectations are derived from the literature to guide our empirical analysis.

Macro level: Public management models and co-creation
In the public management literature, main ideal-typical models have emerged over time, such as Traditional Public Administration (TPA) and its modern variant the neo-Weberian state; New Public Management (NPM); and New Public Governance (NPG), known also as Network Governance (NG) (Ferlie and Ongaro, 2022;Osborne, 2006;Pollitt and Bouckaert, 2017).These models present a national level perspective on strategic van Gestel et al.
management and each model provides a different context for promoting co-creation as an integral part of public administration (Torfing et al., 2021).
The TPA model is based on Weberian principles (Ferlie and Ongaro, 2022;Pollitt and Bouckaert, 2017) with hierarchy as the main coordination mechanism, as opposed to the co-creation goal of developing horizontal relationships 'ideally on an equal footing' (Ansell and Torfing, 2017: 216).TPA has a central role for the government in solving societal problems.Solutions to social problems are designed by public servants as main experts, driven by a professional ethos based on equity and public interest.Neither external stakeholders nor individual citizens have a (formal) position in TPA other than voting for politicians, who play a central role in policy development and decision-making on their behalf.
NPM has become a leading model in the public sector, spread around the world since the 1980s, but with varying influence given differences in national history, constitution and political culture (Pollitt and Bouckaert, 2017).Unlike TPA's focus on the legal rules and procedures in policy-making and service delivery, NPM is outputoriented (Alford and Freijser, 2018) and emphasizes private sector principles, such as competition and customer focus, for better public sector outputs (Hood, 1991).Management plays an overarching role in NPM, introducing performance targets to increase efficiency and productivity (Aucoin, 1990).NPM provides few incentives for collaboration between organizations and their internal units as silo mentality and territorial wars emerge (Alford and Freijser, 2018).It sees citizens as customers or clients, a more passive role than co-creators.NPM's 'business-like' management focuses more on vested interests than on less powerful citizens and civil society organizations (Diefenbach, 2009).
Around the turn of the century, an alternative ideal of collaborative or networked organizational forms (NPG) became popular, also called Network Governance (NG) (Ferlie and Ongaro, 2022;Osborne, 2006).Many countries became interested in moving away from traditional public administration and the market/management mix that characterizes NPM reforms, to adopt models that provide more room for collaboration and participation (Emerson et al., 2012).NPG is a pluralistic model of public governance, based on collaboration between governments and private and non-profit stakeholders in networks and partnerships (Bryson et al., 2015;Osborne, 2006).The role of citizens in NPG is less defined; they can have an active role in co-developing (as co-producers) aspects of public services (Pestoff, 2018), but the active involvement of lay actors (e.g.users, citizens and local communities) in NPG is not self-evident (Torfing et al., 2021).
While it is often assumed that co-creation and co-production best suit NPG, McMullin (2021) showed for third sector organizations they can also occur in macro contexts dominated by TPA (France) or NPM (UK).Thus far, there are few empirical studies of co-creation beyond single case studies (Baptista et al., 2020;Voorberg et al., 2015), especially in investigating the relationship between the various public management models and co-creation (Voorberg et al., 2017).Further comparative studies are needed to understand how the different strategic models at the macro level (TPA/

4
Public Policy and Administration 0(0) NPM/NPG) stimulate or hinder co-creation at the meso (organization/network) and micro level (Ongaro et al., 2021).This is not an easy task as TPA, NPM and NPG are ideal-typical models that have influenced strategic management in various combinations.However, while no country is a pure example, the characteristics of one model may be more dominant than another.While NPM has influenced government reform worldwide, the model has been transformative in Anglo-Saxon countries such as the UK.Continental states such as France, Croatia and Slovenia have been more resistant to its adoption and are better described as TPA/neo-Weberian than NPG (Pegan, 2023;Pollitt and Bouckaert, 2017;Vrbek and Kuiper, 2022).Scandinavian countries like Denmark have been characterized as consensual democracies, more open to networks with stakeholders and participation, as in the NPG model (Torfing et al., 2021).This also applies to the Netherlands as a decentralized unitary state with a tradition of consensus-oriented policy-making (Kickert et al., 1997).Nonetheless, both the Netherlands and Denmark also have a legacy in legalism as seen in TPA, which could limit the application of co-creation (Voorberg et al., 2017).While acknowledging such hybridity (Denis et al., 2015), we use the three macro models as a guideline when analysing our data.
Expectation 1. Public management in countries known for their relatively strong TPA background (in our study, France, Slovenia and Croatia) hinders co-creation, while in countries characterized by a tradition of NPG (here Denmark and the Netherlands) public management promotes co-creation.We assume that the NPM setting of the UK takes an intermediate position: open to self-government (Baptista et al., 2020), but less positive to horizontal relationships between government and external stakeholders (Alford and Freijser, 2018;Pestoff, 2018).

