Policy labs on the fringes: boundary-spanning strategies for enhancing innovation uptake

Abstract

Governments around the world are increasingly launching policy labs to break the alleged inertia of public bureaucracies. Operating on the fringes of their parent organizations, however, public policy labs often live a secluded life with little or no subsequent organizational adoption of lab-designed policies. This research examines how lab managers enhance policy innovation uptake through boundary-spanning strategies and practices connecting public policy labs to their bureaucratic parent organizations. Based on a qualitative study of three Danish public policy labs, it coins and conceptualizes three boundary-spanning strategies for enhancing policy innovation uptake: agenda linking, arena linking, and actor linking. Analyzing how the lab managers carry out each of these strategies in practice, the study shows their respective strengths and weaknesses in building political-administrative support for lab-designed policies on multiple levels of the public bureaucracy. The study findings suggest that lab managers must strike the right balance between under- and over-integration of policy labs in the public bureaucracy in order to enhance innovation uptake, connecting the lab to bureaucratic agendas, arenas, and actors, without compromising the integrity of the lab as a creative space for the experimental co-design of public policies. Conceptualizing managerial strategies and practices for bridging the gap between policy labs and public bureaucracies, the study contributes not only to policy lab theory and practice, but also to the mounting collaborative governance, management and leadership literatures on the tensions and dilemmas arising at the interface between hierarchical government and interactive governance.

Keywords:

• Policy labs
• public bureaucracy
• boundary-spanning
• policy innovation
• innovation uptake
• co-design

1. Introduction

Governments around the world are increasingly launching public policy labs to spur collaborative solutions to complex problems that cut across administrative silos, sectors, and levels of government (Fleischer and Carstens 2022; Fuller and Lochard 2016; Manzini and Staszowski 2013; Wellstead, Gofen, and Carter 2021). Since the early 1990s, public policy makers have established policy labs to break the alleged inertia of public bureaucracies (Ferreira and Botero 2020; Whicher 2021). More recently, public organizations have made increasing use of public policy labs that involve public sector employees, managers, and civil servants in the co-design of new administrative policies cutting bureaucratic red tape and improving the internal workings of the public administration (Cole 2022, 167; see also Lewis 2021; McGann, Blomkamp, and Lewis 2018).

Public policy labs seek to offer “safe spaces for innovation” (Bason 2018, 132) that suspend organizational hierarchy and evade the risk-averse culture of public bureaucracies (McGann, Blomkamp, and Lewis 2018; Mulgan 2014; Torvinen and Jansson 2022). Accordingly, scholars have described them as “islands of experimentation” (Tõnurist, Kattel, and Lember 2017, 8), “distinct organizational forms that have an outsider position” (Lewis 2021, 246–247), and “foreign entities” (Bason 2014, 233), purposely estranged from the day-to-day operations, roles, norms, and procedures of their bureaucratic parent organizations. However, this “safe space” approach appears to be the quintessential delusion of policy labs; simultaneously their greatest strength and greatest weakness. On the one hand, their autonomy as creative spaces of experimental co-design of new policies makes them excel in ideation and invention (Bason 2018, 36–37). On the other hand, they easily become so detached from their parent organization that they end as short-lived, secluded, and isolated projects with little or no subsequent policy innovation uptake and dissemination beyond their confined local contexts (Bason 2014, 34–39; Blomkamp 2018; Cole 2022, 167; Lewis 2021; McGann, Blomkamp, and Lewis 2018; Olejniczak et al. 2020; Torvinen and Jansson 2022; Tõnurist, Kattel, and Lember 2017). This is particularly problematic for labs attempting not only to build competencies and capacities for policy innovation, but to design new policies for improved public processes and procedures that ultimately produce better outcomes for citizens.

Without viable ways of addressing the problem, the policy lab strategy has arrived at a theoretical and practical impasse. The steadily growing policy lab literature has identified the problem without tackling it head-on in an attempt at identifying and developing possible solutions. In practice, lab managers continue to muddle through but still seem to lack sound managerial strategies for bridging the gap between public policy labs and public bureaucracies. Without viable ways of addressing the issue, public policy labs are set up for recurring failure, broken promises, and disappointment following their retrenching capacity for actual policy innovation.

