shifting cultivation, contentious land change and forest
TRANSCRIPT
Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=fjps20
Download by: [202.62.17.171] Date: 06 November 2017, At: 06:55
The Journal of Peasant Studies
ISSN: 0306-6150 (Print) 1743-9361 (Online) Journal homepage: http://www.tandfonline.com/loi/fjps20
Shifting cultivation, contentious land change andforest governance: the politics of swidden in EastKalimantan
Gregory M. Thaler & Cut Augusta Mindry Anandi
To cite this article: Gregory M. Thaler & Cut Augusta Mindry Anandi (2017) Shifting cultivation,contentious land change and forest governance: the politics of swidden in East Kalimantan, TheJournal of Peasant Studies, 44:5, 1066-1087, DOI: 10.1080/03066150.2016.1243531
To link to this article: http://dx.doi.org/10.1080/03066150.2016.1243531
Published online: 05 Jan 2017.
Submit your article to this journal
Article views: 267
View related articles
View Crossmark data
Shifting cultivation, contentious land change and forest governance: thepolitics of swidden in East Kalimantan
Gregory M. Thaler and Cut Augusta Mindry Anandi
Swidden has historically been one of the most widespread land uses in upland SoutheastAsia. In recent decades, swidden systems across the region have undergone rapidtransformation. While most analyses focus on swidden as a livelihood practice, wedirect attention to the political nature of swidden. Drawing on qualitative fieldworkand household surveys from two villages in East Kalimantan Province, Indonesia, weexamine the politics of swidden along two key dimensions. First, at the householdlevel, we describe swidden as a land control strategy. We identify territorializationand speculation as drivers of ‘contentious land change’ in swidden systems underpressure from expanding plantations and mines. Second, at the village and districtlevels, we examine the politics of swidden within new forest governancearrangements. Control of swidden has provided a focus for multi-stakeholder forestgovernance, but with ambivalent effects, developing village land management andlivelihoods at a cost of temporary increases in swidden clearing and with minimalimpact on deforestation for industrial land uses. Our analysis suggests forestgovernance efforts will be ineffective in eliminating contentious land change orreducing district-level deforestation until they address plantation and miningexpansion as the dominant direct and indirect drivers of forest conversion.
Keywords: swidden; territorialization; environmental governance; deforestation;REDD
Swidden, or shifting cultivation with fire (Mertz et al. 2009), has for centuries been a domi-nant component of the agricultural systems of upland Southeast Asia. For almost as long,swidden systems have been under pressure from external actors, including governments,companies and conservation groups, who have sought to control or eliminate shiftingcultivation (Scott 2009). The transformation of Southeast Asian swidden systems in recentdecades has been especially intense, with widespread extinction or alteration of swidden prac-tices due to a confluence of political, economic and ecological factors, including governmentresettlement and land privatization policies, infrastructure expansion, environmental conser-vation initiatives, and the landscape-scale expansion of industrial agriculture (Padoch et al.2007; Fox et al. 2009). Understanding the current status of swidden is thus of criticalimportance to a broader comprehension of the socio-ecological dynamics of contemporarySoutheast Asia.
© 2017 Informa UK Limited, trading as Taylor & Francis Group
This article was originally published with errors. This version has been corrected. Please see Erratum(https://doi.org/10.1080/03066150.2017.1362781).
The Journal of Peasant Studies, 2017Vol. 44, No. 5, 1066–1087, https://doi.org/10.1080/03066150.2016.1243531
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Most analyses focus on swidden as a livelihood strategy, examining the effects ofdrivers such as population growth, road access, market development and public policy(van Vliet et al. 2012). This contribution focuses instead on the political nature ofswidden (cf. Fox et al. 2009). We use original fieldwork and household survey data fromtwo villages in Berau District, part of East Kalimantan Province in Indonesian Borneo,to explore the politics of swidden along two key dimensions.
First, at the household level, we examine decision-making regarding whether and whereto clear new swiddens. We find that swidden clearing has become increasingly important asa strategy for villagers to assert control over land under conditions of tenure insecurity (cf.Peluso 2005). We describe political-economic dynamics of territorialization and specu-lation that shape swidden systems through what we identify as a form of ‘contentiousland change’ (Aldrich et al. 2012).
Second, at the village and district levels, we examine changes in swidden dynamicsthrough the construction of new forest governance arrangements. ‘Governance’ movesbeyond government by the state towards the exercise of authority by heterogeneous politicalactors (Jordan, Wurzel, and Zito 2003; Himley 2008). Berau is the site of a district-level forestconservation program involving an international environmental non-governmental organiz-ation (NGO), and government, corporate and community actors. Limiting swidden areahas been one component of this program, and we explain that forest governance has coalescedaround swidden control because compensated reductions in swidden can reconcile commu-nity livelihoods with industrial forestry and government development priorities. The concen-tration of forest governance on swidden systems has had ambivalent effects, however,developing village land management and livelihood activities at a cost of temporary increasesin swidden clearing and with minimal impact on larger scale deforestation for industrial landuses such as oil palm and tree fiber plantations, and coal mining.
A focus on the politics of swidden at the household, village and district levels suggeststhat forest governance efforts will be ineffective in eliminating contentious land change orreducing district-level deforestation until they address plantation and mining expansion asthe dominant direct and indirect drivers of forest conversion.
Swidden in East Kalimantan
In Borneo, as in much of the rest of Southeast Asia, swidden systems are centered on uplandrice production. In the classic model, a forested area of roughly 1 ha is selected for cultiva-tion and trees and brush are felled and then burned to enhance soil fertility. The swiddenplot (called ladang in Indonesian1) is cultivated in rice for one or several years, often inter-cropped with or succeeded by other useful species such as chili, cassava or banana, and theplot is then fallowed for a period of usually not less than five years. Often, fruit trees andhardwoods are tended in the fallows. When sufficient time has passed to restore fertility andreduce the population of weeds and agricultural pests, the plot may again be cleared andreturned to cultivation.2 Swidden plots and fallows are traditionally controlled by the house-hold that originally cultivated them, or by descendants of the original cultivators. This
1The Indonesian word ‘ladang’ is widely used and understood, while terms for swidden plots in localdialects differ.2A classic description of an upland rice swidden system is Conklin’s (1957) work on Hanunóo agri-culture in the Philippines. Descriptions of swidden systems in eastern Indonesian Borneo includeInoue and Lahjie (1990), Jessup (1991), Colfer and Dudley (1993), and Colfer (2008).
The Journal of Peasant Studies 1067
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
swidden model has historically been practiced by settled upland groups, who would engagein trade with hunter-gatherer Punan and with coastal populations.3
Particularly since the end of World War II, this swidden system has been circumscribed,disrupted, and otherwise transformed by a variety of political and economic forces. Centralto this transformation has been the assertion of state control over most of the land in Indo-nesian Borneo (Kalimantan) through the creation of a national ‘forest estate’ administeredby the Ministry of Forestry.4 While the Indonesian forest estate has its legal and institutionalorigins in the colonial period, the expansion of state forest control in Dutch Borneo washindered by scarce resources, difficult terrain and political disputes (Peluso and Vanderge-est 2001). After World War II, and especially from the 1970s onwards under the Suhartogovernment, the Indonesian state sought to assert more direct control over Kalimantanterritory through commercial logging and through ‘development’ schemes, which werefacilitated by the relocation and sedentarization of indigenous groups and state-sponsoredtransmigration of non-indigenous settlers (Li 1999; Peluso 2005; Colfer 2008). In responseto state and market pressures and incentives and the encroachment of logging companiesand migrants, most hunter-gatherer groups have become more sedentary, usually takingup shifting cultivation and other agriculturalist livelihood practices (Sitorus et al. 2004).Indigenous agriculturalist communities have responded to shifting political and economicincentives in a variety of ways, including through expanded production of cash crops suchas rubber (Dove 1993) and pursuit of off-farm employment with timber companies. Therapid expansion of oil palm plantations in East Kalimantan in the last two decades hasdramatically altered the upland landscape, and many Dayak now also cultivate oil palmor work as laborers on an oil palm plantation.
