Understanding How Policy Actors Improvise and Collaborate in the Great Barrier Reef

ABSTRACT

Complex policy frameworks guide the management of the Great Barrier Reef (GBR) at multiple decision-making levels. Mounting pressure on its outstanding universal value suggests that further improvements in governance are required. There has been little examination of the role of policy actors in addressing complex governance challenges in large scale marine protected areas (LSMPAs) involving multi-layered governance conflicts across diverse contexts. Framed by street level bureaucracy, pragmatic planning theory, and lessons from MPA governance, this paper examines how policy actors improvised and collaborated to advance outcomes in the GBR LSMPA. We assessed practice-focused oral histories with experienced policy actors who negotiated agreements and achieved policy outcomes in the GBR between 1985 and 2016. These policy actors were skilled improvisers, alternating between roles as democratisers, mediators, and negotiators. They used collaboration and facilitative leadership to advance policy in the face of governance problems. This enabled them to adapt policy across multi-level decision systems, address power and information imbalances, and generally deal with conflict and uncertainty. A deeper understanding of improvisation will assist governments and others to address complex challenges in LSMPAs.

KEYWORDS:

• collaboration
• governance
• great barrier reef
• improvisation
• large scale marine protected areas

Governance challenges and taking action in large scale marine protected areas

Large scale marine protected areas (LSMPAs) are a growing marine conservation trend (Gruby et al. 2016) that pose several important governance challenges compared to smaller management units. Introduced by centralized governments often involving top-down processes and legislation (Gruby et al. 2016), LSMPA models are different from community-based, adaptive co-management models for marine conservation that are known to deliver social and ecological marine conservation at smaller scales (Lane 2001; Cinner et al. 2012). LSMPAs are driven by the centralized governments in partnership with national and international conservation non-government organizations and donors (Campbell et al. 2016). Top-down governance processes that neglect or delay stakeholder involvement can create conflict with groups who need to be involved in marine conservation decisions (Gruby et al. 2016). Conflict can also exacerbate non-participation problems and lead to ecologically focused decisions and poor social justice outcomes for local people over large areas (Ban et al. 2017; Bennett et al. 2017; Mitchell 2017).

The governability of large sites has also been questioned (Campbell et al. 2016). Larger sites involve many stakeholders, multiple levels of decision-making, and complex governance interactions (Dale et al. 2016; Gruby et al. 2016). Adequate consultation is more challenging (Gruby et al. 2016), and large scales pose particular problems for contextualizing policy and engaging local people in monitoring and enforcement (Ban et al. 2017; Mitchell 2017). The blanket application of MPA tools to large marine sites is also highly problematic. Though the blanket application of MPA tools may give the impression that conservation targets have been achieved (i.e., to have 10% or 20% of marine waters covered by a protected area), it does not lead to long-term marine conservation outcomes (Gerhardinger et al. 2011; Agardy, Claudet, and Day 2016). In smaller MPAs, experience has shown that poorly contextualized policy creates its own problems. Marine policy that pays limited attention to local displacement problems, for example, can create the perception that some resource users are unfairly singled out (Agardy, di Sciara, and Christie 2011). Rather than resolving conservation problems, poorly contextualized policy can in fact exacerbate conflict and compliance issues (Agardy, di Sciara, and Christie 2011; Agardy, Claudet, and Day 2016), and prohibit conservation outcomes across diverse coastal and marine contexts (National Research Council Staff 2001; Warner and Pomeroy 2012; Leenhardt et al. 2013). Contextualization problems are acute in large sites involving diverse socio-political, economic, and institutional values and processes (Gruby et al. 2016; Ban et al. 2017; Mitchell 2017).

There has been no examination of the role of policy actors in improvising governance in large MPA contexts. How do policy actors negotiate and advance policy outcomes in large MPA sites? What barriers do they face? What strategies do they employ in complex (multi-dimensional, multi-stakeholder, and multi-scalar) situations to take action? The Great Barrier Reef (GBR) offers an opportunity to learn about how to take action in a very large, complex, and highly contested large MPA (Agardy, Claudet, and Day 2016). Prominent conflict and a strong conservation campaign led the Commonwealth to introduce the Great Barrier Reef Marine Park Act 1975 (the Act) (Day 2016). Since then, GBR policy has been adapted at different spatial locations and decision making levels by government and non-government stakeholders. Unlike LSMPA sites in remote areas (Agardy, Claudet, and Day 2016), the GBR catchment is highly populated and socio-economically diverse (Marshall et al. 2014). The governance system is institutionally complex, fragmented over multiple levels of decision-making and across policy domains (Dale et al. 2016). GBR policy conflicts occur over values and priorities for the use of catchment and marine resources (Day 2016). GBR policy has been adapted and contextualized, leading to innovations in representative zoning (Day 2016), catchment water planning (Dale et al. 2013), and paddock to reef monitoring (McCosker and Northey 2015). Policy advancements have enabled some progress towards marine conservation outcomes and further action is required to address massive threats to the GBR's future (McCook et al. 2010; Hughes, Day, and Brodie 2015; Waterhouse et al. 2016).