The broader collaborative leadership literature offers valuable insights on boundary-spanning, suggesting managerial strategies for connecting collaborative arrangements to public bureaucracies in ways that garner vital support from sponsors and champions (Aldrich and Herker 1977; van Dorp 2018; Crosby, Bryson, and Stone 2015), but the theory and practice of public policy labs has yet to adopt and adapt these insights for the purpose of innovation uptake. To break the deadlock and contribute to the development of effective strategies and practices for facilitating the organizational uptake of lab-designed public policies, this article addresses the practice-oriented research question: How do managers of public policy labs enhance policy innovation uptake through boundary-spanning strategies and practices?

I proceed in four steps. First, I develop a theoretical understanding of the bureaucratic detachment problem of public policy labs and the boundary work of lab managers. I then introduce the case of the Danish Fremfærd (“Moving Forward”) labs, which involve public administrators, managers, and frontline employees in the co-design of new administrative policies that cut bureaucratic red tape. Next, I propund the research methods applied in a qualitative study of three Fremfærd labs. After that, I present the study findings, coining and analyzing three boundary-spanning strategies induced from the study. Finally, I conclude by discussing the study findings and their implications for public policy lab research and practice.

2. Policy labs and public bureaucracies: a complicated relationship

Both public and private organizations apply the lab strategy for multiple purposes (Cole 2022). For example, labs are initiated as creative platforms for employee-driven ideation and buy-in; as change partners for transforming public and private organizations; and as co-creation hubs for social innovation (Carstensen and Bason 2012). Labs may be government-controlled, i.e. based within and wholly funded by government; government-led, i.e. based within but only partly funded by government; government-enabled, i.e. based within non-government organizations but relying significantly on government funding; or independently-run, i.e. based within and funded by private or third-sector organizations (Lewis 2021; McGann, Blomkamp, and Lewis 2018). Public policy labs focusing on the co-design of new administrative policies that cut bureaucratic red tape are typically either government-controlled or government-led labs that are both based within and at least partly funded by central and/or local government (Bason 2018, 144).

When initiating public policy labs, governments typically issue a mandate stipulating an overall purpose, a set of goals, a timeframe, and the organizational responsibility for designing and managing the labs (Bason 2014, 230–231; see also Krogh 2022). Even for government-controlled and government-led labs, however, relative lab autonomy from the public bureaucracy is crucial for their capacity to function as policy innovation incubators. Based on the principles of co-design theory and design thinking, the lab approach prompts the participants to co-examine policy-design problems and their interrelated causes and effects before developing and refining possible solutions by prototyping and testing them in confined settings (van Buuren et al. 2020). Valuing explorative and open-ended processes based on experimental reasoning, the institutional logic of labs differs significantly from the conventional technical-bureaucratic appreciation of rational, logical, and deductive policy analysis and design (Olejniczak et al. 2020). In fact, the very reason for establishing public policy labs is to provide an alternative space for policy innovation that suspends, circumvents, and disrupts the conventional wisdom, rules, and procedures of the administrative machineries of government (Lewis 2021; Mulgan 2014; Torvinen and Jansson 2022; Tõnurist, Kattel, and Lember 2017).

As much as public policy labs need distance from the public bureaucracy, however, they also depend on it to translate their prototypical solutions into actual policy innovation if they are to improve organizational practice. Public policy lab managers therefore face a bureaucratic detachment/attachment conundrum: They must shield the lab from the technocratic logics of bureaucracy while avoiding lab isolation from the bureaucracy. This relates to what Mulgan (2014) describes as “the radical’s dilemma”: on the one hand, the process of developing innovative alternatives to the status quo requires distance to the established power structure; on the other hand, that very distance may result in ostracism and marginalization. Due to their need for relative autonomy from their bureaucratic parent organizations, public policy labs often suffer from a lack of wider legitimacy among political-administrative decision-makers, which encumbers the organizational adoption of lab-designed policies (Bason 2018; Fleischer and Carstens 2022).