Study villages in Berau District
In this paper, we focus on the Dayak villages of Gunung Madu and Long Kelay in Berau Dis-trict, East Kalimantan (Figure 1). (The village names are pseudonyms.) Both villages arelocated in the Kelay watershed. The 56 households of Gunung Madu consist primarily ofDayak Lebbo, and the village has been on its current site for several generations. Thevillage is abutted by a timber concession, an oil palm plantation, and a protected forest areacovering a karstic limestone escarpment, as well as by neighboring villages to the north andwest. Gunung Madu has only been accessible by road since 2012, when company accessroads reached the settlement. The roughly 35 households of Long Kelay are predominantlyDayak Punan, who until recently were semi-nomadic and have traditionally relied heavilyon hunting and gathering. The village has moved several times in the past decades, mostrecently from one side of the river to the other in order to avoid flooding and to have betteraccess to a logging road. Long Kelay is abutted by timber concessions, as well as byupriver and downriver villages and a neighboring inland village of Dayak Kenyah, whosettled in the area around 2008. For the purposes of our argument, the villages offer a contrast
3‘Dayak’ and ‘Punan’ are generic terms for indigenous agriculturalist and hunter-gatherer groups,respectively, and encompass a variety of different ethnicities of indigenous Christian or animistpeople, though distinct from the Muslim Kutai and Malay populations (C.J.P. Colfer pers. comm.,Dounias et al. 2007). Settled Punan may also refer to themselves as ‘Dayak Punan’, and we referin this paper to all settled, upland, predominantly Christian or animist indigenous groups as ‘Dayak’.4In late 2014, the Ministry of Forestry merged with the Ministry of Environment to form the newMinistry of Environment and Forestry.
1068 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
between a landscape of timber concessions with smallholder agriculture (Long Kelay) and alandscape undergoing conversion for plantations and mining (Gunung Madu).
Swidden in both Long Kelay and Gunung Madu conforms to many of the patternsdescribed in the broader literature. We find that especially in Gunung Madu, however,swidden is playing a critical political role as a form of establishing land claims andcontrol for villagers who feel pressured by competing land uses driven directly andindirectly by the plantation and mining sectors. Rapid expansion of mining and oil palmconcessions heightens perceptions of tenure insecurity among villagers, leading to specu-lative and contentious land clearing.5
Both villages also participate in forest governance programs with The Nature Conser-vancy (TNC), an international environmental NGO. When it arrived in Indonesia in the1990s, TNC initially focused on protected area management, especially at Lore LinduNational Park in Sulawesi (Li 2007). During the 2000s, the organization shifted its focustowards East Kalimantan, where it promoted sustainable logging and orangutan conserva-tion, and in the late 2000s, with the advent of projects for reducing emissions from defor-estation and forest degradation (REDD) on the international agenda, TNC joined with thedistrict government of Berau to launch the Berau Forest Carbon Program (BFCP).
BFCP was designed as a jurisdictional REDD program in order to develop an integratedapproach to land change at the landscape level and to facilitate interaction with government
Figure 1. Berau District, East Kalimantan, Indonesia.
5While contentious land change in swidden systems has received little attention, much has beenwritten elsewhere on the relationship between insecure land tenure and deforestation. See forexample Deacon (1994), Alston, Libecap, and Mueller (1999), Geist and Lambin (2002) and Margulis(2004).
The Journal of Peasant Studies 1069
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
institutions and initiatives (Fishbein and Lee 2015; Berau REDD+ Working Group 2011).The program comprises multiple components, including corporate engagement withlogging concessions, protected area conservation, development of REDD institutional fra-meworks, and community-based natural resource management. Under this last component,TNC has targeted Long Kelay and Gunung Madu as pilot villages for participation inBFCP, and limiting forest clearing for swidden has been one component of village-levelactivities. The villages were selected after an evaluation of 20 villages in Berau based oncriteria including reasonable accessibility for TNC staff, substantial remaining forestcover and positive community interest. Long Kelay and Gunung Madu were chosen inpart to provide a contrast between working with timber concessions (Long Kelay) andworking with protected forest area (Gunung Madu), according to a TNC manager. WhileBFCP has so far generally failed to engage the major corporate actors involved in defores-tation and has struggled with lukewarm commitment from the government, we find that theprojects in Gunung Madu and Long Kelay have been successful in constructing multi-sta-keholder collaboration, and we explore how and why the limitation of swidden has becomea focus for development of forest governance.
Methods
This study is based on eight months of fieldwork in Indonesia by the first author in 2014–2015,including interviews and participant observation at the national, provincial, district and villagelevels. Field visits to Gunung Madu and Long Kelay by the first author took place in March–April 2015. We also make use of both internal and publicly available documents from TNCand government agencies, and media reports. Additionally, we draw on data collected bythe Center for International Forestry Research (CIFOR) as part of its Global ComparativeStudy on REDD+ (Sunderlin et al. 2015), which has carried out household surveys inBerau in 2012 and 2014. Gunung Madu and Long Kelay are both included in the CIFORstudy, and CIFOR data collection in these villages was coordinated by the second author.
In the following sections, we describe the swidden systems in Long Kelay and GunungMadu, and we focus especially on the political dimension of household decision-makingregarding swidden clearing in Gunung Madu. We then examine village- and district-levelswidden politics under recent forest governance initiatives.
Swidden in Long Kelay and Gunung Madu
Long Kelay
The Dayak Punan of Long Kelay previously depended on sago as a dietary staple, though asthey have become more sedentary over the past several decades their diet has shifted to hillrice, with sago rarely consumed outside of ceremonial occasions.6 Elders still speak of atime ‘before rice’, but today the villagers are primarily swidden rice farmers. A 2013report for TNC prepared by the French Institute for Sustainable Development and Inter-national Relations (IDDRI) found that 77 percent of villagers farm swiddens, and for 61percent of villagers it was their primary livelihood activity. The most common secondarylivelihood source was artisanal gold mining, which provided income for over half of thevillage households (Pirard and Lapeyre 2013). In the swidden cycle, villagers say they
6Unless otherwise cited, details in this section on Long Kelay come from the first author’s field notes,April 2015.
1070 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
would not return to a fallow plot for a minimum of five years, and they prefer at least aseven-year fallow. The CIFOR survey found that between 2010–2012 and 2012–2014the percentage of households clearing forest remained constant at around 75 percent, andthese households cleared an average of 1 ha/year in the first period, decreasing to 0.6 ha/year in the second. In both periods, clearing for ladang occurs primarily in ‘natural’(mature) forest area (i.e., mature old-growth forest or advanced secondary succession), asopposed to in more recent fallows (Anandi et al. 2014, 373), resulting in expansion ofthe total agricultural area of the village. In the longer term, the IDDRI report describes atrajectory of expanding swidden area in Long Kelay, from roughly five plots total perhousehold in 2003 (including both fallows and actively cultivated fields) to eight plotsper household in 2013, with villagers anticipating further expansion to 11 plots per house-hold by 2023 (Table 1). Part of the reason for this expansionary dynamic in the Long Kelayswidden system is the relatively recent establishment of the village in its current location.Villagers are opening ladang for yearly rice production, but also with a view to establishingfruit groves and rubber or cacao gardens, which are longer term land uses they have not yetfully developed in their current area. Indeed, the IDDRI study finds villagers anticipate afourfold expansion of their tree gardens over the coming decade. Villagers also seek toexpand the land base under their control as a legacy to their descendants.
While the village is located within logging concessions, the logging companies do notoperate within a 1–2 km radius of the village, and they respect the villagers’ right to clearswidden areas. There are some boundary conflicts with neighboring villages, particularlythe inland village of recent Kenyah migrants, but these conflicts generally do not appear toaffect the Long Kelay villagers’ swidden areas. Swidden plots are usually located alongthe river or on the logging road leading to the village. The Kenyah village is located onthe road rather far from Long Kelay, so conflicts pertain to hunting or illegal clearing bythe Kenyah in the Long Kelay forest area, but occur beyond the radius of Long Kelayswidden land. While logging activities around Long Kelay are not driving the villagers toengage in territorializing or speculative clearing, the conflicts with the Kenyah represent anindirect effect of plantation expansion, since the Kenyah emigrated from the village ofLong Segar in neighboring Kutai Timur district due in part to land conflicts caused by oilpalm concessions (cf. Elmhirst, Siscawati, and Colfer forthcoming).
