Narrative policy analysis and decision making

The goal of this article is to examine the utility of narrative policy analysis in integrating public involvement in decision making. This issue is at the heart of deliberative democracy and how public preferences are given due consideration. This can be achieved through the juxtaposition of lay and expert opinion and consideration of how power differentials influence the expression of such opinion through examination of their discourse expressed in narrative form. This exercise aims to show how narrative policy analysis can contribute to the enterprise of deliberative and discursive democracy, as expressed by Dryzek (1990, 2002). The article aims to show how discursive policy analysis can facilitate the fundamental enterprise of facilitating the expression of agreement and disagreement within deliberative democracy. Guttman and Thompson (1996, p. 3) argue that “citizens should try to accommodate the moral convictions of their opponents to the greatest extent possible, without compromising their own moral convictions”. It will be argued that narrative policy analysis achieves this goal.

The integration of public preferences with policy analysis has been addressed by some writers. For example, Walters et al. (2000) describe a process of discovery, education, measurement, persuasion and legitimisation. Discovery is important when people do not have well formed opinions on new issues and helps develop a common language for discussing the criteria of evaluation. Education of stakeholders might be required to deal with technical issues. Public participation is most warranted when issues are poorly structured. Narrative policy analysis is of particular use in analysing the structure of issues when there is controversy and delineating the language in use by protagonists.

Narrative policy analysis consists of the identification of narratives which describe policy dilemmas. Roe (1994) argues that narrative policy analysis is useful when policy issues are uncertain, complex and polarised. The process begins with the identification of dominant narratives, which express uncertainty and complexity and non-stories and counter-stories, which are contrary to the dominant narrative. The policy analyst then generates a meta-narrative derived from the comparison of stories, non-stories and counter-stories. Meta-narratives are “policy narratives in a controversy that embrace, however temporarily, the major oppositions to a controversy without in the process slighting any of that opposition” (Roe 1994, p. 52). The meta-narrative may make the issues more amenable to decision making. In a policy controversy the formulation of a meta-narrative is an alternative to seeking consensus and provides an altogether different story, which is amenable to policy intervention. “The meta-narrative is, in short, the candidate for a new policy narrative that underwrites and stabilizes the assumptions for decision making on an issue whose current policy narratives are so conflicting as to paralyze decision making” (Roe 1994, p. 4). Roe maintains that policy narratives are not insignificant in situations which are not divisive, uncertain or complex. They will always be found stabilising the assumptions made in decision making. This schema of narrative analysis is applied to drinking water quality policy and planning scenarios. The first scenario was chosen because it suggests how the development of a meta-narrative could have been facilitated through public debate about a controversy. This situation was of political importance to the water industry and to interest groups. The second scenario provides an example of the development of a meta-narrative through the juxtaposition of opposing views about planning options. It demonstrates how a meta-narrative can evolve when there is open debate.

A drinking water treatment policy scenario

In order to meet the needs of an expanding population, in the context of a drought, a local council in Toowoomba, south east Queensland, Australia, had proposed the addition of a small percentage of recycled water to the potable supply. Their interest in this situation was ostensibly the assurance of a long-term supply of potable water. This scenario was chosen to suggest how not facilitating open debate between protagonists may have contributed to a lack of resolution. Engineers and social scientists alike were awaiting the outcome of this referendum as it represented a test of public opinion on water reuse for potable purposes. There were several towns in Australia that were considering water reuse for potable purposes dependent upon the outcome. The federal government had agreed to assist with funding if the public supported this use of recycled water. For the dominant narrative this case study is derived from telephone communication with the Council marketing manager (Marketing-manager 2006, personal communication), telephone communication with a Council administrative officer (Council-Officer 2006, personal communication), Council websites (Toowoomba-City-Council 2009), Federal government websites (Turnbull 2006) and for the counter-narrative from newspaper articles and public websites (e.g. Smith 2006), which had assumed an important role in public discourse as it became one of the main means by which the public could express its opinions. These different information sources provided triangulation of the data sources for each narrative (Yin 2009).

The council embarked on a vigorous public relations campaign in support of recycled water and shied away from public engagement (Marketing-manager 2006, personal communication). ‘The campaign’, as it was known internally, was designed to convince the public of the benefit of recycled water, rather than air concerns of the public. The dominant narrative was that recycled water could be safely used for drinking. The rhetoric of the council was that this was a war against irrational forces within the public who opposed the use of recycled water for drinking.

Experts with engineering and academic backgrounds were utilised from technical areas to strengthen the case for recycled water. They were not readily made available to the public to answer concerns about contamination from recycled water. The public was effectively quarantined from open dialogue with Council’s experts. The Council shied away from the use of experts in public participation method (Marketing-manager 2006, personal communication). In response to this approach, a public interest group developed to support the ‘no’ vote. Council response was to spend more on marketing the ‘yes’ vote. The final outcome of the referendum was a defeat of the ‘yes’ vote.

