A view through the lens of policy formulation: the struggle to formulate Swedish moose policy

ABSTRACT

Policy formulation refers to how problems identified in the agenda-setting phase transform into government programs. As the process of designing policy alternatives expresses and allocates power among different interests, policy formulation affects both implementation and outcomes. This paper examines the Swedish moose policy of 2010, revealing that the policy portrays the issue of moose as an ecological problem, while motivations for adopting policy measures are largely described in economic terms. Because of this incongruity, the policy may not achieve its goals. Furthermore, implementation principles stemming from different design strategies clash in the attempt to incorporate both local and ecosystem-based levels of management into a single system, leaving the policy implementation with many uncertainties and tensions. To deliver the policy’s goal, the government could consider clarifying the operationalization of the ecosystem-based management approach and identifying the prerequisites necessary for building capacity, dealing with strong stakeholders, and ensuring fair representation of key actors. Future research should further explore the consequences of policy imbalances in relation to intended goals, the importance of understanding the rationales and design strategies underpinning implementation principles, and the need to discuss operationalization of EBM in relation to different types of objects of management and to issues of scale.

KEYWORDS:

• Policy design
• policy formulation
• moose management
• strategies
• natural resources

Introduction

Policy formulation, or design, is an important stage in the policy cycle (Howlett, 2019; Schneider & Ingram, 1997; Wegrich & Jann, 2007; Weible & Sabatier, 2018). It refers to how identified problems in the agenda-setting phase transform into government programs in terms of styles, patterns, and outcomes (Wegrich & Jann, 2007, pp. 48–51). In contrast to related concepts such as policy reform (Patashnik, 2003) and policy layering (Daugbjerg & Swinbank, 2016), policy formulation does not primarily concern policy- or institutional change, but rather how policy is constructed and the consequences of such constructions (Schneider & Ingram, 1997).

As the process of designing policy alternatives allocates power between social, political, and economic interests (Sidney, 2007, p. 79), policy formulation affects both the implementation phase and the outcomes of policies (Schneider & Ingram, 1997, pp. 2–10). Past research has mainly focused on standard approaches, including policy- or institutional change, evaluation of outcomes of the policy process or policy instruments, and policy mixes (Orach & Schlüter, 2016; Rogge & Reichardt, 2013; Weible & Sabatier, 2018). The policy literature, particularly in relation to natural resources, often focuses on ways of knowing (especially evidence-based implementation), but tends to miss the strategic aspects, or what is politically constructed at the beginning of the policy cycle (Beland Lindahl et al., 2016; Crabb & Leroy, 2012; Orach & Schlüter, 2016; Owens, 2015). However, interest in policy formulation is increasing, as it is the stage in the cycle where some of the most important decisions are taken (Howlett & Mukherjee, 2017). Turnpenny et al. (2015, p. 6) argue that policy formulation may even constitute ‘the very essence’ or the ‘missing link’ needed for policy analysis to explain why policies succeed or fail. This paper makes a contribution to this growing field of research.

We apply a policy formulation lens to a particular case of natural resource conflict: the highly contentious new moose policy in Sweden (introduced in 2010). This policy aims to handle the conflict between strong economic interests such as forest owners, who argue that a large moose population negatively affects forest economic value due to browsing damage, and hunters with strong sociocultural traditions of hunting and wildlife management, who claim that a rich (in numbers, types, and distribution) moose population improves hunting opportunities and contributes to biological diversity and ecosystem services (for example, meat and recreation) (Dressel, 2020; Ljung et al., 2012).

Despite various studies on Swedish moose management focusing on the implementation phase (Bjärstig et al., 2014; Dressel, 2020; Dressel et al., 2018; Lindqvist et al., 2014; Sandström et al., 2013; Sjölander-Lindqvist & Sandström, 2019), the ideas and rationales of the policy have not previously been comprehensively analyzed. Moreover, the hunter and landowner groups could be assumed to share traditional underlying ethics (or values) commonly related to natural resources (Dobson, 2016, p. 21-ff), as many hunters are landowners and vice versa. This case therefore serves as an example of a social conflict over a natural resource where traditional value patterns are not sufficient to understand the conflict. By examining the 2010 Swedish moose policy in terms of how the moose is constructed as an object of government policy and the measures suggested for its governance, we aim to gain insights into the ongoing conflict, highlight potential inconsistencies, and discuss the consequences of this for the implementation of the policy.

Insights from the application of a policy formulation lens to the Swedish moose policy will also add to other policy formulation studies on natural resource conflicts, that share similar characteristics. The Swedish moose policy represents an attempt to move from single-species management to multi- and ecosystem management – a challenge shared with many other cases of natural resource- and wildlife management policies across the world. In a broader sense, we can find similar dynamics in any policy domain struggling to resolve conflicts and implementation problems across several sectors and (groups of) actors. Through our focus on policy formulation, particularly the ways in which objects of policy are socially and politically constructed, we will contribute also to this wider field of research.

Policy formulation – knowledge and power

Theoretical foundations

The key concern in this paper is policy formulation or policy design – the stage following agenda-setting in the public policy process (Weible & Sabatier, 2018), in which problems are transformed into coherent programs. The two terms are used interchangeably here and in much of the literature (Sidney, 2007, p. 79). Policy formulation is not always instrumental and knowledge-driven, but can also be strategic and opinion-driven. To understand the policy and trade-offs in a complex policy context, we need to consider what knowledge informs the process. We also need to recognize that some design contexts are dominated by politics (Howlett & Mukherjee, 2017; Owens, 2015), while others are science- or profession-based (Howlett & Mukherjee, 2014).