Meso level: Strategic management models and co-creation in organizations and networks
At the meso level of public organizations or public-private networks, strategic management is needed for co-creation to expand beyond single projects (Torfing et al., 2021;Van Gestel, 2020).Change at a strategic level is necessary because a systemic turn toward co-creation within public organizations will not happen automatically.Strategic management in organizations differs from operational management in important respects: it has a long-term rather than short-term focus; holds an organization-wide approach more than a focus on a single project or department, and involves all relevant components for operations, from finance to HRM and implementation (Ferlie, 2021).
An overview and discussion of several schools of strategic management at the meso level of organizations and networks is presented by Ferlie and Ongaro (2022).Many van Gestel et al. models the authors present were originally developed for the private sector but can also be relevant for public organizations.Among the models are the classic design school of strategy, the strategic planning school, and the strategic positioning approach associated with Porter; all developed during the same period as the TPA model and characterized by similar principles of hierarchy, top-down planning and high-level professional expertise.Strategic planning is known for rational and formal steps and procedures to achieve organizational goals, although some define it as a more deliberative, yet disciplined effort (Bryson, 2011).Ferlie and Ongaro (2022) also describe a group of alternative strategic management models that focus on corporate culture, history, process and organizational learning, for example, the Mintzberg approach or the cultural school.Later schools of strategic management are also often process-based, emphasizing plural knowledge, value diversity, collaborations between different actors and a gradual, bottom-up emergence of strategy.Examples are the resource-based view of the organization, the strategy as process school, collaborative network approaches and the public value model (Ferlie and Ongaro, 2022).Ongaro et al. (2021) examined three strategic management models that were useful in supporting a shift to co-creation in government agencies.Pegan (2023) explored strategic planning in Croatia and Slovenia.Since these studies are limited to one case or model, there is a clear need for more comparative research (Ferlie, 2021;Baptista et al., 2020).
In this article, we investigate which strategic management models at the meso level can be recognized in the co-creation cases in our study.For the purpose of this study, we divide the meso level strategic management models into two types: strategic planning models and process-based models (see Table 1).Expectation 2. In countries with a strong background in TPA (in our study France, Slovenia and Croatia) strategic planning models are used for co-creation, due to a corresponding emphasis on hierarchy and a top-down approach, while cases in countries with a tradition in NPG (here Denmark and the Netherlands) will lean on process-based strategic management.We assume that in the UK NPM setting mainly strategic planning is used to steer and measure the performance of co-creation.