There was meagre open dialogue on the issue of how to augment or constrain the potable water supply. Proponents of the ‘yes’ and ‘no’ votes did not engage in constructive dialogue on the issue of a long-term potable supply. The ‘no vote’ public was construed as irrational and unreasonable and not amenable to the arguments of experts. Their interest in this situation was the avoidance of contamination in the water supply. There was however consideration of recycled water (Smith 2006). Although there was meagre support for drinking recycled water there was consideration of its use for non-potable purposes. The Federal minister stated that public participation had been upheld (Turnbull 2006) because the public’s wishes were heeded, however the debate on which the decision was based on was limited and divisive. Council officers lamented the lack of time for engaging in public education leading up to the referendum (Council-Officer 2006, personal communication). This scenario provides an example of how a meta-narrative did not evolve because of a lack of dialogue between lay publics and experts. The power in this situation was closely guarded by the local council and they were not prepared to share it with the public. This is reflected in the divisive narratives which were prevalent in the community. Roe’s treatment of unequal power relations, which are reflected in the narratives prevalent in the town, are dealt with later in this article. In contrast with this scenario is that of a local manager of a utility who was prepared to promote open dialogue within public participation on water treatment and where it will be argued, a meta-narrative evolved. This planning issue will be dealt with in more detail.

A drinking water treatment planning scenario

The quality of potable water in the Illawarra region of New South Wales, Australia, had been problematic for the public and hence for the water utility for considerable time. The policy of the state government and its water utility focussed on a decision to improve potable water quality for all of the reservoirs and reticulation systems under the control of the government utility rather than specific concerns about the Illawarra area. The utility had resolved to improve the quality of potable water and had initially decided to develop a centralised water treatment and filtration system. The main form of consultation adopted was the production of an Environmental Impact Statement (EIS), which involved presentation of information on a proposed local scheme with subsequent evaluation and analysis of feedback from specific community stakeholder groups. When the EIS was presented to the community there was considerable opposition to the proposed location for a water treatment plant. This scenario was chosen to illustrate how a meta-narrative can evolve spontaneously through interaction between protagonists. This case-study data is triangulated with the following documents. The dominant narrative is supported by the original EIS document (PPK-Consultants 1992), meetings with the engineers, and the counter-narrative is derived from meetings and correspondence with the community action group (Farmborough-Heights-Resident-Action-Group 1993, personal communication) and audio recordings of the workshops.

Following from the EIS evaluation, the utility’s dominant narrative was that the preferred site for a water treatment and filtration plant was based on a ‘good’ engineering solution for receiving the water from the reservoir, treating and filtering the water and distributing it through gravity feed to local reservoirs throughout the region. The site was owned by the utility, was readily available and was directly positioned under the reservoir. The community adjacent to the preferred site was opposed to the plant being developed in their area. Their opposition centred on whether it was safe for it to be located in close proximity to a residential community. The dominant narrative for the community (as assessed from discussion with and correspondence received from the community action group) was that the transport of chemicals through the neighbourhood and storage and use of chlorine in the water treatment process would be dangerous to the community and the construction of the plant would create land instability. Their narrative expressed general opposition to the entire project and a denial of the need for improved water quality. This denial of the need for treatment became a major focus on the community group’s campaign against the treatment plant.

Due to the considerable community opposition to the proposed site for the plant, the utility decided to embark on a public participation program in order to determine whether the public wanted improved drinking water quality and if so what was the preferred method for water quality improvement. The local manager of the utility determined that the outcome of the public consultation would determine whether the treatment plant would go ahead. The power in this situation was ostensibly given to the community. The participation process was designed to bring together the plethora of voices grappling with potable water quality problems in the region. The main issues for deliberation were whether water treatment was necessary and if it was deemed necessary, where a possible treatment plant should be located. Experts on water treatment and opponents to treatment and site locations were brought together with sectors of the community who suffered poor water quality.

The participation process consisted of two series of workshops. The first series of six workshops examined water quality problems experienced by customers. Participants were obtained from a list of people who had suffered potable water quality problems. All of the people on the list were contracted and approximately 60% responded. Sixty-one people participated in these workshops, which varied in size from 7 to 25. The dominant narrative derived from analysis of audio recordings of the dialogue of these groups was that poor water quality had created a myriad of domestic problems for many years. The second series of six workshops used the same list of participants from which 83% responded with 84 attending. The workshops explored these customers’ preferences for water quality improvement options. Workshop sizes varied from 25 to 9.

