John Ive and Doug Cocks

CSIRO Wildlife and Ecology

GPO BOX 284, CANBERRA, ACT, 2601, AUSTRALIA

Prolonged public controversy continues over how the dwindling temperate hardwood forests of eastern Australia should be used. The basic disagreement is over the proportion of forests to allocate to productive uses based on some form of harvesting (intensive harvesting for pulpwood remains particularly controversial) versus protective uses based on conservation and various low-impact uses which minimally disturb (a) the recreational and aesthetic and (b) the natural functions (such as gene conservation, gas exchange, water quality maintenance) of forest areas. In Burton’s¹ terminology, struggles over forests and other natural resources are disputes rather than conflicts.

The proportions and pattern in which forests are allocated between productive and protective uses are ultimately set by political decisions. Unfortunately, there is no widely accepted ‘technical’ criterion for judging whether use rights or property rights have been ‘correctly’ allocated between stakeholders. In a democratic pluralist society the pragmatic test of a good decision is the extent to which society collectively accepts the decision as legitimate. Such general acceptance implies that the diverse stakeholders interested in that decision also accept it, perhaps grudgingly, as legitimate.

This, in turn, requires that the allocation process be recognised as genuinely attempting to satisfy collectively the demands of various interests (including future generations) in a comprehensive, equitable and efficient manner.

In addition, there is a public demand for procedural legitimacy or ‘due process’. Here, a prime criterion is that stakeholders play an active and significant role in the allocation process. Until recently, this has meant public participation in decision-making on resource allocation matters, e.g. by seeking submissions. Now, the concept of procedural legitimacy encompasses either direct or mediated negotiations between stakeholders. Mediated negotiations use a mediator, that is, a non-stakeholder who actively and impartially assists stakeholders reach agreement on partitioning the resources at issue between them.

This note summarises the construction and demonstration of a mediation and negotiation support system for helping stakeholders and, ultimately, governments produce legitimate decisions about the future use of extensive tracts of native forests. Further details of how the approach was implemented and trialed at a workshop attended by stakeholders representing conservation and timber industry interests have been published previously^2,3.

Wildlife and Ecology in Australia's Commonwealth Scientific and Industrial Research Organisation (CSIRO) has considerable experience in developing and applying procedural theories and computer systems for helping an individual, or an agency, with a putatively integrated set of values to produce land use plans. The SIRO-Plan procedural theory or ‘recipe’⁴ and the LUPIS software package⁵ are the Division's main products in this area. Together they have been used in some thirty studies in Australia and elsewhere covering many resource allocation problems from marine to agricultural situations.

The need in forest allocation (and indeed other natural resource) disputes is for tools that will support multi-party mediation and negotiation, rather than single-party plan making. SIRO-Plan and LUPIS were adapted for this purpose. The resultant product was named SIRO-MED, short for CSIRO Mediation and Negotiation Support System.

Concurrently, a 3500 sq. km area of state forests and national parks near Batemans Bay on the south coast of New South Wales, Australia was selected for prototype development and demonstration. What resulted was a real exercise in the sense that it was done as realistically as possible within a self-imposed time frame. The stakeholders participating were ‘real’ foresters and conservationists, a full spectrum of issues was addressed and good empirical and synthetic supporting data acquired. However, it was never an exercise in making a plan to be implemented by some agency. While relevant agencies willingly participated by providing data etc. CSIRO has no charter for acting as a ‘super agency’ with responsibility to comprehensively plan the disputed area. In that sense, what is being reported here is a demonstration exercise albeit most realistic.

In modifying SIRO-Plan into SIRO-MED, established mediation principles were sourced⁶ and followed. These principles include the necessity to scope the dispute so that a range of outcomes is available rather than locking the process into a single format. Constant attention should be given to identifying and extolling points of agreement between the parties before addressing points of disagreement. Most importantly, stakeholders need to be encouraged to feel they ‘own’ the results of the process at every stage.

The SIRO-MED ‘recipe’ for seeking consensus on resource use comprises four main steps; each supported by relevant analytical aids and computer capabilities.