To understand opportunities for overcoming governance challenges in LSMPAs, we examine improvisation and collaboration by policy actors in the GBR LSMPA. The evaluation frame combines: street-level bureaucracy (from Lipsky 1980) to consider the impact of discretion on policy action in the GBR, pragmatic planning (from Forester 1982, 1987, 1989) to examine how experienced policy actors improvised in complex and contested MPA governance contexts, and governance lessons from the diverse MPA literature to examine requirements for policy contextualization. We draw on oral histories with experienced policy actors in the GBR to explore how they mediate, formulate, and negotiate outcomes in response to conflict. Our subjects were low- to mid-level policy actors in government and non-government organizations who played a role in GBR governance. Policy actors interacted directly with diverse government and non-government stakeholders and the general public to advance change in the GBR marine park and the adjacent GBR catchment. They were involved in plan making (setting objectives, gathering and analysing information, formulating and evaluating alternatives, implementing policy, monitoring and evaluating progress) and political decision-making (i.e., advocacy, coalition building, collaborative processes) (cf Albrechts 2003). The framework is described below, leading into the methods and application to the GBR.

Analytical framework

To examine the problems policy actors confront and how they take action, we have applied two separate and complementary bodies of theory to the LSMPA context. The first body of theory, street-level bureaucracy (SLB), considers public policy actors and how discretion can act as a barrier, limiting the application of policy to specific situations or to provide opportunities for action. The second, pragmatic planning theory, considers how improvisation can help to negotiate outcomes in complex and contested contexts. We introduce SLB and pragmatic planning theory as a way to understand how to address governance problems identified in the MPA literature.

Policy actors and discretion—street level bureaucracy

MPA policy actors are a type of street-level bureaucrat. They are at the frontline of policy and deal with the gap between policy “in theory” (i.e., legislative goals) and “in practice” (policy outcomes) (Loyens and Maesschalck 2010). Policy actors have considerable discretion to constitute and apply policy in different management settings (Lipsky 1980). Policy success (i.e., goal attainment) depends on the capacity of policy actors to apply general and abstract ideas to specific situations and the individual will of policy actors to take action (Lipsky 1980; Loyens and Maesschalck 2010; McLaughlin 1987). Organizations often have ambiguous, contradictory, and frequently changing goals (Weatherly and Lipsky 1977); lack standards to interpret policy in practical contexts (Lipsky 1980); and find it difficult to supervise and monitor how policy actors perform their role (Brehm and Gates 1997). Conflicts between personal and organizational (or broader) values and interests (Jones 1991; Kaptein and Van Reenen 2001), relationship expectations (i.e., between senior managers and policy actors/policy actors and other stakeholders), and accountabilities (to organizations, the general public, self) create moral dilemmas for policy actors (Wark and Krebs 2000). Policy actors respond to these dilemmas by using coping mechanisms (i.e., stereotyping, screening others by providing different levels of support, making it difficult to access services, and voluntarily withholding information) to survive in their role (Weatherly and Lipsky 1977; Brehm and Gates 1997). Critics argue that professionalism shapes policy actors' discretion (Evans and Harris 2004). This depends on how professionalism is enacted (van der Aa and van Berkel 2015). To address dilemmas, policy actors need to make personal judgments to resolve conflicts over resources and goals, inner motivations, and struggles. These judgments must balance concerns over what is expected and what is realistically achievable from their performance (McLaughlin 1987; Brodkin 1997; Vinzant and Crothers 1998).

Improvisation to overcome governance problems—pragmatic planning theory

Pragmatic planning theory offers guidance about how policy actors in complex, uncertain, and overwhelmingly political environments can improvise to overcome governance problems and advance policy (Forester 1989). In these contexts, policy actors face “great imbalances of power; and multiple, ambiguous, and conflicting political goals” (Forester 1987, p. 303). Planning theory considers how policy actors use information and practical judgment to take action. It finds that no single strategy is desirable in all circumstances; thus, policy actors must improvise—by compromising, negotiating, and mediating—to achieve outcomes in complex situations of uncertainty (Forester 1999; Laws and Forester 2015). This can occur through strategies of “facilitative leadership” to give effect to planning and public deliberation (Forester 2013). Effective policy actors use information and their discretion to make practical judgments about whether to behave as either mediators or negotiators to take action among stakeholders or negotiate compromises between conflicting powers (Forester 1987, 1999, 2013). Effective improvisers are highly skilled and are:

•	pragmatic, technically skilled, and politically able to act as organizing and democratizing forces;
•	able to work in contexts of bounded rationality (incomplete information); and
•	able to behave politically and use discretion to balance and create fair and shared solutions (Forester 1984, 1989).