In the first stage of the second workshop each option for water quality improvement was discussed by participants. The account given by the utility in the workshop followed the technical analysis provided in the EIS. It was comprised of statements from the representative engineers that centralised water treatment would provide an effective strategy for improved water quality throughout the region and that the site, in proximity to a residential area, provided an ideal engineering and economical solution for transporting water from the reservoir to a central treatment plant and then to the entire region as the location would allow gravity feed. They discounted the other options as not providing a comprehensive and viable solution to region wide water quality problems.

A public participation consultancy group from the local University, of which the author was a member, was employed to organise the workshops and had suggested that the community interest group be represented in discussions. These consultants were politically committed to principles of deliberative democracy. Therefore, representatives from the opposing community were asked to address each workshop about their concerns. Their counter story, as told by the participants, was that centralised water treatment was unnecessary, as the utility had overstated the regional water quality problems. Their counter story for the treatment plant also portrayed the plant as creating risks for any residential area. Their solution to water quality problems, which they considered to be minimal, was that the reticulation system should be cleaned or replaced where necessary. The utility then made statements, which reinforced the dominant narrative, that water quality problems across the region were substantial and that considerable funds had been spent over a prolonged period on pipe cleaning and replacement and that this strategy had not been successful.

A turning point in the workshops occurred, when in the fourth workshop residents of the opposing community acknowledged that residents of the adjacent community were suffering severe water quality problems. This came about through hearing the full stories of the difficulties experienced by the community. The community participants initiated acceptance of the story provided by the “dirty water sufferers”. They realised that the community members were not simply pawns of the utility but had interests which needed to be considered. Participants stated that they had not realised how extreme the problem had become for some residents and hearing their concerns had convinced them of the need for centralised water treatment. This was a fundamental shift in their attitude from one of opposition to support for the project. The immediacy of hearing the opposing community’s stories and concerns about having the treatment plant located near their neighbourhood also influenced the people suffering water quality problems and they stated clearly that their preference was for the plant not to be located near any residential area. The community participants initiated this rapprochement without prompting by the facilitators. The private vote on a preferred option reflected this with 19 participants preferring centralised treatment and four participants preferring increased flushing and cleaning of pipes. When asked to vote for preferred sites for a treatment plant, none of the participants voted for the residential area, seven participants preferred an industrial area and 16 participants preferred a site adjacent to the reservoir. This voting pattern was similar to the preferences expressed in the other five workshops except that in two workshops a larger majority of participants preferred centralised treatment and preferred the industrial area as a site for a plant.

The meta-narrative, which developed within the workshop was that the treatment plant was necessary for the region but that it should not be located within close proximity to any residential community. This was a resolution of the controversy for the communities affected by poor water quality and the proposed treatment plant site. The need for regional water improvement generally had been accepted by the opposing community group rather than having this vocal and politically active group opposing any form of centralised water treatment.

The success of the process was due in part to enabling the affected parties to have a voice so that they could state their concerns in a situation where they would be heard. This was dependent on the manager’s willingness to share power with the community. The meta-narrative which evolved was in part due to providing a situation where the dominant narratives could be juxtaposed and responded by opposing parties. Without a sharing of narratives the protagonists would have probably continued to construe the situation according to their preferred narrative. This case study characterises the way in which narrative policy analysis and planning can articulate the issue of central concern in a process where there is a multiplicity of views to deal with. Narrative policy analysis guided the inclusion of the community activists in the consultations and fostered the identification, juxtaposition and disagreement between narratives and fostered the eventual evolution of the meta-narrative. The utility officials were pleased with the consultation process in that a way forward was created. The utility’s interest in centralised filtration was maintained and they were the dominant players in the scenario. This suggests that the dominant narrative was maintained; it was however significantly modified.

The role of the meta-narrative

This article focuses on the use of the meta-narrative in decision making and the usefulness of the concept to policy analysis and planning. The concept of narrative has attracted considerable attention in the field. Kaplan (1986) first wrote about criteria for good analytic narratives in policy analysis. They are characterised by truth, richness, consistency, congruency and unity; such narratives can be used to convey situations and have enhanced credibility. Scenarios are stories of future action and consequences. They can be used to describe possible implementations of programs or to share future visions. Narrative policy analysis has continued to be used in portraying a broad array of policy issues. For example, Cassiman (2006) uses narrative policy analysis to develop a counter-narrative to a welfare dependence narrative. She articulates the development of this dependency narrative and how social work has supported this dominant narrative. She then creates an alternative narrative which recognises the economic impacts of poverty and trauma.

The methodology involved in narrative policy analysis is primarily qualitative. However, attempts have been made to combine qualitative and quantitative methods in the practice of narrative policy analysis. McBeth et al. (2007) use a mixed methodology to examine the narratives in documents about the Greater Yellowstone Area. A criticism of Narrative Policy Analysis is that it is difficult to falsify as a research approach and their use of mixed method is an attempt to address this criticism. They quantify categorisation of narratives in documents and derive statistically significant differences between interest groups in their use of narratives. They note that problem definitions within narratives are strategic and are constructed as a part of political manoeuvring.

