What went right? A collaborative process to prepare a city forest management strategy

ABSTRACT We analyze a multi-stakeholder process that succeeded in creating a joint forest management strategy for the city of Jyväskylä, Finland. The analysis draws on the participants’ own account of the process, elicited via interviews and questionnaires. We attend to critical context and process factors to account for the success of the collaborative process and evaluate the effectiveness of the agreement in terms of ecological and social outcomes. The process created a practical agreement, which increased the share of protected forests and introduced new biodiversity protection measures for commercial and recreational forests. It also created innovative solutions, like the new concept of a nature value forest, which helped the parties to negotiate around their differences. However, disagreement over the impacts of forest management practices, especially continuous cover forestry, remained. The crucial contextual conditions contributing to the agreement were strong initiating leadership and political mandate, which motivated the participants to engage in collaborative dialogue and stick with the process. The key process factors were a third-party facilitation and joint fact-finding. Most importantly, the process was not constrained by a pre-defined agenda or assumptions related to the status quo, but the participants were granted considerable influence over decisions and outputs.


Introduction
An ongoing trend in environmental management is a shift from primarily state-initiated, regulatory strategies toward collaborative governance, which relies on dialogue and cooperative relations between governmental bodies, non-governmental organizations, and private interests (Margerum & Robinson, 2016).Ansell and Gash (2008, p. 544) have defined collaborative governance as 'a governing arrangement where one or more public agencies directly engage non-state stakeholders in a collective decision-making process that is formal, consensusoriented, and deliberative and that aims to make or implement public policy or manage public programs or assets.'It is an umbrella term that covers diverse inclusive approaches, including deliberative governance (Healey et al., 2003), consensus-building (Innes, 2004), and communicative planning practices (Forester, 1999).
The proponents of collaborative governance maintain that it can minimize destructive conflicts by providing an orderly forum for interest articulation and create innovative and efficient solutions for public policy problems (Ansell & Gash, 2008;Innes & Booher, 2003).It is anticipated to build trust and social capital and allow the participants to pool their knowledge and resources in an effective manner (Emerson & Nabatchi, 2015;Healey et al., 2003).However, collaboration is not a panacea, and cooperative processes can go wrong in several ways.While token and half-hearted participatory processes are doomed to fail, serious efforts can also come to nothing (see, e.g.Booth & Halseth, 2011;Walker & Hurley, 2004).As Forester (2006) points out, participatory processes can be messy, unpredictable, and uncertain, and they are threatened by inequalities of power, income, and information.Some critics are concerned about the ability of collaborative processes to include weaker groups and insurgent voices (Dryzek, 2000;Hillier, 2003), while others caution that collaboration can produce watered-down compromises that jeopardize ecological sustainability goals (Koontz & Thomas, 2006;Singleton, 2000).Given the precariousness of collaborative processes, one of the most intractable questions that concerns researchers is to determine whether and how collaborative processes produce improved outcomes across ecological, economic, and social domains (Lindgren et al., 2021).
In this paper, we analyze a multi-stakeholder process to prepare a forest strategy for the city of Jyväskylä, Finland.The cooperation group succeeded in creating a joint forest program that all participants approved.However, the agreement among a group of contentious stakeholders, representing forest sectors actors, environmental organizations, businesses, and residential as well as recreational interest groups, was at no point guaranteed; the process had several critical moments when it was about to fall apart.We zoom in to these critical instances and explore the forest program process from the inside, through the participants' voices (see Forester, 1999), to understand why and how the collaborative effort was sustained.We also attend to critical context and process factors to account for the success of the process and evaluate the effectiveness of the agreement in terms of ecological and social outcomes, drawing on recent advances in collaborative governance evaluation studies (Ansell & Gash, 2008;Emerson & Nabatchi, 2015;Innes, 2004;Reed et al., 2018).Our narrative in-depth account of the forest program process aims to add nuance to our understanding of how collaborative processes can be carried out for the better or the worse.We also provide new empirical evidence of the outcomes of collaborative environmental governance, suggesting that well-conducted collaborative processes can indeed create innovative solutions and lead to improved environmental outcomes.Furthermore, we illustrate the challenges of engaging ordinary citizens in multi-stakeholder dialogue processes and propose ways to include diverse citizen interests in collaborative environmental governance practices.