The role of science in policy design varies depending on the political risks and opportunities afforded by political leaders, as well as the cohesion between science- and professional perspectives (Schneider & Ingram, 1997, p. 6). Policy designs may be determined by what Howlett (2019) terms instrumental factors of design, and Schneider and Ingram (1997, p. 78-ff) call scientific design: knowledge-based, logical, experience-based, and analytical, with instruments chosen or designed to maximize effectiveness and efficiency based on objective knowledge claims. By contrast, policy designs may be based on so-called ‘non-design strategies’ (Hood, 2010), mainly pursuing self-interest: political calculation, bargaining (Sidney, 2007), accommodations, political gains, and blame-avoidance calculations (Howlett & Mukherjee, 2014), strategic use of power, and manipulation of social constructions (Schneider & Ingram, 1997, p. 78-ff). The empirically relevant question then becomes: What is constructed and manifested in the policy?

This paper starts from a critical perspective, emphasizing how social constructions underpin policy design. Social constructions refer to the shaping of reality (Berger & Luckmann, 1967) and reflect social settings, mental structures, and historical circumstances through formulation. Constructions are intersubjective, and thus subject to contestation; that is, different people may interpret, expect, and understand an object differently. In relation to policymaking, where many people are involved in the process, there is a struggle to find a common construction or understanding for certain problems, groups of people, events, etc. (Schneider & Ingram, 1997, pp. 73–74). Different forms of power and power relations in the policy making process are key contextual characteristics, determining whose views will prevail (Schneider & Ingram, 1997, pp. 75–76).

With respect to potential consequences of social constructions in policy design, much of the traditional policy literature centers on material effects related to the distribution of resources in society and economic costs and benefits among groups (Cairney, 2012). Critical perspectives instead tend to focus on interpretive effects: how a policy design may shape individual understandings, experiences, values and attitudes of democracy, justice, problem-solving capacity, citizenship, or political participation (Dryzek, 1990; Schneider & Ingram, 1997, pp. 79–81; Mettler & SoRelle, 2018, pp. 118–119). Since policy designs are highly context-dependent, it is difficult to predict the consequences of a particular design. However, to illustrate potential consequences, we use the link between design elements and subsequent translation dynamics – that is, the interpretive process by which actors experience a policy design’s positive or negative impact on their interests or actions (Schneider & Ingram, 1997).

Underlying constructions of design elements reveal the formulators’ conceptions of the policy subject, and thus highlights the design (or non-design) strategies employed by decision-makers (Schneider et al., 2014; Schneider & Ingram, 1997). The elements confer benefits and burdens for different actors. By uncovering constructions (based on values and conceptions), we can assess whether, and how, the design elements balance different values. If there is an imbalance, there is a risk of disappointment in the formulator (in this case, the government) (Schneider & Ingram, 1997, pp. 82–84) and weak implementation mechanisms leading to unachieved goals (Beland Lindahl et al., 2016).

Transferring this logic to natural resource conflicts, we argue that effects of policy design primarily do not include citizens in general, but specifically actors who are directly affected by the policy and actors who are involved in implementation, including state officials and private stakeholders. In moose policy, social constructions link to design strategies by the prioritization of certain values, through the use of particular knowledge, and through the ways in which the policy manages to balance stakeholder groups. These constructions could then have effects on implementation in terms of shaping the actors’ understandings of balance or sustainability of the resources (mainly moose and forest), as well as understandings of the actors; that is, the construction of the social conflict (Beland Lindahl et al., 2016). It will therefore subsequently encourage or discourage certain actions (as prescribed by the translation dynamics) (Schneider & Ingram, 1997, p. 79).

To capture the constructions, we focus on the concrete content of policy – the design elements – thus expanding a literature where procedural aspects have previously received the most analytical attention (Howlett & Mukherjee, 2014). Rather than focusing on actor processes and who the policy formulator/designer is (Howlett & Mukherjee, 2014; Schneider & Ingram, 1997; Turnpenny et al., 2015), we will add a new piece to the empirical puzzle by examining the formal policy content. Our approach allows us to focus on aspects that risk being overlooked by approaches not explicitly engaging with the ways in which policy formulation affects both policy implementation and policy outcomes. By identifying the underlying constructions of Swedish moose policy, treating these as a determinant, we are able to deconstruct policy design and discuss its consequences in practice. We are also able to discuss potential theoretical contributions to the growing research field focusing on policy formulation, where analyses of the construction of objects of government policy may increase our understanding of policy implementation and its successes and failures.

Analytical framework

Taking inspiration from the policy design framework developed by Schneider and Ingram (1997), combined with components from other policy formulation studies (e.g. Beland Lindahl et al., 2016; Schneider & Ingram, 1990), we use a framework developed to identify and operationalize policy design elements and their relationship to strategies (see Table 1).

Table 1. Framework for analyzing policy formulation. (Mainly adapted from Schneider & Ingram, 1997).