A broad use of the concept of narrative has been undertaken by Stone (2002). Stone’s contribution focuses on how narrative is used in the policy process as a tool of political strategy. Stone takes a much broader approach to metaphor than Roe, by focusing on the general use of metaphor. She discusses the use of various forms of metaphor such as the use of war analogies in policy analysis, as was evident in the case study on water recycling in Toowoomba. Metaphors can often be used to make political claims. She concurs that policy problems are often defined in narrative form and that the storyline in policy problems is often hidden. Policy stories use many literary devices that lead to a course of action. There are common metaphors in policy such as likening an institution to an organism or using machines and mechanical devices as the basis of policy metaphors. Ambiguity is an important aspect of metaphors for Stone. It enables unification on the same policy for different reasons and facilitates negotiation and compromise. The ambiguity of symbols enable collective decisions to be derived from individual strivings. In this respect ambiguity functions like a meta-narrative in that individual narratives are maintained in its construction. Roe’s work on meta-narratives is emphasised in this article as it focuses on the process of dealing with a controversy in a methodical manner.

Roe (1989) initially concentrates on the reduction of uncertainty through the use of the narrative form in policy analysis. Roe (1994) utilises the concept of meta-narrative and its usefulness has since been demonstrated in a number of studies. Garvin and Eyles (1997) analyse health policy in relation to skin cancer and the hole in the ozone layer and argue that despite scientific uncertainty a sun safety meta-narrative has evolved. The meta-narrative enables uncertainty to be managed and action to be taken. It became the dominant story in the controversy. Rather than a reliance on absolute proof there was a dependence on argumentation and persuasion. The precautionary principle was bought into play to manage the uncertain situation.

Bridgman and Barry (2002) note that a meta-narrative is like a superordinate frame that joins otherwise incompatible positions. The meta-narrative serves to distance protagonists from their original position. Thematic analysis, which focuses on metaphor can be used to resolve power dynamics within a policy dispute because, they argue, metaphors control action and ultimately change action. They discuss how a prefigured meta-narrative may constrain the examination of alternatives in a dispute. They note that when public deliberation is consistent with existing narratives about deliberation, the process will be more productive.

Harbour (2006) used hermeneutic interpretation along with narrative policy analysis to examine legislative and administrative texts about higher education in Colorado. Hermeneutic interpretation of textual data was accompanied by narrative analysis in order to synthesise textual data. Narrative policy analysis was used to portray some of the data in a new framework using literary devices and producing a narrative which created a new understanding of higher education policy. Harbour created a meta-narrative by producing a rough script based on major themes, annotations and reflections. He then identified acts and scenes and edited and moved them to appropriate places to form the meta-narrative.

van Eeten (2007) discusses how aggregated policy narratives are constructed by the analyst rather than protagonists through methods such as content analysis, network analysis or stakeholder analysis. When the narratives have been reconstructed, the analysis can be undertaken. van Eeten regards the meta-narrative as a story about the comparison of the narratives. The analyst does not search for the correct meta-narrative but the one which allows a way forward. van Eeten notes that different comparisons other than between a story and a non-story might generate a meta-narrative. He has used the Q-sort as a method of eliciting and quantifying policy narratives. van Eeten provides a useful current review of the literature on narrative policy analysis and how the methodology has evolved since Roe’s initial work.

Roe continues to practice narrative policy analysis and outlines the differences between conventional policy analysis and narrative policy analysis in Roe (2007). He focuses more on the counter-narrative but still refers to the concept of meta-narrative which he regards as a way of upholding opposing narratives without detracting from either narrative. He places emphasis on how narratives are used when issues are uncertain and complex and conflicted. Not slighting any of the oppositions is important in the process of developing a meta-narrative. The conditions under which the narrative and counter-narrative hold at the same time constitute the meta-narrative

The resolution of the water filtration planning scenario was not a consensus in that there were a range of options voted for. There was however a way forward pursued in the scenario in that the public and the utility largely agreed to centralised filtration. This is one of the purposes of a developing meta-narrative; to unlock a previously intractable situation.

Similar concepts to meta-narrative have been employed in policy analysis and planning. Gray (2004) examines the concept of frame of reference and how such frames foster or create resistance to collaboration. In some respects these frames of reference are similar to the metaphors utilised by coalitions in that they constrain or encourage the dialogue occurring between groups. They are social constructions that help each group in a conflict to make sense of the situation. Interests might be reframed to reflect a frame of reference where collaboration is encouraged. Further discussion of meta-narrative focuses on how policy and expertise can be collaboratively considered.

