The politics of making Finland an experimenting nation

ABSTRACT

As atop-level national policy agenda, the Experimental Finland initiative (2015-2019) opens up an opportunity to investigate the politics of the nationally promoted experimental turn. We examine how the state launched agovernmental-level project aiming for the whole nation to become experimental. The Experimental Finland initiative was Prime Minister Juha Sipilä’s governmental Spearhead Programmes 2015-2019. This paper explores the underlying politics revealed by the implementation of Experimental Finland by asking:1) what kind of evidence concerning the barriers and obstacles of experimental culture was included and excluded in the national evaluation of the Experimental Finland programme, and2) how did the understanding of those barriers differ when compared to local level experiments? We argue that as experiments are expected to facilitate learning, they instead cause ambiguities in the organizational routines and imbalances in the existing power relations between different actors and actor groups.

KEYWORDS:

• Experimentation
• experimental governance
• policy analysis
• experimental Finland

Introduction

The whole experimental culture and agile projects are just a surface. They are not at the deep core of this organization. What I have learned when working at this organization [City of Helsinki] is how a single pilot or an experiment turns into a real everyday practice. And the answer is: not easily. The system is very rigid and inflexible. It is not enough that someone is interested in experiments; it does not mean that it will turn into practice. (Senior public servant, 20 October 2016)

The OECD has highlighted Finland as an innovative example of implementing experiments and experimental culture on a national level. The 2015–2019 government of Finland had the ambitious aim of turning the country into a benchmark case of an experimenting nation. This goal raised interest worldwide and was regarded as an opportunity for Finland to become a leader in experimental policy making (Mannevuo 2019). However, the national ambitions and international hype on experimenting met the harsh local-level reality, as the quotation above illustrates. In this article, we concentrate on exploring the sources and contested explanations for this contradiction. By doing so, we aim to increase the understanding of the obstacles faced when implementing experimental policies in practice.

Finland is not the only country devoted to experimentation. Innovative societal experiments have recently become more and more popular on a global scale (Li and Van Ryzin 2017; Bulkeley, Broto, and Edwards 2015; Kivimaa et al. 2017; Scoones and Stirling 2020; Hildén, Jordan, and Huitema 2018; Sørensen, Mattsson, and Sundbo 2010). Related policies are being actively promoted by prominent international organizations, such as the World Bank, the OECD, and the EU (Ylöstalo 2019). For example, the European Union has encouraged experimentation through its research programs, such as different living lab schemes (Kronsell and Mukhtar-Landgren 2018; Liedtke et al. 2012). Experimenting as a new and emerging form of public policy has involved altering the roles and responsibilities of service production, strengthening the role of private-sector actors, and profit-making (Powells and Blake 2016). On the other hand, experimentation has also been justified by the urgent need to engage novel actors in innovative solutions to mitigate global challenges, such as climate change, and to boost broader socio-technical change toward sustainability (Bulkeley, Broto, and Edwards 2015). The Experimental Finland initiative opens up an interesting opportunity to investigate how the different ambitions and goals of experimental governance become contested in the politics of the nationally promoted experimental turn.

The Experimental Finland program was launched by the Finnish government, and thus it executed national politics directly. The initiative was part of the broader vision of the Center-Conservative government to dismantle the welfare state by diminishing the role of public authorities. The Finnish government of 2015–2019 started to operate in a context wherein the previous government (2011–2014) had included parties such as the Social Democrats and the Left Alliance, who both wanted to protect the ideals of the welfare state (Hallitusohjelma 2011). As the winners of the 2015 elections, the Center-Conservative government wanted to cut restrictive legislation and create easier opportunities for business-life actors to operate in Finland. The government’s strategic program wanted not only to secure a new social contract, but to continue welfare reform that included an intensification of and experiments with workfare (Adkins et al. 2019).

As part of this program, the 2015–2019 Finnish government promoted innovative policymaking and introduced a culture of experimentation. This included the implementation of an experimentation program consisting of extensive trials (including the basic income experiment) and several smaller experiments (Antikainen et al. 2019; Ylöstalo 2019; Adkins and Ylöstalo 2018). The policy design of the culture of experimentation in Finland leaned on the global trend of experimental ‘nudging’ techniques (Ylöstalo 2019) and was a systematic attempt to give rise to a new, experimentalist governance.

Discussion on experimentalist governance arose at the beginning of the 2010s (James, Jilke, and Van Ryzin 2017; Overdevest, Bleicher, and Gross 2010). Experimentalist governance is understood as an observed or proscribed commitment to a process of trying something new, evaluating the results, and revising practices based on what has been learned (e.g. Laakso, Berg, and Annala 2017; Overdevest and Zeitlin 2014; De Búrca, Keohane, and Sabel 2014). The idea may be interpreted as a profound move toward a new way of thinking. However, learning from experiments is not a trouble-free process. For example, according to a comprehensive literature review, Kivimaa et al. (2017) claim that the connection between experiments and aimed institutional change remains unclear, thus questioning the power of governing by experimenting. The fragility of experiments (Romjin et al. 2010), combined with the path-dependent nature of institutions (Unruh 2000;), suggests that experiments seldom lead to major changes of governance.