Micro level: Public actors and relationships in co-creation
Theories of co-creation emphasize the active involvement of state and non-state actors, such as private and non-profit organizations and service users in the design and delivery of public services and the development of public policy (Torfing et al., 2021).Public actors seen as strongly involved in the co-creation process with citizens are the professionals (Pestoff, 2018;Zacka, 2017).In classic literature, professionals are characterized by a (technical) knowledge base, a service ethic and autonomy (Wilensky, 1964).Examples of so-called 'classic' or 'full' professionals are medical doctors or lawyers (Noordegraaf, 2015).We adopt an inclusive definition of professionals, also studying groups that are seen as 'semi-professional' (Etzioni, 1969), such as police officers or social workers.They may have less control over their work than 'full' professionals but a decisive role in direct interactions with citizens and other stakeholders in co-creation.Professionals can therefore be viewed as mediators between the macro level of the state, the meso level of organizations and networks, and the micro level of co-creation in practice (Hendrikx et al., 2022).
The success rate of implementation and the quality of service innovation would be enhanced by involving frontline workers in policy design (Ansell and Torfing, 2017;Osborne, 2006).The likelihood of involvement of professionals, in turn, depends on the roles and values that public management models place on them.Zacka (2017) highlights the different, competing demands on professionals that arise from macro public management models (TPA/ NPM/NPG).Emphasizing competences and standards, professionals are the main guarantors of service quality in TPA (Pestoff, 2018).But national policies and organizational and managerial control may also challenge professionalism.NPM is often blamed for undermining professionals' discretion by striving for short-term efficiency and control rather than strategic goals and collaboration with 'users' (Brodkin, 2011;Diefenbach, 2009).
Public leaders, in particular politicians and top executive managers, are another group of important but less explored actors in the co-creation literature (Sørensen et al., 2021).In TPA, policy-making is primarily a matter for politicians and the government divides tasks and responsibilities based on specialized expertise (Peters, 2018).In traditional models of representative democracy, the emphasis is on acquiring and retaining power between established political actors (e.g.parliamentary parties) rather than sharing it with the electorate and other societal actors beyond what is necessary for (re-)election.NPM has further strengthened the traditional split between policy-making and implementation (Pollitt and Bouckaert, 2017).Given the potential benefits of cocreation as part of a government strategy, few have considered what co-creation means for public leadership (Sørensen et al., 2021).Although co-creation might reduce the influence of traditional key players in representative democracy, politicians and public van Gestel et al.
administrators can also benefit from the knowledge and support that arise from cocreation in the formulation and implementation of public policy (Ansell and Torfing, 2017;Sørensen et al., 2021).
Expectation 3. Cases from countries with a stronger TPA profile (here France, Slovenia and Croatia) have a limited role for public leaders in enhancing co-creation, while in countries with a tradition in NPG (in our study Denmark and the Netherlands) public leaders stimulate co-creation and engage professionals in co-creation.We suppose that in the UK case, given the NPM setting, the decision to take up co-creation is led by market principles rather than political commitments.
We use these three broad expectations resulting from our literature review to determine the type of strategic management at the macro and meso level, and what enables or hinders co-creation at the micro level.

Case selection
A comparative case study approach is chosen in this study (Eisenhardt 1989).The cocreation cases were selected for similar characteristics to allow comparison: they include multiple (state and non-state) actors and different policy domains (Bryson et al., 2015); aim to address problems not solved by traditional policies, and seek new solutions (Ansell and Torfing, 2017;Torfing et al., 2021); aim to create more integrated services and coordinated governance (Mavrot and Hadorn, 2021); and look for building shared ownership and democratic legitimacy (Sørensen et al., 2021;Torfing et al., 2021).The 14 cases were further selected for balancing geographical location to represent European countries (Croatia, Denmark, France, Netherlands, Slovenia and the United Kingdom/ England).They cover broad policy domains (e.g.social policy, environmental policy); and include public organizations and public/private networks or third sector entities.As a condition, we sought consent for interviewing professionals and managers directly involved in co-creation.The criteria were successfully met.The 14 cases reflected similar characteristics of co-creation projects in a variety of countries (see Table 2), to compare how the macro context may (or may not) influence meso strategies and micro practices; and which meso strategies were developed.

Data collection
The data was collected from November 2019 to November 2020.Case studies were based on qualitative data: focus group sessions, semi-structured interviews, and documents.Respondents were professionals and their managers involved in cocreation projects, in direct contact with citizens and other stakeholders.With the consent of respondents, the focus groups/interviews were recorded with anonymity and confidentiality guaranteed.Focus group sessions were led by a discussion leader from each participating country, with one other researcher taking notes.Most group   sessions were carried out online during the pandemic, taking between 90 and 120 min in length.Individual interviews took 45-60 min.All interviews followed a shared interview guideline investigating the national (macro) and organizational (meso) strategic reforms, and the (micro) case history, drivers and barriers in co-creation, the work and role of professionals and supervisors in co-creation, and prospects of the co-