Public involvement in policy development and planning

In recent years the field of policy analysis and planning has been replete with enquiry into the process of public involvement. This analysis has focussed on topics such as balancing expert and public knowledge (Eden 1996) and the political interchange of diverse coalitions advocating for their preferences. In the past, public consultation has often been appended to the policy and planning process, in an incremental fashion, as political crises have demanded public involvement.

Policy is assumed to be oriented to governmental levels of dealing with an issue, which may be limited in scope and with a focus on political expediency (Eden 1996; Patton and Sawicki 1993). In contrast, planning can be regarded as operating in a more comprehensive manner, with a longer term focus; giving consideration to a wider array of alternatives and being less concerned with political pressures. The rational-comprehensive model and the incremental model are often referred in the fields of policy analysis (Doyle and Kellow 1995) and planning (Kaiser et al. 1995). Within the rational-comprehensive model (Doyle and Kellow 1995) an option chosen from a range of alternatives has various benefits and costs some of which will be public support or opposition. Public values and preferences may be calculated as a benefit or cost depending upon whether they are congruent or incongruent with an option. The cost of public rejection will be social upheaval and political damage for the elected officials. There may also be public values which are more complex than can be conveyed in simple unitary monetary terms. If public values are not incorporated in the assessment of options within the rational comprehensive model the most efficient and effective alternative may flounder politically. There are variants on the traditional rational-comprehensive model, such as Kaiser’s et al. (1995) model of community planning discourse, which advocates the need for extensive public participation in the planning process.

Lindblom’s alternative model of incremental planning was put forth as a more realistic portrayal of planning in practice (Taylor 1988). In the incremental model (Doyle and Kellow 1995) social values can be ignored and the search for policy options is limited and reliant on past experience. Small policy steps, away from existing policies, are taken in order to gauge consequences as they occur. If policy implementation does not go well, the previous policy can be reverted to. This strategy is risky if public opposition is politically damaging or the results of the policy action are irreversible, as may often be the case with environmental issues.

The narrative policy analysis method of identifying narratives and counter-narratives and developing meta-narratives is a potentially useful analytical process for planning as it provides a structure for understanding and working with the narratives encountered by planners when engaged in consultation and participation. The importance of narratives occurring in planning has been highlighted by Forester (1999). Forester recounts planners’ experiences of public participation in terms of the stories they tell of the planning process. He argues that such stories should be listened to and understood for their insights and implications for professional practice. Forester (1999, p. 29) writes that “in planning practice, these stories do particular kinds of work—descriptive work of reportage… political work of identifying friends and foes, and the play of power in support and opposition and… deliberative work of considering means and ends, values and options”. Throgmorton (1991) argues that planning is a form of persuasive and competitive storytelling of diverse and often opposing views, which are antagonistic to one another. These notions of planning stories are similar to Roe’s concept of asymmetrical and opposing policy narratives. The usefulness to planning of Roe’s approach to policy analysis of juxtaposing narratives and seeking a meta-narrative is demonstrated in the planning scenario concerning a drinking water treatment controversy. The community, mentioned in the case study, had called into question the government policy of water treatment to meet new national guidelines and was opposed to a plan to construct a treatment plant adjacent to a particular neighbourhood. The local utility manager was prepared to listen to the community’s various stories concerning water treatment. This contrasts with the lack of meaningful public participation and the evolution of a meta-narrative in a policy choice in a region considering the use of recycled water for its potable water supply.

Participatory policy analysis and planning

Traditional models of policy analysis and planning do not advocate constructive and systematic public involvement. Public preferences may be incorporated for the purposes of political expediency at a stage of policy development which precludes comprehensive public participation. More recently commentators have discussed the benefits of participatory policy analysis. Referring to Dryzek, Durning (1993) argues that participatory policy analysis requires analysts to contribute to policy deliberation in a way which challenges entrenched thinking. Participatory policy analysis purports that all affected parties to a policy decision should, through the means of discursive democracy, have a political voice and “should be heard without prejudice or advantage” (deLeon 1994, p. 88). The task of the policy analyst is to identify effected groups, educate them about the issues and extract the essence of discussion in an impartial or balanced manner. deLeon (1994) argues that involving citizens on a nominal basis through such means as administrative hearings or public surveys is insufficient; citizens need to be directly involved in the design of programs that affect them. Open policy forums (deLeon 1994), scenario workshops (Mayer 1997), and citizen panels and juries (Kathlene and Martin 1991) are some of the deliberative methods which can be utilised.

Forester’s (1999) accounts of participatory planning processes provide insight into the flexibility and creativity required to implement a deliberative approach. These methods enable participants to express diverse views and develop either a consensual view of an issue or a collective view, which maintains the diversity of opinions. Whether participants are able to achieve integration or collective expression of their diverse views depends on the effectiveness of the facilitation process used to guide these deliberations. It is also incumbent on the analyst and planner to maintain the integrity of the diversity of viewpoints in the representation of these views to decision makers.