Despite the critics, there exists a strong discourse from the proponents of the experimental turn, who trust experimentation will produce innovations that will improve social services and advance nationwide goals. This is based on the assumption that the feedback from the experiments will provide evidence and learning from best practices to local and national organizations (Schram and Silverman 2012, 132). In addition, the need to engage a broad range of actors for societal renewal through experimenting has been raised within the recent sustainability transitions debate (Powells and Blake 2016). Learning and innovating through experimenting has been seen as a key driver for the socio-technical changes needed to tackle global sustainability challenges (Van de Walle 2017). The Experimental Finland program – launched as part of the conservative coalition governmental program – intertwined the two topical issues of experimenting and evidence-based policy and integrated them with the ethos of societal renewal. In this thinking, evidence-based policy was supported by an active, experimental ethos, constantly looking for new and better ways of doing things (Ylöstalo 2019; Mannevuo 2019).

However, while implementing evidence-based policies, a number of questions arise as to how knowledge is recognized and mobilized in the policy process (Standring 2017). Recent research has emphasized that the evidence-based policy discourse is inherently naive about the realities of public policy (Wesselink, Colebatch, and Pearce 2014; Cairney 2016). The ‘evidence’ itself in policy processes can be highly ambiguous, and the different preferences policy actors hold for particular sorts of evidence, or evidence from particular disciplines, help to underpin the clashes among the competing narratives on the issue (Wesselink, Colebatch, and Pearce 2014, 341):

The assumptions here seem to be that the only relevant factor is the inherent quality of evidence that the evidence “speaks for itself” and that “good evidence” will be a conclusive guide to action. However, evidence cannot speak: it has to be introduced in some way by the participants as part of their framing of the policy problem. Much of the discussion about the use of evidence in policy making revolves around the interplay of quality, context and discourse. (Wesselink, Colebatch, and Pearce 2014, 341)

For example, the law-like idea that evidence of success in one context will mean success in another has been questioned in policy studies several times (Cairney 2016, 39; Cartwright and Hardie 2012).

In this paper, we explore the controversies of the evidence-based policy approach in justifying experimental governance during the implementation of Experimental Finland by asking the following questions:

• 1) What kind of evidence concerning the obstacles of experimental culture was included and excluded in the national evaluation of the Experimental Finland program?

  2) How did the understanding of those obstacles differ when compared to local-level experiments?

We were interested in how problems were framed differently on the national and local levels, and particularly in what these differences can teach us about the politics of implementing experimental governance. Following from this, our analysis focuses on the creation and interpretation of evidence as part of the practices of evaluating the national policy program and implementing local-level experiments. This practice-oriented approach (Laws & Hajer 2006; Wagenaar & Cook 2003) enables us to bring up the tensions and ambiguities that arise when evidence is simultaneously created and evaluated as part of implementing experimental governance. We claim that defining the results of the national program, that is, the ‘evidence,’ is part of the policy process, and heavily dependent on the context.

2. The practice-oriented approach to evidence-based policy

We have chosen a practice-based approach for analyzing the politics of the Experimental Finland program (Freeman, Griggs, and Boaz 2011; Laws & Hajer 2006; Wagenaar & Cook 2003). The concept of practice allows grasping the ‘units’ at which experimentation takes place, where we can locate the flexibility and robustness of experimenting action (Laws & Hajer 2006, 421). What we aim at is a fuller picture of policy making as an act. The practice-based approach emphasizes bottom-up translations of policy initiatives by local actors, providing alternative ways of explaining diversity or conflicting outcomes. Thus, we want to enhance our understanding of the roles of local civil servants, professionals, and citizens, and to recognize the multiple and overlapping contexts within which the policy is enacted (Freeman, Griggs, and Boaz 2011, 130–131).

According to Lipsky (1980) it is the decisions of street-level bureaucrats, the routines they establish, and the artifacts they use to cope with uncertainties that become the public policies they carry out. As we see it, analyzing the local actors and their daily encounters sheds light on the problems of both the implementation and goals of the Experimental Finland program.

Practices generate artifacts, which in turn structure practices. The Experimental Finland program is an artifact of practice within the field of public administration, the product of a certain amount of administrative processing in support of various policy aims. We agree with Cook and Wagenaar (2012), who argue that practice itself is not passive but active, and that among its active traits is that it gives shape to knowledge and context. The contents of knowledge and context are accepted, sustained, and modified or rejected through practice.

When we think of Finnish city administrations where most experiments are conducted, we must acknowledge that not only the context and institutional structures play a role. Analytical practices must address the constraining effects of history. Actors are locked inside a perspective – that of their own personality and experience, that of their group, and that of their milieu. Although these perspectives allow the actor to act, they also constrain him or her in interactions with others (Cook and Wagenaar 2012).