This produces results, further shaping relationships and positively influencing others' abilities to act and organize, and develops the power and capacity of policy actors to take action over time (Forester 1989).

Governance lessons from the MPA experience

Though diverse in its theoretical orientation, the MPA literature would seem to agree that examining discretion and improvisation could help to better understand governance problems and ways to address them in large scale sites. A lack of effective contextualization of MPA targets and policy to LSMPA governance situations is a problem for advancing conservation outcomes in three related ways.

First, absolute and inflexible MPA policy targets undermine MPA policy development and implementation (Agardy, di Sciara, and Christie 2011; Agardy, Claudet, and Day 2016). Second, poorly contextualized MPA policy undermines its implementation (Warner and Pomeroy 2012). Policies tailored to social, economic, institutional, and governance conditions at the community scale are most likely to advance marine conservation outcomes, particularly in LSMPAs (Sevä 2014; Warner and Pomeroy 2012; Carcamo, Garay-Fluhmann, and Gaymer 2014; Horigue et al. 2016).

Third, multi-level policy actors within and outside of government must be involved in adapting policy to resource, capacity, and administrative situations (Haward and Vince 2009; Bastari et al. 2016; Batista and Cabral 2016; Berdej and Armitage 2016; Gurney et al. 2016; Friedlander et al. 2016). Participation improves the legitimacy of MPA agreements, implementation, and enforcement, particularly across sovereign borders (Leenhardt et al. 2013). Policy actors can constrain financial resources, and administrative and legal capacities, severely limiting implementation (Ferraro et al. 2011; Gerhardinger et al. 2011; Jentoft et al. 2012; Ramirez 2016; Sandström et al. 2016). These problems highlight the importance of understanding how to improvise and adapt policy to be responsive to diverse conservation needs, stakeholder contexts, and to overcome multiple and diverse governance barriers limiting action across large sites.

Methods

We used Forester's (Forester 2009, 2013; Laws and Forester 2015) method to develop practice stories from oral histories with 11 highly experienced and professionally esteemed GBR policy actors. Our subjects were prominent reef policy makers with between 20 and 40 years experience in conservation, coastal, and marine policy. Over this time their roles in policy development, plan making, and political decision-making changed considerably. Four had held positions at multiple levels of government (state and Commonwealth), seven had worked for government and non-government organizations (e.g., universities, regional NRM¹ organizations, consultancies, and industry), and eight had worked in different Australian states and overseas. Three subjects focused mostly on marine policy, one focused mostly on catchment policy, and seven focused on both. Their roles in the GBR were diverse and changed considerably over time. Our subjects had developed plans, drafted policy, led consultations, crafted multi-stakeholder partnerships, built consensus, led political advocacy, managed programs, built monitoring and evaluation programs, delivered assessments, secured resources, represented their organizations externally (i.e., to government and community stakeholders), and faced parliament. Participants were asked to recount their involvement in achieving a major policy outcome in the GBR during 1 h, face-to-face, semi-formal interviews. Table 1 summarizes the major policy outcome that underpinned each interview and participants drew on lessons from earlier in their careers in retelling what they did to achieve an outcome. They were asked to critically reflect on how they had acted to handle dilemmas and overcome obstacles, what they had learned, what they had found challenging, what they had found surprising, and how they had adapted. Data were collected between October 2014 and December 2016 and included some fieldwork and data collected for a soon-to-be published work (Vella and Forester 2018).

Table 1. Interviewee profiles.

Code interviewee	Policy focus underpinning the interview
I1, I9	Marine park zoning and statutory plans of management
I2, I7	Strategic planning for the GBR
I2, I8	GBR region strategic assessment
I4	Environmental impact assessment of development in the GBR marine park
I3, I5, I6, I10, I11	Catchment-based natural resource management and water quality improvement

Governance and policy in the GBR

The 11 policy actors we interviewed worked at multiple levels and achieved policy outcomes through diverse GBR policy mechanisms. Policy action in the GBR is fragmented across government and non-government jurisdictions. The Great Barrier Reef Marine Park Authority (GBRMPA) is a federal level agency introduced to manage the Great Barrier Reef World Heritage Area. The GBRMPA and the Queensland Government, through the Queensland National Parks Department, operate a joint day-to-day management program in the GBR. The Queensland Government has primary jurisdiction for resource management in catchments. Over the last 20 years, the Australian and Queensland Governments have increased funding to community-based regional NRM organizations and agricultural peak bodies² to work with farmers and local stakeholders and improve the quality of water entering the GBR.

Since the Commonwealth introduced the Act, management policy has changed substantially (Day 2016). The first policy instruments introduced under the Act included marine zoning plans and regulations. They managed use and entry to zones (e.g., for fishing) and established a permitting process to evaluate impacts on the GBR (Ketchington and Day 2011). Statutory plans of management were later introduced to manage conflicts and impacts (e.g., tourism) at finer scales in intensively used areas of the marine park (Ketchington and Day 2011). They managed use, pressures, values, and conflict in specific spatial areas within the GBR.