1. Seek stakeholders' agreement on the components of the problem including delineation of the area under dispute and identification of the current and pending resource allocation issues relating to that area. The stakeholders are helped to draft allocation guidelines that seek adoption of best practice resource management to these issues. Various established procedures may be used for identifying issues, allocation guidelines and data requirements⁷.
2. Acquire and process data needed to operationalise guidelines. The critical role guidelines play in efficiently defining and delineating minimal but sufficient data requirements cannot be over-stressed⁸.
3. Generate and evaluate stakeholder plans. The LUPIS software package supports this stage by assisting each stakeholder to generate value-driven and issue-focussed suitability maps for each land use of interest as precursors to generating a preferred land use plan.
4. Blend and mutually adjust stakeholders’ plans. The LUPIS software provides capabilities for producing various types of negotiation-support ‘aids’. For example, relative land value maps for each party in a free-form format encourage stakeholders to explore in a non-committed fashion, either individually or collectively, the implications of the numerous resource allocation options available.

Readers seeking greater detail on these steps should refer to the cited references^2,3,4. Suffice here to say that to complete SIRO-MED steps 1-3 for the Batemans Bay study involved stakeholders drafting 148 guidelines for the 10 land uses recognised. Eighty-four supporting data items were required for each of the 3439 land parcels making up the disputed area to enable the stakeholders to operationalise the guidelines for developing their own preferred land use plans. The preferred land use plans of individual stakeholders provided the input to step 4. Mediation and negotiation aspects are central to the final step of the SIRO-MED four-step recipe from which a consensus plan was developed. This step is now elaborated.

The implicit incentive for stakeholder participation in a SIRO-MED exercise is the recognition that, if they do not reach agreement, in all likelihood government will impose an allocation on the disputed area that is insensitive to the multi-facetted and spatially-differing relative values held by competing stakeholders.

SIRO-MED deliberately avoids suggesting any single path for leading stakeholders from their individual preferred plans to an agreed plan. Rather, it offers a range of data-based mediation and negotiation support aids for encouraging and assisting stakeholders jointly to develop a consensus or balanced plan from their individual plans. These aids include:

A ‘start-up’ plan is a first attempt at compromise with the most important features (in terms of resource values expressed by stakeholders) of each stakeholder’s plan being incorporated into a single plan. This plan then forms a starting point for negotiation between the stakeholders, negotiation directed to improving everybody's positions. Plan-blending means formally selecting and combining parts of independently produced stakeholder plans. Various plan-blending approaches to producing start-up plans are recognised and supported under SIRO-MED⁹.

Plan-adjusting is the formal step-by-step procedure for modifying a start-up plan in accordance with stakeholders’ preferences. Various plan-adjusting approaches are also recognised and supported by SIRO-MED⁹. While plan-adjusting can be used on any intermediate plan, it is logical to begin with the start-up plan as it recognises for each stakeholder, the resource allocation criteria identified as most important. In a sequence of adjustments, stakeholders seek progressively and incrementally to improve collective acceptance of the plan in terms guideline achievement (see below). Democracy decrees that it is most unlikely that any stakeholder will achieve all the values identified in their individual plan, consequently the final (consensus) plan represents a negotiated compromise around the different resource-allocation values expressed by stakeholders. Compromise comes with mutual understanding and appreciation of the values important to the other participating stakeholders; gaining such understanding is assisted by the following computer generated aids.

Any suggested allocation of forest lands can be evaluated in terms of how well it satisfies the range of land allocation guidelines summarising the preferred response by different stakeholders to the issues that made planning necessary. At any stage, LUPIS allows plans to be modified to increase achievement of specifically nominated guidelines. Because any plan generated by LUPIS is a Pareto efficient plan, an increase (decrease) in achievement for any guideline can only be gained at the expense of decreased (increased) achievement for at least some of the guidelines not nominated.

A stakeholder’s land value map shows the relative value placed upon each parcel in the area when it is used for the stakeholder’s preferred use. That is to say, if land were to be traded in the market place, these maps reflect the relative price each party would tender for the right to use each parcel for their preferred use. Relative value for a parcel is proportional to the overall suitability of the stakeholder's preferred land use on that parcel. Land value maps flow from the quantification of the stakeholder’s guidelines.

Once stakeholders have established the relative value they place upon each parcel (as reflected in land value maps) a dispute indicator map can be constructed. Dispute is likely to be most marked on those parcels valued highly by a number of stakeholders for uses that are incompatible. Dispute is less likely on those parcels where, although preferring different land uses, the values held by stakeholders for their respective land use preferences differ markedly. Obviously dispute on parcels where all parties prefer the same land use is unlikely. In the case of the Batemans Bay study, the dispute indicator map established that dispute was confined to less than 10 per cent of the study area.

