CIAT Research Online - Accepted Manuscript Alternatives to land grabbing: exploring conditions for smallholder inclusion in agricultural commodity chains in Southeast Asia The International Center for Tropical Agriculture (CIAT) believes that open access contributes to its mission of reducing hunger and poverty, and improving human nutrition in the tropics through research aimed at increasing the eco-efficiency of agriculture. CIAT is committed to creating and sharing knowledge and information openly and globally. We do this through collaborative research as well as through the open sharing of our data, tools, and publications. Citation: Cramb, Rob; Manivong, Vongpaphane; Newby, Jonathan C.; Sothorn, Kem; Sibat, Patrick S.. 2017. Alternatives to land grabbing: exploring conditions for smallholder inclusion in agricultural commodity chains in Southeast Asia. The Journal of Peasant Studies 44(4): 939-967. Publisher’s DOI: https://doi.org/10.1080/03066150.2016.1242482 Access through CIAT Research Online: http://hdl.handle.net/10568/83343 Terms: © 2018. CIAT has provided you with this accepted manuscript in line with CIAT’s open access policy and in accordance with the Publisher’s policy on self-archiving. This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License. You may re-use or share this manuscript as long as you acknowledge the authors by citing the version of the record listed above. You may not use this manuscript for commercial purposes. For more information, please contact CIAT Library at CIAT-Library@cgiar.org. PROOF COVER SHEET Author(s): Rob Cramb Article Title: Alternatives to land grabbing: exploring conditions for smallholder inclusion in agricul- tural commodity chains in Southeast Asia Article No: FJPS1242482 Enclosures: 1) Query sheet 2) Article proofs Dear Author, 1. Please check these proofs carefully. It is the responsibility of the corresponding author to check these and approve or amend them. A second proof is not normally provided. Taylor & Francis cannot be held responsible for uncorrected errors, even if introduced during the production process. Once your corrections have been added to the article, it will be considered ready for publication. Please limit changes at this stage to the correction of errors. You should not make trivial changes, improve prose style, add new material, or delete existing material at this stage. You may be charged if your corrections are excessive (we would not expect corrections to exceed 30 changes). For detailed guidance on how to check your proofs, please paste this address into a new browser window: http://journalauthors.tandf.co.uk/production/checkingproofs.asp Your PDF proof file has been enabled so that you can comment on the proof directly using Adobe Acrobat. If you wish to do this, please save the file to your hard disk first. For further information on marking corrections using Acrobat, please paste this address into a new browser window: http:// journalauthors.tandf.co.uk/production/acrobat.asp 2. Please review the table of contributors below and confirm that the first and last names are structured correctly and that the authors are listed in the correct order of contribution. This check is to ensure that your name will appear correctly online and when the article is indexed. Sequence Prefix Given name(s) Surname Suffix 1 Rob Cramb 2 Vongpaphane Manivong 3 Jonathan C. Newby 4 Kem Sothorn 5 Patrick S. Sibat Queries are marked in the margins of the proofs, and you can also click the hyperlinks below. Content changes made during copy-editing are shown as tracked changes. Inserted text is in red font and revisions have a red indicator . Changes can also be viewed using the list comments function. To correct the proofs, you should insert or delete text following the instructions below, but do not add comments to the existing tracked changes. AUTHOR QUERIES General points: 1. Permissions: You have warranted that you have secured the necessary written permission from the appropriate copyright owner for the reproduction of any text, illustration, or other material in your article. Please see http://journalauthors.tandf.co.uk/permissions/usingThirdPartyMaterial.asp. 2. Third-party content: If there is third-party content in your article, please check that the rightsholder details for re-use are shown correctly. 3. Affiliation: The corresponding author is responsible for ensuring that address and email details are correct for all the co-authors. Affiliations given in the article should be the affiliation at the time the research was conducted. Please see http://journalauthors.tandf.co.uk/preparation/ writing.asp. 4. Funding: Was your research for this article funded by a funding agency? If so, please insert ‘This work was supported by ’, followed by the grant number in square brackets ‘[grant number xxxx]’. 5. Supplemental data and underlying research materials: Do you wish to include the location of the underlying research materials (e.g. data, samples or models) for your article? If so, please insert this sentence before the reference section: ‘The underlying research materials for this article can be accessed at / description of location [author to complete]’. If your article includes supplemental data, the link will also be provided in this paragraph. See for further explanation of supplemental data and underlying research materials. 