As understanding about the threats to the GBR has grown, policy has increasingly sought to address conflicts in the GBR over resource conservation and use, external threats from catchment activity on the quality of water entering the GBR, and threats to the GBR posed by fragmented policy. This has led to a raft of strategic planning, improved zoning, and catchment planning innovations (summarized in Table 2). These underpin the practice stories we now explore.

Table 2. Timeline of key marine and catchment planning policy evolutions in the GBR.

Year	Instrument and features of the planning policy evolution
1994	25 Year Strategic Plan
	• First comprehensive plan for the entire GBRWHA;
	• Aimed to address pressures on the reef and better coordinate policy across agencies/user groups;
	• Developed by the Great Barrier Reef Marine Park Authority (GBRMPA) as an instrument of the Australian Government;
	• Preparation involved 60+ user groups, stakeholders and agencies, many with conflicting views;
	• Took almost two years of intensive engagement to reach agreement on a vision.
2003	The Great Barrier Reef Marine Park Zoning Plan
	• Was a key recommendation of the 25 Year Strategic Plan;
	• Reviewed and changed existing zoning plans to increase the protection of representative biological communities across the GBR to act as source areas, reference areas, and reservoirs of biodiversity and species abundance. All sections of the GBRMP amalgamated and zoned consistently;
	• Some planning stages were highly contentious due to concerns about resource-use restrictions and locations of marine national park (‘no take’) zones. Over 31,540 written submissions were received by the GBRMPA during the formal planning process (Day 2016);
	• Intensive consultation, research, and negotiations with community and industry stakeholders by the GBRMPA produced a zoning plan that met objectives to protect the range of biodiversity across the GBR and minimised impacts on users (Fernandes et al. 2005; Day 2016)
	• Enacted in 2003, came into force on July 1, 2004.
2003	Reef Water Quality Protection Plan 2003
	• Introduced by Queensland and Australian Governments following scientific/public concern about sugar cane/grazing impacts on GBR health/resilience;
	• Goal: halt and reverse “the decline in water quality entering the Reef within 10 years” (State of Queensland and Commonwealth of Australia 2003 p5);
	• Proposed voluntary “actions, mechanisms and partnerships to build on existing government policies and industry and community initiatives” (State of Queensland and Commonwealth of Australia 2003, 2) and promoted implementation though low cost measures, as “good planning” and assistance to “landholders in adopting best management practices that are both profitable and environmentally sustainable” (State of Queensland and Commonwealth of Australia 2003, 2).
2009	Reef Water Quality Protection Plan 2009
	• Continued immediate goal “to halt and reverse the decline in water quality entering the Reef by 2013” and established a long-term goal “by 2020 the quality of water entering the Reef from adjacent catchments has no detrimental impact on the health and resilience of the Reef” (State of Queensland 2009, 14);
	• Established water quality and management practice targets, identified priority areas for action and established implementation mechanisms;
	• Implementation mechanisms included the US$200 million Reef Rescue Program (Commonwealth) and the Reef Protection Regulations (Queensland).
2013	Reef Water Quality Protection Plan 2013
	• Monitoring and report cards showed progress in addressing water quality (Great Barrier Reef Marine Park Authority 2014). It also found that progress was not sufficient to build reef health and protect it for future generations (Brodie et al. 2013);
	• Introduced higher water quality/land and catchment management targets to reduce diffuse pollution from agricultural sources and was linked to the Reef 2050 plan (State of Queensland 2013);
	• Modified the long-term goal: “To ensure that by 2020 the quality of water entering the reef from broad scale land use has no detrimental impact on the health and resilience of the Great Barrier Reef” (State of Queensland 2013: p 20);
	• Implementation: Australian and Queensland governments increased funding to improve land and catchment management practices. Reef regulations framework was deprioritised.
2015	Reef 2050 Plan
	• Response to request by UNESCO's World Heritage Committee to show how the “outstanding universal value of the GBR” is being adequately protected and managed (Commonwealth of Australia 2015);
	• Coastal and Marine Strategic Assessment Reports conducted and formed the basis of stakeholder consultation in 2013–2014;
	• Developed a bilateral plan by Australian and Queensland Governments during 2013 and 2014;
	• The scope expanded to include the adjacent catchments in addition to the World Heritage Area;
	• Consultation mechanisms occurred through key organisations and industry partnership groups, targeted traditional owner consultations, local marine advisory committees, and public consultation processes. 6,809 submissions were received on the draft plan during the formal consultation process (Australian Government, 2016);
	• Key elements added: progress objectives, implementation arrangements, monitoring, reporting, and review arrangements.