Table 1. Swidden agriculture in Long Kelay and Gunung Madu.
LongKelay
GunungMadu
Agriculture as primary livelihood activity, % of householdsa 61 50Agriculture as secondary livelihood activity, % of householdsa 16 23Area of natural forest cleared, ha/household/year (average), 2010–2012b,c 1.0 0.9Area of natural forest cleared, ha/household/year (average), 2012–2014b,c 0.6 0.6Number of ladang plots/household (average), 2003–2004a,d 4.8 3.1Number of ladang plots/household (average), 2013a,d 7.8 3.6Number of ladang plots/household (average), 2023 (estimate)a,d 10.9 4.3
aPirard and Lapeyre (2013).bData from CIFOR Global Comparative Study on REDD+, Module 2 on Subnational Initiatives.cAverage among households reporting any forest clearing.dTotal of both fallows and actively cultivated plots.
The Journal of Peasant Studies 1071
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Gunung Madu
Gunung Madu7 village has been in its current location for several generations and is sur-rounded by a more diverse mosaic of land uses than Long Kelay, resulting in substantiallydifferent swidden dynamics. Gunung Madu villagers have overall more diversified liveli-hood strategies than villagers in Long Kelay. The IDDRI study, while only managing tosurvey 60 percent of GunungMadu households, found that while 73 percent of those house-holds still farmed ladang, it was the primary livelihood activity for just 50 percent of them.Harvesting edible birds’ nests, either individually or as a worker for the private company PTWalesta, which holds a concession to caves in the nearby karst escarpment, providedincome for 44 percent of households. Residents may also work for the logging company,and they are widely known for harvesting large amounts of forest honey.
Residents of Gunung Madu have historically tended to clear less ladang area than vil-lagers in Long Kelay. IDDRI data show that while in 2003 Gunung Madu villagers usuallyplanted around 0.8 ha/year, in 2013 they were planting only 0.6 ha. Gunung Madu villagersalso had comparatively fewer total (cultivated and fallow) swidden plots, an average of 3.6in 2013, which they expected to grow only modestly over the following decade to anaverage of 4.3, according to IDDRI (Table 1). Gunung Madu villagers have received gov-ernment and TNC support to plant rubber trees, and IDDRI found that most villagers had atleast one tree garden and expected to add several more in the coming decade. The morelimited practice of swidden in Gunung Madu is explained in part by the longer presenceof the village in its current location (meaning households have had time to build up aland bank for non-swidden land uses such as fruit groves), and in part also by the greaterreliance on tree crop and non-agricultural livelihood sources relative to Long Kelay,where villagers are more dependent on swiddens for subsistence. In recent years, total clear-ing in Gunung Madu has increased, however. CIFOR data show that while in 2010–2012only 36 percent of surveyed households cleared forest area, in 2012–2014 that figurejumped to 70 percent. The average annual clearing per household declined slightly,similar to in Long Kelay, from 0.9 ha in 2010–2012 to 0.6 ha in 2012–2014 (Figure 2).These recent changes result from a combination of governance incentives and contentiousland change processes, and are discussed further in the following sections.
As in Long Kelay, villagers in Gunung Madu still cleared almost exclusively mature forestareas when opening new ladang in 2010–2014, as opposed to returning to recent fallows. Onevillager stated succinctly the logic, saying, ‘We prefer to open ladang in forested areasbecause it’s more fertile and we get new land’. Returning to fallows is cheaper, but then theyoften experience problems with weeds, and they lose the chance to establish possession over anew area. Many fallows have also been planted with fruit trees, which villagers prefer not tofell. The desire to establish rights over new land through swidden clearing is motivated inLong Kelay by the need to establish a land bank for planting fruit groves and providing alegacy for future generations. In Gunung Madu, however, swidden is a relatively smaller com-ponent of the livelihood strategies of most households, and the area cultivated through swiddenhas been and is expected to remain fairly stable. Villagers anticipate expanding their area of treecrops, especially rubber, which is one motivation for opening new land, but there are two otherimportant, non-agronomic motivations: the use of swidden to establish territorial control for thevillage, and the use of swidden to establish private land rights as a form of economic speculation.
7Unless otherwise cited, details in this section on Gunung Madu come from the first author’s fieldnotes, March 2015.
1072 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Swidden as a form of territorial control is used to counter competing claims and encroach-ment by neighboring villages. Gunung Madu villagers have ongoing boundary disputes withMerapun, the Dayak Lebbo village downriver from Gunung Madu. Encroachment byMerapun villagers has been exacerbated by the establishment of an oil palm plantation sur-rounding Merapun village, which has limited their forest area at the same time as newroads have enhanced access to Gunung Madu territory, making inter-village land conflict anindirect effect of oil palm expansion. Gunung Madu villagers may choose to open swiddensin contested areas in order to establish territorial control for their village. At the same time,both Merapun and Gunung Madu villagers have private speculative interests in openingladang in contested areas because of the expectation that an oil palm concession may beissued and villagers whose ladang fall within the concession would be able to make claimsfor compensation. Similar speculative clearing by Gunung Madu villagers is said to have
Figure 2. Households clearing forest in previous two years (A); average annual forest clearing perhousehold (B) of households clearing forest in previous two years.
The Journal of Peasant Studies 1073
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
occurred in response to a coal company survey, as villagers sought to position themselves togain compensation from any land that would eventually be taken over by a mine.
Household decision-making: contentious land change in a swidden system
The strategic use of swidden as a form of land control has received relatively little attentionin the literature on drivers of change in swidden systems. It receives no direct mention, forexample, in van Vliet et al.’s (2012) meta-analysis of 111 publications. While in Dayakcommunities, swidden clearing confers land rights under customary law, tree crops suchas rubber and cacao have historically been more likely than swidden to be recognized asdenoting a land claim, especially by Indonesian authorities or private companies (Dove1993, 142). In recent years, however, government and corporate actors have increasinglyrecognized managed swidden fallows (i.e., fallows within a certain radius of a villageand marked with hardwood or fruit trees) as village property, even as these same actorshave sought to restrict the practice of swidden overall. The possibility of external recog-nition increases the salience of land control considerations for villagers in locatingswidden plots. The case of Gunung Madu highlights two separate land control functionsof swidden – territorialization and speculation – that help determine both the location ofswidden plots and the continued expansion of swidden into new forest areas.
The territorializing function of swidden presents a paradox, or ‘problem of composition’,since rights to a swidden plot are customarily held by the household that cultivates it, yet thathousehold also belongs to a village, and so its plot may be considered a marker of village ter-ritory (cf. Dove 1983). Most villages in Berau do not have official territorial boundaries demar-cated by the government. The village as a formal administrative unit with defined cartographicboundaries only began to be imposed in Kalimantan in the last 35 years (Peluso 2005), andmuch more recently in upland East Kalimantan (C.J.P. Colfer, pers. comm.). Rather, multipleinstitutional layers define overlapping boundaries that are negotiated and contested amongcommunities, concessionaires and government agencies.
Dayak communities practice swidden within a customary territory (wilayah ulayat) that ismarked by totems, sacred areas such as graveyards, and the planting of fruit trees and hard-woods in swidden fallows. These customary areas are recognized by neighboring Dayakcommunities, but may not be known to outsiders, and they may fluctuate over time as com-munities migrate. Designation of forest use categories in Kalimantan occurred without refer-ence to village boundaries, but the Ministry of Forestry now recognizes village areas withinthe forest estate as a point surrounded by a 1-km buffer. These village ‘enclaves’ do notconcede tenure, but they proscribe exploitation by concessionaires to avoid conflict withthe community. Before beginning operations, concession companies should recognizevillage boundaries in their management plans, and these boundaries will be known to the gov-ernment and to at least village elites. Government agencies will also make use of villageboundaries drawn by the Central Statistics Bureau (BPS), though these boundaries are deter-mined without community input and often fail to reflect realities on the ground. Finally,village land-use planning has included participatory mapping, led by TNC in Long Kelayin 2012 and in Gunung Madu in 2013. These mapping exercises indicated village boundarieswith global positioning system (GPS) points that were agreed to by neighboring villages, butthese community maps may differ from BPS boundaries or those used by concessionaires. Inthe absence of formally demarcated cartographic boundaries and under conditions of rapidlandscape change driven by migration and industrial expansion, possession may be ‘nine-tenths of the law’, in that the occupation of an area through swidden clearing and the sub-sequent planting of hardwoods or fruit trees by a household may serve to reinforce village
1074 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
territorial claims that are contested across government and company boundaries, NGO landuse plans and customary areas. These claims are important not just for determining the spacesfor villagers to carry out their livelihood activities, but also for determining collective pay-ments made by companies operating in the village territory.