The viewpoints of situationality and context dependency are important when analyzing the implementation of a policy. In each new situation or circumstance, a policy is repeated but with a new interpretation. In this way, practice is also always about improvisation as well (Freeman, Griggs, and Boaz 2011, 129). The purpose of a policy – such as spreading experimental culture in Finland – is to shape and order practice, and evidence is one way of doing so (Freeman, Griggs, and Boaz 2011). Evidence may be derived from practice or designed to inform it.

Evidence-based policy has been in some approaches interpreted as a direct tool to help public policies to meet their objectives (Newman 2016). However, this interpretation has often been criticized and questioned (Standring 2017; Lennon 2014; Wesselink, Colebatch, and Pearce 2014). In implementing evidence-based policy, the aim is to look for neutral, context-free, and universally applicable evidence. Expecting a linear relationship between evidence and policy is problematic, as there are multiple forms of knowledge at play, the processes are multi-voiced, and the mobilization of evidence is part of an interactive process (Wesselink, Colebatch, and Pearce 2014). Evidence-based policy may seek to aim for neutrality in the knowledge produced, but it is as personal, contextual, and institutionally intertwined a practice as any other form of knowledge gathering.

Our aim is to enrich the understanding of the context dependency of evidence creation when analyzing the Experimental Finland program. This is done with a thick empirical analysis of the actors’ interpretations of the obstacles of experimental culture at two different levels of policy implementation, i.e. the national and local. Instead of offering a clear, coherent story of the development and evaluation of the Experimental Finland program, we offer a more polyphonic picture of the practical world where civil servants and citizens are trying to implement the actual experiments. Practice implies that these people, while experimenting, are always positioned as parts of larger networks of relations, conventions, and obligations (Wagenaar & Cook, 2003). This makes the comparison of perspectives between the national- and local-level actors important. Focusing only on the narrow community which runs the national program may lead to undervaluing such local-level experiences that have not become recognized as important information in the framework of national-level program evaluators. As we see it, practice-based policy analysis is the tool for illustrating that the people implementing national policies are active game changers and not solely instruments for top-down policies to take place (Freeman, Griggs, and Boaz 2011, 134).

The case of experimental finland: data and methods

The Experimental Finland initiative was launched by the new coalition government, consisting of the Center, Conservative, and nationalist populist parties, after the parliamentary elections in 2015. In the governmental program, Experimental Finland was linked to the goal of increasing the efficiency of the public sector and enhancing entrepreneurship. One key target was to overarchingly promote and establish an experimental culture in the country.

The program identified three levels on which experiments were thought to operate and to have an impact: 1) The strategic level – pilot studies selected by the government, such as pilots for basic income, service initiatives, and local government trials; 2) The pooled pilots and partnerships level – pilot studies that promote the objectives of the government program, executed in the regions, NGOs, and the business environment; and 3) The grass-roots level – civil society initiatives (kokeilevasuomi.fi). One of the most popular and visible strategic-level experiments was the basic income pilot, where 2,000 people each received 560 euros per month for two years. An example of a regional partnership was the digital municipality experiment, which promoted the digitalization of municipal services. The objective of the smaller, grass-roots level experiments was to generate innovative solutions for improving services and creating new jobs. The total budget for the Experimental Finland program was 3 million euros (Antikainen et al. 2019, 7).

The data for our analysis are threefold. First, we have analyzed the material provided by Experimental Finland on its website, including the national evaluation report of the program and the background literature provided online. The second data set consists of an analysis of the program’s seven evaluation events, entitled ‘Morning coffee’ (March to September 2018, duration of each 90 minutes) and the final seminar of the program (13 November 2018). These data were video recorded by the organizers, and we transcribed and thematically categorized them. The third data set includes 22 interviews with actors from the local-level experiments in the cities of Helsinki and Tampere. The interviews were conducted between September 2016 and May 2017, and their duration varied from 1 to 2.5 hours. The interviews were first transcribed and then thematically categorized. The local-level experiments were case studies that we followed closely during an Academy of Finland-funded research project (2016–2019). For this paper, we selected from the interview data the themes that dealt with the problems in implementing experiments and the tensions the interviewees had recognized during the experimental action.

The interviews with people who conducted local-level experiments – i.e. the ‘street level actor’ perspective – were important, as most of the experiments of Experimental Finland program were conducted in cities, and the governmental program particularly mentioned municipalities, grass-roots organizations, and citizens as key actors to be empowered through experimental culture (Prime Minister’s Office 2015).