Results

The interviews identified a complex array of obstacles confronting policy actors in the GBR. Drawing on the practice stories, we identified the dilemmas and strategies they employed to negotiate policy outcomes.

Adapting policy to community contexts

The introduction of rezoning in the GBR highlighted that working with the community helped to contextualize policy and negotiate outcomes that stakeholders could live with in high conflict situations. To resolve conflicts over values and engage local people in resource stewardship, policy actors have to mediate between community and government interests and negotiate compromises:

One thing I've learned is you have to compromise… You need to say there'll be compromises. You can't have something as radical as the Great Barrier Reef zoning without everyone losing a little bit (I1).

Compromises took different forms across different policy mechanisms in the GBR. Another policy actor reflected on compromising by agreeing to get involved and take policy action. In this case, local tourist operators wanted the GBRMPA to mediate and address a local conflict over water skiing and jet skiing in the GBR. The GBRMPA was “reluctant to get involved … because [the conflict] was more about social and economic impacts” (I9) than ecological impacts, and there was no clear process. However, the policy actor also knew that to engage the community in stewardship, it was important to build relationships with stakeholders and between stakeholders and that the policy outcomes would help the GBRMPA make permitting decisions. The policy actor also wanted to establish “community ownership and control of management” (I9). The policy actor compromised and established dialogue between stakeholder groups. Collectively they trialled and evaluated management options, leading to the development of a plan of management for the Whitsundays, which took over ten years to negotiate. This built local ownership of the plan and allowed stakeholders to negotiate workable management options. Once local community stakeholders were satisfied, they then lobbied ministers to have the plans approved.

This need for stakeholder support was comparable to the 2003 Great Barrier Reef zoning plan, which also required community support for parliamentary approval. A key difference with the zoning plan is that the timeframes were shorter. It was an action out of the 25 Year Strategic Plan (Great Barrier Reef Marine Park Authority 1994), and as one subject stated, “I remember saying, ‘Look, if we tell them it's going to take three of four years, we won't be allowed to start it. So, tell them it will take two to three years.' As it turned out, it was five” (I1). Achieving the representative areas program through rezoning required policy actors to negotiate compromise agreements. The buffer zone for example—effectively a no take zone “for everyone except for the charter fishers” (I1)—was introduced to keep the charter fishermen on side. Though policy actors described participation as one of the most important strategies for taking action, it is not effective without a “willingness to be flexible” (I9) and reach compromise.

Linking policy and action across disjointed governance structures³

In 2015, the Reef 2050 Plan (Australian Government 2015) was introduced jointly by the Australian and Queensland Governments in response to mounting pressure and a request by UNESCO's World Heritage Committee to show how the outstanding universal value of the GBR is being adequately protected and managed. This plan is the most far reaching in scope in the GBR, covering management of the GBR and its adjacent catchments and policy and action at multiple decision-making levels (Australian Government 2015). It also includes progress objectives, implementation arrangements, and monitoring and reporting and review arrangements (Australian Government 2015). Implementing the Reef 2050 Plan requires co-ordinated governance across disjointed decision-making at national, state, regional, and local levels. As one policy actor noted, good governance is a key challenge to be considered “at the start of planning processes” (I2) to build ownership and the capacity to implement actions across government and non-government agencies.

Several policy actors we interviewed (I2, I3, I5, I6, I10, I11) identified the importance of systemic governance approaches in “transforming” (I5) policy change to address problems associated with catchment water quality. They spoke of natural resource management approaches that introduced leadership at the regional scale and advocated for outcomes by:

…bringing people together, [achieving] consensus, mobilizing effort, and working fundamentally closely with state-federal frameworks (I5).

These collaborative regional approaches provided “power to establish positive and partnership-based regional solutions” (I5), helping stakeholders to implement on-farm management practices and improving the quality of catchment water entering the GBR. Policy actors were able to scale-up regional NRM approaches and take leadership in developing reef-wide water policy. An alliance between regional bodies, agricultural industries (cane, horticulture, grazing), and the World Wildlife Fund helped broker new supportive reef programs with the Australian and Queensland Governments (i.e., Reef Rescue, Paddock to Reef Monitoring Program) to take action under the Reef Water Quality Protection Plan (State of Queensland and Commonwealth of Australia 2003).

As one policy actor reflected, the partnership began through discussions about coordinating monitoring “because that's a safe place for people to agree on objectives” (I10). Diverse policy actors who had held conflicting positions in the past used conversations about monitoring to share information, develop trust, and build relationships. Though this process provided space for formal meetings, it was the informal interactions around these meetings that allowed relationships to be built and deliberations to occur:

We'd have these formal meetings where everybody would discuss very interesting things apart from what are they going to do about [water quality]. And then you'd have these little hallway quiet conservations about, ‘Oh, you can do this…’, ‘You should talk sense out of it all..’, ‘Be good if you…’, ‘You know – I didn't tell you about…’. And that's often where the real business [happens] and the formal process is a bit of theatre (I10).

