With Conservation International at Merauke
Margules reenters this story in 2008, by which point he had become Vice-President of CI for the Asia-Pacific region. CI had become one the biggest, richest, and most powerful non-governmental conservation organizations since its founding in 1987. But we also knew that CI was perceived very negatively by a wide array of conservation groups in the South. So, while my laboratory was excited by the prospect of applying our methodologies in the field, we approached the potential collaboration with trepidation. As it turned out we were justified in both our excitement and our trepidation.
In 2008, CI had contracted with the Medco Foundation, established by the Medco group (an Indonesian conglomerate founded by Arifin Panigoro), to devise a land use plan for an industrial forestry plantation concession obtained by Medco in the Merauke region of Papua Province in Indonesian New Guinea. The area was tropical savanna, more like northern Australia than the more famous wet evergreen forests found elsewhere in New Guinea. We would soon be practicing formal epistemology in this tropical savanna.
Medco intended to grow trees for pulp on its concession. However, it claimed to want to do so sustainably and while conserving biodiversity. According to CI personnel, the goals were to achieve sustainability of forestry production, conservation of biodiversity, maintenance of ecosystem function, and satisfaction of the interests of the indigenous communities using the habitat. These were nine Marind communities: Baad, Buepe, Kaiza, Kaliki, Kaptel, Koa, Senegi, Wapeko, and Wayau, all of which had traditional lands that intersected with the concession area. Medco made an initial commitment to exempt 40 percent of the concession area from plantation fanning; however, this 40 percent included land used by the nine communities. Margules wanted to use our new multi-criteria analysis techniques to develop a portfolio of spatial plans that incorporated all the goals. This portfolio was then going to be presented to Medco for a final choice.1
The planning process began in Jakarta in December 2008 with a meeting that included Medco, CI representatives, other stakeholders, and members of my laboratory as decision analysts. At the Jakarta meeting, all stakeholders identified by CI were present except, critically, representatives of the nine communities affected by Medco’s proposed development who (in Cl’s judgment) could not be included at this early stage for logistical reasons. However, there were supposed to be further iterations of the Jakarta meeting with their participation—meetings that never happened. CI personnel took it upon themselves to represent the communities’ views.2
The absence of systematic and routine engagement with local stakeholders led me to worry' whether Medco, aided and abetted by CI, was engaged in greenwashing. Although I never had meaningful contact with Arifin Panigoro, long conversations with several members of his family and other associates led me to give Medco the benefit of the doubt. Two reasons were most important. First, in the Indonesian context, Medco seems to have no motivation for greenwashing. There were strict environmental regulations (for instance, the protection of all wetlands, which constrained our plans severely, as will be seen below). But, beyond that, there was no public environmentalist constituency that needed appeasement through greenwashing. Second—and this reason was particularly compelling for me with my Indian cultural background—upper echelon Medco personnel seemed genuinely concerned about doing something beneficial for the Indonesian environment. For them it was a matter of national pride.
Returning to our task of decision support, the first task as decision analysts was to understand the context and to chart the goals and values of the stakeholders. For formal multi-criteria analysis, this involved the construction of an objectives hierarchy (OH) that established the fundamental objectives of the analysis and the subsidiary objectives under each of them. In many ways, this was the most interesting part of the process since it had to be done through group deliberation. Most of the time at the first meeting in Jakarta was spent in explaining the process and developing what we took to be a very preliminary version of the objectives hierarchy.
Given our original briefing by CI, we expected the fundamental objectives to be sustainability, biodiversity, ecosystem functioning, and community interests. To our surprise, at the instigation of Medco representatives, sustainability morphed early into production suitability of a patch of habitat for plantation farming (though, admittedly, farming strategies throughout were supposed to be sustainable in the long run). Even more surprisingly, at the insistence of a variety of local stakeholders, ecosystem functioning was subsumed under biodiversity conservation. It was clear that these stakeholders had a conception of biodiversity that was less to do with entities (as in most of the North) than with processes. Spatial configuration was added as a fundamental objective while community interests remained unaltered.
As external analysts we observed and recorded these discussions but did not participate in them except to clarify technical issues about multi-criteria analysis when asked. A point that we emphasized repeatedly was that we were not stakeholders because we were not part of the local communities affected by the decisions (and not even directly affected in some other capacity). We did promote biodiversity conservation in general but, for me, biodiversity is a local value and not a matter of global heritage—for more on this, see Sarkar (2012b). Even though we were philosophers, because of this intentional disengagement with any attempt to influence the outcome of local discussions of values, we did not enter into normative discussions that we witnessed. Perhaps we would have done so had there been an anthropologist present in the group. One of the lessons that I learned from this work was that, in contexts where there are wide cultural differences, we must have collaborations in place with anthropologists.
As expected, the stakeholder discussions generated complex hierarchies of sub-objectives under each fundamental objective. For instance, under community interests, there were nine sub-objectives at the next lower level, one to embody the interests of each of the nine communities. Below these were the goals of each community and these diverged across the set of communities. For instance, while all the other communities valued grassland within their areas, the Buepe did not. Under biodiversity, ecosystem services were incorporated using a single lower sub-objective: maintenance of all wetlands (which was a hard constraint in the sense that it was required under Indonesian law).
The most interesting structure that emerged was the hierarchy under the spatial configuration fundamental objective. This underwent modification after communities’ views had been canvassed by CI. There were two sub-objectives: the first was based on biological criteria and promoted both by CI and the communities; the second, guided by economic criteria, was promoted by Medco which wanted areas slated for production to be as close to transportation links as possible. The biological criteria included standard ecological ones such as the size of individual conserved patches and the connectivity between them. However, an unexpected preference emerged from community discussions: the communities wanted conserved areas to be as close to habitation as possible. Though we did not challenge this preference while incorporating it, I pursued it further in informal discussions: the preference reflected the practice of older inhabitants from villages walking to forests to collect non-timber products. Nearby forest persistence was an important community goal.
Shortly after the Jakarta meeting, CI began surveys of the concession area and organized regular field trips to Merauke. This work was supervised by Neville Kemp of CI who had decades of field experience in Indonesia. On the basis of the field trips, which mostly involved discussions with focal groups in each of the communities, CI decided that a reiteration of the Jakarta meeting including community representatives was unnecessary: except in the case of the spatial configuration fundamental objective, there was minimal change to the original objectives hierarchy. However, this meant that all the stakeholders were never assembled together in the same place, a decision about which I continued to remain uneasy. During 2009 I spent more than a month in Indonesia spread over three long visits. Language barriers prevented me—or any others from the University of Texas—from meaningful contact with the communities.
An even greater challenge was to elicit weights from the communities on the elements of the objectives hierarchy in such a way that the process was transparent. After trial and error, Kemp hit upon a reliable (that is, stable) method of eliciting weights. Since the relative weights across sub-objectives at any single level of the hierarchy added up to one, 100 pebbles were given to each focus group which then deliberated upon how these would be distributed across the objectives. The weight was simply the number of pebbles (divided by 100). Back in Texas, Dyer observed that this method of eliciting weights appears to be novel in the literature of decision analysis (and also that it assumes that preferences must be compounded additively, a point to which I return below).
By the middle of 2009 data collection and treatment were complete. What remained to be performed was the computational multi-criteria analysis. From my perspective, the project was progressing smoothly at Merauke. In particular, local communities appeared to be supportive of Medco’s efforts, especially because they had been promised schools, roads, and other facilities to which they had not previously had access.