Helling and Thomas (2001) contrast traditional and more recent methods of promoting community dialogue and note that more recent approaches emphasise bottom up approaches where the public set the agenda. Traditional approaches minimise community dialogue and public involvement may be limited to publicity and public education. They ask the question as to whether more recent collaborative methods do any better. One of the aims of this article is to demonstrate that narrative policy analysis and planning can achieve public participation in decision making as the process of participation and planning and policy analysis is unified through the juxtaposition of diverse views and the creation of a policy and planning alternative.

Consultation or participation?

In the public engagement literature the terms consultation and participation have often been used interchangeably. Public participation has been traditionally distinguished by whether public preferences are taken into account in a decision making or are merely acknowledged (Arnstein 1969). Roberts (1998) argues that involvement programs should be distinguished according to the degree to which the public will influence the final decision rather than just be consulted for their views on what the outcomes should be. Consultation without influence on the final decision is distinguished from a participation program where there is a clear commitment to participatory democracy. Participation requires a different policy process to the situation where public preferences will merely be taken into consideration. It is argued that narrative policy analysis is of particular use when there is a commitment to upholding public preferences in a decision. The question of whether public preferences are taken into account in a decision is dependent upon the commitment of decision makers. A clearer conceptual analysis of this commitment is undertaken with reference to Goodin’s (1993) work on democracy and paternalism.

Goodin (1993) states that democracy is a matter of respecting peoples’ preferences and sets up a typology of how preferences are treated directly or indirectly and reflectively or unreflectively. The fullest extent of participation in decision making involves public preferences being directly incorporated into the policy-making process in an unreflective manner without qualification and is termed populist democracy. Referenda are an example of direct populist democracy but can be expensive, time consuming and time lags may make the process unwieldy. There are alternatives such as sample surveys, which have a referendum structure (McDaniels 1996).

Direct and reflective respect for preferences represents deliberative democracy and characterises the common practice of public participation where public preferences are incorporated into a decision but are rarely accepted without some form of modification or compromise.

Democratic elitism involves indirect and reflective respect for public preferences and epitomises consultation programs where information about public interests is often required to strategically manage public reaction to a development. Goodin (1993) regards this type of public involvement as paternalistic. Attempts might be made to justify such paternalism through an argument that decision making requires the comprehension of complex technical information. Rather than fostering paternalism, participation practitioners and agencies responsible for development projects can educate the public on pertinent issues and foster participatory relationships between technical experts and the public in order to assist understanding of complex issues.

Indirect democracy may be characterised as traditional democracy whereby voters indicate their preferences for a set of policies through their election of representatives (Goodin 1993). This means that direct expression of preferences is limited and citizens are dependent upon their representatives’ linkage with the public to understand their actual preferences and the determination of that representative to be accountable to the public.

Goodin’s distinctions of direct and indirect and reflective and unreflective provide a basis for clarifying the type of public involvement being sought in a policy and planning process. Public involvement practitioners can utilise these distinctions when formulating an involvement program so that the objectives of the program are clearly articulated and the methodologies utilised articulate with the level of involvement in decision making. Narrative policy analysis is of particular use when there is a commitment to direct and reflective upholding of public preferences as it provides a process for managing diverse views. The case study on water recycling can be characterised as indirect and unreflective as there was no immediate incorporation of public preferences into the policy process. The situation with water filtration was one of direct and reflective respect for public preferences. Narrative policy analysis and planning is suitable for deliberative democracy where the goal is not necessarily agreement and consensus but discussion and agreeing to differ.

Participatory expertise

Sharing of expertise is critical to participatory policy analysis and planning. Government reluctance to incorporate public preferences directly into policy has been attributed to a tendency for governments to rely on expert opinion in the development of public policy. Eden (1996) notes that in the UK that responsibility for implementation of Agenda 21 was assigned to the public, whereas responsibility for decision making in policy development was taken by the government with insufficient public participation, in favour of a reliance on scientific accounts. Extending science is required and involves the use of local knowledge grounded in experience and observation, contextual knowledge which links data collection and interpretation and knowledge which relates practical actions to situations. Publics also make reference to morals, ethics, culture and behavioural issues, which are not considered in scientific analyses. Such issues are critical to policy analysis and planning and are vital to the public and ultimately to governments, which must contend with public sentiments which have not been considered by scientists.

Anderson (1993) argues that scientists have in the past been used to bolster power elites and now need to form collaborative relationships with citizen groups in order to facilitate public understanding of technical issues. The scientist needs to help the public explore the assumptions, biases and generalities hidden behind expert opinion. This participatory approach to research enables an interaction between public values and scientific knowledge which is used in the decision-making process. It assists the process of decision making if scientists are able to work with public values and address public concerns in their provision of information.