The policy actors involved described that they did not have formal positions of leadership, and no individual organization or individual had responsibility to take action. Rather, “…different people in different places played big roles” (I5). Key policy actors assumed leadership and worked together in a collaborative alliance to develop the science base. The relationships they built with the regions, agricultural industries, the World Wildlife Fund, and with the Queensland and Australian governments allowed critical conversations to be had. This led to the design of an implementation framework in the 2009 Reef Water Quality Protection Plan (State of Queensland and Commonwealth of Australia 2003) and these workable strategies allowed agricultural management practices to be improved.

Policy actors did not have power in a hierarchical sense; however, they all had agency—the ability to “affect positive change” (I5). This takes partnered leadership, absolute passion, and humility:

Understand you're not the solution to everything. You've got to bring everyone with you. You need to help build consensus, you build on people's lives, and you bring the science in. And all those things eventually lead to significant political change (I5).

Like policy actors working in partnership with the community, building collaborative partnerships and influence across governance systems is time consuming. Partnerships need to be built well in advance, because when opportunities to take action occur: “…you then don't have time to go away and develop proposals or strategies or relationships. You have to have those things in place” (I10).

Using contested science in contested policy processes

Policy actors in the GBR delivered strategic evaluations and impact assessments and worked with stakeholders to develop, introduce, and implement strategic plans, zoning plans, plans of management, industry best practice programs, and regulations in highly contested political contexts involving “a lot of uncertainty and imprecise science” (I4). Though science has been a powerful driver of public opinion and reef policy activism in the GBR, leading to its inclusion in the World Heritage List, the role of science in the GBR is complex and contested. Policy actors identified that incomplete and uncertain science linked to activist agendas can create obstacles for stakeholder involvement. This was highlighted as a problem for regulating fisheries access, agricultural practice improvement, and issuing permits (I4, I10).

To address problems regarding incomplete science and uncertainty, policy actors democratize science using “consensus statements” (I6) and “synthesis” (I10) products to make scientific information available and accessible to local stakeholders. Policy actors adjudicate conflict by drawing together key areas of agreement in the peer reviewed science literature to provide a basis for policy negotiation with local stakeholders. Policy actors used this method to negotiate the Reef Rescue Program (Eberhard Consulting 2011) and set targets for agricultural practice improvements (I10). Policy actors also used this method to build a case for introducing on-farm regulations in the GBR catchment (I6). In this case, information was sent out to cane and grazing industry stakeholders “…based on the science consensus. It was very soundly worked up and presented very coherently to the affected parties, particularly to the cane growers and the graziers and to the peak groups. Because the evidence was so well documented, so well published, and so well publicized, it was difficult for people to come in and say ‘this is wrong’.” (I6).

When negotiating outcomes, policy actors also mediate how science is used in policy making. Policy actors (I9, I1, I4, I6) learned that, when enacting policy within parliamentary processes (i.e., introducing catchment or marine regulations, such as plans of management, zoning, on-farm requirements), information that is uncertain, contestable, and not widely supported can exacerbate conflict. This can disempower policy actors seeking to negotiate compromise agreements and if not addressed, scientific uncertainties can be fatal for taking action: “You only have to have one fact that you've put up be knocked down and proven to be incorrect. It could undermine your whole exercise; at least, the competence of the minister” (I9). Another reflected that the approach they took to introduce the on-farm reef regulations was based on “very sharp lessons” (I6) from the spread of misinformation during the koala habitat planning process, which resulted in the loss of our “…minister and three seats in the next election” (I6) and “…we were never able to recover that poor start” (I6). The policy actor learned from this experience how critically important it was to democratize the science base by tailoring scientific information to different stakeholder interests and making this accessible to all stakeholders very early in the policy process. Policy actors found that this was essential to prevent stakeholders from forming negative attitudes and to reduce the potential that fear campaigns could take hold politically (I6, I1).

Though policy actors mediate science to engage stakeholders and reach policy compromise (to negotiate agreements), they can draw criticism from the science community, particularly if outcomes do not reflect scientific views. In two different policy processes—the 25 Year Strategic Plan (Great Barrier Reef Marine Park Authority 1994) and the high profile decision to allow dredge material to be disposed of in the GBR—policy actors reported contesting with scientists over whether scientists should “independently” (I4) make policy decisions to set targets and allocate permits. As one policy actor explained, “Scientists are far more deeply embedded in the socio-political system than they think they are” (I4), and achieving action in contested contexts requires more than just “good science” (I4). Policy actors handle this differently depending on the policy context, by bringing science into community and multi-stakeholder dialogic spaces (I10, I7) on one hand, and by limiting the involvement of science to balance stakeholder power on the other (I4).