Formulation elements	Definition	Questions asked to the material
Problems to solve	What needs to be changed.	What problems are identified? How are they addressed?
Goals	What needs to be achieved. Explanations/motivations for justification of policy.	What goals are expressed? How are goals derived and justified? Which ideas are important guiding principles in determining policy goals?
Target groups	Individuals, groups, or organizations targeted by the policy, who will be receivers of e.g. costs and benefits.	Who/which groups does the policy target, concerning tools and rules? How are they described? (E.g. positively/negatively/demands?)
Implementation	Actors who are a part of the formal governance structure and responsible for implementation of policy. Structures/relationships between actors and/or levels. Means for changing and/or steering behaviors as well as explanations/motivations.	Which actors are responsible for implementation? What are their assignments and responsibilities? What is their agency? What means/mechanisms are initiated to affect target groups or actors? How are efforts/solutions legitimized, derived or justified?

The first element is problems to solve. This identifies the intentions and motivations driving the policy. By asking how problems are addressed, we acknowledge that problem descriptions are social constructions. The second element is goals. Like problems, these are social constructions, describing what is to be achieved. They may be clear or unclear, realistic or unrealistic, opposing or complementary (Schneider & Ingram, 1997, pp. 83–84). The third element is target groups. These are actors whose behaviors the policy aims to affect. Target groups are usually assigned benefits or burdens, and such constructions have implications for society in terms of justice, citizenship, support for institutions, and democratic problem-solving. Through these constructions, the policy may strengthen or undermine the support of the state (Schneider & Ingram, 1997, p. 85). The fourth element is implementation. It includes actors who oversee implementation, and the capacity, authority, and discretion they are assigned in the policy content (Schneider & Ingram, 1997, p. 82), their interrelationships, and the tools and/or strategies that are used to change behaviors or capacity of the actors and target groups (Schneider & Ingram, 1997, pp. 93–95).

Methods for analyzing policy formulation

Case selection

In 2010, the Swedish Parliament passed new legislation on moose governance and management. The law departs from the Convention on Biological Diversity and the associated ecosystem approach, thus representing an attempt to move away from single species management to multi- and ecosystem management – a challenge that many countries are facing. A new institutional level, Moose Management Areas (MMA), was added to the institutional landscape. Existing levels included the Swedish Environmental Protection Agency (SEPA), which has overarching national responsibility for management together with the Swedish Forest Agency (SFA), and the County Administrative Boards (CAB), which make decisions about regional goals and management plans in collaboration with Wildlife Management Delegations (WMD). At the local level, Moose Management Units (MMU) consult with local landowners and people with hunting rights to suggest a local management plan, with local hunting teams constituting the lowest level. The new level is situated sub-regionally, corresponding to the ecosystem level of the moose. Its responsibilities include coordinating an area in terms of inventory, statistical data, and education, and suggesting a management plan for the area based on the plans originating from the local level. Many actors consider this change an important tool to manage conflicts between different groups of stakeholders unable to agree on best practices for more socially and ecologically sustainable management (Dressel, 2020; Sjölander-Lindqvist & Sandström, 2019), and thereby a means for the Swedish state to meet various global and international goals and targets (Sandström, 2012).

The initial implementation phase proved difficult, revealing uncertainty concerning mandates and issues related to actors’ willingness and ability to act as stipulated by the legislation (Bjärstig et al., 2014). Recent research on the Swedish moose management system reveals that target achievement is quite high (Dressel, 2020) and that many stakeholders involved in the new system find collaboration desirable (Bjärstig et al., 2014; Lindqvist et al., 2014; Sandström et al., 2013; Sjölander-Lindqvist & Sandström, 2019). However, a prevailing problem is that there is no simple relationship between the number of harvested moose and a decrease in browsing damage (Pfeffer et al., 2018; Spitzer et al., 2019). Managing moose solely via harvesting targets does not seem to be sufficient to decrease browsing damage, nor to reduce social conflicts.

The conflicts surrounding moose management in Sweden are not new, but have escalated during recent decades. The situation is complex, and several aspects are specific to the Swedish context – and to moose. Sweden has one of the densest moose populations in the world (Wallgren et al., 2013). Intensified forest production, which affects moose grazing areas and is itself vulnerable to browsing damage, is increasingly promoted as a key instrument for climate change mitigation (Swedish Forest Agency, 2020). Moreover, Sweden has large socio-ecological variations across the country in terms of species composition, property right structures, and hunting traditions (Dressel, 2020).

Policy analysis, and specifically analysis of policy formulation, will make important empirical contributions to the understanding of Swedish moose policy. First, the moose policy has not previously been studied through a policy formulation lens. It has been analyzed from a variety of perspectives, including ecological and biological aspects of management (Allen et al., 2016; Felton et al., 2016), governance and implementation (Bjärstig et al., 2014; Dressel, 2020; Lindqvist et al., 2014; Sandström et al., 2013), and economic and social values of hunting (Ljung, 2014; Mattson et al., 2014; Mensah & Elofsson, 2017). However, the specific policy content, the underlying ideas and rationale, and the political design strategies (which may contain important elements that shape subsequent events), have not previously been subject to analysis. Second, in relation to the traditional underlying values commonly found in relation to and explaining natural resource conflicts, that is, anthropocentrism and ecocentrism (Dobson, 2016, p. 21-ff), the values of both hunters and landowners are supposedly aligned to anthropocentrism – including an instrumental view of nature and a utilitarian approach to resources. This implies that value patterns are not sufficient to understand the conflict. In accordance with policy theory (Howlett & Mukherjee, 2017; Schneider & Ingram, 1997; Sidney, 2007), we assume that the way the government constructs problems, goals, and objects of policy affects the implementation and success of the policy.