Local knowledge incorporated into policy may temper expert knowledge by emphasising the uncertainty and indeterminacies of expert knowledge (López Cerezo and González García 1996). This can create a more cautious approach to decision making. The inclusion of local knowledge also prevents political manipulation of public opinion. Local knowledge provides useful information about the social system and cultural perspectives and physical environment in which policy is going to be developed and new perspectives on unexpected social and environmental impacts of a policy.

Fischer (1993) also suggests the need for the democratisation of policy expertise. In order for public views to be melded with scientific management information there must be the opportunity for interaction with experts on a participatory basis. This will provide education for the public concerning scientific knowledge and will educate the scientist on public concerns, which may not be addressed in the limited methodological arena of science. This process involves the collaborative exploration of the assumptions and biases of expert opinion. Fischer suggests that discussions between expert and citizen need to be structured so that dialogue is constructive.

Narrative policy analysis and planning has the potential to incorporate expert and local knowledge in a participatory planning process. The analytical process treats expert and local knowledge in the same way it would any narrative. Expert knowledge is likely to be part of the dominant narrative in a controversy. The development of a meta-narrative may see a rapprochement between expert and local knowledge.

The use of expert information is contrasted in the two case studies that began this article. In the policy scenario concerning recycled water, expert information was reified for the public through the pronouncement by the council of providing ‘the scientific facts’ in presentations by these experts in public forums. There was little opportunity for the expert information to be considered as another narrative in the controversy. In contrast, local engineers who attended the second set of workshops, in the drinking water treatment scenario, participated in the same manner as the citizens in the way that they provided explanation of technical information. Their account of the drinking water treatment options was treated as another narrative in the interplay of narratives. The public was able to question and accept or reject the expert information provided. The engineers eventually appreciated the concerns of the community and their local knowledge concerning the development.

Narrative policy analysis and planning as an integrative approach

Narrative policy analysis and planning has the potential to integrate these concerns about level of participation and sharing of expertise and the integration of participation with policy and planning.

The process of narrative policy analysis requires that a plurality of voices be listened to including those that are marginalised. If this is done it is more likely that a meta-narrative will become apparent. All parties need to have equal access to resources such as detailed information, particularly in a controversy of high uncertainty. Differential access to information creates asymmetrical stories and hegemony (Roe 1989). In a controversy where one narrative is coherent and the counter-narrative takes the form of a critique, ambiguity and uncertainty are increased. This may lead to the development of a meta-narrative, which highlights uncertainty and risk. Roe (1989) demonstrates how in a controversy, a proponent view with a complete argument and sub-plots can seem more credible than an opposing view when it is little more than critiques of the other without a viable and complete alternative solution. If this opposing view had provided more complete arguments it may have been more persuasive. In the Toowoomba water recycling controversy experts and lay public had little opportunity for constructive interaction but the no vote was bolstered by a privately funded information campaign. Power in this situation was equalised through the ‘no vote campaign’ on the hazards of drinking recycled water.

Roe maintains that narrative policy analysis enables power and politics to be analysed in uncertain and complex controversies. The presence of asymmetrical narratives indicate that unequal power relations are being worked through. It is important for the analyst to recognise when unequal power relations are occurring so that the views of parties who may be subordinate in such relations are not obscured by the completeness or complexity of the major view. “Unequal power relations work themselves out through the competition and opposition of stories, storytelling, and other policy narratives that get people to change their own stories when conditions are complicated, full of unknowns, and divisive in the extreme” (1994, pp. 13–14). In the Toowoomba situation the no campaign mounted a vigorous assault on the safety of drinking recycled water and that equalised the competition between narratives. Meta-narratives, on the other hand, do not stifle the difference between opposing views in a controversy. They do not promote consensus or agreement but rather a different agenda, which allows opposing parties to move on. This is what occurred in the water treatment controversy where an alternative solution to the situation evolved.

Narrative policy analysis and public participation

Fischer (2003) argues that Roe’s meta-narrative policy analysis is compatible with participatory democracy but that the implications for the latter are not drawn out by Roe. Fischer points out that narrative policy analysis is useful for integrating public views on issues by juxtaposing pubic narratives on issues with policy narratives that are created to make sense of an issue. It can move disputants beyond policy impasses. Fischer (2003) is critical of Roe’s narrative policy analysis approach because a meta-narrative is usually constructed by a policy analyst and does not rely on the participation of actors involved in a controversy. On the contrary if opposing parties in a controversy jointly deliberate on an issue, they may intentionally or unintentionally develop a meta-narrative in their interaction, which can then be used by policy analysts or planners to promote equitable or constructive decisions. The process of competition and opposition of stories, as described by Roe, is critical to the deliberate or spontaneous development of such meta-narratives. This is what is argued, happened in the scenario concerning regional water treatment. These criticisms and differences between Roe and Fischer have been summarised by van Eeten (2007).