To contribute to policy innovation, lab managers must therefore engage in boundary work between the lab and the parent organization. Boundary-spanning can be defined as externally oriented gatekeeping that works to acquire and maintain resources for, and the legitimacy of, an organizational unit (in this case the lab) and its results (Fleischer and Carstens 2022). Boundary-spanning practices regulate what transfers between an organization and its environment; in this case, between the lab and the bureaucratic parent organization. Boundary work selects, transmits, and interprets information and affects who gets what information, when, and how. When used strategically, it provides information to specific client groups, specially adapted for them, so that they will feel their interests are being represented (Aldrich and Herker 1977, 218ff). Research has shown how administrative leaders continually engage in boundary-spanning activities in relation to their authorizing environment of political office holders and other top-level decision-makers (upwards); their executive environment of subordinate managers and frontline staff (downwards); and their external environment of stakeholders, societal partners, and competitors (outwards) (van Dorp 2018).

Collaborative leadership research has shown the importance of boundary-spanning for garnering support from a wide range of organizational sponsors and champions on different organizational levels, which in turn is crucial for the adoption and institutionalization of collaborative innovation. As Crosby, Bryson, and Stone (2015, 654) observes, “collaborations that are dependent in some important way on a public bureaucracy need one or more consistent sponsors and champions who are embedded near the top of the public bureaucracy.” Sponsors are decision-makers who use their formal position and authority to provide organizational resources for effectuating collaborative solutions, while champions mainly deploy their informal authority and process skills to build support for the collaborative endeavors (Crosby, Bryson, and Stone 2006, 2015). Policy lab research has further nourished the hypothesis that effective boundary-spanning ensures that policy labs do not “fall out of favor with government” and “lose buy-in” from vital sponsors and champions, which may otherwise result in lab closure (Brock, 2021, 237).

Accordingly, lab managers must continually foster interest in the lab, its work processes, and its results to enhance policy innovation uptake. This requires a managerial focus on not only internal relationships and interactions within the lab, but also its relationship to the surrounding organizational environment(s). Agranoff and McGuire (2001) identify four management behaviors that aid multi-actor policy formation and implementation, which provide possible avenues for lab managers to develop productive relationships to sponsor and champions in the public bureaucracy. These include: a) identifying and activating persons with relevant skills, knowledge, information, expertise, money and other pertinent resources; b) framing their perceptions by creating and celebrating a shared purpose or vision that align their engagement; c) inducing them to make – and keep – a commitment to the joint undertaking by “selling the idea” and forging agreements on its purpose(s), objectives, role and scope; and d) integrating the strategies of participants and external decision-makers by facilitating their interaction, promoting information exchange, and mediating their communication. As a supplement to these primarily person-oriented management behaviors, lab managers may also forge institutional linkages between the labs and formal decision-making arenas at various levels of the organization (cf. Krogh 2022). Institutionalizing the connections between the policy labs and their parent organization through structural amalgamation make them more robust toward changes in leadership as original champions and sponsors move on to other positions (cf. Crosby, Bryson, and Stone 2006).

3. The Fremfærd policy labs

In 2017, the Danish public sector development fund Fremfærd (“Moving Forward”) launched a two-year policy lab programme that provided funding and consultancy to municipal public policy labs aimed at cutting bureaucratic red tape at the local government level (samskabtstyring.dk). During the course of the lab programme, all Danish municipalities could apply for funding to establish policy labs that would contribute to public policy innovation in municipal administrations and public institutions in public service areas such as eldercare, childcare, homecare, primary schools, and social services for citizens with special needs. Once funding was obtained, the Fremfærd consultancy provided guidance on lab design and management, supporting the municipal lab managers in involving relevant public administrators, managers, and frontline employees and facilitating policy co-design processes in the labs. By May 2018, eight local governments had obtained Fremfærd funding for public policy labs addressing problems of bureaucratic red tape in various public service areas.

To utilize the labs to develop research-based knowledge on lab management, Fremfærd decided to fund a two-year empirical research project, upon which the author of this article embarked in 2018. It was agreed that the research was independent in terms of research questions, theories, methods, analysis, results, and conclusions, and that the researcher should disseminate the results in scientific journals and in oral presentations targeting the municipal lab managers and participants as well as the Fremfærd consultancy. This article reports on the results of the empirical research conducted on three Fremfærd policy labs that had completed their lab processes in late 2018. The focus on boundary-spanning strategies and practices emerged as a salient research topic in the early stages of the research process. The three labs were thus selected as cases for research before the decision to focus on boundary-spanning, whereby their specific practices and outcomes were unknown before the research was conducted.