We also argue that the Swedish moose policy is a representative case of natural resource conflicts where social constructions and design strategies based on knowledge and/or power are present; for example, in its attempts to define ecosystem-based management operationally (Lindqvist et al., 2014). Insights from the study of this case can contribute to the increased understanding of other cases, that share similar characteristics and that similarly need to handle conflicts and implementation problems across several sectors and (groups of) actors – both in natural resource management, and in other policy domains.

Material and data collection

The empirical material consists of policy documents. We view policy formulation as a product, and the policy documents as expressing problems, goals, instruments/strategies, and potential outcomes, reflecting the process of design/formulation (Schneider & Ingram, 1997). Formal documents such as governmental records of investigations, government bills, and plenary debates present the formal version of the policy, its design, and its intentions (Halperin & Heath, 2017). They do not account for individual perceptions, but represent a collective notion of a certain issue – including the policy content as well as the state’s arguments in favor of certain choices, and its reasons for goals and instruments. Such documents, therefore, represent appropriate material to capture the constructions and assumptions underlying the policy design. It is important to be aware of the gap between formulation and actual implementation of the policy, and it cannot be assumed that the policy mirrors events in society.

The selection of empirical material followed the Swedish legislation process, which commonly starts with a knowledge-based inquiry by a government-appointed committee or investigator, commissioned to explore the conditions for the government’s implementation plans. The resulting report (including the analysis and drafts for new legislation or amendments to existing laws) is referred for consideration to relevant authorities, organizations, municipalities, and other stakeholders before the government’s final consideration. After the inquiry and referral process, the responsible ministry drafts the proposed law and submits it to parliament (often after referral to the Council on Legislation, to ensure coherence with existing legislation). One or more parliamentary committees may submit comments on the proposal before parliament votes on it.

The moose policy is a framework legislation, and the agencies in the Swedish political system are independent (Government Offices, 2015). This implies that the policy is formulated to create discretion for the authorities and the actors involved. Furthermore, the government chooses which referral responses to bring into the final bill, to provide arguments for its suggestions. Such content may be important as it shapes the formulation. To find relevant policy documents, we traced references in the scientific literature to ‘white papers’ and searched government websites and databases. Table 2 summarizes the material analyzed, listed chronologically to reflect the legislative process.

Table 2. Selected policy documents for analysis.

Title of document (English, reference code in brackets)	Context and relevance	Origin
Directive 2008:63 (JO 2008:03)	The initiating document stipulating the official problem and appointing a committee of inquiry.	Government
Final inquiry report (SOU 2009:54)	The final report from the investigation initiated by the directive. Conducted between 2008 and 2009.	Investigator
Compilation of referral responses (JO 2009/1932)	Document compiling all responses to the inquiry report 2009:54, from stakeholders and other actors.	Ministry of Agriculture^a
Memorandum (JO 2010/1120)	Memorandum outlining proposal on law text, submitted to the Council on Legislation and sent out for response.	Ministry of Agriculture
Compilation of referral responses (JO 2010/1120)	Document compiling all responses to memorandum JO 2010/1120, from stakeholders and other actors.	Ministry of Agriculture
Statement of opinion ‘Älgförvaltningen’, 2010-06-10	Document outlining the opinions of the Council on Legislation regarding law text in memorandum 2010/1120.	Council on Legislation
Government Bill 2009/10:239	The bill stipulating proposals for moose governance and management in Sweden.	Government
Parliamentary Committee Report (2010/11 MJU6)	Output of parliamentary committee deliberation.	Committee of Environment and Agriculture (Parliament)
Motion 2010/11 MJ2	Motion from representatives of the Social Democratic Party concerning the bill.	Matilda Ernkrans (S), et al.
Motion 2010/11 MJ395	Motion from representatives of the Moderate Party concerning the bill.	Krister Hammarberg (M), et al.
Minutes 2010/11:25	Meeting minutes from parliamentary debate. There was no debate concerning moose management.	Parliament
The Swedish Hunting Act 1987:259	Framework legislation regulating hunting and management of moose. Main legislative document.	Parliament
Hunting Ordinance 1987:905	As above, but with subordinate bylaws for further clarifications of legislation.	Government

^a Now: Ministry of Enterprise and Innovation.

Analytical method

The analytical method used is a qualitative content analysis (Elo & Kyngäs, 2008; Ritchie et al., 2014). The first author categorized and coded the material according to the theoretical elements and empirical questions presented in Table 1, and the other authors subsequently validated the first author’s coding. For example, to analyze the policy’s construction of target groups, the texts were searched for mentions of by whom suggested tools would be used or to which groups proposed rules would apply, and for descriptions of these target groups. We use quotes generated through the coding structure (translated from Swedish to English by the first author) in the results to highlight, exemplify, or clarify specific elements. Data management and coding was performed using QSR International’s NVivo 12 software.

Some elements have explicit expressions in the studied texts, while others are implicit. To capture implicit content, we searched the material for expressions of ideas, value statements, and/or general assumptions about behaviors, actors, or similar. To support our interpretations of the implicit content, we describe the line of thought and provide quotes or paraphrasing to increase credibility of the analysis (Lincoln, 1995; Morrow, 2005).

We processed, categorized, and coded the material in its entirety, with increasing level of detail in each step of the process. In the coding process, codes were edited, merged, or deleted depending on their relevance to the research questions (see Table 1). Some of the texts were analyzed in more detail, as they provided more substantial material for the analysis of, for example, problem- and goal formulation, but to understand the process of policy formulation as a whole, we found it necessary to include and process texts from all stages of the legislative process.