6. The CrossRef database (www.crossref.org/) has been used to validate the references. Changes resulting from mismatches are tracked in red font. QUERY NO. QUERY DETAILS AQ1 The reference “Cramb 2016” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ2 The reference “Hall, 2004” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ3 The reference “Li, 2002” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ4 The reference “Li, 2014” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. QUERY NO. QUERY DETAILS AQ5 The reference “Shiel et al. 1999” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ6 The reference “Mahanty and Milne 2006” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ7 The reference “Mohns and Laity 2010” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ8 The reference “Ducourtieux et al. 2005” is cited in the text but is not listed in the references list. Please either delete the in- text citation or provide full reference details following journal style. AQ9 The disclosure statement has been inserted. Please correct if this is inaccurate. AQ10 Funding details have been taken from information supplied with your manuscript submission. Please correct if this is inaccurate. AQ11 Funding details have been taken from information supplied with your manuscript submission. Please correct if this is inaccurate. AQ12 The CrossRef database (www.crossref.org/) has been used to validate the references. Mismatches between the original manuscript and CrossRef are tracked in red font. Please provide a revision if the change is incorrect. Do not comment on correct changes AQ13 Please provide missing city for the reference “Byerlee and Haggblade 2014” references list entry. AQ14 The reference “McCarthy and Cramb 2009” is listed in the references list but is not cited in the text. Please either cite the reference or remove it from the references list. AQ15 The reference “Mahanty and Milne 2016” is listed in the references list but is not cited in the text. Please either cite the reference or remove it from the references list. AQ16 Please provide missing page range for reference “Manivong and Cramb 2008b” references list entry. AQ17 The reference “Sheil et al. 2009” is listed in the references list but is not cited in the text. Please either cite the reference or remove it from the references list. AQ18 Note 1: Please provide page number for the quotation “adverse incorporation”. AQ19 “at least 5–10,000 ha”: If this number range means 5 ha to 10,000 ha, it may stand. However, if it is meant to be 5000 ha to 10,000 ha, please change this to “at least 5000– 10,000 ha”. AQ20 Note 11: The reference “Floresteca Group website” is cited in the text but is not listed in the reference list. Please either delete the in-text citation or provide full reference details following journal style. If retained, please cite by author (or organization) and year. AQ21 Please provide page number(s) for the quotations “everyday processes of accumulation and dispossession” and “small-scale land grabs”. QUERY NO. QUERY DETAILS AQ22 Table 4: Please change both instances of the table text “Land and Forest Allocation managed by village” to either “Land and Forest Allocation programme managed by village” or “Land and forest allocation managed by village”, according to your intended meaning. How to make corrections to your proofs using Adobe Acrobat/Reader Taylor & Francis offers you a choice of options to help you make corrections to your proofs. Your PDF proof file has been enabled so that you can edit the proof directly using Adobe Acrobat/Reader. This is the simplest and best way for you to ensure that your corrections will be incorporated. If you wish to do this, please follow these instructions: 1. Save the file to your hard disk. 2. Check which version of Adobe Acrobat/Reader you have on your computer. You can do this by clicking on the “Help” tab, and then “About”. If Adobe Reader is not installed, you can get the latest version free from http://get.adobe.com/reader/. 3. If you have Adobe Acrobat/Reader 10 or a later version, click on the “Comment” link at the right-hand side to view the Comments pane. 4. You can then select any text and mark it up for deletion or replacement, or insert new text as needed. Please note that these will clearly be displayed in the Comments pane and secondary annotation is not needed to draw attention to your corrections. If you need to include new sections of text, it is also possible to add a comment to the proofs. To do this, use the Sticky Note tool in the task bar. Please also see our FAQs here: http://journalauthors.tandf.co.uk/production/index.asp. 5. Make sure that you save the file when you close the document before uploading it to CATS using the “Upload File” button on the online correction form. If you have more than one file, please zip them together and then upload the zip file. If you prefer, you can make your corrections using the CATS online correction form. Troubleshooting Acrobat help: http://helpx.adobe.com/acrobat.html Reader help: http://helpx.adobe.com/reader.html Please note that full user guides for earlier versions of these programs are available from the Adobe Help pages by clicking on the link “Previous versions” under the “Help and tutorials” heading from the relevant link above. Commenting functionality is available from Adobe Reader 8.0 onwards and from Adobe Acrobat 7.0 onwards. Firefox users: Firefox’s inbuilt PDF Viewer