Data analysis
Each case was reported according to a common template to facilitate comparative analysis.The template was built on a literature review on public management models and the role of public organizations and professionals in co-creation, with a pilot case for case reporting (Hendrikx et al., 2020).Six case reports were composed by the authors of this article and eight by colleagues in the larger research project (see Acknowledgements).In total, 14 case reports were delivered across six countries, based upon 15 focus group sessions with 107 professionals, and interviews with 26 public managers (see Table 2).
The case reports served as input for a systematic overview of the main findings in a summary and evaluation report (Van Gestel, 2020).To illustrate some of our findings, we quote directly from the unpublished case reports (further described in Van Gestel, 2020).
To ensure the reliability of the study, we held two meetings with the larger group of researchers during the study period, discussing case selection criteria, interview templates, respondent types, and case reporting methodology.The first author shared the draft summary and evaluation report with all case authors and the larger research team and used their feedback to finalize the analysis (Van Gestel, 2020).The triangulation of input by researchers for case selection, templates for interviews and case reporting, and sharing findings contributed to the reliability of the study.
For this article, we analysed the 14 case reports in two steps, based on our research questions.First, the case reports were coded to reflect elements of the TPA, NPM and NPG models-as incentives or barriers to co-creation.Based on the literature on public management models (see theory section), excerpts in the case reports that emphasized hierarchy and formal regulation in their co-creation practice were coded as TPA.Equally, passages reflecting NPM emphasized performance, efficiency, competition and financial control.Codes for NPG referenced passages stressing interprofessional collaboration and network arrangements.In a second step, we detected the type of strategic management at the meso level of organizations and networks in the co-creation cases, using the characteristics of strategic planning approaches versus process-based strategies (see Table 1).In the next section, we present our findings.

Findings
Based on the 14 cases in six countries, our findings allow for a comparison, refinement and discussion of the strategic management models, as applied to co-creation.We use the three expectations developed in the theory section to discuss strategic management at the macro, meso and micro level in relation to our cases.
van Gestel et al.

The macro context of strategic management
The first expectation was that countries known for strong TPA backgrounds (in our study France, Slovenia and Croatia) are less likely to promote co-creation, while countries characterized by a NPG tradition (in our study Denmark and the Netherlands) will do so.The UK, as a prominent NPM example, is supposed to take an intermediate position on co-creation: positive towards more focus on users, but less positive towards horizontal relationships.
The French cases (Veloux Nova space and Aix Marseille University) and the Slovenian cases (National and Local public administration in Ljubljana) indeed demonstrate the impact of a traditional hierarchy more clearly than elsewhere.These four cases have a strong adherence to TPA but also NPM rather than to NPG and co-creation.As illustrated by the Nova Velaux case: 'The French local authorities are administrations with rigid and bureaucratic structures [and] the latest reforms show that NPM is still very present in French public organisations'.Although the managers and professionals in the French and Slovenian cases intend to compensate citizens for their problems, especially during COVID-19, co-creation as defined in the literature is not enacted: defining citizens' needs is a one-sided approach by government actors and collaboration with external stakeholders is limited.The Slovenian case report on the Ministry of Public Administration concluded: 'Both the focus group and the interview revealed a strong inclination towards a legal positivist approach, which does not leave much leeway for professionals to experiment or include external stakeholders beyond what is stipulated by the law'.An exception in this group of TPA/NPM led countries is the RLPP case in Croatia; a longterm co-creation program initiated in 2005 for strengthening civic values.Respondents highlighted that citizens' involvement in policy-making and the frequency of cooperation with external stakeholders 'is greatly increasing'.However, a TPA/NPM legacy is also visible as the RLPP case started with 'a more liberal approach to citizens, but later […] led to more rules to be followed by them'.
Limited co-creation is not restricted to our cases in France or Slovenia, but also includes some country cases which we expected to be open to NPG.For example, in Copenhagen, Denmark: '[…] the Climate Change Collaborative does not succeed in making the homeowners capable of co-creating a plan for the climate adaptation of their private common areas'.In addition, two Dutch cases show limited co-creation.In the Maastricht case, hierarchy in policy-making tends to limit options for co-creation: 'There is a firm tendency to standardize and protocolise in the wider organization.Hence, striving for client-centred, integrated services (NPG) goes hand in hand with a focus on standardized performance and output (NPM)'.Also, the strategic management in the Pijnacker-Nootdorp case has a TPA/NPM focus rather than NPG.A third Dutch case in the same (social) policy domain (Drechtsteden), however, strongly refers to NPG in embedding co-creation, emphasizing client-centeredness and effective internal and external collaborations to reach policy aims.The manager explains: 'Professionals are expected to not only work within legal frameworks (what is 'allowed' and legitimate), but to look explicitly at what is needed to help clients get their lives together'.Thus, there is strong local variation within the same macro (NPG) context; even within the same policy 14 Public Policy and Administration 0(0) domain, which contradicts the notion that macro trends will be visible across countries and policy domains.
The UK case (Tyne Housing) takes an intermediate position; it resembles a macro strategy based on NPM, with increasing government demands to demonstrate value for money, proving data security and increasing contractual performance (vertical relationships).But it also shows that users (a residency consultation group) are involved in strategic plans with their input being sought after, valued and implemented.The UK case suggests that co-creation in an NPM setting is easier with user groups than with organizations, as they are potential competitors for government grants.
We conclude that our first expectation is only partially supported.Indeed, the strategic management models (TPA/NPM/NPG) point to different national contexts for the cocreation cases, in line with our expectation.But some cases show opposite findings (RLPP in Croatia; Climate policy case in Denmark), and we also found variation within the same country (Denmark), and even within the same policy area (Netherlands).These findings imply that explaining co-creation from strategic management models at the macro level (TPA/NPM/NPG) is not the full story, so we need additional explanations.The next section deals with strategic management at the meso level.