Results

The result section has been structured in accordance with the analytical framework (see Table 1). In a first step, we analyze three of the formulation elements: problems to solve, goals, and target groups. This is followed by an analysis of the fourth formulation element, implementation.

Goals, problems, and target groups

Following the logic of the Swedish tradition of developing laws as frameworks, government bills are often formulated as a set of principles. Starting with the overarching goal of the proposed moose policy, we interpret it as including two parts: one focusing on the object of the policy (what needs to be achieved, and why), and the other on how to implement the policy. The goal is:

Today’s moose management system should be changed to create a high-quality moose population in balance with pasture resources. The management system should consider important public interests such as large carnivores, counteracting traffic accidents with moose, damage to forest and impact on biodiversity. Tomorrow’s moose management should be characterized by collaboration between the actors that affect the moose population. (Gov. bill 2009/10:239, p. 16)

The goal is further refined, with principles for implementation, in the following statement:

… moose management should be locally anchored, and ecosystem-based. The goal is a viable, high-quality moose population that is in balance with pasture resources and a production-based moose hunt. The administration of moose hunting should be simplified. The moose hunt should be more goal-oriented than today. (Gov. bill 2009/10:239, p. 20)

To deconstruct the content of this goal, we consider its origin and justification, and the important guiding principles. We found that the policy formulation articulates two groups of problems serving as explanations for a lack of effective steering – ecological problems and economic problems, respectively.

The goal implies a need to come to terms with the current imbalance between the number of moose and available food resources; that is, the policy expresses a need to restore a degraded ecosystem to be able to improve the quality of the moose population. This also includes predator-prey relationsimpacts in relation to traffic, forestry, and biodiversity. The policy defines the concepts of ‘viability’ and ‘high quality’ in relation to low risk of extinction. It identifies life expectancy and skewed sex ratio, as well as poor management, as factors that determine viability. It generally assumes that quality can be assessed by factors that have little dependence on heredity, e.g. number of calves born per cow, slaughter weights and antler sizes (Gov. bill 1986/87:58, p. 37; Gov. bill 2009/10:239, p. 22). We interpret this as an ecological focus in the goal. In terms of management, the policy states that the goal could be achieved through production-based moose hunting where ‘the size of the moose population is adapted to pasturage, land-based industries, and traffic safety through the use of regulated culling’ (Dir. 2008:63, p. 2; Gov. bill 2009/10:239, p. 22).

However, it appears both in the preparatory work and in the bill that ecological aspects are not the main driver for the goal. Rather, we find that the core rationale is the need to address the economic loss to forestry and agriculture caused by moose browsing damage, including negative effects on timber quality and production as well as negative consequences for biological diversity. It is the economic balance that needs restoring, rather than the ecological. The policy briefly mentions positive aspects of moose presence, such as a source for recreation and provisions, including meat from hunting (Gov. bill 2009/10:239, p. 17), but these aspects receive little attention compared to the negative ones.

To illustrate these drivers, we investigate the group of problems gaining the most attention in the policy: the negative effects on timber production and timber quality due to browsing (Gov. bill 2009/10:239, p. 18). The policy starts from an estimate of annual financial losses in the forestry sector that ranges between SEK 500 million and 1.3 billion, depending on the damage level (Dir. 2008:63, p. 2). It identifies the biological consequences of this as loss of growth, decreased tree vitality, and competition between tree species (Gov. bill 2009/10:239, p. 18). The policy describes how some forest owners, to minimize the risk of financial loss, have chosen to plant spruce instead of pine, even on land that is more suitable for pine. Although this approach reduces grazing pressure and damage within the stand, since spruce is less sensitive to browsing damage than pine and moose prefer pine over spruce, the policy associates this strategy with a certain loss of economic value (Gov. bill 2009/10:239, p. 19). In addition, the policy describes the approach as leading to ‘sprucification’ of the landscape, with negative effects on biological diversity (Gov. bill 2009/10:239, p. 18).

Furthermore, spruce plantations are described as having negative effects on the moose population because food production is comparatively lower in them (Gov. bill 2009/10:239, p. 19). The moose will concentrate on the remaining pine stands, thus increasing the browsing damage. This problem has also been described in previous policies (Gov. bill 1991/92:9, Dir. 2005:142, Inquiry report 2006:81, and Inquiry report 2007:63). In the preparatory work, this is explained by a lack of quality assured information on which to base management plans and decisions. This, in turn, is described as mainly due to a lack of inventory methods, both in terms of access and interpretation competence, and a lack of trust in the methods used (Inquiry report 2009:54, p. 133; Gov. bill 2009/10:239, pp. 50–52).

Several of the referral responses explain the imbalance between moose and their food resources as a result of a lack of power among landowners. For example, the Federation of Swedish Farmers argues that the landowner has little room to influence the size of the moose population – due to the set-up of the (old) management system – at the same time as he/she faces the consequences of browsing damage. They argue that the moose population controls what tree species develop into forests, not the landowner (the sprucification), and that this negatively affects biological diversity and economic returns (Gov. bill 2009/10:239, pp. 16–17). We interpret this as the landowners being dependent on hunters to achieve the overarching objectives of the policy. The inquiry report also brings up the argument of landowner dependence on hunters, stating that forestry- and agricultural actors have identified that the hunters (who have a voluntary obligation to carry out hunting and reach culling quotas) do not consider the problems of high levels of damage when they hunt (Inquiry report 2009:54, p. 126). We find this illustrative of a tendency with the inquiry report to yield political priority to economic aspects and to include hunters in the problem formulation.