The competitive telling of stories by opposing parties is more likely to lead to a meta-narrative if the deliberative process between participants is facilitated so that different viewpoints are heard, particularly those of parties who may be in a subordinate position in a controversy. Public participation practice is a useful source of methods for facilitating this process. A wide variety of facilitation techniques have been developed in this field for eliciting public preferences and integrating divergent views (Creighton 2005). Narrative policy analysis has commonalities with public participation method in that both approaches encourage the expression of a diversity of views in a controversy. The resolution of controversy is dependent upon the opportunity for opposing parties to participate in debate in order for consensus to occur or for the development of understanding and mutual respect for different viewpoints. The provision of local and expert knowledge on an issue is critical to this process. The views of experts are likely to be represented in the dominant narrative as the proponent is likely to have the resources for completing this narrative with technical complexity. Local knowledge may be represented in the counter story presented by a community. Narrative policy analysis adds to the methods of public participation by suggesting that the development of a meta-narrative is dependent upon the views of all parties being expressed without distortion and with respect for diversity. This is an alternative to the somewhat idealistic goal of producing consensus in a public participation process. The development of the meta-narrative also enables uncertainty and risk to be exposed and recognised. The provision of resources enabling access to information is critical to avoiding inequity and the prevention of an adequate meta-narrative being developed. This process is eminently suitable for that characterised by deliberative democracy.

An important benefit of the narrative policy analysis approach for public participation is that the discourse of decision makers in a controversy can be readily included in the analysis. This can occur because their narrative of the issues at stake often forms the basis for the development of counter-stories, which may be expressed by public groups. As the decision makers’ narrative may be represented by expert opinion, employed to propound their view, their narrative is more likely to be the well formed and argued version of the controversy from which counter stories are juxtaposed. The development of a meta-narrative is then formed from the reciprocal expression of the decision makers’ and the publics’ stories on a controversy. As Roe suggests, oppositional stories might change as they are competitively expressed, and this might facilitate the development of a new policy and planning direction. The inclusion and juxtaposition of the decision makers’ perspectives with the publics’ perspectives is therefore more likely to lead to the development of a meta-narrative which provides a way forward in a policy and planning dilemma as the direction taken might be mutually beneficial and therefore more likely to succeed politically.

The type of interaction and involvement engendered by the narrative policy analysis and planning approach is more likely to promote direct and reflective respect for public preferences. This is likely to occur even when the nature of public involvement is indirect and unreflective as bureaucratic decision makers can be brought into the process of expressing policy positions and counter positions. This is due to the potential for the competitive telling of stories to lead to changes in perspectives and appreciation of others’ perspectives.

Narrative policy analysis and planning as direct and reflective

Narrative policy analysis and planning is an approach which facilitates the expression of views from divergent parties in a controversy and provides a method of analysis which can suggest a way forward in a dilemma. The process allows for the juxtaposition of expert and local knowledge as the views of experts and local participants are included in the stories and counter stories. If reporting on the process retains the discursive context of the stories and counter stories expressed by participants the possibility of preferences being taken out of context may be minimised. Reporting may be subject to distortion and bias but this can be minimised if the narratives and meta-narratives developed in the process are checked for validity and accuracy by participants. In the planning case study reported, this process was undertaken with the initially opposing community who provided written feedback that their voice had been heard and their concerns heeded.

One advantage of this process was that participants determined the nature of the discourse and brought issues determined by their own concerns to the participative forum. Bringing people together to debate an issue facilitates the process of facilitating disagreement within democratic procedures, without slighting the viewpoints of protagonists (Guttman and Thompson 1996). In this process participants’ preferences may develop and form part of the rhetorical discourse they contribute to. Rationalist and structured methods are more dependent on an analyst or facilitator for processing and this may limit the extent to which participants’ concerns are addressed and evolve through interaction. Such methods may be efficient but may also stifle discourse and public debate of a policy and planning issue. Their systematic nature may appeal to public officials and decision makers as they are likely to provide succinct and ranked alternatives. The method of narrative policy analysis and planning may prove in some situations to be advantageous by providing participants with the freedom to have their say.

Does this process make a contribution to the policy sciences? In Pielke’s (2004) consideration of this field, narrative policy analysis and planning could be regarded as a systematic approach to analyzing narrative which occurs spontaneously in a policy and planning process. It casts order on the process of analysing narratives in practical decision making. It can be embellished with social science methodology to improve validity and reliability, as has been done by McBeth et al. (2007). It provides a policy scientist with an orderly methodology with which to analyze the plethora of dialogue occurring in a controversy. It is ideally suited to the quest for democracy which epitomises the policy sciences (Farr et al. 2008).