The ideological underpinnings and inspiration of the program was introduced in the list of publications provided on the program’s website for its justification and motivation (kokeilevasuomi.fi). These publications were divided into two categories. First, there were papers concentrating on transition management theory (Schot and Geels 2008; Bos & Brown 2012;), which indicates that in addition to serving the goals of diminishing the role of the public sector and enhancing entrepreneurship, experimental governance was also introduced as a means to invite a broad range of actors to address sustainability challenges. The literature list includes a considerable number of background publications on climate change experiments (Bulkeley, Broto, and Edwards 2015; Heiskanen et al. 2015; Laakso, Berg, and Annala 2017). The second category of literature links experimental governance to the need to diminish the role of the public sector due to fiscal pressures and the need to tackle challenges such as aging, climate change, and migration (e.g. Mulgan 2014). The literature in this category emphasizes the reconceptualization of the role of the state as a facilitator of entrepreneurial activities (Bakhshi, Freeman, and Potts 2011; Thaler and Sunstein 2009). We used the literature list provided by the program as background data for our study.

The data were analyzed with qualitative content analysis (Mayring 2004). As we wanted to explore the controversies of the evidence-based policy approach in the implementation of Experimental Finland, we started our analysis by thematically coding the evidence provided by the national-level evaluation process on the barriers and obstacles of experimental culture. Next, we did similar coding with the interviews of local-level actors and compared the results of these two levels. We were particularly interested in the different interpretations of the nature of the problems of experimenting in Finland. When collating the data, we were able to illustrate how the evidence, the results of the national programme, was heavily dependent on context and framing. In the next section, we will first go through the national perspective on the bottlenecks of experimenting in Finland and then move on to the local perspective.

The bottlenecks of experimental culture in Finland: National evaluation

Along the Experimental Finland program, the potential for upscaling experiments and experimental culture received enthusiasm in the government cabinet and among the civil servants working in the program. The opportunities to expand the impact of the program were discussed in the evaluation events called ‘morning coffee sessions.’ These events also elaborated on the hindering elements for the expansion of experimental culture. Next, we go through the three most discussed obstacles, which we have named: 1) reluctance to take risks, 2) inability to learn and transfer knowledge from experiments, and 3) inability to integrate knowledge produced in experiments into strategic decision-making.

1) Reluctance to take risks

In the national evaluation, one of the key differences between governance practice and experimental culture was claimed to be a willingness to take risks and fail versus risk aversion, playing it safe, and using considerable efforts to plan carefully. The willingness to take risks was seen as some type of mind-set and discussed in terms of attitudes. Johanna Kotipelto, project leader in the final seminar on 13 November 2018, had this to say as her opening remark: ‘This is a journey to something new, no one can say what happens when we dive into murky waters.’

In the evaluation discourse, ‘courage’ was framed to be mainly an individual quality, whereas ‘careful,’ ‘slow,’ and ‘risk averse’ were adjectives attributed to the organizational culture. This interpretation, which was again affected by the underlying idea of the bureaucratic administration, was questioned in the evaluation debates.

About failures and risk taking: I think it often is an individual’s responsibility. It is a challenge, because on the organizational level we cannot manage that. There isn’t a mechanism that would allow the organization to take the risk and carry the responsibility. It all falls on the individual. (Civil servant, 18 September 2018)

The discussions on evaluation events highlighted the individual’s inability to take risks several times. The courage needed to take risks was framed as an individual attribute. This was recognized as an obstacle to experimentation, and in the national discussion the participants pondered how to find ways to make the situation less risky for individuals – or at least that the people considering implementing an experiment would not get stuck on the notion of risk.

However, the discussions did not deal with the reality that risk is embedded in liability and responsibility in the bureaucratic model of local municipalities. The real reason for the unwillingness to take risks was the fact that in any local municipality, officials perform their duties while subject to liability for any acts. For the municipal official, this means criminal liability and liability for damages. Thus, the threshold for experimenting in everyday work may be quite high and acting outside the box is not common sense.

2) Inability to learn and transfer knowledge from experiments

According to the evaluation report, the essence of experimental culture was seen to be in the ability of public organizations and society to learn from experiments (Antikainen et al. 2019). In this context, the claims that there is a widely experienced inability in the field to systematically evaluate experiments, learn from individual experiments, and spread the knowledge learned were regarded as being critical, and they were referred to as basic obstacles for experimental culture. In addition to the evaluation report, the problem of losing knowledge gained from experiments was also emphasized in the evaluation events. In all of the eight sessions analyzed, the issues of learning and transferring the results from the experiments to other contexts were regarded as important. In the panel discussion of the final seminar of the Experimental Finland program, learning from experiments was still an unresolved problem. One reason for this was that often the people conducting the experiments were not the same as those who had set the goals of transferring the knowledge from one project to another. As one of the panelists argued, there were no clear receivers of the knowledge produced in the experiments

If we notice at the end of an experiment that all the knowledge is falling into emptiness, it means that the users of the knowledge weren’t participating at the beginning of the project. (Person working for National Research Institute, 13 November 2018)

The evaluators and civil servants working for the Experimental Finland program addressed the problem in two different ways. Firstly, it was seen as a result of a lacking communication channel devoted to cross-experiment learning and knowledge distribution. As a remedy for this shortfall, the Prime Minister’s Office and several projects established a virtual platform to assemble and distribute knowledge from the separate platforms. The role of the virtual platform was introduced in one of the morning coffee events by a project researcher

We are currently building this website, and I hope to receive input from you [the audience] on what kind of knowledge there needs to be for the website to be a useful platform. (Researcher giving a briefing for a project, 7 June 2018)

From the practice-oriented policy analysis viewpoint this solution was inadequate. The discussion focused on developing the virtual platform as a user-friendly environment instead of concentrating on what actually happens when a group of people try to learn from the results achieved in a different context. The conception of how learning takes place and knowledge transfers from one local context to another did not include the viewpoints of the situationality and context dependency of the experiments.