Due to their embeddedness in the same lab framework, the three Fremfærd policy labs are largely similarly designed. However, an explicit discretion for locally adapted lab designs gave way to some variation. In all three labs, the lab managers were involved in co-designing the labs, albeit other municipal actors such as the city council and the directors of services had considerable leverage over important aspects of the lab design, including what to address as the main lab purpose and whom to include as lab participants. From the perspective of the lab managers, the lab design was thus a relatively stable and only partly modifiable structure, which their boundary-spanning practices had to work within and work around. Table 1 provides an overview of the design of the three labs: their specific lab purpose, public service areas, number of lab days and period of duration, policy innovation phases, and involved lab participants. For context, the table also displays the population size, the size of municipal staff, and the predominant socio-geographical setting of the municipalities in which the labs took place.

Table 1. The design and context of the three Fremfærd policy labs.

	LAB A	LAB B	LAB C
Purpose	Simplify administrative policies and procedures	Improve performance documentation policies and procedures	Improve performance management policies
Public service areas	Primary schools, childcare, and homecare	Childcare and homecare	Social services for citizens with special needs
Lab days and duration	Two lab days within a 6-week period	Five lab days within a 10-month period	Nine lab days within a 3-month period
Policy innovation phases	Problem identification and idea development	Problem identification and idea development	Idea development
Lab participants	The city manager, director of schools, director of eldercare services, and managers and employees from two schools, one childcare institution, and three homecare and nursery teams	One homecare team and one childcare institution	The director of social services and managers and employees of all municipal caseworker teams and homecare teams
Lab managers	Municipal development consultant	Municipal development consultants	Municipal development consultant
Size of municipal population	Approx. 50000 inhabitants	Approx. 60000 inhabitants	Approx. 20000 inhabitants
Size of municipal staff	Approx. 5000 employees	Approx. 6000 employees	Approx. 1500 employees
Predominant socio-geographical setting	Suburban	Rural	Rural

4. Research methods

The aim of the qualitative study of the three Fremfærd policy labs is to understand how the lab managers span boundaries between the labs and their bureaucratic parent organizations and to analyze how these boundary-spanning practices affect policy innovation uptake in and through their capacity of building political-administrative support for lab-designed policies. Investigating largely similarly designed labs in the context of Danish municipalities, the lab study does not set out to analyze the influence of different contextual conditions for boundary work, but rather to bring diverse boundary-spanning practices to light and analyze their significance for the innovation uptake of lab-designed policies.

The study triangulates written material and qualitative research interviews gathered and conducted shortly after the lab processes had been completed. The written materials include project descriptions, policy briefs, PowerPoint presentations from the lab sessions and meetings in which the labs were discussed, internal evaluation reports, and related documents from the lab process. As a whole, the material provides rich information on lab design, process management, lab results, and boundary-spanning practices and activities. The interview data include 17 semi-structured, in-person, qualitative interviews (four individual interviews, six double interviews, and seven group interviews) with a total of 38 key actors on multiple levels of the municipal organization, ranging from the city manager to frontline employees as well as the policy lab managers themselves. The distribution of the interviewees in each case reflects the empirical distribution of actors involved in the respective labs. Table 2 provides an overview of the interviewees.

Table 2. Interviewees in the three cases.

	LAB A	LAB B	LAB C	TOTAL
Lab managers	1	2	1	4
City manager	1	0	0	1
Directors	2	0	1	3
Managers	1	2	1	4
Employees	6	10	10	26
TOTAL	11	14	13	38

Bold Signifies that it is the sum total.

Due to access constraints, no politicians were interviewed, which limits the direct access to information on feelings of ownership to the labs and their results among the political decision-makers. The assessment of the political support for the lab and the lab-designed policies thus relies on the statements made by the interviewed actors as well as the written material documenting political decisions on policy adoption.