Outcomes and conditions for successful collaboration
Participatory multi-stakeholder processes have become increasingly common in addressing complex and contentious environmental management problems (Lindgren et al., 2021;Margerum & Robinson, 2016).They can include collaborative governance regimes, in which multi-party interaction represents the prevailing pattern of activity (Emerson & Nabatchi, 2015), but they can also be one-time collaborative efforts set up to prepare a joint strategy and/or solve a public policy dispute (Innes & Booher, 2003).
The success of the collaborative processes can be evaluated in terms of their outcome: Did they create organizational or other benefits for the participants; and did they meet the convenors' expectations of the process?And did they contribute to socially just and environmentally sustainable outcomes that were regarded as beneficial and legitimate by political decision-makers and the public at large (Conley & Moote, 2003;Emerson & Nabatchi, 2015)?According to Innes and Booher (2003), a good collaborative process should meet the following outcome criteria: The process ended a stalemate and produced a high-quality agreement in terms of meeting the main interests of all stakeholders.It built new relationships and social capital, which assist the participants' interactions in the future.Perhaps most importantly, the process produced novel and innovative ideas for action and contributed to learning and reflection on the participants' initial beliefs, positions, and interests.
Several authors have presented contextual conditions and process factors that affect the outcomes of collaborative processes and determine their success (Ansell & Gash, 2008;Innes, 2004;Reed et al., 2018).Contextual factors, such as interdependencies and power dynamics between the participants and their background organizations, generate the energy and impetus to begin collaboration (Ansell & Gash, 2008).Incentives to participate are low when stakeholders can achieve their goals unilaterally.Effective processes also need initiating leadership as well as a mandate from policymakers and executives who are committed to respect and honor the outcomes of collaborative processes (ibid).Collaboration is particularly suited for situations that require ongoing cooperation, because they create a motivation to develop rapport and social capital among the participants (Ansell & Gash, 2008).Overall, collaboration is more likely to succeed in open and transparent political systems, in which decision-makers are accountable to their constituencies and need to justify their decisions publicly (Reed et al., 2018).
The process criteria for successful collaboration include balanced engagement of all relevant actors who are affected by or care about the issue; clear ground rules; a self-organizing process, which permits the status quo and all assumptions to be questioned; information that is accessible and fully shared among participants; a face-to-face dialogue, where all are heard and respected; equal opportunities to influence process outcomes; and consensus-oriented decision-making (Ansell & Gash, 2008;Innes, 2004;Reed et al., 2018).Some authors emphasize consensus-based decision-making (Innes, 2004;Susskind et al., 1999), while others grant that consensus is usually unattainable and suggest consensus or compromise-oriented decision-making (Ansell & Gash, 2008).In the presence of significant power and/or resource imbalances between stakeholders, effective collaborative governance requires the empowerment of weaker or disadvantaged stakeholders.Collaboration does not alter the power balance outside the table, but a collaborative process can level the playing field at the table (Innes, 2004).Furthermore, a prehistory of antagonism among stakeholders necessitates positive steps and sufficient time for remedial trust building (Ansell & Gash, 2008).It usually requires the contribution of a trusted third party, such as a neutral facilitator who can address the relationship issues and help the parties to respectfully speak and listen to one another (Ansell & Gash, 2008;Innes, 2004).Finally, effective collaboration requires that the participants are committed to the process and feel ownership of it (Ansell & Gash, 2008).
The outcome criteria, as well as the factors influencing the success of collaborative processes, are presented in Figure 1, which depicts the theoretical framework of our study.