Another interpretation of the claimed inability to learn from experiments was made by a consultant who had previously worked for the Experimental Finland program

I think the problem is that on several levels, on the levels of management, execution, and employers, we have very diverse goals for the same experiment. And what would be the most important is to arrive at an understanding and joint ownership for the whole process from the very beginning to the end of the experiment. (Consultant involved in the Experimental Finland program, 13 November 2018)

According to the evaluation discussions, the Experimental Finland program was struggling with the commitment of people implementing the experiments to advance a broader experimental culture. Following from this, joint ownership and commitment to the national-level policy was not realized on the local level. Thus, knowledge about the ways in which experimenting could advance the various policy and development goals learned from the experiments was not shared. As the local actors’ lack of commitment to promoting the governmental goals of spreading experimental culture was identified as an obstacle, another important issue, the diversity of goals people want to achieve by experimenting and what can be learned from this diversity, was not elaborated on any further in the discussions. The difficulties related to knowledge transfer and learning were also relevant in terms of strategic decision-making that constitutes the third obstacle found from the data.

3) Inability to integrate knowledge produced in experiments into strategic decision-making

The third major obstacle identified was the difficulty in integrating knowledge from experiments into decision-making processes, and particularly into strategic governance. The experiments were claimed to be too random and thus lacking the capability to be more closely intertwined in decision-making arenas. This aspect was raised among others by a civil servant participating in the final seminar of Experimental Finland, who underlined the lacking linkages between experimenting and strategic governance

Well, who is the user of the knowledge? Is it the strategic level that needs to make the changes? This means that the strategic level should participate at the beginning of the experiment and decide or commission already then what is needed to be learned from the experiment. And it is the same strategic level that needs to think where the results of the experiment are to be attached. (Civil servant, 13 November 2018)

During the national-level discussions, the tensions between the top-down-initiated strategic experiments and bottom-up-organized experiments became clearly articulated. As one member of the audience participating in the evaluation events pointed out, the decision-makers should be engaged to guarantee the strategic usefulness of the experiments:

Before we start to make experiments, we must think what the need is for the experiment. Where do we want to have an impact? That way we would have the will to utilize the results from the experiments. Not like it is now, where we have random experiments popping up like mushrooms in the rain and afterwards we are thinking, “who should we actually inform about these experiments?” (Individual citizen from the audience, 18 September 2018)

According to the evaluation discussions, a certain balance between strategic-level steering and bottom-up learning by doing is needed to spread experimental culture. The two-way relationship was expressed in the evaluation report by pointing to the possibility that decision-makers might use experiments as a tool to justify their own interests. This was raised by a participating citizen in the morning coffee event in September 2018

I am interested in how to strengthen the relationship between experiments and decision-making. Currently, it seems that experiments are used as a political tool where the results are not that important but certain people can say that they support experiments. If there are no clear rules, the political sphere just uses experiments. (Individual citizen from audience, 18 September 2019)

The Experimental Finland program itself was a large experiment strongly supported by the Finnish government. The rhetoric promoting national experimental culture was ahead of the actual practice of experimenting. Thus, the problem of delivering relevant knowledge from the experiments to the national level and committing the decision-makers was not solely the fault of the individual experiments. The national-level rhetoric generated enthusiasm to conduct individual experiments in a situation where the national level was not ready to receive and integrate the knowledge into strategic decision-making, as a former employee of the Experimental Finland program explained:

As I was working in this project, there was this pressure coming from the field to do experiments. The problem was that there were national reforms going on and Experimental Finland could only say: “yeah, you just go on and experiment,” but at the same time, the national level went its own way and couldn’t take into consideration the results coming from these experiments. (Civil servant, 18 September 2018)

This is an interesting quote, as most of the problematic issues in the national-level discussion were approached as problems of individual experiments or local-level actors’ inabilities. The actors implementing the Experimental Finland program on the national level failed to understand the challenges from the local-level viewpoint.

Local-level obstacles to experimentation

Next, we will introduce the four main obstacles to practice experimentation at the local level brought up in the interviews with municipal actors. Three obstacles out of four addressed similar types of issues as identified at the national level, but the municipal actors gave very different interpretations for them. The fourth obstacle raised in the local-level interviews was lacking from the national evaluation.