The speculative function of swidden is a response to the rapid and poorly regulatedexpansion of industrial land uses around Gunung Madu. Most of the landscape fallswithin the forest estate, whose management is under the authority of the central govern-ment’s Ministry of Forestry, while even in lands outside the forest estate there is rarelyany formal recognition of community land uses (Rights and Resources Initiative 2015;Contreras-Hermosilla and Fay 2005). The resulting tenure insecurity means communitieshave few options for resisting the conversion of village territory to mines or plantations(Myers and Ardiansyah 2014). In addition to providing collective payments or subsidiesto communities that fall within their footprint, companies may also directly compensatehouseholds whose ladang or tree gardens are taken by a concession. Compensation prac-tices vary, but in the case of oil palm plantations, compensation most often takes theform of certificates that entitle a household to revenue from the production of 1 or 2 haof plantation land. Companies deduct costs related to planting and management of theareas, so during the first five to six years a household holding a certificate may expect toreceive IDR 100,000–300,000/month.8 In circumstances where not all households in avillage receive certificates, those whose swidden lands have been taken by the companymay be able to make stronger claims to the company or village elites in order to receive com-pensation. Facing the possibility of expropriation of lands within the village territory, anduncertain of how compensation may be administered, villagers speculate on future companyactivities by locating swidden in possible concession areas to establish personal land claimsthat will enhance their prospects for compensation when the lands are expropriated.9
The expansion of swidden as a land control strategy for territorialization and specu-lation represents a form of contentious land change. Aldrich et al. (2012), working in theAmazon, argue that while land change science typically analyzes land cover change asthe result of decisions made by individual rational actors, a more complete understandingof land change recognizes its social nature. They assert that ‘a component of deforestationin Amazônia results from contentious social interactions aimed at land possession inaddition to what we refer to as agronomic deforestation undertaken to plant crops orpasture as an economic activity’ (109, original emphasis). Aldrich et al. were unable toobtain statistical estimates of the amount of deforestation in their case study that wasspecifically due to contentious processes, and any individual land-clearing decision willtypically combine both social and agronomic considerations. Nonetheless, the recent devel-opment of a village spatial plan under TNC’s project in Gunung Madu allows for the identi-fication of specific ladang plots as examples of land-use change predominantly determinedby contentious processes. These clearings are explored in more detail in the followingsection, which examines the transformation of the swidden systems of Long Kelay andGunung Madu through the evolution of new structures of forest governance.
8This sum is not insignificant, considering the national poverty line is set at just over IDR 300,000/person/month.9In Gunung Madu, speculation is evident through the location of swidden clearing, not the totalamount of clearing by a household. Households are still clearing and planting plots of roughly1 ha, but in some cases they are siting these plots strategically based in part on possible future com-pensation. Household resources for clearing and planting to establish land claims are limited, sorunaway speculation involving larger scale clearing has not been observed.
The Journal of Peasant Studies 1075
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Village and district-level forest governance
The Nature Conservancy’s conditional payments agreements
TNC has interacted with the Dayak communities of Kelay since the early 2000s, at firstthrough support for orangutan protection in the Lesan forest area and for the developmentof the Upper Kelay Village Forum (Forum Kampung Hulu Kelay) as an interlocutor to thelogging companies in support of ‘collaborative forest management’. After launching BFCPwith the district government in 2009, TNC began to look for ways to integrate communitiesinto REDD (Moeliono et al. 2010), including piloting incentive schemes that would com-pensate villagers for emission reductions (Pirard and Lapeyre 2013). More information onTNC’s community incentive agreements can be found in Lapeyre, Hartanto, and Pirard(2015). After a participatory planning period, both Gunung Madu and Long Kelaysigned agreements in late 2013 declaring their participation in BFCP and their commitmentto protecting the forest and promoting environmentally friendly livelihood activities.
The village declarations were accompanied by Conditional Grant Agreements withTNC. Under these agreements, villagers’ performance in meeting specified targets duringthe first year of the contract would determine the level of support provided by TNC inthe subsequent year, with support reduced in the case of failure to meet targets. Theinitial TNC grant to the communities, IDR 239 million each (about USD 20,000), includedsupport for forest patrols and forest management, rubber cultivation, chicken raising,vegetable gardening, honey production, fish farming and capacity building, among otheractivities. In exchange, villagers agreed to ‘strengthening the enabling conditions’ and‘[climate change] mitigation and management of natural resources’ (TNC-Indonesian.d.). Activities under enabling conditions generally relate to village financial managementand capacity building. Under mitigation and management, villagers agreed to carry outforest patrols and biodiversity and ecotourism surveys, and to limit their practice of shiftingcultivation. Villagers agreed to open not more than 1 ha/household/year of ladang, and onlyto open ladang in fallows as opposed to in new forest areas. New forest areas could beopened exceptionally in the case of new households that had not yet established a landbase. In Gunung Madu, villagers committed to a maximum of four ladang plots per house-hold (active or fallowed), for a total of 4 ha of swidden land each (‘Kesepakatan DanKomitmen Masyarakat Kampung Gunung Madu’ 2013). In Long Kelay, the communityagreed to a maximum of seven ladang per household (active or fallowed), totaling 7 haof swidden land. Additionally, TNC assisted each village in the development of a villageland-use plan, specifying areas for settlement, swidden farming, rubber gardens, agrofores-try and reserve land.
These agreements would appear to constitute an important change in the swiddendynamics of the two villages. Where prior to the agreements villagers were predominantlyopening ladang in new forest areas, they should now only open ladang in fallows. It wasexplicitly understood that this commitment was dependent on alternative livelihoodsupport from TNC, which has focused most heavily on the planting of rubber gardens,and small animal husbandry (Pirard and Lapeyre 2013).10 In fact, these agreements stillallow for the clearing of new forest areas under several provisions, and the immediateimpact in the first year of the agreements has been to support new clearing, albeit underthe aegis of the village land-use plan and ostensibly as a one-time occurrence.
10Dove (1993) points out that rubber may not actually be understood as an ‘alternative’ to rice, sincethere remains in many Dayak communities a strong cultural emphasis on rice cultivation and the sub-sistence sector.
1076 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
First, households that have not yet reached the maximum swidden land base of four orseven plots are permitted to clear new forest area. In addition to migrants to the village,these households include young adults from the village who are starting new households.While their parents may already have a full land allotment, these younger households arelicensed to clear forest for new ladang up to their own allotment. Under the agreements,as population grows so will clearing, but in a fixed proportion of land per household.Second, TNC is supporting the establishment of 2 ha of rubber garden per household.While Kalimantan Dayak have a long history of rubber cultivation, there was previouslylittle rubber in Long Kelay due to the recent establishment of the village in its currentlocation and the more limited agricultural experience of the Punan inhabitants. Therewas previously some rubber production in Gunung Madu, but still of a limited nature, inpart because of the reliance of Gunung Madu villagers on birds’ nest harvesting. Therubber gardens supported by TNC are not required to be established on fallows, so in thefirst year of the agreement in Gunung Madu, many households opened 1 ha of ladang inforest area along a new roadway, explaining the increase in the number of households clear-ing forest land. After the first year’s rice planting, they will turn the plot over to rubber andrepeat the process, so that after two years they will have established their 2 ha rubber allot-ment. According to the land-use plans, after obtaining 2 ha of rubber land and 4 or 7 ha ofswidden plots, a household should not engage in any new forest clearing. It will take severalyears to determine whether that limitation has occurred. CIFOR data show that while in2010–2012 all new clearing in Gunung Madu was devoted to rice cultivation, in 2012–2014, 35 percent of households planted rubber in new clearings (Figure 3).11
While the TNC agreements thus permit limited expansion of swidden and rubber gardensinto new forest areas under certain conditions, the organization’s strategy is to control andlimit the clearing through village land-use and development planning. We focus here particu-larly on Gunung Madu, because of the importance of contentious land-change processes inthe village and because the first author was a participant observer at the assembly to evaluatethe first year of the conditional payments agreement in March 2015.