The second group of problems underpinning the goal concerns the complex management system, which, according to the policy, serves as another explanation for lack of effective steering. The system is described as difficult to understand and to oversee, and as lacking interaction and communication between different levels (Inquiry report 2009:54, pp. 130–131). The bill argues that the way the management system has developed over time has led to complexity:

The current system contains so many different types of licenses and forms of management that it is difficult to oversee and understand. It also causes a lot of administrative work with re-registration between different types of license areas. During the years 2006–2008, the total number of de-registrations, re-registrations and new registrations was just over 15,000 [cases]. This means that when three years have passed, half of all existing license areas have been changed and re-registered. (Gov. bill 2009/10:239, p. 19)

This pluralism of management forms is said to contribute to a considerable administrative burden for the County Administrative Boards (CABs), which, it is argued, is very costly for the state. The policy also identifies a lack of overview and state control as another aspect of this problem. For example, the analyzed texts point to a lack of state oversight of local management plans, and the considerable variation between CABs regarding monitoring and follow-up, arguing that this leads to questions such as who has responsibility for setting up and implementing ecosystem-based management (EBM) (Inquiry report 2009:54, p. 131; Gov. bill 2009/10:239, p. 20). Finally, the policy identifies the size of management areas as a problem. Some management areas are too small to manage their own moose population, making it difficult to reach the goal of a production-based moose hunt with a viable, high-quality moose population, i.e. social-administrative systems do not match ecological systems (Inquiry report 2009:54, pp. 130–131).

Implementation: actors, structures, strategies

Based on the goal, and in order to come to terms with the identified lack of effective steering, the government proposes that the future governance of moose should rest on five overarching principles: (1) the ecosystem approach, (2) adaptive management, (3) ecosystem-based local management, (4) simplicity and rule of law, and (5) a holistic perspective. The governance system should be based on the same building blocks throughout the country, and the same regulations and systems should apply. To implement these principles, the government introduces a new multi-level governance system.

The core rationale underpinning this new system is the ecosystem approach. The bill states that Sweden is bound by its commitments to the Convention on Biological Diversity (CBD), which stipulates EBM as the way for signatory states to work with nature conservation and biodiversity (Gov. bill 2009/10:239, p. 21). The CBD argues that EBM is based on the application of appropriate scientific methodologies focused on levels of biological organization, which encompass the essential structure, processes, functions and interactions between organisms and their environment (Secretariat of the CBD, 2004). Furthermore, the bill describes how EBM starts from a holistic perspective and includes a principle of subsidiarity (decision-making at the lowest appropriate level) where it is important to ensure participation of relevant actors in goal formulation and in management practice. Finally, the precautionary principle should apply. This approach requires adaptive management to deal with the complex and dynamic nature of ecosystems and the absence of complete knowledge or understanding of their functioning, i.e. to formulate goals, monitor, and adjust management practices, all in a continuous learning loop (Inquiry report 2009:54, pp. 119–120; Gov. bill 2009/10:239, pp. 21–23).

In the moose policy, the Swedish government operationalizes EBM as local EBM. Starting from the idea of ‘lowest appropriate level’, the suggested system contains two levels – a revised form of the previous local level, and a new sub-regional level: Moose Management Areas (MMA), administered by Moose Management Groups (MMG) consisting of three representatives of landowners and three representatives of hunting.¹ The MMA is presented as an attempt to match the ecosystems of the moose to a political-administrative system, and the policy argues that each MMA should therefore cover its own huntable moose population.

We find three main rationales for proposing the introduction of MMAs as a new, sub-regional level. The first is to ensure the principles of EBM. Using larger management areas is described as a way of practicing holism and remedying the issue of a lack of overview, in order to improve the steering of moose hunting to reach the balance goal (Inquiry report 2009:54, p. 131, 136; Gov. bill 2009/10:239, p. 20). A second rationale is to ensure effective use of CAB resources by decreasing license areas and reducing costs, as MMGs are now tasked with compiling and analyzing data at the ecosystem level. At the same time, the state retains control by having the CAB as final decision-maker for the management plans (Gov. bill 2009/10:239, pp. 23-ff.). The third rationale, according to our interpretation, stems from political pressure (from concerned actors and the inquiry report) to increase the influence of landowners in matters concerning moose and hunting. For example, forestry actors have expressed dissatisfaction with hunters having great influence on management (JO 2010/1120). The inquiry report supports this, concluding that landowners ‘must be able to control’ moose culling, to reduce the negative impact of the moose and to be able to exercise landowner responsibilities in accordance with the Forestry Act (1979) (Inquiry report 2009:54, p. 131).

On the local level, Moose Management Units (MMU) have been in place since 1992. Within the MMUs, hunters and landowners co-manage moose, by suggesting appropriate hunting quotas based on inventory data. The bill proposes changing the administration and size of the MMUs, including larger areas to allow for culling of ten adult moose, and to increase dialogue between levels by moving responsibilities for inventory compilation and data collection to the MMGs. The stated intention of this change is to provide incentives for local hunters and landowners to collaborate, and to incentivize hunters to reach culling goals. Using adaptive management in the implementation structure is assumed to increase capacity in the system and to improve the inventory and knowledge base on which more sound decisions can be made. Hunters are expected to present reliable information in order for the system to work, while nothing is mentioned about expectations in relation to landowner responsibilities (Gov. bill 2009/10:239, p. 34). The rationale for these changes is the recognition that voluntary hunting collaboration is an important foundation for the stakeholder parties and for management (Gov. bill 2009/10:239, p. 31).