is set to the default; please see the following for instructions on how to use this and download the PDF to your hard drive: http://support.mozilla.org/ en-US/kb/view-pdf-files-firefox-without-downloading-them#w_using-a-pdf-reader-plugin Alternatives to land grabbing: exploring conditions for smallholder inclusion in agricultural commodity chains in Southeast Asia Rob Cramb, Vongpaphane Manivong, Jonathan C. Newby, Kem Sothorn and Patrick S. Sibat Smallholder engagement with export commodities in Southeast Asia potentially offers a more inclusive development pathway than large-scale plantation production, which has been associated with the phenomenon of land grabs. This raises three questions which we explore in this paper: What are the agro-economic factors favouring or obstructing smallholders relative to plantations? What are the incentives for agribusiness firms to contribute to smallholder production other than by direct control of land? Can smallholder production be broadly inclusive in the face of internal differentiation and encroachment by external investors? We compare smallholder involvement with four cash crops which have experienced strong market demand – rubber, oil palm, cassava and teak – based on fieldwork in six Southeast Asian countries. We conclude that smallholder production can be a viable and inclusive strategy, contingent on the case- by-case confluence of a number of key factors which we enumerate. Keywords: smallholders; agribusiness; plantations; cash crops; value chains; agrarian differentiation; Southeast Asia Introduction Much of the conflict over land acquisitions in Southeast Asia has been associated with the expansion of large-scale agribusiness investment in cash crops, such as rubber and oil palm, for which export demand has been increasing (Hall 2011; Hall, Hirsch, and Li 2011; Borras et al. 2011; Deininger 2011). Though rent-seeking, speculation, and amassing ‘land banks’ to impress stock exchanges are part of the motivation for land acquisitions, the fundamental driver is the underlying profitability of producing these export commodities, given strong market conditions (Byerlee 2013; Cramb 2013a; Fairbairn 2014; McAllister 2015; Baird and Fox 2015). The state has frequently underpinned this profitability with policies that provide land and labour at low prices to favoured agribusiness companies, with minimal regulation (Hall 2011; Cramb and McCarthy 2016), but the long-term growth in demand for the end products, particularly from China and India, is the driving force behind land investment. The expansion of investment in these cash crops has been associated with a resurgence in the plantation or estate mode of production throughout the region (Byerlee 2014; Cramb and McCarthy 2016), hence the perceived need for the state to constrain or displace existing small-scale landholders to make way for large plantations. It is thus the association between the growing regional demand for cash crops and the plantation mode that is at the heart of the ‘land grabbing’ phenomenon in Southeast Asia. © 2016 Informa UK Limited, trading as Taylor & Francis Group The Journal of Peasant Studies, 2016 http://dx.doi.org/10.1080/03066150.2016.1242482 5 10 15 20 25 30 35 40 45 FJPS1242482 Techset Composition India (P) Ltd., Bangalore and Chennai, India 10/11/2016 Yet, despite the political dominance of these plantation interests, in both the ‘first era of globalisation’ under high colonialism and in the ‘second era of globalisation’ of the last few decades there has been an upsurge in smallholder production of export commodities (Byerlee 2013). The widespread smallholder impulse to engage in cash crop production in Southeast Asia offers a potentially more inclusive development pathway than large- scale plantation production (Hayami 1998, 2010; Byerlee 2013). By ‘inclusive’ we mean a process in which large numbers of small farm households are enabled to benefit finan- cially from participating in a profitable commodity chain, without being locked into low productivity, incurring burdensome debt, or risking the loss of their land assets.1 In this paper we temporarily put aside the question of the political economy of planta- tions (Hayami 2010; Cramb 2016AQ1 ¶ ) in order to focus on a series of issues affecting the via- bility of a smallholder-oriented development strategy: . What are the agro-economic factors favouring or obstructing smallholder modes of cash crop production relative to large-scale production entities (estates, plantations, land concessions)? . What are the incentives for agribusiness firms to contribute to smallholder cash crop production through roles other than direct farm management? . Can smallholder cash crop production be made broadly inclusive in the face of ten- dencies towards agrarian differentiation and the operation of emerging land markets? To address these questions we compare a variety of modes of smallholder participation in commodity chains for four different cash crops – rubber, oil palm, cassava and teak – and illustrate this with examples and case studies from six Southeast Asian countries – Myanmar, Cambodia, Laos, Vietnam, Malaysia and Indonesia. These crops were selected because of their contemporary importance to agrarian change in major parts of Southeast Asia, differences in their agro-economic characteristics, the availability of a good compara- tive literature, and our