Methodology
Our research approach is a single case study analysis (Eisenhardt & Graebner, 2007) using a combination of neutral, third-party evaluation and participatory evaluation (Conley & Moote, 2003).One of the authors observed the forest program process as part of her MSc thesis while assisting the facilitators.The participant observation approach allowed us to detect group dynamics that the interviews alone could not have captured.The case is a critical case in a sense that it is, in many ways, a best-practice example of collaborative governance.As Innes (2004) points out, the real potential of collaboration can only be evaluated through cases in which the conditions for authentic dialogue are satisfied.
The data consists of a situation assessment report prepared by the facilitators in the beginning of the process, facilitators' memos (n = 18) of the cooperation group meetings, responses to an online questionnaire by the participants (n = 8) right after the process, and in-depth interviews (n = 9) with the most active or otherwise key participants, who were willing to be interviewed.Three of the 12-person cooperation group either declined the interview or could not be reached.The interview guide and questionnaire are included in the Appendix.The interviews were carried out in August 2018, except one additional interview in April 2020.We also interviewed the senior facilitator of the process in March 2019.The interviews lasted from one to two hours, and they were taped and transcribed.The quotes from the interviews in the text are translated from Finnish and lightly edited for the purposes of readability and anonymity of the interviewees.The interviewees who could be identified were asked to read the manuscript and give their consent to use of the quotes.
The data is analyzed in an interpretative and narrative fashion (Miles & Huberman, 1994) to describe the process as the participants experienced it, paying specific attention to the critical moments of the process that either jeopardized or saved collaborative dialogue.The analysis is informed by the analytical categories derived from the theoretical framework (Figure 1).

The starting point
The city of Jyväskylä is a medium-sized city in central Finland with around 140,000 inhabitants.As a major forest owner, with around 9000 hectares of forest land (Metsäohjelma, 2018), the city faces the challenge of balancing a traditional orientation toward timber production with the newly arising demands for multifunctional forestry, including biodiversity and recreational values.Ecologists have called attention to the role of intensive forestry in decreasing the habitats of endangered species, especially old-growth forest, and pointed out the importance of decayed wood, which is vital for one-quarter of all forest species in Finland (Mönkkönen, 1999).The forest sector has responded to these demands by developing sustainable forestry guidelines and forest certification standards, which ensure protection of the most important habitats and a certain number of retention trees in logging sites.Furthermore, continuous cover forestry was included as an appropriate forest management practice in the amended Forest Act in 2014.Yet, the forestry sector still largely rejects the idea of continuous cover forestry and maintains that it is less productive than periodic forestry with clear cuts (Mäntyranta, 2018).
These arguments were at play also in Jyväskylä, where the environmental actors advocated for continuous cover forestry and demanded that the city should follow the United Nations Biodiversity Agreement and protect 17% of its forest land by 2020.According to the situation assessment preceding the collaborative process, some environmental non-governmental organizations (ENGOs) even proposed that the city should set an example and completely renounce targets for return of assets from publicly owned forest.The calls for giving up clear cuts were echoed by residential organizations, which have frequently protested individual logging sites near residential areas or popular outing areas.The forest sector emphasized the role of good forest management in ensuring safe and pleasant forests for recreational users, and in ensuring economically viable forestry in commercial forests.In their view, the citizens are best served if old hazardous trees are removed, and the forest are kept relatively open and easy to access.They also maintained that adopting new, untested forest management practices is risky and should not be adopted uncritically on a large scale.
To address the conflicting demands, the Jyväskylä Urban Planning and City Infrastructure Committee (UPCIC) commissioned a situation assessment to see whether there is a scope for and interest in, among the key stakeholder groups, participation in a collaborative process and preparation of a joint forest strategy for the city.The initiative came from the city land use planning agency, which was keen to try out new approaches to engage citizens and stakeholder groups, and also to facilitate city interagency collaboration between forest management and environmental authorities.The timeline of the process in presented in Figure 2.

Mapping the baseline situation and designing the process
In the situation assessment stage, a facilitation company interviewed 24 stakeholder representatives to map out the key interests and concerns related to city forest management as well as their willingness to engage in a collaborative process.The assessment showed that there is indeed the scope for cooperation, and the UPCIC decided to set up a cooperation group with a task to prepare a multifunctional forest strategy for the cityowned forest.Given the history of antagonism between the stakeholders, the UPCIC hired the facilitator company to coordinate and run the forest program process.
All interviewees were invited to a kick-off meeting in January 2017.The main task of this meeting was to select 15 members to the cooperation group.A 15-member group was considered small enough to allow indepth dialogue but large enough to be representative of the different perspectives, covering forest management and timber production, biodiversity protection, urban planning, recreational use of forests, and residential interests.The selection of the group members was consensual, and all interested parties could join the group.The group size was later reduced to 12, because some people changed jobs or withdrew for other reasons.
During the design phase, the group formulated a working plan, including schedule, ground rules, and the goal for the process.At this point, the UPCIC promised to ratify the forest program proposed by the cooperation group, assuming that it was unanimous.