The policy justifies this multi-level construction by three main arguments. First, management by stakeholder groups is argued to be cost-effective and easier for the state to administer, e.g. by decreasing registration of license areas. Secondly, if this system is successful, it is argued that it will provide the foundation for multi-species management (which is another state intention). Thirdly, changing the incentive structure for hunters (by providing better hunting conditions if they voluntarily collaborate) and landowners (by providing them with formal representation and the casting vote) is assumed to increase collaboration and foster a greater mutual understanding between the two stakeholder groups. It is assumed that this will encourage hunters to increase culling (solving the problem of low goal fulfillment) and make landowners take more responsibility for land management, in terms of forestry practices (easing the problem of browsing damage).

The referral actors are, in general, pleased with the system proposed in the bill (JO 2009/1932; JO 2010/1120). However, some criticize the justifications for being overly simplistic, and for lacking understanding of drivers of hunter and landowner behaviors. As the CAB of Västerbotten puts it, ‘the proposal does not include any clear incentive to hunt within an MMU’ (JO 2009/10:1120, p. 14), implying that the goals of simplified administration and goal-oriented steering will not be achieved as described. Both the CAB of Västerbotten and the Swedish Association for Hunting and Wildlife Management (SAHWM) argue for a clarification of incentives for local cooperation in the MMU (JO 2009/10:1120, p. 17). They argue that as the MMU is voluntary, it may not be formed at all, and therefore the problems of complexity in the system and heavy administration will only partly be met (JO 2009/10:1120, p. 14).

Referral actors also suggested that maintaining the MMU and introducing the MMA will mean a decrease of local power, moving power vertically in the system from the local level (MMU) to the sub-regional level (MMG) (Inquiry report 2009:54, p. 139-ff). The hunting representatives (SAHWM, and the National Association for Hunting), the CAB of Västerbotten and Sweden’s Landowner Association all emphasize the importance of MMUs as self-governing bodies. Some of the major landowner representatives (Federation of Swedish Farmers, Swedish Forest Industries, Sveaskog AB, Swedish Church) and most of the CABs oppose the proposal to make MMGs advisory bodies, not decision-makers, arguing that MMGs have no real power (Gov. bill 2009/10:239, pp. 23–24). How power is distributed in practice thus remains unaddressed in the policy.

Concluding discussion

We examined the social constructions underpinning the Swedish moose management policy of 2010 to increase understanding of the ongoing conflict. Our results show that the policy contains imbalances in relation to its intended goal, partly by the operationalization of local EBM, and partly by how it manages different stakeholder groups.

Our analysis shows how the policy initially portrays the problem of moose as an ecological imbalance, while drivers and justifications are largely economic and aim for economic balance for landowners. Constructing moose as an economic problem, where the policy emphasizes economic consequences of browsing damage rather than ecological knowledge (very little is said about how to achieve a high-quality moose population), the government’s strategy aligns to a considerable degree with the landowner perspective, as expressed in e.g. referral responses from the Federation of Swedish Farmers and Sveaskog AB. The government’s position might be explained by economic incentives, as a response to criticism, or because it shares the problem description with the forestry actors, all stemming from the fact that the forestry sector constitutes an important part of the Swedish economy (Statistics Sweden, 2019). We argue that this serves as an illustration of why not only knowledge use, but also political strategies, deserve attention when considering a particular formulation (Howlett & Mukherjee, 2017).

In their referral responses, the forestry actors argue that the negative economic consequences of browsing are due to lack of influence of landowners in decision-making and that landowners are dependent on hunters for hunting. The character of these opinions is largely practical/technical, in contrast to the hunters’ associations, who highlight societal values and ask how the different resources are valued. By following the economic rationale and stating a goal of balance without incorporating forestry practices (how to manage forest in a way that can decrease browsing damage) (Felton et al., 2016), or the role of other ungulate species as regards browsing damage (Pfeffer et al., 2018), we conclude that the policy fails in its ambition to implement a holistic perspective through EBM.

Furthermore, implementation principles (stemming from two different design strategies) clash with each other while attempting to incorporate both local and ecosystem-based levels of management into a single system. By introducing the MMAs, the government chooses a design strategy focused on scientific knowledge (Howlett & Mukherjee, 2017; Schneider & Ingram, 1997, pp. 67–68) – EBM – to improve the quality of the moose population and improve decision-making. At the same time, the MMU is local, that is, a strategy where the interests of strong stakeholders and consideration of the institutional landscape (for example, landownership rights) in Sweden come into play (Howlett & Mukherjee, 2017; Schneider & Ingram, 1997). This leaves the management system with many uncertainties and tensions. As in previous research, we identify a need for a deeper discussion about how to operationalize EBM in relation to moose (Hoffman & Flø, 2017; Lindqvist et al., 2014) and the fragmented stakeholder groups (Ezebilo et al., 2012). Because of the imbalances, the policy will have difficulty achieving its goals unless it is adjusted to consider more than just economic concerns.

In attempting to find the ‘lowest appropriate level’, the local and the sub-regional levels are at risk of competing. The new, sub-regional governance level may appear to give landowners more power, but actual decision-making power still lies with the CAB and the MMU (the local level, at which management takes place operationally). Asserting power can be broader than decision-making, and landowners as a stakeholder group still carry a large influence, but this lack of clear and fair power-sharing is important to consider.