Evaluating the Gunung Madu agreement: contentious clearing versus forestgovernance
In Gunung Madu, the recent construction of the logging road to the village opened up alarge area of land that was previously less accessible. Land along the road would almostcertainly have been deforested in any case, but under the village land-use plan, the landalong the road to the northwest of the village was zoned for rubber gardens and dividedinto 2-ha plots, which were allocated to each household in a random drawing. The resultingarrangement asserts village control over the area and allows for expanding livelihood activi-ties, but does so in an ordered way that aims to avoid the conflicts and inequities that oftenaccompany the opening of a new land frontier.
The village agreement, conditional payments contract, and land-use plan have not fullyeliminated contentious land change processes from the Gunung Madu swidden system,however. The one-year evaluation of the conditional payments agreement took the formof a village assembly facilitated by TNC staff, where villagers discussed the activities
11A similar shift to rubber occurred in Long Kelay, where the percentage of households plantingrubber in new clearings went from 0 to 35 percent. Planting in new clearings is only indicative oflarger land use shifts towards rubber, as these statistics do not reflect the planting of rubber inplots previously devoted to other crops.
The Journal of Peasant Studies 1077
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
that had taken place and evaluated themselves according to the agreed-upon targets. Whilethe villagers and TNC agreed that financial management, training, and forest patrol targetshad been met, the issue of swidden clearing emerged as a clear dilemma for participants. Atthe prompting of a TNC staff member, villagers acknowledged that four households hadcleared new swidden plots along the road to the southwest of the village in the areazoned as reserve land under the village plan.12 Discussion then ensued about whetherthis clearing constituted a violation of the village agreement.
Figure 3. (A–B) Crops planted in new clearings.
12Reserve land in principle would be set aside for management by future generations.
1078 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
While none of the four households that had cleared in the reserve zone were in attend-ance at the evaluation meeting, the possible violation concerned all villagers because oftheir collective commitments and the potential for violations to affect the amount offunding provided to the village in the coming year. Opinion among the villagers wasdivided. Those who viewed the clearings as a violation pointed to the letter of the agree-ments, while those on the opposite side advanced a variety of reasons why these clearingsshould not be considered a breach. The first justification offered was that people had openedin the reserve land in order to keep outsiders from encroaching. Villagers from Merapunhave been challenging Gunung Madu’s control in this area, so swidden was in this caseviewed as a means of territorialization. It was also noted that those opening in thereserve land had not cleared more than 1 ha nor had they exceeded their total land allot-ments. Other villagers pointed out, however, that if the transgressors were intending tokeep away Merapun villagers, they had opened their swiddens on the wrong side of theroad. A second justification suggested was that the plots distributed for rubber were notall equally desirable, and those who had drawn worse plots might prefer to open land else-where. Discussion focused primarily on the territorial justification. On another occasion, avillager explained to the first author that people in Gunung Madu were no longer openingladang in new forest area except when there was a boundary conflict with another commu-nity, in which case they would clear forest to claim the land. The territorial function ofswidden is clearly still considered a legitimate determinant of clearing by a significant pro-portion of the village population.
This episode illustrates the encounter of competing territorializations. Territorializationthrough swidden clearing conflicts with territorialization through spatial planning. Thespatial planning approach has the additional attribute of helping the village to elaboratevillage development plans, through which the village can access funding from the Indone-sian government, according to a TNC manager. Enforcement of village territorial control isalso supported under the spatial planning approach through forest patrols. Nonetheless, inan area where control is contested with another village, spatial plans and occasional patrolsdo not have the materiality or cultural weight of a swidden plot. This episode also illustratesthe paradox of swidden as a form of territorialization, however. In the view of one of theTNC staff members most familiar with the village, the households opening ladang in thereserve land were motivated not by collective territorial control, but by private speculativeinterest. He noted that there is already an area near Gunung Madu, near the reserve land andMerapun, which has been zoned as conversion forest by the central government (meaning itcan be converted to non-forest uses), and there is an area (where most of the reserve landlies) that is still classified as production forest (meaning it should remain under forest uses).Gunung Madu villagers fear that if people from Merapun continue entering and degradingthe production forest, it will be rezoned to conversion forest, and then oil palm will enterand the village government has no authority to prevent it. The four transgressing house-holds do not expect the village government to keep out the oil palm plantations. Byopening swidden in the contested area and establishing land claims now, they may atleast receive compensation if the land is expropriated for oil palm.
Speculative and territorial drivers of household decision-making intersect in this case,and because these clearings happened contrary to the village forest governance agreements,we can identify contentious land change in Gunung Madu as the cause of roughly 4 ha ofnew deforestation. In the CIFOR sample of 33 Gunung Madu households, 70 percentcleared forest land in 2012–2014, averaging 0.64 ha/household/year of clearing. Extrapo-lating to the village population of 56 households, we estimate that a total of 25 ha was
The Journal of Peasant Studies 1079
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
cleared in 2014.13 We therefore estimate that 4 out of 25 ha of clearing, or 16 percent of the2014 deforestation in Gunung Madu, was determined through contentious land-change pro-cesses. While the four transgressors would likely have cleared ladang within the properzone had they not opened ladang in the reserve land, the rubber zone was slated for conver-sion, while the reserve land clearing constitutes deforestation that would not otherwise haveoccurred were it not for these specific processes of contentious land change.
In the end, community members voted 8 to 5 that the clearings in reserve land were not aviolation, and they gave themselves full points for compliance with the swidden clauses ofthe agreement. This episode highlights the tensions between different economic systems indetermining household clearing decisions. While an economic logic might attempt toexplain these decisions by comparing the benefits of TNC’s conditional payments to thebenefits of oil palm compensation, the TNC agreement and oil palm represent two distinctvillage economic systems (a smallholder economy versus a plantation economy) and cannotbe effectively analyzed independently from those systems. In considering the costs andbenefits of TNC payments versus oil palm compensation, villagers may consider the fullrange of costs and benefits associated with a smallholder versus plantation economy.Most villagers in Gunung Madu would prefer not to lose their lands to oil palm. Nonethe-less, our account illustrates that under conditions of high tenure insecurity, hedging againstfuture expropriation may be rational despite the risk of forgone payments from TNC. Thepractical negotiation of these tensions does not flow directly from a calculus of formal, insti-tutionalized costs and benefits, but rather meanders through the moral economy of thevillage and villagers’ interactions with TNC, other villages, the oil palm company andthe government (Scott 1976). Villagers who clear speculatively have moral expectationsabout land claims and compensation that may not be commensurate with the compensationpractices of companies, who are likely to compensate the village collectively as opposed toindividual landowners. Speculative clearing may enhance the claims of individual house-holds vis-à-vis village elites in order to obtain a share of the collective compensation,though elites may also appropriate compensation to themselves and their cronies. In thiscase, speculating households hedged against future expropriation and the village did notlose any payments due to their clearing, but the village authorities subsequently penalizedthose households by excluding them from further benefits under the conditional paymentsagreement.
Swidden as a focus for multi-stakeholder governance
The attempted re-ordering of swidden practice in Gunung Madu and Long Kelay is notlimited to the interactions between TNC and the communities. Rather, it intersects withgovernment policy and corporate practice as a nexus for the emergence of ‘multi-stake-holder’ forest governance in Berau. Multi-stakeholder collaboration is a key feature of neo-liberal governance regimes (Castree 2010; McCarthy and Prudham 2004), as well as acentral tenet of TNC’s conservation strategy, which the organization describes as a ‘non-confrontational, pragmatic, market-based’ approach to conservation based on partnerships
13New clearings usually average about 1 ha in size, but not every household clears land every year,explaining the average annual clearing of less than 1 ha/household. Clearing in 2014 may have beenhigher than in 2013, given incentives for rubber production. If all 39 households estimated to havecleared forest in 2012–2014 cleared new ladang in 2014 specifically, then a lower bound for the pro-portion of contentious clearing would be 4 out of 39 ha, or 10 percent.