Obstacle 1: No organizational support for risk taking for individual actors

The requirement to take risks by experimenting was high on the agenda in the national evaluation. Looking at this question from a local viewpoint, the issue boiled down to one question: who takes responsibility for risks and possible failures? The local municipalities and other actors working with city officials raised this concern several times. Thus, avoiding possible risks was a real barrier for the experiments at the local level. This was not only a question of liability and the existing rules and norms. It was also the case that the organization’s structure did not allow the crossing of regulatory boundaries or stepping into the mandate of a neighboring office. The lack of clear local-level backup for taking bold, unusual action was a clear hindrance to experimenting, as indicated by a civil servant working in Helsinki

The people planning experiments should have courage, and so should the administration behind the permits; it should be brave enough to adapt. (…) For example, in one of the latest cases, the fire department became paralyzed. They announced that they could not interpret this case any more loosely than the other cases, because if somebody dies, it would be the fire department who was responsible for it. (Civil servant in Helsinki municipality, 20 October 2016)

The quote articulates the contradiction: experiments require risk taking, but taking a risk is against the principles of good public administration. If experiments put the liability of any public official into question in terms of safety or security, it is very reasonable for the civil servants not to take action.

Obstacle 2: Disrupting power positions

The national debate briefly touched on the issue of power positions when an audience member pointed out that ‘we need to see if there is real willingness to change the power relationships when encouraging experiments’ (Evaluation event in May 2018). In the local-level interviews, this issue was encountered several times and seen as a fundamental challenge to any experimentation. For example, a civil servant in Helsinki underlined the difficulty of experiments dismantling the current power positions:

There is also the question of power in presenting issues in the municipal system. In the case of temporary spaces, new ideas need to be represented by building control. But if they presented changes to the current situation, they would give up some of their decision-making power over these buildings. So why would they present anything that takes power away from them? (Civil servant in Helsinki municipality, 20 October 2016)

In several local-level experiments (2016–2019), we witnessed how some parts of the city organization were immune to the goals presented by another department within the same organization. The policies that the cities proposed and supported on a strategic level were later rejected by another actor within the same organization. The ethos of acting leaned on a rationalistic tradition, which, in addition to responsibilities, distributes power. The established power positions efficiently blocked any shifts toward experimental culture.

Obstacle 3: Misfit between experimenting and rationalistic urban governance ideals

As the national-level report and the public discussions focused on the inability in the field to systematically learn from individual experiments and spread knowledge, this problem was also approached on the local level, but from a different, practice-oriented perspective. Instead of focusing on the systematic evaluation of experiments, the local-level actors brought up the general friction between rationalistic urban governance and the frequent difficulty of experimental processes. Many of the local-level experiments that emerge from the bottom up are characteristically ambiguous, involve contradictory interests, and have evolving goal settings. Public actors are usually only one partner among others in these experiments, or they may even be mere bystanders, meaning that the long-term development goals of public actors are not necessarily easy to pursue through this kind of evolving collaboration (Kronsell and Mukhtar-Landgren 2018). This causes problems with the traditionally rationalistic administrative culture of cities based on the idea that the development targets should be clearly specified in a political process before choosing the most efficient implementation strategy.

Our data contain examples where city officials have been alienated from experimenting due to unclear project management protocols and ad hoc decision-making during the implementation. In the interviews, the city officials expressed concerns over poorly organized experiment projects resulting in information gaps, poor budgeting and documentation, and the unclear roles of actors. This was argued, for example, by the representative of a public utility company who had taken part in an experiment in the city of Tampere

I think that this is a textbook example of a project where the directors are enthusiastic about inventing, but then all the normal issues are overlooked: Who is responsible and for what? What is the timetable? What are the proper plans? … Enthusiastic talk at the outset, “Yes, yes, let’s do this,” but then the documentation and allocation of responsibilities are forgotten. … I have learnt the hard way that responsibilities and timetables must be set precisely, and that requires a lot of work. (Representative of public utility company, 2016)

There were cases in which city departments or municipally owned companies had withdrawn from the experiment because of unclear project management. As the energy and resources went into the implementation of the experiment itself, there were usually no resources left afterward to organize the evaluative learning processes or upscale the experiment.

Obstacle 4: Inability to collaborate cross-sectorally in routinized organizational practice

The fourth obstacle raised by local-level actors was not directly addressed in the national evaluation report. The idea of experimenting and learning together easily clashes with the siloed urban administration, where there are clear and separate mandates for different officials and administrative departments. This is not a question of the attitude of individual civil servants. The division of work is stabilized with existing regulations and norms, which assign certain tasks and responsibilities to particular offices. Siloed urban governance is both embedded in the regulatory system and a deep-rooted habit. The everyday practices of municipalities keep up the routinized work, regardless of the hierarchical structure of the municipality. This was illustrated in an interview with one city official who had worked in the neighboring cities of Helsinki and Espoo