Joint fact-finding phase
The joint fact-finding phase included information sharing and presentations between the group members, an expert seminar, field visits, and a map-based questionnaire to the city residents on their forest management preferences.
A one-day open seminar was organized in August 2017 to address the open questions identified by the cooperation group.The presentations covered landscape ecological planning, insect damages, continuouscover silviculture, health and well-being benefits from forests, as well as timber provisioning a.Most interviewees regarded the seminar as helpful, but some forestry actors complained that conservation biology questions received too much attention compared to themes, which they regarded as 'actual' forestry considerations.
The field visits were organized jointly by the forestry and environmental actors.The sites selected by the former included examples of good forest management practices or failed experiments, like one area where small-scale loggings had a negative impact on timber growth, while the sites selected by the latter were those in which forestry operations had negative impacts on nature.The field visits were mentioned by several interviewees as a critical juncture, as they provided the participants with a concrete idea of the impacts of forest management practices on the ground.As one environmental actor said, 'The field visit was probably one of the most important events in this work, and it served as a reference point in the later stages.'The view was echoed by a forestry actor, 'It was a real eye-opener, as many participants had not seen what the different forest management practices proposed during the process actually meant in practice.' The aim of the map-based questionnaire was to find out public preferences for forest use.According to the results, citizens regarded landscape values and recreation as the most important goals of the city forest management, and they viewed biodiversity protection more important than revenue from timber sales.Furthermore, protected areas were the most popular outdoor activity areas.The interviewed group members found that the results provided important input to the process.However, some forestry actors felt that the economic benefits from forestry might have been underestimated because the questionnaire contained a somewhat technical term 'revenue from timber sales.'They maintained that the results might have been different if the income from timber sales was formulated in more concrete terms, like more money for the city to organize health services.However, the forestry actors also acknowledged that the key message was that citizens value forests.

Negotiation stage
At the negotiation stage, the cooperation group agreed on numerical targets for protected forests and decided on the designation of forest areas for different purposes.This stage was described by several participants as the toughest one.The negotiations were aided with consensus-based decision-making, in which the participants could signal with 'traffic lights' if they approved a proposal, were not entirely comfortable with it, or rejected it outright.The approach was viewed as a relatively quick and simple way to measure the level of agreement in the room.It also helped the participants to understand better how their ideas were received by the group: It was really informative, the more yellows other people raised, the more people had to think that there must be something wrong with this.And it also made you think that ok, they have given me yellow a few times, so at some point, I have to [propose something that also they can approve].
However, by this point, some participants were tired with the lengthy process, while others were frustrated that there was still no concrete output in sight: 'There was a period, around half a year before the finishing line, when everybody was quite strained and we wondered if anything will come out of this.' Some participants were also increasingly uncomfortable with the confrontational tone of the discussions: 'This was very quarrelsome and harrowing, not at all constructive, but people were deep in their trenches and kept shooting at their opponents.'One participant admitted that they used 'quite strong' language to make their case.However, they would have also liked the other participants to say things directly: 'It would be very important for this kind of process that people would speak their minds and tell what they really think instead of mincing words and trying not to offend anyone.' A critical juncture during this phase was a meeting which was facilitated by a junior member of the facilitation team and organized as a plenary without break-out groups due to the attendance of fewer participants than usual.This meeting was described by several participants as chaotic or terrible, and it eroded relationships built in the earlier stages of the process.A further source of trouble was that the composition of the facilitation team varied during the long process, and at one point, some participants felt that a facilitator team member had an environmental bias.One participant recounted the experience of another group member: 'They felt that the facilitator opposed their views, "You cannot think that way", and they were really insulted and said that they were not coming to the meetings anymore.' To get the process back on track, the composition of the facilitation team was altered.The team also met all participants in person to give them a proper hearing and elicit suggestions to improve the atmosphere.Furthermore, they organized 'an appreciation round' in which each participant said something positive about the others and their contribution to the process.This was described by one participant as a complete success: 'The comments were really good, and some people were almost moved to tears.'This was a particularly important step for the city forest sector representatives who had constantly felt that their work was questioned and criticized during the process.These interventions helped to improve the relationships, as indicated by the comment below: 'The consultant [facilitator] was working hard in the background so that we reached a situation in which we could talk to each other like decent human beings, respecting and appreciating one another.It was a huge effort.'