Tradeoff graphs summarise how guidelines implying competition for the same resources are achieved differentially under different plans. In the case of conservation and forestry stakeholders, a tradeoff graph shows the path in guideline achievement for each guideline in moving progressively from the conservationists’ plan to the foresters’ plan (Figure 1). The tradeoff relationship between two resource-competing guidelines monotonically increments in opposite directions at a rate that is resource dependent. Each step along the axis from the conservation plan to the forestry plan corresponds to a unique plan. The challenge for the mediation process is to identify that step which yields the plan that is collectively most acceptable.

While SIRO-MED is a serious attempt to develop a science-based social technology for supporting the mediation process, it remains to be fully evaluated in a contentious real-world context. Notwithstanding, at the workshop organised to present and trial SIRO-MED, participants acknowledged that it had considerable promise. Why?

1. SIRO-MED allows large quantities of issue-relevant information and supporting data to be progressively condensed to a final plan while still allowing original data and intermediary derived information to be retrieved and studied. Stakeholder plans are one end-point of this process. In total, the Batemans Bay study generated an electronically catalogued library of over 2500 hierarchically linked maps displayable on demand.
2. SIRO-MED allows stakeholders to feel that they are full participants in the decision-making process at all stages (although data collection will commonly be left to the technically qualified).
3. SIRO-MED provides a non-confronting interface that encourages negotiation because increased stakeholder understanding is assisted by expressing all stakeholders’ demands in comparable terms including guideline achievement levels and land use plans.
4. SIRO-MED provides an opportunity to explore systematically the implications of a multitude of resource allocation options in terms valued by the stakeholders.

Despite these positive perceptions, any serious attempt at information-based mediation of an extensive forest dispute is likely to take at least a year to complete and perhaps cost several million dollars. Certainly the cost of a SIRO-MED exercise is likely to be high compared to the immediate cost of current practice where it can be claimed that resource analysis under-pinning political decisions is poorly focussed and superficial.

Conversely, it might be low compared with the dollar and non-dollar value of the resources at issue. Also, if information-based mediation does produce ‘legitimate’ allocation decisions that are accepted (even grudgingly) by stakeholders then the dollar cost is likely to be low compared with the savings made from avoiding prolonged social disputation and industry disruption. The gains from achieving acceptable solutions to forest allocation problems ‘first time round’ are obvious.

At a technical level, SIRO-MED and LUPIS foster a ‘learning’ approach to multi-party multi-criteria decision making⁹ which can be contrasted with conventional mathematical programming approaches where preference tradeoff rates and utility functions are assumed as given. Further, in contrast to the traditional rational decision making approach of comparing a handful of alternative plans (e.g. Janssen¹⁰), SIRO-MED encourages the development, comparison and refinement of numerous alternatives- in the case of the Batemans Bay study this amounted to some 55 plans.

To date, computer support for the mediation process has been mainly adopted to increase the efficiency of implementing established approaches¹¹. With today’s powerful personal computers there is a challenging opportunity to re-conceptualise the task of mediation support so that this power can be employed in more innovative ways¹². Central to such innovative systems must be a capability to derive and present highly complex options quickly and in readily understandable ways. As such, this requires action beyond recognition and accommodation of multiple goals, multiple alternatives and multiple relations between goals and alternatives. Rather the more demanding, but rewarding challenge is to capture such perceptions in a system that neither usurps nor threatens the perceived role of stakeholders or decision makers; on the contrary such a system needs to foster innovative input and involvement. The use of LUPIS to support multi-party multi-criteria decision making is a step in the evolution of such approaches.

Coping with complexity is not an end in itself. It is a means to creating more comprehensive, equitable and efficient compromises wherein stakeholders can see how the full range of their values is being collectively recognised and maximally satisfied.

SIRO-MED has been presented here as a ‘state-of-art’ procedure for assisting the resolution of forest allocation disputes. From our own, and others, years of experience with its predecessor, SIRO-PLAN, we believe it should be equally applicable to other spatial resource allocation problems. Evidence of this is the adoption of SIRO-MED and LUPIS as central components in two large externally-funded land use studies seeking a sustainable future for extensive areas in Australia’s vast rangelands, one centred on the North-east Goldfields area of Western Australia and the other the Western Division of New South Wales. Further applications could include for example, coastal resources, local government zoning, park and forest management planning. The basic requirements for undertaking such studies are that (a) the value to competing stakeholders of the resources at issue differ spatially and (b) numerous ways of allocating resources amongst stakeholders exist. To the extent that resource allocation processes are similar in all pluralistic democracies, the approach is in no way limited to Australian problems.

Figure 1: Tradeoff graph showing the relationship in terms of guideline achievement (pc) between two resource-competing guidelines (1 and 28) when moving from the Conservationists’ plan to the Foresters’ plan.