The individual interviews lasted about one hour and the group interviews roughly one and a half hours. The interviews followed a common interview guide focusing on the (perceived) purposes and outcomes of the labs as well as the role of the lab managers in facilitating and supporting the lab processes. To create a space for sharing both positive and negative lab experiences, the interviewees were promised full anonymity, including a clause that subsequent publication based on the interview data would involve no direct quotation. However, the group interviews that included both the local manager and employees could not guarantee the anonymity of the employees vis-à-vis their manager. In these interviews, the employees generally expressed their negative experiences in more positive terms, for instance by stressing how future lab processes might be designed and managed to produce better outcomes, which required greater sensitivity to the interview setting in the subsequent interpretation of the interview data (Hollander, 2004).

Inspired by the Gioia methodology (Gioia, Corley, and Hamilton 2013), the data material was inductively coded in three steps. First, the various empirical expressions of boundary-spanning practices were captured in numerous first-order codes. Second, patterns in the first-order codes were detected and aggregated into analytical second-order themes. Third, the themes were conceptualized into three general boundary-spanning strategies. The subsequent section defines the three boundary-spanning strategies and analyses the boundary-spanning practices associated with each strategy, focusing on practices in cases where it was done well while also examining their drawbacks and limitations.

5. Three boundary-spanning strategies for enhancing innovation uptake

The managers of the three labs span boundaries between the labs and their bureaucratic parent organizations in different ways and to various degrees. However, they all engage in boundary work that can be analytically tied to three boundary-spanning strategies, captured by the concepts of agenda linking, arena linking, and actor linking. Agenda linking denotes a boundary-spanning strategy of connecting the lab purpose, goals, processes, and results to the political-administrative agendas, strategies, goals, and priorities of the bureaucratic parent organization. Arena linking refers to a boundary-spanning strategy of connecting lab processes and activities to decision-making arenas at various levels of the parent organization. Actor linking refers to the strategy of connecting lab initiation, execution, and evaluation to influential decision-makers at various levels of the parent organization. Figure 1 displays a graphic representation of the three boundary-spanning strategies.

Figure 1. The three boundary-spanning strategies.

The following sections analyze the empirical manifestations of the strategies and their associated boundary-spanning practices.

5.1. The agenda linking strategy

In their initial framing of the lab—and throughout their continual facilitation of the lab—the lab managers played a key role in keeping the goals, processes, and results of the labs in line with the given political-administrative lab mandate and the strategic priorities of the bureaucratic parent organization. In Lab C, for instance, the city council mandated the policy lab to develop new, more economically viable performance management policies for the provision of services to citizens with special needs that would a) increase the economic accountability of the homecare teams and their managers; b) decrease the number of performance targets and enhance professional discretion; c) enhance the homecare workers’ performance orientation; and d) streamline the documentation of results. It was also linked to a broader political-administrative agenda of advancing trust-based management, co-creation, and active citizenship in the municipality. The lab managers structured the lab around the politically mandated requirements and repeatedly evoked them in the process. Ultimately, the city council found that the lab-generated policies met the requirements and accepted them tout court for implementation in the coming year.

In reporting the lab results, the lab managers also facilitated the policy innovation uptake of lab-generated solutions through their framing and specific wording of the results. In Lab B, for example, the policy lab was launched with the purpose of aligning the performance documentation of local institutions with the municipality’s so-called “maxims of desirable documentation.” These maxims defined “desirable documentation” as relevant and important, meaningful, effective, interdisciplinary, and collaborative documentation with a clear and communicated purpose and value for citizens. In their internal evaluation report, the lab managers stressed how most of the new solutions aligned with the maxims. As they themselves saw it, they did so because alignment with the maxims was pivotal for the political-administrative decision on whether to terminate, continue, or expand the lab strategy to involve other public institutions. They thus actively linked the lab and its solutions to the municipal agenda of “desirable documentation” through the reporting of results in order to mobilize the support of central decision-makers.

While the analysis shows how boundary work connecting the lab purpose, goals, processes, and results to political-administrative agendas, strategies, goals, and priorities paved the way for political-administrative support for the lab results—as well as for the lab strategy as such—it also shows how excessive agenda linking can compromise lab autonomy, eroding ownership among decentral lab participants and resulting in retrenching the transformative capacity of the labs. In Lab C, for example, the efforts of the lab managers to align the lab goals, processes, and results with the political-administrative lab mandate and the strategic priorities of the parent organization resulted in a relatively strong steering and tightly structured lab process, which made some of the participants question the legitimacy of the lab and its results. They expressed experiences of the lab process as a “deceptive show” without actual employee influence and participation, limiting their ownership of the lab results.