Writing up the program
After the outlines were agreed on, the last step was to put the strategy in writing.This stage also had some critical moments, as the main writing task fell to the city authorities, who were used to working with consultants who deliver reports, not facilitators who only run the process.The authorities found the writing task burdensome and were particularly offended when their work was unduly criticized for purposefully omitting some issues: 'Suddenly we had to pull everything together and start almost from scratch.And then we were blamed for not presenting correctly what had been agreed on in the meetings.It did not feel very nice.' In a similar way, a group member who also contributed to writing the programavailable for the whole group in a shared drivewas upset, as their contribution was accused of distorting the strategy and adding details that were not agreed on.In their words, I was very busy at work […] so I wrote the text during the night, and [then some people claimed] that I had tried to smuggle in my views in the darkness of the night.[…] It was at that point when [I thought to give up and] say that ok, keep your program.
The process could have shipwrecked at the finishing line if the facilitators had not intervened and put an end to the destructive email exchange.They reminded the participants about the ground rules and urged them to appreciate the input of those members who provided a significant contribution to the joint document.

The agreement
The cooperation group met altogether 19 times over two years and eventually reached an agreement on the forest program and its implementation plan.The UPCIC approved the program with no changes in June 2018.The feedback from the public hearing of the strategy in May 2018 was very positive.
The new program increased the percentage of protected forest surface area from 13.5 to 17 and the percentage of recreational forests from 36 to 44.Furthermore, the implementation plan specifies several measures to maintain biodiversity in all forests, including 20 retention trees per hectare, which is twice the amount required by the Forest Stewardship Council certification system.Other planned means to increase the amount of decayed wood in the forest are, for example, to retain all dead fallen or standing trunks, and making artificial rotten standing trees in all logging sites.Furthermore, the diversity of tree species will be boosted by increasing the share of deciduous trees in all forests, in the long term, by 5%.In economic terms, the program amounts to around 300,000-euro losses from timber sales for the city.
The implementation plan details the forest management goals and measures in different types of forests.The timber production forests will be managed for economic profit and even-aged management and clear cuts are the main forest management methods.The recreational forest, including forests nearby residential areas, will be managed in a way that gives priority to maintaining landscapes and easy access to forests, as well as ensuring the safety of people in the forests.The aim is to create forests with a diverse structure using continuous cover forestry as well as small-scale clear cuts.Clear cuts can be carried out in old spruce forests to create forests with a layered canopy.Decayed trees will be cleared from the vicinity of hiking trails.The protected forests will be managed for biodiversity protection and left to age without any forest management activities.Only hazardous trees will be taken down, but they are left in the forest.
All participants approved the forest program as indicated by the following comments: It was a heavy process, but the outcome is good, we are all pleased with it.
Well, it did not entail anything that is not impossible to implement […], one just must look at things from a little different perspective now.I fully stand by it.
I think it is quite good.It has several elements that an ambitious city forest management plan should have […], although the level of ambition was lower than I would have wanted.
According to an independent audit, released in February 2023, the implementation of the forest program was proceeding according to the plans in an agreed upon schedule.The only critical comment was that there is a need to better document the amount and location of the retention tree groups.