Using discretion in sluggish⁴ bureaucratic contexts

Above, we reported on how policy actors use discretion to take action and make progress against water quality outcomes in the GBR, regardless of their role or leadership in a hierarchical sense (I5). Policy actors also encountered the discretion of others as barriers to taking action, particularly in large “sluggish” government bureaucracies. One policy actor described negative attitudes toward change and resistence to taking action to overcome governance barriers within government itself. As this policy actor described, some government officials are more concerned about protecting their positions and lack the will to resolve policy conflicts, preferring to prevent “…their budget blowing out, or they don't want their lives to be hard” (I3).

In the GBR catchment, reducing water quality pollution entering the GBR from agricultural sources is primarly addressed through community-based regional natural resource management solutions. Though this provides opportunities to work regionally within the GBR to tailor targets and action to different agricultural landholders and has had an impact on diffuse pollution (Waterhouse et al. 2016), maintaining government support for community-led action has been difficult and inconsistent over time (Dale et al. 2016). To provide support to community based action, policy actors had to overcome barriers posed by senior and mid level government officials. They learned that the way to do this was by establishing collaborative processes outside of government to gain broad agreement about regional action and use this to create a space for action within government (I3). A policy actor learned that by creating a collaboration, stakeholders external to the government (e.g., local government and regional NRM bodies) would engage the minister in discussions about regional natural resource management plans and their implementation (I3). The minister requested departmental briefings, which engaged senior government officials and limited the ability of mid-level officials to stymie action. This actor learned that collaborative spaces created opportunities for lower level staff to provide written information, advice, and recommendations direct to ministers and senior government officials, leading to agreed decisions (I3). This gave momentum to policy initiatives, and it was also difficult for mid-level officials to stop or slow progress. The fast-moving policy environment actually created new lines of communication between policy actors, ministers, and senior officials, creating a safeguard against efforts by mid-level officials to simplify their workload (I3).

Establishing direct communication with ministers and senior officials helps policy actors to anticipate the extent to which they can take discretionary action to negotiate an outcome:

If your only accountability is back to your director-general and your minister and you've got a fairly clear idea of what they are thinking, then you can implement that discretion with a high level of confidence (I6).

Collaborative action built the power of regional stakeholders and limited the discretion of government executives. Taking this kind of action involved personal judgment and risk (I3, I7, I4), and policy actors addressed this in different ways. Some decided that they just “wanted to make a difference” (I3, I7, I6), while others relied on informal networks to give comfort and help manage disappointment (I2, I13, I6).

Mediating tensions about program assessment

Another coping mechanism explored in the interviews related to tensions over the framing and outcomes of government program assessment. Government bureaucrats are substantial custodians of data relevant to judging the efficacy of GBR management arrangements, have power over evaluative processes, and can create barriers to avoid negative performance assessments that recommend changes to decision-making arrangements, resourcing, and workloads. One policy actor was involved in conducting an evaluation to inform the strategic assessment in response to a request by UNESCO that Australia undertake a comprehensive strategic assessment of the GBR. This individual reported the evaluation was source of heated contestation “…with some of the bureaucrats from the Commonwealth Government who were coming from a different perspective” (I8). As this policy actor described, some government officials wanted the assessment to be influenced towards particular outcomes, for example, “We had a staff member who was pressuring us to make the assessment or their area of responsibility look better” (I8) at the same time another official argued “…that things were worse than we assessed” (I8).

To manage tensions over program assessment, a policy actor used a framework that had credibility in another setting and tailored it to the GBR (I8). Government staff were engaged collaboratively through interviews, workshop discussions, and presentations of draft findings. They helped to assemble the evidence indicators and provide feedback (I8). To balance staff input, findings were benchmarked against national and international cases to strengthen the assessment and limit the extent to which it could be influenced by individual agendas. The rigor of the assessment model and processes that were used meant that the senior executives were “…able to take on board some pretty bad news, and use that to develop responses when most governments would've tried to bury it” (I8).

Discussion

The first application of street-level bureaucracy and pragmatic planning theory in the GBR finds that policy actors face complex governance problems when taking action in a large scale MPA. Policy actors routinely dealt with diverse barriers emanating from conflicting stakeholder values, conflicting perceptions of stakeholder roles, lack of trusted relationships, lack of stakeholder support for MPA policy outcomes, fragmented and disjointed decision making across scales, uncertain and contested science, resistance to change within bureaucracies, dealing with multiple accountabilities (self, stakeholders, officials), and dealing with self-interested attempts to exhort influence and control. Addressing these barriers depends on the individual will of policy actors to take action (see Lipsky 1980; McLaughlin 1987; Loyens and Maesschalck 2010). Policy actors in the GBR took action to overcome governance barriers in three ways: 1) by using discretion, 2) through participation and collaboration strategies; and 3) by improvising.