Moreover, the analysis reveals an inherent limitation of agenda linking, stemming from the volatility of organizational agendas. In Lab A, the city council named 2018 the Year of Simplification to signal the strategic priority of solving problems related to bureaucratic red tape. Echoing the municipal agenda, the policy lab set out to do away with “administrative bother” (“bøvl”) in public primary schools and childcare institutions. According to the lab managers and lab participants, the link between the lab purpose and the municipal simplification agenda lend support and legitimacy to the lab, bolstering interest in its outcomes among participants and non-participants at both the local and central levels. Implicitly putting a time cap on the simplification agenda, however, the gesture of naming a specific year the Year of Simplification had the backlash of setting an expiration date on the long-term strategic relevance of the lab and its results, making managers and employees question whether the politicians and administrative executives would have any interest in the lab-generated solutions beyond the relatively limited one-year timeframe. The initial agenda linking thus had the immediate advantage of garnering attention and interest in the lab but did little to support the continued relevance of the lab work beyond the initial phases of the policy innovation process.

5.2. The arena linking strategy

In all three labs, the lab managers connected the lab processes and activities to the so-called co-committees (‘MED-udvalg’), a three-tier institutionalized structure for management–employee cooperation in all Danish municipalities, including local co-committees, area co-committees, and a main co-committee in each municipality. In the co-committees, managers and shop stewards deliberate matters related to working conditions, working environment, and management–employee relations.

The lab managers connected the labs to the committees in various ways with different implications. In Lab B, the lab managers used the main co-committee as a platform for lab initiation and summative evaluation but did not connect the lab and the committee during the lab process. This hands-off approach granted the lab significant autonomy in the process while limiting the committee ownership of the lab results. According to the lab managers, the co-committee showed interest in the lab as a space for manager–employee collaboration but did not play an active role in supporting the implementation and dissemination of results. In Lab C, on the other hand, the lab managers linked the lab to the area co-committee, using it actively as a platform for lab preparation and the continuous monitoring of lab activities and results. It contributed toward significant ownership and support for the lab and its results from formal and informal leaders in the municipal organization but also fueled the experience of some of the participants that the lab process was heavily steered and monitored by formal leaders in the bureaucratic organization.

In addition to the committee-based linking, the lab managers also linked the labs to staff meetings, team meetings, and other arenas for local deliberation. In Lab A, for example, every school staff meeting had lab progression on the agenda during the lab process. After the formal termination of the lab, they continued to discuss emergent problems and their possible solutions, suggesting that the boundary-spanning practice facilitated a transfer of lab-related methods into the daily organizational structure. In Lab B, the childcare institution also decided to discuss lab activities in their staff and team meetings, thereby involving all employees in the lab process. The lab work therefore quickly became an integral part of the professional conversations in the childcare institution. The homecare team, on the other hand, did not connect the lab work to any local arenas, contributing to a local lab isolation. Lab participants thus reported that they had no lab-related conversations with nonparticipating colleagues during the lab process. This indicates how local-level arena linking matters for the broader mobilization of interest in lab results, fertilizing the ground for policy innovation uptake.

While arena linking broadens the ownership of lab results and stimulates interest and support at multiple levels of the public bureaucracy, the study results also show how a lack of available organizational arenas challenges the boundary work of lab managers in ways that call for institutional entrepreneurism. In settings where appropriate decision-making structures do not already exist, arena linking requires design and effectuation of new project-specific arenas in the public bureaucracy. Such arenas provide organizational support, albeit in a more temporary and less sustainable manner. In Lab A, for instance, the homecare teams lacked a suitable organizational arena for discussing lab related issues with a broader circle of organizational stakeholder. To provide organizational support for the lab activities during the lab process, they created a new forum that involved all municipal home care employees. A project-specific add-on to the existing structure, however, it perished by the end of the lab process. The lack of workable arenas already embedded in the public bureaucracy made it inherently more difficult for the lab managers to ensure enduring anchorage of lab-related activities and results in the home care services.