The outcome of the process
The forest program process was successful with respect to the key outcome criteria: reciprocal agreement, relationships, creativity, and learning (Figure 1).In terms of agreement, the process ended a long-standing stalemate and created a jointly approved strategy which accommodated the main interests and concerns of the disputing parties.The environmental actors were particularly pleased with the 17% protection level, which is in line with the United Nations biodiversity targets.One clear advancement from their perspective is that continuous cover forestry is included as a main forest management method in recreational forests.Furthermore, being involved in selecting the nature value areas is a far more proactive strategy than protesting individual logging sites one by one.The city forestry actors were content with the fact that the agreement allows them to carry out standard forest management activities in commercial forests and light safety-improving operations also in recreational forests.Importantly, the agreement provides them an undisturbed operating space with fewer protests and complaints, causing costs and delays.The city land use planning authorities, who initiated the process, felt that the process succeeded well in its aim to balance the different interests and concerns related to the city forests management.Also, the UPCIC approved the forest program, indicating external legitimacy of the process outcome.To be sure, the participants were not entirely happy with the outcome.The environmental actors had to relinquish the ambitious goal of protecting all city-owned forests, while the forestry actors pointed out that the losses from timber revenue for the city were considerable.
The participants nevertheless felt that the cost and benefits associated with the program were distributed equitably enough across the beneficiaries and that it accommodated ecological, social, and economic goals in an even-handed manner.
The process also performed well in terms of building new relationships and social capital among the parties.The antagonisms ran high, especially at the beginning of the process, but eventually the parties succeeded in building sufficient trust and rapport which helped them to come to an agreement.The participants also learned what the issues meant to the others (see Innes & Booher, 2003).For example, some forestry actors noted that they now better understand the importance of species diversity for some stakeholders, 'it really matters to them.'One forestry actor even joined a local birdwatching organization 'to better understand how they see issues.'The ENGO representatives, too, felt that the process developed working relationships that will assist their interactions with the forestry actors in the future.The city land use agency observed contently that the process improved communication between the city agencies and provided a model for collaborative dialogue to be used in future land use planning processes.
Importantly, the process created novel and innovative solutions to break the impasse.To meet the 17% protection target, the participants developed a new forest category 'nature value forest' under an existing class 'value forest,' which previously covered forests with significant landscape value.This solution prevented the parties from getting locked in a debate over statutory protection, a typical 'hot button' in forest management conflicts.The nature value forests are effectively protected, as no loggings are permitted in them, yet they are not defined as protection areas.Another example of innovative solutions was the decision to use map-based questionnaires to scan out the citizens' preferences for forest management practices in the future, in regular intervals.
Finally, the process instigated learning and reflection among the parties.The parties did not come to an agreement on the most fundamental question concerning the feasibility of different forest management strategies (clear-cuttings vs. continuous-cover forestry), but they reconsidered their views on citizens' preferences for management of city-owned forests.Forest sector actors acknowledged that the map-based questionnaire showed quite unequivocally that people value forests and do not like clear cuts, while some environmental actors admitted that it was a surprise for them that some people actually support landscape loggings.As one ENGO representative put it, 'I understand now that people also want trees to be cut down, to see a lake or something.It has been a difficult thing for me to understand, because I find forests as such a positive thing.'The participants also learned how to work with each other effectively.As one participant noted, 'The process taught us [in the city administration] to be more attentive to diverse views.'

Factors contributing to the agreement