1080 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
with governments, companies, multilateral organizations, other NGOs and ‘local stake-holders’ (TNC 2013). In Berau, TNC attempted to establish BFCP as a program led bythe district government with participation from the provincial and national government,companies, NGOs and communities. The Berau REDD+ Task Force (Pokja REDD),which supported the development of BFCP and functions as a de facto secretariat for theprogram, counts within its membership the district forestry department and environmentalbureau, two major timber concessionaires, a local environmental NGO and TNC.
Conspicuously, for a program aimed at reducing emissions from deforestation, there hasbeen virtually no participation in BFCP by oil palm or tree fiber plantation companies,which are the industrial sectors most responsible for conversion of natural forests inBerau (Griscom et al. 2016), or the mining sector, which is responsible for relativelylittle direct deforestation, but has large secondary impacts that have not yet been effectivelyquantified (Scrivener 2013). Prior to 2015, TNC had also largely failed to target thesesectors in its Berau program. Engagement with the oil palm sector was delayed in partby changes in BFCP organization and negotiations between TNC and funders, a TNCstaffer explains. An appropriate model for reining in plantation-driven deforestation hasalso been lacking. ‘Land swaps’ that would allow companies to trade forested areaswithin a concession for degraded lands currently within the forest estate were promotedas a model during the early 2010s by the World Resources Institute (Gingold et al.2012), but TNC has found the model infeasible in Berau because most forestryconcessions are active and there is little degraded area suitable for oil palm within theforest estate, according to provincial- and national-level TNC staff. Limiting oil palmexpansion is also politically sensitive given the centrality of oil palm to district and provin-cial development plans and the payoffs that district officials receive from oil palm compa-nies. In 2015, TNC launched a new oil palm program with four years of funding from theGerman Federal Environment Ministry with the aim of supporting community developmentplanning and improved siting and mitigation of oil palm plantations, among other goals, butTNC staff acknowledged that the oil palm sector had been a major gap in their Berauprogram until the time of research.
Community initiatives and sustainable logging, meanwhile, have been the areas of mostsubstantial on-the-ground activity by TNC and other environmental organizations workingin inland Berau. Aside from TNC, the most significant non-state environmental actor is theGerman government’s Forests and Climate Change Programme (FORCLIME). Activitiesin the forestry and communities sector include TNC and FORCLIME support forlogging companies to adopt reduced-impact logging methods and achieve sustainabilitycertifications; logging company and community participation in ‘collaborative forest man-agement’, where companies respect community areas and compensate communities fortimber extraction while communities help monitor logging concessions for illegal activities;and the government’s establishment of a Forest Management Unit (KPH) for western Berauto coordinate on-the-ground activities including monitoring, enforcement and conflictresolution, as well as supporting the establishment of ‘village forest’ (hutan desa) rights.In 2014, Gunung Madu was granted a village forest area over part of the protected forestland within village boundaries, becoming the first village in Berau to obtain a hutandesa designation.14
14The Gunung Madu hutan desa petition was facilitated by TNC, who led the participatory mappingprocess with the community and liaised with government bureaucracy. For more on hutan desa, seeAkiefnawati et al. (2010).
The Journal of Peasant Studies 1081
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
The regulation of swidden practice is also a cornerstone of the intersecting governanceinitiatives in the community-forestry sector. The beginning of industrial logging in EastKalimantan in the 1970s coincided with government efforts to resettle and sedentarizeupland populations, which on the one hand served to make those populations ‘legible’ tothe state (Scott 1998), while on the other it cleared space for industrial resource extraction.Conflicts between logging companies and communities have sputtered over the decades,but currently logging companies in Berau tend to respect communities’ rights to cultivatea certain amount of swidden area within their traditional territories.15 Further delimiting thepractice of swidden remains a common cause across the government and NGO actors in theforestry sector. While most do not view swidden as a major environmental problem, theynonetheless see reform of swidden as desirable and progressive. Both a FORCLIMEemployee and a district REDD coordinator claimed that ladang is the largest source ofillegal deforestation in Berau, since conversion for plantations and mining occurs under gov-ernment licenses. This ‘illegality’ of ladang speaks to the government’s desire to suppressshifting cultivation, and facilitates the confluence of government and NGO interest inmaking ladang ‘more effective’, in the words of a local NGO director, or offering ‘alternativeeconomies’, in the words of the REDD coordinator (cf. Dove 1983). Government officials inthe district forestry department and environment bureau also mention the use of fire as adanger of swidden cultivation, though Berau has not suffered fires of the same severity asparts of East Kalimantan farther to the south. In one case, the district government is reportedto have made shifting from swidden to permanent agriculture a requirement for confirming avillage’s administrative status (Anandi et al. 2014, 374).
From the environmentalist side, control of swidden has been central to NGO commu-nity work as a strategy of climate change mitigation and a counterpart to the promotion ofalternative livelihoods. Like TNC, the FORCLIME program supports REDD ‘demon-stration activities’ in upland villages. FORCLIME likewise stipulates that communitiesshould not clear new forest areas, and encourages rubber planting, small animal husbandry,and commercialization of non-timber forest products such as honey and herbal tea. OneFORCLIME employee imagines the possibility of a moratorium on swidden, saying,‘People deforest for their stomachs, for rice. If we pay for their rice, they wouldn’t defor-est’.16 He contemplates a strategy of hill rice intensification and the development of wet riceproduction, leading eventually to a moratorium on shifting cultivation. TNC’s model ofvillage land-use planning and alternative livelihoods development has also been codifiedby the organization as ‘SIGAP REDD+’ (Hartanto, Yulianto, and Hidayat 2014), andTNC is now directing around USD 10 million from a US government debt-for-natureswap to support local NGOs to implement SIGAP REDD+ in Berau, where swidden limit-ation is one among numerous options for communities to mitigate carbon emissions.
The effect of these various interlocking initiatives is that control of swidden – forclimate change mitigation, for state administrative control and for industrial timber extrac-tion – has become a central component of an emergent, loosely institutionalized forest gov-ernance apparatus that integrates government, logging companies, NGOs and communitiesin managing a landscape of timber concessions and small-scale agricultural and hunting and
15The development of ‘collaborative forest management’ between communities and timber companiesin Berau came in the early 2000s and was also supported by TNC. It marks an important chapter in thedevelopment of forest governance in the district, but space constraints preclude more detailed discus-sion here.16This statement entirely neglects the cultural value of swidden, a point we owe to Carol Colfer. Formore on the cultural value of swidden, see Colfer (2008), Dove (1993, 1998) and Gönner (2000).
1082 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
gathering activities. The relatively minor contribution of swidden to overall deforestation inBerau makes it particularly noteworthy that swidden has received such emphasis in thedevelopment of multi-stakeholder forest governance. One TNC manager explains thefocus on communities, affirming that while the emissions reductions of community activi-ties may not be great in themselves, the impact of community work is substantial because itserves as a way to attract government attention and commitment, to engage with companiesand to enhance the capacity of local NGOs. This multi-stakeholder apparatus could osten-sibly then expand to develop environmental governance in other sectors.
While TNC’s new oil palm project may bring change to that sector, at present the com-munity-forestry environmental governance apparatus is being undercut by the expansion ofindustrial mining and especially oil palm plantations, which introduces new ‘stakeholders’and contentious processes that upset the community-forestry consensus. Across Kaliman-tan, oil palm plantations are expanding at the expense of forest (Carlson et al. 2012), andcontentious land change ripples out from their expansion. In Long Kelay, where loggingis the primary industrial land use, swidden area has been expanding primarily due to agro-nomic reasons, but conflicts with the neighboring Kenyah village herald indirect impactsfrom oil palm. In Gunung Madu, where oil palm and mining threaten to enter the villageterritory, the agronomic drivers of swidden expansion are more limited, but territorializationand speculation are driving additional swidden clearing in natural forest areas.