First, policy actors used their discretion to create spaces for deliberation and to manage information processes in the face of power and conflict (see Forester 1987, 1999). They excercised judgment to resolve conflicts and contextualize policy in the face of conflicts. Though it might be interpreted that policy actors were not accountable to MPA goals, objectives, standards, or policy frameworks, the results do not indicate this. To advance the outstanding universal value of the GBR, policy actors made decisions to address barriers, mediate imbalances, and engage stakeholders.

Second, participation and collaboration strategies helped policy actors to overcome barriers and created opportunities for action. Policy actors used collaboration to democratize and apply science and GBR policy in the context of specific problem situations. For instance, policy actors addressed water quality problems with agricultural stakeholders by synthesizing science and deliberating through collaborative processes (I4, I6, I10). Policy actors advanced stewardship programs by engaging with industry in deliberative processes to trial, evaluate, and adapt policy (I9).

Collaborative processes helped policy actors to mediate conflicts and address power imbalances. This was powerful in enabling leadership to emerge at local and regional scales, and this leadership helped pioneer innovative reef water protection plan programs (i.e., Reef Rescue, Reef Regulations, Paddock to Reef) (I2, I3, I5, I6, I10, I13). It also helped large, slow-moving, and multi-layered government departments to take action (I3, I4, I7, I8). Policy actors also used collaborative processes to negotiate compromises with stakeholders. They did this to introduce new policy with stakeholder support for implementation. This was critical for securing the representative zoning (I1), the plans of management (I9), and strategic plans (I2, I7) within parliament.

Collaboration also helped policy actors to cope with the inherently political policy environment of the GBR. Unlike street-level bureaucracy coping mechanisms that aim to reduce complexity and workloads and avoid taking action (Weatherly and Lipsky 1977), GBR policy actors used collaboration to help make complex decisions, secure resources, and mobilize effort. Collaborative relationships helped policy actors to take risks and handle disappointment when things did not go to plan (I2, I6, I13). Multi-stakeholder collaboratives overcame efforts to stymie action (I5, I6).

Third, the results highlight that GBR policy actors improvised in astute ways to overcome governance barriers and advance marine conservation outcomes (see Forester 1987, 1989, 1999; Laws and Forester 2015). Policy actors alternated between roles as democratizers, mediators, and negotiators to address specific dilemmas in specific contexts. They improvised by balancing participation to prevent individual agendas from usurping deliberation, assessment, and decision processes (I8, I11, I9). Policy actors improvised by building powerful systemic processes to allow stakeholders to work together and negotiate outcomes (I5, I6 I10). Policy actors also improvised by adapting their roles to fit the socio-economic, political, and governance situations at hand (I8). Policy actors improvisational skills were improved by experience in multiple government and non-government roles (I5, I6, I10). This helped them understand how to build relationships with diverse stakeholders (I1, I6, I9), how to balance power (I8, I9), and how to bring diverse stakeholders along (I1, I5, I9).

Conclusion

The governance complexities of LSMPAs pose particular challenges for advancing marine conservation outcomes. LSMPAs involve multiple-stakeholders, fragmented decision-making, scientific uncertainty, and inherent conflict. It is little wonder that the governability of LSMPAs has been called into question. However, the growing trend toward LSMPAs means that governments and conservation stakeholders must find ways to overcome governance barriers and adress diverse conflicts.

The key findings from our application of street-level bureaucracy and pragmatic planning theory in the GBR LSMPA reveal important lessons about how policy actors can take action to overcome governance problems and overcome conflict. Effective policy actors were skilled improvisers who adapted their action to advance policy in different contexts. They: 1) used discretion to create spaces for action, 2) employed collaborative strategies and facilitative leadership to develop trusted relationships, resolve conflict, and negotiate agreements; and 3) adapted their role continually to respond to different types of conflict and problems at hand. We contend that finding and supporting policy actors with good improvisational skills, and a will to take action, could help address complex governance problems in LSMPAs. This could be a way to contextualize marine policy locally across large, complex, and contested LSMPA governance systems. Further exploration of improvisation in other MPAs and LSMPAs will provide a deeper understanding about how improvisation can be used to overcome governance barriers and advance marine conservation policy.

Acknowledgments

The authors would like to thank John Forester for providing guidance to this research and to Patrick Christie, Rebecca Gruby, Noella Gray, and the anonymous reviewers for their constructive feedback on the manuscript.

Notes

1. Regional natural resource management (NRM) organizations are community based non-statutory organizations that deliver on-the-ground sustainability action (Vella et al 2015)

2. Agricultural peak bodies are agriculture industry associations representing farmers and agriculture in Australia.

3. “Disjointed decision structures” refers to uncoordinated, poorly integrated decision-making processes that often encompass several organizations.

4. “Sluggish” refers to government contexts where impediments and excesses in formal structure and relationships translate to slow-moving or inactive procedures, thus affecting the achievement of policy and management outcomes.