Meso level: Strategic management at the level of the organization/network
Our second expectation was that cases from countries with a more traditional/neo-Weberian public administration profile (in our study, France, Slovenia and Croatia) will use strategic planning models, while countries with a tradition in NPG (as Denmark and the Netherlands) hold process-oriented types of strategic management for co-creation.We assume that in the UK as a typical NPM example, our case relies on strategic planning to guide and measure the performance of co-creation.
To begin with, our findings highlight that six co-creation cases are explicitly designed by municipalities or public-private networks to take lessons in order to develop cocreation in the wider organization/network. Hence, the primary aim of these organizations/networks is strategic -rather than operational management.Examples are Holbaek (Knastrup, Denmark), RLPP in Croatia, the Climate policy case in Copenhagen, the municipal case in Slagelse (Denmark), and the Dutch cases of Drechtsteden and the Netherlands Enterprise Agency.The other eight cases have a less prominent role as a potential beacon for widening co-creation (the French and Slovenian cases, the English case, two Dutch cases and one Danish case).
Three out of the six 'strategic management cases' -Holbaek (Knastrup) in Denmark, RLPP in Croatia and Drechtsteden in the Netherlandsare 'strategic planning' examples of co-creation.All have a relatively long history of co-creation (10-15 years) and a strong political backing.Co-creation in these cases was a political initiative to avoid (future) social problems.In the case of Knabstrup village, the larger Holbaek municipality initiated co-creation with the village community because of population ageing with a future negative impact on continuing public services for education, leisure and housing.However, as concluded in the case report, the strategic planning approach is seen in line with NPG rather than TPA/NPM: 'All in all, the co-creation model in Holbaek van Gestel et al.
municipality seems highly inspired by the NPG-approach'.In Rijeka (RLPP, Croatia), the Mayor took the initiative for digital co-creation with citizens because he believed the transition to a democratic society in a post-communist country needed the strengthening of civic values.In Drechtsteden (Netherlands), local public leadership recognized problems of social inequality and joined forces with (semi-)public and private organizations and citizens to develop a strategic agenda for a flourishing social and economic region.So, in these three strategic co-creation cases the initiative came from the top of the organization/network, in line with formal strategic planning models but in different macro contexts (TPA/NPM/NPG).
The approach in the other three 'strategic management' cases resembles a processbased and more bottom-up approach, with an emphasis on experiments and learning.In Slagelse (Denmark), the municipal strategic ambition is to develop a viable Danish welfare model through co-creation rather than by top-down planning.In a bottom-up process, professionals are encouraged to carry out 'trial-and-error-acts' in Local Development Plans for community support, e.g.child nursery.The case context, however, is a mixture of NPG, NPM and TPA, as concluded in the case report: 'The self-management and co-creational re-organization of Slagelse municipality is in line with the principles of NPG, emphasizing distributed leadership and horizontal collaboration with internal and external actors', but 'the CEO of Slagelse municipality underpins he's also "a great fan of NPM." […] [and] the success of self-governance and co-creation depends also on the ability of professionals to navigate in a traditional top-down hierarchy'.Also the Climate Change Partnership in Copenhagen was set up as a bottom-up co-creation process to develop adaptation plans for preventing damage from flooding after heavy rainfall and cloud bursts.To conduct experimental co-creation with homeowner associations, an organization called Energy and Water was appointed in the City of Copenhagen as the process facilitator of co-creation.Conversely, the homeowners felt little engaged in the process, but were satisfied with a traditional, top-down solution from the professional experts in this 'technically complicated project'.In the case of the Netherlands Enterprise Agency, a new strategy for sustainable development goals based on the public value model was discussed and adopted in bottom-up staff meetings.Societal challenges (climate change, energy transition, green agriculture) required new solutions, with a coherent, holistic approach from the executive agency rather than separate tasks executed in silos.Here the bottom-up co-creation strategy ran into implementation barriers because the Netherlands Enterprise Agency remains dependent on funding it receives for the fulfilment of TPA/NPM assignments.As the focus group participants saw it: 'The organization is primarily focused on efficient processes and production on time and within budget; the current organizational structure therefore does not facilitate the intended strategy change'.All three process-based strategic management cases thus show that cocreation is influenced by a mix of TPA, NPM and NPG, creating tensions for implementation.
Two voluntary sector organizations in our study also followed a process-based approach, with organizational core values as a higher order influence over how decisions are made in organizations (cultural school).The two cases -INSP (Roskilde, Denmark) and Tyne Housing (England) -have much in common here.In both cases the aim is to assist 16 Public Policy and Administration 0(0) vulnerable people: homeless youngsters or adults with multiple problems.Co-creation is used as a tool to support self-governance.In contrast to other cases, they are not initiated by government or public agencies, although there are close relationships via regulation and funding.The two cases differ in that Tyne Housing is mostly financed through government funds and regulated by law, while INSP is paid for by a range of actors (municipality, private actors and INSP-organized activities) and government regulation is weak, although administrative requirements tend to increase.In both cases, 'external' accountability requirements are perceived adverse to the organizations' aims and approach to improve the life of homeless people.
In sum, our second expectation cannot be confirmed.At the meso level of organizations and networks, we saw strategic planning cases as well as process-oriented strategic management, but they were not necessarily linked to TPA/NPM versus NPG-type countries.Third sector organizations in our study are indeed a special category, with similar characteristics at the meso level despite a different macro context (UK and Denmark).Another notable finding is that strategic planning cases, contrary to expectations, are not merely found in TPA-type countries but especially in NPG-type countries.Thus, the meso level analysis of strategic management provides additional and sometimes adverse explanations over the macro level analysis.