The city of Helsinki has high organisational structures and no culture of [the different departments] doing things together. The city of Espoo has a lower organisational hierarchy and still no culture of doing things together. (Civil servant working for Helsinki municipality, 20 October 2017)

As local experiments are put into practice, they simultaneously intervene in the usual proceedings and institutional mandates. The idea of experimenting was several times characterized as a ‘light’ and ‘quick and easy’ way of trying new things in the national-level discussions. When analyzing the local practices, the everyday working environment appeared quite the opposite. Introducing novel collaborations was not always warmly welcomed, as one civil servant devoted to experimenting explained in an interview

Inside the city organization, there have for decades been these strong actors. … And I think it is good to dismantle the strong actors and units who tend to live their own lives. We need to avoid this development, the creation of separate silos within the city. … I have tried to come closer to the colleagues who are working in other areas of the city, and how should I put it? If I am totally honest, the collaborative ideas haven’t been received very well on the other end. (Civil servant working for Tampere municipality, 16 March 2017)

This quotation illustrates how difficult it can be to create novel collaborative ideas and try to put them into practice. Implementing an experiment on the local level seemed to require years of stamina and determination from the local actors. These questions, either clearly outspoken or quietly influential in the background, hinder both the experimenting and the upscaling possibilities of the experiment.

The obstacles at the local level were the real-world challenges of doing experiments, including the important barriers to risk taking, collaboration, and power shifting. In the interviews, the people explicating the problems had distant or no relation to the national-level Experimental Finland program. Their objective was the development of local-level practices, and their viewpoints were bound to situationality and the context of their milieu.

Experimenting nation and the challenge of learning

By juxtaposing the national- and local-level discussions on experimentation in Finland, it becomes obvious that although similar types of issues arise, these debates do not fully meet each other (see Table 1). The perspectives, goals, and aims of experimentation differed between the national program and local-level actors, and due to this only some parts of the local tensions and complex organizational practices preventing the spread of experimental culture were recognized and used as evidence in the program evaluation, as indicated in Table 1. Furthermore, although similar things were identified as obstacles at both levels, the explanations for the causes of these obstacles differed. The national-level debate focused more on intangible cultural issues and attitudes, whereas on the local level, the actors emphasized the importance of concrete practices, regulations, and resources.

Table 1. Obstacles preventing the spread of experimental culture in Finland.

OBSTACLES	IDENTIFIED CAUSES
NATIONAL LEVEL EVALUATION
Reluctance to take risks	Lack of courage Organizational culture discourages risk taking
Inability to learn from experiments	Lack of communication channels Old, bureaucratic administrative culture Lack of commitment at the grass roots level to advance experimental culture
Inability to integrate knowledge to decision-making	Loose connections between experiments and strategic level decision-making Weak strategic steering of experiments
LOCAL LEVEL EXPERIENCES
No organizational support for risk taking for individual actors	Risk taking is against the principle of good governance and questions the reliability of administration Civil servants are subject to liability of their actions
Disrupting power positions	Experimentations questions stabilized power positions
Misfit between experimenting and rationalistic urban governance ideals	Messy and ambiguous nature of experimentation Lack of experience in cities to collaborate with loosely governed grass roots networks All resources on local level go to hands-on managing of experiments, no resources for upscaling
Inability to collaborate cross-sectorally in routinized organizational practice	Civil servants are guided and restricted by the existing regulations, norms and routines

The national-level discussion on disseminating experimenting culture in Finland was led by a small number of people. This critique was also raised in the program evaluation report, where Experimental Finland, according to the interviews, was considered a ‘project for a small circle of actors in the capital of Finland’ (Antikainen et al. 2019, 23). The report briefly acknowledged previous public policy pilots that had been conducted for decades, well before the Experimental Finland program was launched, but these pilots were not reflected on during the process. Neglecting the tradition of developing the administration through experiments in some municipalities raised frustration among the participants of the evaluation workshops. These responses were not discussed in the report and the information did not have an impact on the interpretation framework, which reflected the underlying assumption of an existing division between the old bureaucratic administrative culture and a new regenerative experimental culture.

Thus, the governmental program focused only on some parts of the existing knowledge and omitted most of the recent history of experimenting in municipalities. This indicates that the policy goal of the governmental program to create a novel administrative culture shaped the information collection process and how knowledge was turned into evidence. This finding is in line with the critics of evidence-based policy, who claim that because information is turned into ‘evidence’ as part of the policy process, it is rarely neutral but heavily dependent on the context (Wesselink, Colebatch, and Pearce 2014; Standring 2017). The inability to value the local-level experiences of developing the municipal administration through experimentations led to a situation where important insights were lost. Moreover, it deepened the division between ‘the people from the capital’ leading the governmental program and those doing the hands-on experimenting in the provinces.

In terms of policy learning, the tension was similar to that expressed by Schram and Silverman (2012, 133), as the reports and efforts to publish best practices on the national level functioned as discourse generalization, suggesting that ‘what works there can work here too.’ However, this clashed with local communities that had different needs, populations, and resources. The evaluation report and evidence gathered in Experimental Finland denigrated local characteristics and tried to generalize experimental innovations across them (ibid. 133–134).