involvement in localised case studies of smallholder production of each crop which has stimulated our thinking on these larger issues. Including other crops (e.g. sugarcane, banana and shrimp) would have been desirable but was not feasible within the limits of one paper. The analysis is exploratory in nature and attempts to build on the contributions of Hall (2004AQ2 ¶ , 2011), Hall, Hirsch, and Li (2011), Li (2002AQ3 ¶ , 2014 AQ4 ¶ ), McCarthy (2010) and others who have in different ways pursued a conjunctural approach to one or more of these ques- tions. This approach involves systematically examining the way a number of elements (from the agro-ecological to the political) come together at particular conjunctures to influ- ence agrarian structures and trajectories, and the way in which these conjunctures influence subsequent developments through path-dependence. Conjunctures are complex, dynamic and difficult to unravel, but they are not random (Li 2014, 16). Hence, it is possible to build up an understanding of key elements and interactions through comparative case studies. We first analyse the potential roles of the different actors in the production and market- ing of export commodities and examine how various institutional arrangements combine these roles in different ways. For each of the four commodities we then give (1) a brief over- view of the modes of production used in Southeast Asia and (2) a case study derived from our own fieldwork, illustrating some of the ways these alternative arrangements work out in 1These latter features are characteristic of situations described as ‘adverse incorporation’ (McCarthy 2010).AQ18 ¶ 2 Rob Cramb et al. 50 55 60 65 70 75 80 85 90 practice. Following this we provide a comparative analysis of the four cases in which we identify a large number of factors related to agro-economic characteristics, agribusiness incentives and tendencies to agrarian differentiation that we suggest can contribute to favourable or unfavourable conjunctures for inclusive smallholder development. The analy- sis is thus necessarily ‘under-determined’ – there are too many factors and too few cases to derive definitive causal explanations. However, we put forward this analysis as a contri- bution to understanding the conditions for smallholder modes of production to provide an alternative to the large-scale land investments associated with ‘land grabbing’. Smallholders, largeholders and in between The ability of small family farms to out-compete large-scale centrally managed production operations has long been argued – though with widely differing interpretations (Chayanov [1966] 1925; Netting 1993; Allen and Lueck 1998; Hayami 1998, 2010; Eastwood, Lipton, and Newell 2010; Bernstein 2010, ch. 6; Byerlee 2013; Van Vliet et al. 2015). The advan- tages derive from considerations such as the seasonal variability and complexity of agricul- tural production, the intrinsic motivation of family labour to support the household as a unit, the flexibility with which family labour can be deployed in space and time (both on- and off-farm), the low supervision costs (family workers are also typically joint managers), the greater local (site-specific) knowledge of family workers, and the diversity of household livelihood activities, promoting greater economic resilience. Hence (for example), in tra- ditional settings where land is unequally distributed, large landowners have typically pre- ferred dividing their land among share-croppers to hiring and supervising labour with the attendant costs and risks (Stiglitz 1986). However, large-scale entities may have an economic advantage in activities where there are significant economies of scale and financial risk (Hayami 2010). These may include mechanised land development, mobilising finance, acquiring (high-quality) production inputs, and of course processing and marketing. None of these in itself necessitates large-scale production or long-term control of land. For example, land development on the ‘agricultural frontier’ exhibits economies of scale where it involves mechanised land clearing, terracing, drainage, soil improvement (e.g. heavy doses of lime and phosphorus in acid upland soils), road construction and finance. However, as Byerlee, Masters, and Robinson (2015) have highlighted, there are historical examples in both developed and developing countries where this function has been the precursor to sub-dividing and selling the developed land to small-scale farmers who then take over the management of production (essentially a ‘sites-and-services’ approach to rural development). The small- holder settler schemes for rubber and oil palm pioneered by the Federal Land Development Authority (FELDA) in Malaysia from the 1960s to the 1980s also illustrate this approach, though with a state agency rather than a private firm acting as the land developer (Pletcher 1991; Sutton 2001). Likewise, economies of scale in processing often favour large-scale agribusiness enti- ties with the necessary capital, technology and skills. Whether this also favours large-scale production depends on the stage of development of the industry as well as policy decisions (Hayami 2010). In order to achieve the economies of scale, processors need a reliable throughput of raw material; hence, in the pioneer phase it may make economic sense for them to invest in large-scale production to assure this supply. However, if