The emergence of forest governance through control of swidden and more broadlythrough the conciliation of community livelihoods and industrial forestry, and the failureto integrate mining and plantations into the forest governance apparatus, may be explainedby the fact that community livelihoods and timber harvesting can be compatible withnatural forest cover, while oil palm, fiber plantations and mining (which in East Kalimantanis primarily open-pit coal mining) require the conversion of forest to an alternative landcover. Government actors are happy to limit the area under swidden as a way of sedentar-izing villagers and promoting the production of commodities such as rubber, but officials atboth the district and provincial levels view industrial plantations, in particular, as the key toeconomic development. They have no intention of limiting the overall expansion of planta-tions in favor of forest conservation. In the words of one FORCLIME employee, ‘oil palmis from the government, so we can’t do anything to stop it’. Controlling swidden is a form offorest governance that does not challenge the expansion of industrial land uses, so it hasbecome a site for successful multi-stakeholder coordination, while governance of themore significant drivers of deforestation in Berau remains elusive.
Conclusion
Swidden systems in East Kalimantan have been dramatically transformed over the last fourdecades, initially through government sedentarization programs and the advent of industriallogging, and more recently through the expansion of mining and plantations and thedevelopment of forest conservation programs. We use a study of the swidden systems oftwo Dayak villages in Berau District to examine the politics of swidden at the household,village and district levels. At the household level, dynamics of territorialization andeconomic speculation contribute to the expansion of swidden area through processes ofcontentious land change. At the village and district levels, control of swidden hasbecome a focus for the development of forest governance within a network of village,NGO, government and company actors. Efforts by these actors to limit swidden clearingand promote alternative livelihoods aim to define and minimize community agriculturalarea and eliminate contentious land change. At the same time, the limitation of swidden
The Journal of Peasant Studies 1083
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
plays an instrumental role in clearing space for industrial land uses (logging and oil palmconcessions) and promoting the deeper integration of villagers into the market economy.The omission of these industrial land uses from the forest governance regime, however,undercuts efforts to limit contentious clearing and results in the failure to reduce district-level deforestation.
The expansion of the mining and plantation sectors undermines the viability of thesmall-scale rural livelihoods on which the existing forest governance apparatus is anchored.Even communities whose territory remains forested are challenged by landscape-levelforest fragmentation (which affects biodiversity and species abundance), hydrologicaland regional climatic changes caused by large-scale forest clearing, and socio-economictransformations that trigger resource conflicts and drive contentious land change. Thecurrent status of swidden as a focus of nascent forest governance structures in Beraumay not herald the long-term viability of forest-dependent livelihoods so much as it indi-cates the inability of forest governance efforts to otherwise check the juggernaut of coalmines, tree plantations, and oil palm currently steamrolling the remaining forests of Indo-nesian Borneo.
AcknowledgmentsThe authors wish to thank the inhabitants of Long Kelay and Gunung Madu, the partners who sup-ported implementation of CIFOR surveys, and the staff of TNC-Indonesia. The first author thanksZachary Anderson for collaboration on fieldwork in East Kalimantan. We are grateful to CarolColfer, Christine Padoch, William Sunderlin, Kai Thaler and two anonymous reviewers for their com-ments, and to Eka Rianta for assistance with the map. The authors assume responsibility for anyremaining errors. An earlier version of this contribution was presented at the conference ‘Land grab-bing, conflict and agrarian-environmental transformations: perspectives from East and SoutheastAsia’ in June 2015.
FundingThe first author was supported by the American Institute for Indonesian Studies under an AIFISResearch Grant, the United States-Indonesia Society under a USINDO Travel Grant, and the USNational Science Foundation [under grant number DGE-1144153]. This research is part ofCIFOR’s Global Comparative Study on REDD+ (www.cifor.org/gcs). The funding partners thathave supported this research include the Norwegian Agency for Development Cooperation(Norad), the Australian Department of Foreign Affairs and Trade (DFAT), the European Union(EU), the International Climate Initiative (IKI) of the German Federal Ministry for the Environment,Nature Conservation, Building and Nuclear Safety (BMUB), the Department for International Devel-opment (UKAID), and the CGIAR Research Program on Forests, Trees and Agroforestry (CRP-FTA), with financial support from the donors contributing to the CGIAR Fund.
Disclosure statementNo potential conflict of interest was reported by the authors.
ReferencesAkiefnawati, R., G.B. Villamor, F. Zulfikar, I. Budisetiawan, E. Mulyoutami, A. Ayat, and M. Van
Noordwijk. 2010. Stewardship agreement to reduce emissions from deforestation and degradation(REDD): Case study from Lubuk Beringin’s Hutan Desa, Jambi province, Sumatra, Indonesia.International Forestry Review 12, no. 4: 349–60.
1084 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Aldrich, S., R. Walker, C. Simmons, M.M. Caldas, and S. Perz. 2012. Contentious land change in theAmazon’s arc of deforestation. Annals of the Association of American Geographers 102, no. 1:103–28.
Alston, L.J., G.D. Libecap, and B. Mueller. 1999. Titles, conflict, and land use: The development ofproperty rights and land reform on the Brazilian Amazon frontier. Ann Arbor: University ofMichigan Press.
Anandi, C.A.M., I.A.P. Resosudarmo, M. Komalasari, A.D. Ekaputri, and D.Y. Intarini. 2014. TNC’sinitiative within the Berau Forest Carbon Program, East Kalimantan, Indonesia. In REDD+ on theground: A case book of subnational initiatives across the globe, ed. E.O. Sills, S.S. Atmadja, C.De Sassi, A.E. Duchelle, D.L. Kweka, I.A.P. Resosudarmo, and W.D. Sunderlin, 362–379.Bogor, Indonesia: CIFOR.
Berau REDD+ Working Group. 2011. Berau Forest Carbon Program, 2011–2015: Berau’s contri-bution to the world. Indonesia: Tanjung Redeb. http://tfcakalimantan.org/wp-content/uploads/2012/12/BFCP-English_Compile-Done.pdf.
Carlson, K.M., L.M. Curran, G.P. Asner, A.M. Pittman, S.N. Trigg, and J.M. Adeney. 2012. Carbonemissions from forest conversion by Kalimantan oil palm plantations. Nature Climate Change 2,no. 10: 1–5.
Castree, N. 2010. Neoliberalism and the biophysical environment 1: What ‘neoliberalism’ is, andwhat difference nature makes to It. Geography Compass 4, no. 12: 1725–33.
Colfer, C.J.P. 2008. The longhouse of the tarsier: Changing landscapes, gender and well being inBorneo. Phillips, ME: Borneo Research Council.
Colfer, C.J.P., and R.G. Dudley. 1993. Shifting cultivators of Indonesia: Managers or marauders ofthe forest? Rice production and forest use among the Uma’ Jalan of East Kalimantan. Rome:Food and Agriculture Organization.
Conklin, H.C. 1957.Hanunóo agriculture: A report on an integral system of shifting cultivation in thePhilippines. Rome: Food and Agriculture Organization.
Contreras-Hermosilla, A., and C. Fay. 2005. Strengthening forest management in Indonesia throughland tenure reform. Washington, D.C.: Forest Trends.
Deacon, R.T. 1994. Deforestation and the rule of law in a cross-section of countries. Land Economics70, no. 4: 414–30.
Dounias, E., A. Selzner, M. Koizumi, and P. Levang. 2007. From sago to rice, from forest to town:The consequences of sedentarization for the nutritional ecology of Punan former hunter-gatherersof Borneo. Food and Nutrition Bulletin 28 (2 SUPPL.): S294–S302.
Dove, M.R. 1983. Theories of swidden agriculture, and the political economy of ignorance.Agroforestry Systems 1, no, 2: 85–99.
Dove, M.R. 1993. Smallholder rubber and swidden agriculture in Borneo: A sustainable adaptation tothe ecology and economy of the tropical forest. Economic Botany 47, no. 2: 136–47.
Dove, M.R. 1998. Living rubber, dead land, and persisting systems in Borneo; indigenous represen-tations of sustainability. Bijdragen 154, no. 1: 20–54.