Several studies of organizational and institutional changes (Bovey and Hede 2001; Gilley, Dixon, and Gilley 2008; Oreg, Vakola, and Armenakis 2011), have shown that such cultural shifts, as aimed at by Experimental Finland, are likely to cause tensions, reluctance, and uncertainty among the actors. These tensions become clearly pronounced in the interviews of local-level actors, but they were not addressed in the national report or national discussions. One consequence of this negligence was the trivial response the evaluation provided to one of the identified key problems, learning from experiments. The report suggested technical solutions, such as including a knowledge distribution platform. Thus, the national-level discussion did not go into the issue of how learning from individual experiments should take place. As McFadgen and Huitema (2017) have noted, conceptual and empirical work on policy learning and experimentation has been mostly limited to findings indicating that experiments produce new knowledge. As experiments are expected to facilitate learning, little is known about how experiments in fact produce learning, what types of learning, and how they can be designed to enhance learning effects.

A thorough analysis and discussion of the issues at stake in local-level experiments would have highlighted the impossible goal of spreading a national experimentation culture as an abstract phenomenon. Learning and upscaling requires more work than just transmitting one successful experiment to other places. Despite the interest in learning a new administrative culture through experiments, the evaluation of Experimental Finland did not raise the issue of how differently designed experiments could produce different types of policy learning (McFadgen and Huitema 2017). Furthermore, the evaluation discussion did not go into the details of practical accounts of complex, messy experiments and the particularities of given contextual settings (Laws and Forester 2015).

Conclusions: The inherently political nature of experiments

Experimental Finland is a unique case. It can be analyzed as a critical case that has strategic importance in relation to a more general problem (Flyvbjerg 2006, 229). It characterizes itself as creating socially innovative models of governance and aiming to promote and establish an overarching experimental culture. This goal needs to be examined critically. The Experimental Finland program was, as Beck, U (1999, 41) argues, full of unauthorized actors and non-codified practices. These potentially permit socially innovative forms of re-organization and governing, but they also opened up a terrain of institutional ambiguity and potential conflict revolving around institutional power relations (Bäcklund et al. 2018).

We agree with Van de Walle (2017), who sees the problem as lying in the decontextualization of experimentation. Experiments work well when one seeks causal effects in fairly simple and linear situations. The evidence of obstacles presented in the national-level evaluation suffered from decontextualization, which led to the inability to create a meaningful discussion of how to facilitate learning between experiments in different types of institutional contexts and with different, substantial targets and timescales. The data produced in the evaluation workshops were interpreted through the goals of the governmental program assuming a division between an old bureaucratic administration and a new agile experimental mode of governance. The diversity of local-level actors, situated needs and practices, past and present experiences, and tensions between multiple goals and interests related to experimenting was downplayed.

The differing perspectives on the obstacles to experimenting reveal how central actors can trivialize the local obstacles to experimentation and thus provide trivial policy solutions to address those obstacles in productive ways. This observation is important for any large-scale societal transition. Whether it is an experimental culture or other significant transition, the revision is likely to cause not only temporary ambiguities in the organizational routines but also imbalances in the existing power relations between different actors and actor groups. Moreover, from the viewpoint of democracy, experimental culture cannot be delivered as a product to the local level. Rather, it must be created and co-produced with local actors. This is a point that needs to be taken seriously if the national level wants to advance the experimental culture and learning from the experiments on a larger scale.

Acknowledgments

We are grateful to Maarit Särkilahti for releasing part of her research data for our use. We also wish to thank Vanesa Castán Broto for her helpful and insightful comments on earlier drafts of this article, the researchers of the Dwellers in Agile Cities research project, and the researchers of the PONTE group at Tampere University. We are also grateful to the editors and reviewers for their careful reading and thoughtful comments on this article.

Disclosure statement

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

Additional information

Funding

This work was supported by the Academy of Finland Funding Agency under Grant 289691;Academy of Finland [289691,303481];

Notes on contributors

Helena Leino

Dr.Sc. Helena Leino Helena Leino is senior lecturer in Environmental Policy at the faculty of Management and Business in the Tampere University, Finland. She is also adjunct professor at Aalto University, department of Built Environment. She received her PhD in Environmental Policy 2006. Her recent research has focused on practice-oriented policy analysis, sustainable city development as well as participatory knowledge production in urban planning.

Maria Åkerman

Dr.Sc. Maria Åkerman Maria Åkerman is principal scientist at the VTT Technological Research Centre, Foresight and safety culture. She is also affiliated to the University of Eastern Finland as PI of a research project dealing with nutrient recycling and shifting economies of fertilization. She has previously acted as a Professor in natural resources policy at the University of Eastern Finland (UEF). Her current research focuses on food production, urban development, circular economy, bioeconomy and renewable energy.