there are suffi- cient small-scale producers within the catchment of a processing plant, a steady flow of raw material may be forthcoming without incurring the higher costs of centralised pro- duction. The influence of government policy in this commercial decision can be seen in The Journal of Peasant Studies 3 95 100 105 110 115 120 125 130 135 the case of oil palm. In Thailand, mills to process fresh fruit bunches can be established as independent business entities (Dallinger 2011) but in Malaysia a company needs to have a minimum area of oil palm (5000 ha) before it can obtain a licence to establish a mill, thereby effectively mandating large vertically integrated operations, at least at the outset (Cramb and McCarthy 2016).2 Thus agribusiness firms (small and large) and/or state development agencies can make a variety of contributions to commodity value chains, singly or in various combinations, as enumerated in Table 1. The incentives for a firm to contribute any given function vary with the agro-economic characteristics of the commodity, the state of rural infrastructure, and the politico-legal context (Hayami 2010; Hall 2011). Where these combine to provide strong incentives for large-scale, vertically integrated operations, there will be pressure to squeeze out smallholder activity, increasing the likelihood of (though not neces- sitating) a ‘land grab’. However, the incentives might be such that large firms concentrate on just some of these functions (e.g. input supply, processing) and rely on smallholders to undertake (minimally) the management of crop production. These incentives typically shift over time, for both smallholders and agribusinesses, for example as industry infrastructure is developed, as smallholders’ knowledge and skills increase, as the volume of production increases and as government policies change. Such shifts can account for the historical resurgence of smallholder commodity production, as well as the contemporary impetus for smallholders to capitalise on profitable market conditions for cash crops. Various typologies have been developed to capture the ways in which smallholders, agri- businesses and state agencies can combine the roles or functions listed in Table 1 (Cramb and Curry 2012;Byerlee andHaggblade 2014;Byerlee et al. 2014). The typology inTable 2 gives us a useful framework for analysing the cases considered here. The table shows a continuum of modes of production from independent smallholders to fully integrated agribusiness con- cerns. The intermediate typesmay need some clarification. ‘Supported smallholders’ are con- sidered to be farmerswhohave full autonomyover the production process and the sale of their produce but receive support, typically but not necessarily from a state development agency, in the form of inputs (e.g. subsidised high-yielding planting material) and technical advice. ‘Organised smallholders’ are those who, with or without facilitation or incentives from state agencies, organise themselves into groups to acquire inputs, share technical knowledge and/ or to assemble, process and sell their produce.3 ‘Contract smallholders’ enter into a formal or informal contract with a trader/processor in which inputs are provided on condition that the product meets specified criteria and is sold to the contractor (who in this way recovers his/her investment). ‘Outgrowers’ are smallholders (whether the original landowners or new settlers) who are tied to a nucleus estate from which they receive inputs and varying degrees of tech- nical supervision and towhich their produce is committed. ‘Managed smallholders’ are those whose land is managed on a large scale by a state agency while they continue to occupy and operate their individual lots. ‘Joint ventures’ go a step further towards the plantation mode in that smallholders pool their land in a large estate which is managed entirely by a private company in which smallholders hold shares in proportion to the area of land they have contributed. 2Hayami (2010) makes a similar point about the influence of government policy on the early structure of the rubber industry in Malaysia (where the colonial government favoured British plantations) compared with Thailand (where a smallholder policy prevailed). 3These are sometimes referred to loosely as ‘cooperatives’ but do not conform to the traditional defi- nition of a producer cooperative or a multi-purpose cooperative which both require much higher levels of organisation and are rarely successful. 4 Rob Cramb et al. 140 145 150 155 160 165 170 175 180 Smallholder rubber in Luang Namtha province, Laos Rubber (Hevea brasiliensis) began as an estate crop in Insular Southeast Asia (and southern Vietnam) around the turn of the twentieth century but was also taken up spontaneously by smallholders at an early stage, and eventually smallholders came to dominate the sector, despite fierce opposition at times from plantation interests (Barlow 1978; Byerlee 2013; De Koninck and Rousseau 2012). Rubber has agro-economic features that make it suitable for either smallholders or large-scale plantations, depending on the context. While sup- ported smallholders can do well with rubber, there is a strong incentive for agribusiness firms to control the entire operation from land preparation to processing in a single verti- cally integrated operation, provided they can obtain secure tenure over large