Elmhirst, R., M. Siscawati, and C.J.P. Colfer. forthcoming. Revisiting gender and forestry in LongSegar: Divided aspirations in the context of commercial oil palm expansion in EastKalimantan, Indonesia. In Gender and forests: Climate change, tenure, value chains, and emer-ging issues, ed. C.J.P. Colfer, B.S. Basnett, and M. Elias. London: Earthscan.
Fishbein, G., and D. Lee. 2015. Early lessons from jurisdictional REDD+ and low emissions devel-opment programs. Arlington, Virginia. http://www.nature.org/media/climatechange/REDD+_LED_Programs.pdf.
Fox, J., Y. Fujita, D. Ngidang, N. Peluso, L. Potter, N. Sakuntaladewi, J. Sturgeon, and D. Thomas.2009. Policies, political-economy, and swidden in Southeast Asia. Human Ecology 37, no. 3:305–22.
Geist, H.J., and E.F. Lambin. 2002. Proximate causes and underlying driving forces of tropical defor-estation. BioScience 52, no. 2: 143–50.
Gingold, B., A. Rosenbarger, Y.I. Ketut, D. Muliastra, F. Stolle, I.M. Sudana, M. Dwi, et al. 2012.How to identify degraded land for sustainable palm oil in Indonesia. Washington, D.C.:World Resources Institute and Sekala.
Gönner, C. 2000. Resource management in a Dayak Benuaq village: Strategies, dynamics, and pro-spects. Eschborn: GTZ.
The Journal of Peasant Studies 1085
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
Griscom, B.W., P.W. Ellis, A. Baccini, and D. Marthinus. 2016. Synthesizing global and local data-sets to estimate jurisdictional forest carbon fluxes in Berau, Indonesia. PLoS ONE 11, no. 1:e0146357.
Hartanto, H., T.S. Yulianto, and T. Hidayat. 2014. SIGAP-REDD+: Aksi Inspiratif Warga UntukPerubahan Dalam REDD +. Jakarta: TNC.
Himley, M. 2008. Geographies of environmental governance: The nexus of nature and neoliberalism.Geography Compass 2: 433–51.
Inoue, M., and A.M. Lahjie. 1990. Dynamics of swidden agriculture in East Kalimantan. AgroforestrySystems 12 (3): 269–84.
Jessup, T.C. 1991. Persistence and change in the practice of shifting cultivation in the Apo Kayan,East Kalimantan, Indonesia. Bulletin of the Indo-Pacific Prehistory Association 10: 218–25.
Jordan, A., R.K.W. Wurzel, and A.R. Zito. 2003. ‘New’ instruments of environmental governance:Patterns and pathways of change. Environmental Politics 12, no. 1: 1–24.
“Kesepakatan Dan Komitmen Masyarakat Kampung Gunung Madu [Agreement and commitment ofthe population of Gunung Madu village].” 2013, Gunung Madu, Berau, Indonesia, November 13.
Lapeyre, R., H. Hartanto, and R. Pirard. 2015. Designing incentive agreements for conservation: Aninnovative approach. Jakarta: TNC.
Li, T.M., ed. 1999. Transforming the Indonesian uplands: Marginality, power and production.New York: Routledge.
Li, T.M., ed. 2007. The will to improve: Governmentality, development, and the practice of politics.Durham, NC: Duke University Press.
Margulis, S. 2004. Causes of deforestation in the Brazilian Amazon. World Bank Working Paper 22.Washington, DC: The World Bank.
McCarthy, J., and S. Prudham. 2004. Neoliberal nature and the nature of neoliberalism. Geoforum 35(3): 275–83.
Mertz, O., C. Padoch, J. Fox, R.A. Cramb, S.J. Leisz, N.T. Lam, and T.D. Vien. 2009. Swidden change inSoutheast Asia: Understanding causes and consequences. Human Ecology 37, no. 3: 259–64.
Moeliono, I., W.H. Winarno, E.D. Hartono, and T.S. Yulianto. 2010. Masyarakat Kampung DiKabupaten Berau Dan Rekomendasi Keterlibatannya Dalam Skema Pengurangan Emisi DariDeforestasi Dan Degradasi Hutan (REDD). http://tfcakalimantan.org/wp-content/uploads/2013/05/Laporan-Kajian-Keterlibatan-Masyarakat-Kampung-dalam-REDD-di-Berau1.pdf.
Myers, R., and F. Ardiansyah. 2014.Who holds power in land-use decisions? Implications for REDD+ in Indonesia. Info Brief 100. Bogor, Indonesia: CIFOR.
Padoch, C., K. Coffey, O. Mertz, S.J. Leisz, J. Fox, and R.L. Wadley. 2007. The demise of swidden inSoutheast Asia? Local realities and regional ambiguities. Geografisk Tidsskrift-Danish Journal ofGeography 107, no. 1: 29–41.
Peluso, N.L. 2005. Seeing property in land use: Local territorializations in West Kalimantan,Indonesia. Geografisk Tidsskrift-Danish Journal of Geography 105, no. 1: 1–15.
Peluso, N.L., and P. Vandergeest. 2001. Genealogies of the political forest and customary rights inIndonesia, Malaysia, and Thailand. The Journal of Asian Studies 60, no. 3: 761–812.
Pirard, R., and R. Lapeyre. 2013. Building a rapid assessment tool to help design community incentiveagreements in the context of the Berau Forest Carbon Program: The example of 2 model villages(Long Kelay, Gunung Madu).
Rights and Resources Initiative. 2015. Who owns the world’s lands? A global baseline of formallyrecognized indigenous and community land rights. Washington, D.C.: RRI.
Scott, J.C. 1976. The moral economy of the peasant: Subsistence and rebellion in Southeast Asia.New Haven, CT: Yale University Press.
Scott, J.C. 1998. Seeing like a state: How certain schemes to improve the human condition havefailed. New Haven, CT: Yale University Press.
Scott, J.C. 2009. The art of not being governed: An anarchist history of upland Southeast Asia. NewHaven, CT: Yale University Press.
Scrivener, A. 2013. Banking while Borneo burns. World Development Movement. http://www.globaljustice.org.uk/sites/default/files/files/resources/banking_while_borneo_burns_0.pdf.
Sitorus, S., P. Levang, E. Dounias, D. Mamung, and D. Abot. 2004. Potret Punan Kalimantan Timur:Sensus Punan 2002–2003. Bogor, Indonesia: CIFOR.
Sunderlin, W.D., A.M. Larson, A.E. Duchelle, E. Sills, C. Luttrell, P. Jagger, S.K. Pattanayak, et al.2015. Technical guidelines for research on REDD+ subnational initiatives. Bogor, Indonesia:CIFOR.
1086 Gregory M. Thaler and Cut Augusta Mindry Anandi
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017
TNC. 2013. Our partners in conservation. http://www.nature.org/about-us/our-partners/index.htm.TNC-Indonesia. n.d. TNC conditional payment agreement with Gunung Madu Village.van Vliet, N., O. Mertz, A. Heinimann, T. Langanke, U. Pascual, B. Schmook, C. Adams, et al. 2012.
Trends, drivers and impacts of changes in swidden cultivation in tropical forest-agriculture fron-tiers: A global assessment. Global Environmental Change 22, no. 2: 418–29.
Gregory M. Thaler is a PhD candidate in the Department of Government at Cornell University. Heresearches tropical forest governance in Brazil and Indonesia and the role of transnational environ-mental NGOs in environmental politics. In 2014–2015 he was an affiliated researcher with theCenter for International Forestry Research in Bogor, Indonesia. Email: [email protected]
Cut Augusta Mindry Anandi is a coordinator for sustainable land use planning at the United StatesAgency for International Development (USAID) LESTARI project in Indonesia. She works on land-scape approaches for sustainable land use through zoning, spatial planning, strategic environmentalassessment, and licensing tools at the sub-national level in three key landscapes in Indonesia:Aceh, Central Kalimantan and Papua. In 2012–2016 she was a research officer and field researchsupervisor at the Center for International Forestry Research working on REDD+ projects in Acehand East Kalimantan and smallholder oil palm certification in Riau, Sumatra. Email: [email protected]
The Journal of Peasant Studies 1087
Dow
nloa
ded
by [
202.
62.1
7.17
1] a
t 06:
55 0
6 N
ovem
ber
2017