Micro level: Public actors and relationships in co-creation
Our third expectation was that cases from countries with a strong TPA profile (in our study France, Slovenia and Croatia) would have a limited role for public leaders in enhancing co-creation, while countries with a tradition in NPG (in our study Denmark and the Netherlands) may stimulate co-creation and engage professionals in such projects.We assumed that public leaders in the UK, as a prominent NPM example, will be led by market principles rather than political commitments to take up co-creation.
Our findings show that (local) politicians play different roles in the co-creation cases.Three categories of political involvement appear in our 14 cases: politicians taking the initiative and the lead; politicians taking a more passive role but supporting co-creation by rules and finance; and politicians being detached from co-creation.In the first category, we see local politicians as initiators, for example in RLPP in Croatia; but also in Holbaek (Denmark); and in Drechtsteden (Netherlands).In the Holbaek case this leading role is explained as: "the NPG-orientation is largely connected to an ongoing TPA form of governance, where political backing from the top is vital for success".Cases in the second category often -but not exclusively -are initiatives led by civil society.Compared to the first category, these cases display a less prominent role for local public leaders, for example in Tyne Housing (UK) and INSP (Roskilde, Denmark), although these cocreation projects are financially supported by the government and operate in a regulatory framework.In the case of Tyne Housing, regulation is also more institutionalized and expanded than in INSP, where private donors have a greater role.Both third sector organizations also keep local politics at a distance to protect their civic aims.As mentioned by an INSP respondent: 'To be politicized is a delicate issue for us'.Politicians adopting their co-creation project for electoral reasons, 'actually hurt us more than van Gestel et al.
strengthen us, because it brings us into the political game'.The third category of a detached role of politicians is demonstrated in, for example, the Climate case in Copenhagen (Denmark), or the cultural service of the town of Velaux in France.Both cases do poorly in terms of innovation; especially the Copenhagen case is described as a failure of co-creation.Also in the Slagelse case (Denmark), the Aix-Marseille case (France) and the Maastricht case (Netherlands), respondents argue that public leaders are difficult to engage in co-creation.In Slagelse, for example, it was said that politicians 'apparently are less inclined to enter into such dialogical processes and often claim that the political committees are the representative, democratic institutions'.So we see a diverse picture here, with no clear relationship between macro models and the role of public leaders at the micro level.
Many professionals involved in our study are highly motivated for co-creation.They feel it is meaningful to contribute to the life of citizens and social/sustainability goals, and point to the energizing contact with citizens and other stakeholders in cocreation projects.A most prominent barrier to co-creation they mention is limited support from top public leaders.Although a few cases have ambitious politicians with (inter)national networks and resources (RLPP in Croatia, Holbaek (Knastrup) in Denmark and Drechtsteden, Netherlands); in many other cases, the role of public leaders is more detached and less supportive.Public leaders' support is however viewed crucial by professionals and their supervisors in our study.At the same time, it seems difficult to realize in most cases.We note that while public leaders are often assumed to involve professionals in transformation/change, some findings here seem to point to the opposite.
Another barrier frequently mentioned at the micro level is the internal vertical organisation of public organisations.Professionals suffer from a mismatch between horizontal co-creation projects and the vertically organized implementation of rules and funding within government, as in TPA and NPM.As suggested in the Nova Veloux (France) case, 'First of all, it is necessary to strengthen cooperation between municipal services'.Professionals frequently mention tensions in combining NPM type of work (e.g.competing for contracts) with co-creation efforts.The Tyne Housing case (UK) confirms: 'Focus group participants were clear in underlining that cooperation (NPG) is demanded alongside competition (NPM), which they felt were conflicting duties'.Also, administration and managerial work takes up more time given NPM type targets and incentives, also found in NPG-type countries, with the adverse effect of leaving less time to spend on service users and co-creation.Moreover, in many cases, professionals and managers experience austerity, implying limited resources for co-creation at the expense of investments in internal and external relationships necessary to achieve the desired outcomes.Co-creation must therefore be realized in a hybrid environment with multiple requirements, which are burdensome for many professionals.As argued in the Maastricht (Netherlands) case: 'Professionals sometimes feel stuck in their multiple roles because they have to deliver client-oriented services (NPG) that fit within legal frameworks (TPA), and also are cost-effective and efficient (NPM)'.
Thus, our third expectation that public leaders would have a limited role in enhancing co-creation in countries with a strong TPA profile is confirmed by cases in France and