5.3. The actor linking strategy

The analysis shows how the involvement of high ranking decision-makers fuels policy innovation process and subsequent innovation uptake. In Lab A, for instance, the highest ranking civil servant in the municipality, the city manager, backed the initiation of the policy lab and participated directly in the lab alongside subordinate managers and employees. The formal lab endorsement and time investment of the administrative leadership motivated the local managers and employees. According to themselves, it raised their commitment and time investment in lab activities. Signaling high strategic priority at the municipal level, it also bestowed the lab with wider organizational legitimacy and increased the chances of getting policy proposals passed at higher political-administrative levels. According to the lab participants, it contributed toward the fact that the City Council accepted and adopted their lab-generated proposal of terminating mandatory municipal tests in the public schools.

The managers of all three labs involved the second-highest ranking officials, the municipal directors, in various ways, ranging from deep to shallow involvement, which produced different outcomes. In Lab C, the director of social services was deeply involved in initiating, executing and evaluating the lab. In-between lab days, the director summarized the provisional lab results in brief reports that she continually presented to both the lab participants and the executive management in the municipality, making her instrumental for ensuring the central actor linking of the lab and solidifying its strong agenda linking at the higher levels of the public bureaucracy. In the end, the lab-designed policies were all adopted and implemented in the entire organization, indicating the importance of the connective boundary work conducted by the director. According to several lab participants, however, the deep involvement of the director granted her inexpedient leverage over the lab process and its results. In the end, some of them had a hard time recognizing the solutions presented in the lab reports, nourishing a feeling of having a limited say in the process and hampering their ownership of lab-generated solutions. This finding shows how actor linking can become too strong, ultimately harming the decentral commitment to the developed policies.

Conversely, in the other two labs, the involvement of the municipal directors was more superficial, giving way to other problems. In Lab A, the director of schools and director of eldercare services participated in the lab processes, but in a severely curtailed manner. Prompted by the city manager, the lab managers had asked them to remain quiet and simply listen to the local managers and employees during the lab days. Consequently, the directors did not become actively involved in the policy design process, limiting their ownership to the derived solutions. In the end, most of the lab-designed policies did not scale or disseminate to other parts of the municipality, which at least in part can be attributed to the low level of actor linking among the directors.

None of the municipal directors participated directly in Lab B. As an add-on to the lab process, however, the lab managers interviewed several public managers, including the director of childcare services, in order to include their views and experiences of policy problems in the municipality. Only engaging them in problem identification and not in policy design, however, this boundary-spanning practice only procured a relatively weak link to the formal bureaucratic leadership. By the end of the lab process, the lab-generated solutions largely remained local in their reach, enjoying limited ownership and support among key decision-makers in the public administration.

6. Conclusions

The study shows how public policy lab managers mobilize political-administrative support for policy innovation uptake through boundary-spanning strategies and practices of agenda linking, arena linking, and actor linking, which remedies the intrinsic bureaucratic detachment problem of public policy labs by connecting lab purpose, processes, and results to bureaucratic agendas, arenas, and actors at multiple levels of the public bureaucracy. The study findings point to a number of specific actions that lab managers may take to successfully link policy labs to different parts of the public bureaucracy; actions that may be taken separately, but are most likely to be effective when applied in combination.

Firstly, in their initial lab framing, continual lab facilitation, and final reporting of the lab results, lab managers may enhance policy innovation uptake by aligning the goals, processes, and results of the labs with the political-administrative lab mandate and the strategic priorities of the bureaucratic parent organization. Evoking relevant strategic concepts and frameworks that enjoy organizational legitimacy, lab managers may frame the lab and its results in ways that mobilize the interests of central decision-makers. Since organizational agendas are volatile, however, agenda linking is not simply a one-off activity in the initial lab framing, but a continuous endeavor throughout the lab process and beyond. Moreover, lab managers should be wary of excessive agenda linking since too strongly steered and tightly structured lab process can compromise lab autonomy and erode ownership among lab participants, curtailing the transformative capacity of the labs.