Contextual factors, like interdependencies and power dynamics, played an important role in motivating the participants to engage with and stick to the process.Initially, the power distribution was asymmetric, as there were no alternative means for ENGOs and displeased citizens to influence city forest management, other than demonstrations and public objection of planned forestry operations.However, the city land use planning and forest management authorities were keen to reduce the conflicts and increase the legitimacy of city forest management.An important background factor was a heightened political interest in environmental values as a result of the 2017 elected City Council with a Green Party majority.The process also benefitted from strong initiating leadership from the city land use planning agency.Furthermore, the process had a clear mandate from the UPCIC.The need for ongoing cooperation did not motivate the ad-hoc cooperation group to build relationships, but it did influence the city agencies' willingness to engage in constructive interagency cooperation.
The contextual factors created favorable circumstances for the collaborative process, but the forest program process also performed well in terms of the key process criteria for collaborative governance.The process was self-organizing in a sense that the participants could decide the ground rules and the working program.Importantly, the group was not restricted by any predefined logging targets as some other Finnish participatory forest strategy processes (see Saarikoski et al., 2010).As one ENGO representative put it, 'I felt that this process provided a real opportunity to make a difference.Often these [participatory processes] don't do that, the outcome is pretty much predetermined.'Another ENGO representative put it more bluntly: 'This approach was from a different planet in fairness and a real dialogue.' All information generated and collected during the process was fully shared among participants, and the process was based on collaborative dialogue.Not all participants thought that they were heard and respected during all stages of the process, but the corrective moves by the facilitators reinstated more respective dialogue.The joint fact-finding stage was instrumental in sharing information and developing an improved understanding of the basic ecological and forest management issues among the participants.Also, the map-based questionnaire, which was jointly designed by the group, provided important information for the process.These intermediate outcomes provided 'small wins' (see Ansell & Gash, 2008) that sustained and energized the process.
The participants basically had equal opportunities to influence the process outcomes.In practice, it was the partners with the most scientific and professional expertise and skills to write the forest program text who had the biggest influence on the outcome.However, these actors represented both side of the biodiversity vs. forestry debate, and they were trusted by the others to formulate the jointly approved principles and actions in an accurate manner.Furthermore, consensus-oriented decision-making ensured that compromises were reached in situations in which different views could not be reconciled.
Perhaps the single-most important issue accounting for the agreement was the use of neutral third-party facilitators, who helped to design the process, structured the discussions, and strived to build trust and maintain a good working atmosphere.As one participant pointed out, 'We would never have gotten this far on our own.'The importance of good facilitation was most clearly demonstrated by the instances in which it was missing.The process was about to fall apart at the point when some participants felt that a member of the facilitation team was biased, and it took some time and effort from the senior facilitator to re-establish the trust in the process.Also, the stages in which the participants communicated directly with each other via emails were critical, as it was easy to misunderstand each other, and the tone of the conversation quickly turned acrimonious.These critical moments also illustrate the ways in which resentment and hurt feelings, in some instances only a few misplaced words, can get in the way of collaboration and block settlements that are in the participants' best interests.This insight was aptly captured by one interviewee: I learned a lot about forestry.But I also learned how much it depends on people and on each person, how well they get along with each other.That it is not always the facts that count, but also the people and how they talk to each other.
The most important shortcoming of the process was the underrepresentation of residential interests.A residential organization representative dutifully attended the meetings but was largely silent during the discussions.In addition to fortuitous reasons, the place-specific and ad-hoc nature of citizens' protests can make it difficult to find sufficiently engaged representatives for a general residential interestif such thing exists in the first place.The map-based questionnaire partly filled in this gap by bringing in the diversity of residential and recreational user perspectives.