18
Public Policy and Administration 0(0) Slovenia, but not in Rijeka, Croatia.A more prominent role for public leaders in NPG countries is only recognized in Drechtsteden (Netherlands) and Holbaek (Denmark), but not in other cases in a similar macro context.Macro strategies still make a difference with NPM-elements (competition, tendering) dominating the UK case.Co-creation reveals a certain impact on public actor's roles from higher level strategies but with considerable variation at the micro level in our study, both between and within countries.

Discussion and conclusion
Our research shows a large variation between the 14 co-creation projects in different countries and at different levels.We found that the macro models (TPA, NPM, NPG) are relevant in some cases for explaining strategic management at the meso level, but they do not provide a full explanation due to contradictory findings and also country variations.Moreover, at the meso level of organizations and networks, we found that, contrary to expectations, cases of strategic planning occur not only in TPA-like countries, but especially in NPG-like countries.The analysis at the meso level thus provides additional -and sometimes contradictory -explanations compared to the analysis at the macro level.At the micro level of co-creation projects, public leaders are considered critical to the success of co-creation, but few projects show their active involvement, especially in strategic planning cases rather than process-based approaches.
We contribute to theory building and further empirical research into strategic management of co-creation for public services in three ways: First, we find some support that state and governance traditions are influential for cocreation initiatives (also Pestoff 2018;Alford and Freijser, 2018;Voorberg et al., 2017).It is recognized that in countries where the state plays a more central role, such as in France, Croatia and Slovenia, there is less room for other stakeholders to participate in co-creation initiatives; while the UK seems to allow more room for self-organization and the participatory behaviour of citizens (also, Baptista et al., 2020).Greater discretion is seen here as a precondition for co-creation, also emphasized in countries with a decentralized structure, such as the Netherlands and in Denmark, where there is a tradition of participation (Kickert et al., 1997).
Our study adds that macro level models and distinctions do not neatly apply to cocreation cases at the meso level of organizations and networks.Some cases even display opposite findings with co-creation being limited in NPG countries or supported in TPA and NPM contexts.We also found variation within countries, and even within policy areas, showing that only a macro level explanation is unsatisfactory.While Sørensen et al. (2021: 270) treat NPG as 'an umbrella term for different perspectives on interactive governance such as network governance, collaborative governance and co-creation', we show that co-creation is not necessarily linked to NPG as it is also found in TPA and NPM contexts.
Our findings stimulate further research into the fit between macro level strategic management models (TPA/NPM/NPG) with co-creation at the meso level.Despite clear differences in governance traditions, we saw hybridity in most cases (cf.Denis et al., 2015;van Gestel et al. Hendrikx et al., 2022) with values, pressures and tensions from alternative governance models (McMullin, 2021;Torfing et al., 2021).Our findings suggest that barriers and conditions to cocreation need a better understanding of the broader strategic context and its implications.Based on our study, we stress the importance of a comparative analysis of co-creation that includes macro and meso level strategic management in addition to a common micro level focus on single cases (Baptista et al., 2020;Voorberg et al., 2015).