Secondly, lab managers facilitate broader ownership of the lab at multiple levels of the public bureaucracy when connecting the lab to bureaucratic decision-making arenas during lab preparation, lab initiation, lab implementation, and the dissemination of lab results. Putting lab progression on the agenda of staff meetings, team meetings, and other arenas for local deliberation, lab managers may broaden employee involvement and aid the transfer of lab-related methods and results into the organizational structure. In settings where appropriate decision-making structures do not already exist, lab managers should consider designing new project-specific arenas though such project-based arenas are inherently more volatile and less supportive of innovation uptake than their established and bureaucratically ingrained counterparts.

Finally, lab managers fuel the policy innovation process and the subsequent innovation uptake when persuading high-ranking decision-makers to endorse the lab and participate in its initiation. Subsequent involvement of lower-level decision-makers in the execution and evaluation of the lab may also bolster the adoption and implementation of lab-designed policies. However, lab managers should be careful not to grant individual decision-makers too much decision-making power in the lab process since it may narrow the ownership of lab-generated solutions and hamper the decentral commitment to the implementation of the developed policies, ultimately harming the prospect of innovation uptake.

Previous research has shown how public policy labs often rely on individual sponsorship from high-ranking officials, making high leadership turnover in the public sector one of the greatest risks for the sustained use of policy labs in public organizations (Tõnurist, Kattel, and Lember 2017, 20). By the same token, policy labs often lack structural linkages to their parent organization (Fuller and Lochard 2016, 17). In order to successfully amass broad ownership and support, lab managers must strive to combine actor linking with agenda linking and arena linking to ensure strong institutional anchorage at all levels of the public bureaucracy. At the same time, however, the study results also show how lab managers must avoid over-integration of the lab in the hierarchical structures of the public bureaucracy if they are to uphold the integrity of the policy lab as a creative space for experimental policy design. Akin to the art of meta-governing collaborative innovation in relatively autonomous governance networks (Sørensen and Torfing 2017), lab-related boundary-spanning is a delicate act of balancing between too tight and too loose a connection between the policy lab and its bureaucratic parent organization. For example, focusing excessively and exclusively on the agendas and expectations of political and administrative executives brings the lab closer to traditional downstream policy-making, which curtails the principles of co-design embedded in the policy labs. By the same token, lab managers must thus continuously weigh and negotiate the potential tradeoffs between the various boundary-spanning strategies to construct a balanced mix of practices, for instance by combining strategic agenda linking and executive actor linking with more open-ended and adaptive arena linking.

Theoretically, the study has shown the potency of integrating insights from the broader collaborative governance, management and leadership literatures, including established frameworks on collaborative sponsors and champions (Crosby, Bryson, and Stone, 2006, 2015) and conductive managerial behaviors in networked governance arrangements (Agranoff and McGuire, 2001), in the study of managerial strategies and practices for enhancing the uptake of lab-induced policy innovation. Conversely, the study findings also contribute to these literatures by providing empirically grounded conceptualizations and understandings that help answer pertinent questions raised in the mounting body of research on government/quasi-government forms of collaborative governance. First, the results show how public managers may enhance innovation uptake not only through person-oriented boundary work directed at influential decision-makers with formal and informal authority, but also by forging institutional linkages to salient agendas and arenas in the public bureaucracy. Second, the three strategies of agenda linking, arena linking, and actor linking add to existing literature by offering succinct concepts for capturing how public managers may bridge opposing institutional logics of government and governance to overcome tensions between authority and autonomy that otherwise tend to cripple collaborative innovation in such settings (cf. Krogh, 2022; Ambrose, Siddiki, and Brady, 2022). Third, the strategies provide nuanced and empirically grounded answers to the general call for grasping various forms of managerial ambidexterity in cross-sector collaborations that cater to the co-existing needs of innovation and efficiency in public organizations (Page et al., 2021). Collaborative governance, management and leadership scholars may further contemplate how the theoretical contributions and practical implications of the study findings extent beyond policy lab theory and practice.

Geolocation information

Latitude: 55° 56′ 22.83” N. Longitude: 9° 30′ 56.11” E.

Disclosure statement

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

Additional information

Funding

This work was supported by Fremfærd Denmark.