Discussion and conclusions
The Jyväskylä forest agreement demonstrates the potential of collaborative multi-stakeholder processes in settling persistent environmental disputes.The process did not result in a consensus on the best ways to manage city forests, but it created a practical agreement (Laws et al., 2014), which balanced ecological and recreational goals with traditional forest management goals and practices.Retaining the negotiation element in collaborative processes is important, as it makes space for plurality of views and allows cooperation and conflict to coexist in a fruitful way (van den Hove, 2006).The productive role of conflict as a catalyst of social transformation is emphasized by several scholars.Some view societal struggles as openings for new possibilities for collective action (Chambers et al., 2022) while others draw on the notion of agonistic pluralism (Hillier, 2003) and maintain that nonconfrontational practices engender a retreat from radical thinking and innovative environmental solutions (Poncelet, 2001).In the Jyväskylä case, conflict was a driving force in motivating the parties to collaborate but sustaining it simply for the sake of maintaining critical environmental debate would not have served the environmental activists' interests.For them, challenge and resistance are not an end in itself but ways to achieve concrete improvements in environmental protection on the ground, measured in terms of hectares and number of species and habitats.In these terms, the Jyväskylä agreement was quite successful.The results indicating positive environmental outcomes from collaboration align with previous studies that have found a generally positive effect of participation on the environmental standard of governance outputs (Jager et al., 2020;Newig & Fritsch, 2009).The effects were particularly positive in caseslike in Jyväskylä casewhere participants were granted considerable influence over decisions and outputs.The reluctance of vested interest to devolve power to new partners and to allow the questioning of the dominant forest management paradigm (see Ollonqvist, 2002) has been the primary reason why the previous participatory natural resource management efforts have failed to deliver ecologically and socially sustainable outcomes in Finland (Kangas et al., 2010;Raitio, 2012;Saarikoski et al., 2010).
The findings from this case study also illustrate the capacity of collaborative processes to develop creative ideas and innovative solutions, and they provide some support for the joint learning hypothesis (see Innes & Booher, 2003).The concept of nature value forest created during the process served as a boundary object (Baggio et al., 2015), which was sufficiently fluid with respect to the protection status of nature value forests but concrete enough to specify acceptable management practices in such forests.Another innovative solution was the decision to adopt the map-based questionnaire tool as part of the city forest management practices.Adaptive forest management, in which policies and actions are adjusted based on the combination of new scientific and socioeconomic information, is considered essential in supporting reflexive and resilient natural resource management (e.g.McIntyre & Schultz, 2020).
The study confirms the importance of several process-related criteria for the success of collaborative processes.The most crucial factor was the participants' commitment and collaborative capacity to work through their differences, even in the critical moments when the process was about to unravel.Our analysis shows how collaboration can easily get derailed by ineffectual communication even if it is in the parties' best interests to reach an agreement.Another key factor was good facilitation, which enabled genuine dialogue and encouraged the parties to stick with the process despite the difficult moments.The importance of facilitation is found in several empirical studies on collaborative natural resource management (e.g.Lindgren et al., 2021;McIntyre & Schultz, 2020).Conversely, the lack of neutral facilitation has been a major reason for several unsuccessful attempts at collaborative forest management (Raitio, 2012;Vihma & Torikka, 2021).Developing a shared knowledge base on forest management and ecology, and especially citizens' preferences, was also critical.The positive impact of information sharing and joint fact finding is observed also in other studies on collaborative natural resource management (Innes & Booher, 2003;Lindgren et al., 2021).
The contextual factors are equally important in accounting for the process outcome.The forest agency was unlikely to alter its existing practices, ingrained in the dominant Finnish forest management paradigm without the political pressure from the Green-Party-dominated City Council to revisit city forest management strategies and priorities.Additionally, the economic stakes were relatively low compared to conflicts concerning large areas of commercially managed forests (see Zachrisson & Lindahl, 2013).The city forests are a source of revenue, but they are mainly managed for the benefit of citizens.Hence, the most important debates revolved around public interests as well as the epistemic authority concerning good forest management practices.The prospects of finding collaborative outcomes are far more challenging in situations with high economic interests of individual forest owners and/or forest industry (Saarikoski et al., 2010) and national-level pressure to generate forestry revenue to the state (Raitio, 2012).Furthermore, municipal government is likely to be more responsive to citizen concerns than more distant and opaquer national-level policy-making processes.Therefore, the Jyväskylä forest cooperation group was, in many ways, ideally positioned to result in positive environmental outcomes.
Our study affirmed the bearing of process related criteria (see Figure 1) for reaching a practical agreement among conflicting parties.However, it also showed that collaborative processes designed for stakeholder interaction do not necessarily succeed in accommodating the views of the public at large.The cooperation group members made incorrect assumptions of the public's forest management preferences, reflecting their own views instead of those of the average Jyväskylä residents.Furthermore, our experiences show that it is very difficult if not impossible to ensure adequate representation of diverse citizen interests in collaborative groups which are small enough to allow meaningful interaction between the participants.The solution to capture the city residents' perspectives with a map-based survey tool was very effective because it unequivocally showed what people really think and what kind of forest management activities they prefer.Map-based surveys are particularly helpful in providing spatially specific information, often highly relevant in natural resource management situations.Collaborative governance literatures tend to shun one-directional participatory methods which do not allow dialogue and interaction.However, a skillful combination of deliberative and survey-based methods can provide a multifaceted understanding of diverse views, and also address the concerns that stakeholder processes may fail on the grounds of inclusiveness and democratic representativeness (Leach, 2006).
Implementing the forest program and introducing continuous cover forestry to recreational forests can serve as a niche experiment (Kivimaa et al., 2017) that can instigate changes of forest management practices in other local governments, and possibly commercial forests, as well.Also, the social and intellectual capital developed in the process can contribute to novel thinking about city forest management priorities and methods in the future.Single case studies like ours tend to focus on the immediate outcomes and cannot capture the longstanding impacts.Therefore, further research is needed on the long-term effects of collaboration in instigating social learning and promoting sustainability transitions.
Emma Luoma is a doctoral student at the Univerisity of Eastern Finland.Her research interests are in environmental conflict resolution and collaborative capacity building.
Dr. Sanne Bor is postdoctoral researcher at LUT Business School.Her research focuses on inter-organisational collaboration in the face of big societal challenges.
Dr. Pia Polsa is Associate Professor of marketing at Hanken School of Economics.Her current research interests are consumer vulnerability, poverty, and multi-dimensional value in cross-sector settings.

Figure 1 .
Figure 1.The outcome criteria, as well as the factors influencing the success of collaborative processes.

Figure 2 .
Figure 2. Timeline of the forest program process.