Second, we add that strategic planning seems conditional for co-creation, also in NPG countries.A prevalence of strategic planning for co-creation may sound counterintuitive but resembles broader conceptions of strategic planning worked out by Bryson et al. (2018).In this line, strategic planning means a clear realization that strategies are both deliberately set in advance and emergent in practice.It implies an initial focus on a broad agenda and later a more selective action orientation (Ongaro et al., 2021) to effectively turn strategic planning into a process approach that includes gaining legitimacy and involves multiple levels of government and private and non-profit actors (Bryson et al., 2018).Strategic planning might be used to overcome the inflexibility of hierarchy and legalism (Voorberg et al., 2017), especially in TPA/NPM dominated countries, but it can also support the collaborative process in NPG countries.We saw that the strategic management models at the meso level affect the absence or involvement of public leaders in co-creation cases, with a stronger role of politicians in strategic planning cases, implying more (financial, operational) support and long-term establishment than in cocreation cases with a pure bottom-up, process-based approach.Overall, our findings suggest the need for further study of how strategic planning can support sustainable cocreation, in addition to bottom-up, process-based approaches.
A third contribution is that despite different macro contexts and a large variation in strategy and transition at the meso level, the cases of co-creation are remarkably similar in the barriers mentioned by professionals and their managers.This study highlights that even in NPG-type countries, barriers of political and administrative reluctance to 'lose control' (Torfing et al., 2021) are deeply ingrained in the structure and culture of (local) government.Also, a traditional vertical structure goes along with a deeply entrenched hierarchical culture which makes it generally a conservative force in transformation (Ongaro et al., 2021).Moreover, at the micro level of concrete projects, the influences of strategic management at diverse levels come together, and diverse rules and incentives are 'piled up' (Hupe and Buffat, 2014;Zacka, 2017).We saw important differences between what is required in co-creation from public professionals and their managers at 'the streetlevel', and what is offered in terms of conditions in the meso and macro context.Strategic management can thus increase its relevance for co-creation by paying more attention to the agency of frontline actors and their practices, and the condition of internal collaboration in-and between public agencies needed for co-creation.
Our research also has some limitations.First, we conducted our study in six European countries with different public management traditions.Although 14 cases (2-4 cases per country) were studied (except one UK case), future research could compare our findings with a wider variety of co-creation arrangements and methods.Second, given our focus on concrete co-creation projects, the interviewees were frontline public professionals and their managers in direct contact with citizens and other stakeholders.Since the 20 Public Policy and Administration 0(0) construction of co-creation is a collaborative process, more research is needed from different perspectives, namely politicians, citizens and other stakeholders.Third, the purpose of our case selection was to reflect a wide range of co-creation cases in different policy domains, selected for similar co-creation characteristics.We found no immediate evidence that policy domains can be associated with specific forms of strategic management or co-creation, but we left this topic for future analysis.Despite these limitations, our study is, to our knowledge, one of the first to investigate how strategic management at the macro and meso levels influences public professionals and their managers in co-creation, in a comparative study across Europe.We found that macro level changes and organizational strategies (meso level) affect micro level practices, but often in different ways than assumed in literature.Co-creation is both about a vertical relationship between national strategic management, organizational strategy and street-level practices, and horizontal collaboration between public, private and non-profit organizations, each with their own agendas (Mavrot and Hadorn, 2021;Van Gestel, 2020).The aim of making comparative statements about co-creation therefore implies the need to distinguish between contextual strategic influences at different levels of analysis (macro, meso and micro) as we did in our study.

Table 1 .
Strategic models at the organizational/network level.

Table 2 .
Cases and data sources in (semi-)public organisations across Europe.