The division between the local and the sub-regional levels also contributes to the lack of clarity concerning the distribution of tasks and responsibilities, a conclusion supported by Bjärstig et al. (2014). It may contribute to ineffectiveness, as the state pushes responsibilities onto the management actors to deal with uncertainties and political trade-offs (an example would be not to use the casting vote in the MMG). Value conflicts can persist and are pushed onto lower levels, which – as shown by studies on other resources – can be interpreted as a lack of governance and leadership (Beland Lindahl et al., 2016). By choosing this design, the government fails to identify the lowest appropriate level for decision-making as established by EBM.

While arguing for and attempting to promote decentralization and regionalization, the government advocates a nationally comprehensive and unified system of management. There is much literature on governance of social-ecological systems arguing for small-scale solutions working more successfully than large-scale arrangements in relation to natural resources (Fleischman et al., 2014; Ostrom, 1990). In the Swedish context in particular, large-scale management solutions have been identified as problematic due to vast regional ecological and social differences (Dressel, 2020; Dressel et al., 2018). There is also a risk of continued spatial mismatch when moving towards multi-species management (as is the long-term intention), given that the current spatial scale is adapted only to moose (Dressel, 2020, p. 77). This actualizes issues of power to frame the scale (Van Lieshout et al., 2017), that is, whose preferred scale and level is found appropriate, and why.

The strategy of decentralization and increasing stakeholder involvement and collaboration can be explained as a way of incorporating the constitutional right of landowning (SFS 1974:152, Ch. 2 §15), as it cannot be circumscribed in this case. All collaboration and concerted efforts of management need to be based on landowner approval, as they may choose to not participate. This collective action problem is the main reason why the government needs to steer by incentives. It is assumed that the incentives provided by the policy for each of the target groups will encourage the stakeholder groups to undertake the tasks of learning and capacity-building, assuming they have (or will gain by education) the capacity and willingness to do so. The policy is dependent on the actors behaving in accordance with these assumptions (and the assumption that increased knowledge leads to better outcomes), but what happens if these assumptions are deficient?

For example, the policy assumes that the referral actors for forest interests (landowner organizations and/or large-scale industrial forest companies) represent the fragmented structure of landownership and the different types of landowners that exist in Sweden. However, this is not necessarily the case. As shown by Ezebilo et al. (2012), landowner types in Sweden can be categorized in different ways depending on the utility of the land, which leads us to argue that the current policy does not reflect the diversity that exists on a local scale. A similar issue exists with the hunter group. The policy has a superficial treatment of hunter characteristics, incentives, and behaviors, as shown by the explanations for lack of effective steering, which is a problem since the policy puts a clear emphasis on hunters as the policy object needing to be governed. Both landowner fragmentation and hunters as policy objects could be addressed in a more nuanced way in the policy, in order to balance interests.

These conclusions on the case of Swedish moose policy, while context-specific, may be useful for analysis of other cases sharing similar characteristics. Across the world, governments struggle to formulate and implement natural resource- and wildlife management policies that follow multi- and ecosystem management principles (IPBES, 2019). Future research should further explore the consequences of policy imbalances in relation to intended goals, the importance of understanding the rationales and design strategies underpinning implementation principles, and the need to discuss operationalization of EBM (or other multi-sectoral, multi-actor, and multi-scalar approaches) in relation to different types of objects of management and to issues of scale.

In addition, we show that a constructivist perspective helps to open an otherwise instrumental black box, and that analyzing the design elements is useful to entangle and deconstruct the design or non-design strategies at play. This approach highlights how knowledge use and different levels of political involvement can be more or less devised strategies (Howlett & Mukherjee, 2017; Owens, 2015). Based on this perspective, our results constitute an argument that policy analysis should steer away from an instrumental view of policy, as all policy is political (Schneider & Ingram, 1997; Weible & Sabatier, 2018). Focusing on design elements let us explore and draw conclusions on the ways in which problems, goals, and objects of policy are constructed in and through the formulation of policy, thus allowing for a discussion of the balancing of different values and the potential consequences (in terms of implementation and outcome) of a policy. Future research should continue to develop this framework and apply it to other policy areas.

Acknowledgements

We are grateful to the two anonymous reviewers for their useful comments.

Disclosure statement

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

Additional information

Funding

This work was supported by Formas – a Swedish Research Council for Sustainable Development [grant number 251-2011-117] and by the Swedish Environmental Protection Agency (Naturvårdsverkets Viltvårdsfond) [grant number 802-0161-15].

Notes on contributors

Katarina Hansson-Forman

Katarina Hansson-Forman was a PhD student at the Department of Political Science, Umeå University, when conducting this research. Her research explores sustainability and natural resource policy in the Swedish political context.

Elsa Reimerson

Elsa Reimerson is an associate professor at the Department of Political Science, Umeå University. Her main research interests lie in environmental politics and policy, with a focus on collaborative governance, conservation, and the rights of Indigenous peoples.

Therese Bjärstig

Therese Bjärstig is an associate professor and researcher at the Department of Political Science at Umeå University. Her research focuses mainly on collaborative governance, policy instruments, evaluations, and implementation processes in relation to forests, wildlife, natural resource management, rural development, and landscape planning.

Camilla Sandström

Camilla Sandström, professor in political science, Umeå University, has done extensive research on environmental policy and politics as well as on the governance and management of natural resources, including wildlife.

Notes

1 For a detailed and comprehensive description of the moose management system, see Dressel (2020).