land conces- sions and a ready supply of cheap, ‘disciplined’ labour. Hence Fox and Castella (2013, 157) conclude that ‘in places where land rights and support services are provided smallholder rubber cultivation is viable and profitable, and in places where they are not provided small- holders face a much more insecure future’. In the post-war era, though direct linkages with industry remained weak, many rubber smallholders in the traditional growing areas were given significant state support, as well as Table 1. Potential contributions of agribusiness firms and/or state development agencies to production and marketing of industrial crops. Function Details Land development Firm/agency undertakes broad-scale, mechanised land clearing and development, possibly including infrastructure (roads, houses, social infrastructure) Land settlement Firm/agency surveys/allocates individual smallholder lots; agency issues titles, often subject to repayment of development credit and other conditions Production inputs Firm/agency provides planting material, fertiliser and other inputs, typically as credit-in-kind Knowledge Firm/agency provides training and advice to smallholders in crop production, processing techniques and land management Finance Agency provides direct grants to smallholders; firm/agency provides loans to smallholders; firm/agency obtains finance for other functions, e.g. land development Labour management Firm/agency recruits and accommodates plantation labour force, whether locally or internationally Field operations Firm/agency provides direct oversight of production (planting, maintenance, harvesting); work undertaken by plantation labour force Harvesting Firm undertakes contract harvesting of crop, whether mechanised or by harvesting gangs; may involve harvesting fee or purchase of standing crop Acquisition (purchase, assembly, transportation) Firm/agency undertakes purchase, assembly and/or transportation of harvested crop to processor and/or wholesaler Processing Firm/agency transforms raw product into intermediate or final product Marketing (wholesale, retail, export) Firm/agency trades/exports intermediate or final product; agency manages trade/exports to control prices and/or raise revenue Facilitation State agency brokers/enforces agreement between landholders and company for the latter to undertake one or more functions in return for land rent, profit share or other payment The Journal of Peasant Studies 5 185 190 195 200 205 210 215 220 225 Table 2. Division of roles between smallholders and agribusiness firms or public agencies in different modes of production. Function (see Table 1) Independent smallholders Supported smallholders Organised smallholders Contract smallholders Outgrowers (of a nucleus estate) Managed smallholders Joint-venture smallholders Integrated agribusiness company Land development NA NA NA NA + + + + Land settlement NA NA NA NA + + + + Production inputs + + + + + + Knowledge + + + + + + Finance + + + + + + Labour management (+) + + Field operations (+) (+) + Harvesting (+) (+) (+) + Acquisition + + + + + + + Processing + + + + + + + + Marketing + + + + + + + + Facilitation NA + + + + + + NA Note: Shaded cells indicate smallholder role; + indicates agribusiness firm or government agency role; (+) indicates agribusiness firm or government agency may fulfil this role in some cases; NA indicates not applicable. 6 R ob C ram b et al. 230 235 240 245 250 255 260 265 270 benefiting from state and industry investment in rubber research. In Malaysia, FELDA established managed smallholder schemes on state land for landless settlers, and both the Federal Land Consolidation and Rehabilitation Authority (FELCRA) and the Rubber Industry Smallholder Development Authority (RISDA) provided inputs and management for existing smallholders in so-called mini-estates (Barlow 1978). Local Departments of Agriculture also implemented subsidised planting schemes for smallholders. In Thailand, the Office of the Rubber Replanting Aid Fund (ORRAF) has long supported smallholders with high-yielding planting material, technical advice, credit, and assistance with coopera- tive processing and marketing (Chambon et al. 2016). In Indonesia, several types of block planting scheme for rubber smallholders were initiated, culminating in the nucleus estate and smallholder (NES) schemes, with the nucleus estate being established by a state- owned or private plantation company. However, widespread state support for rubber small- holders through extension services and provision of planting material was lacking, resulting in low yields and incomes (Barlow and Jayasuriya 1984; Barlow and Tomich 1991). Rubber remains predominantly a smallholder crop in these traditional producing areas. However, the upsurge in prices from 2002 to 2011 has seen the expansion of rubber by more than 1 million ha in upland regions of Mainland Southeast Asia where rubber trees were not traditionally planted (Fox and Castella 2013), drawing on earlier pioneering work by Chinese agricultural scientists who successfully pushed the geographic limits of rubber cultivation to these higher latitudes and altitudes. This expansion has been under- taken by both independent and contracted smallholders (e.g. in northeast Thailand and northern Laos) and private plantation companies utilising large land concessions, particu- larly in southern Laos, parts of Cambodia, and Myanmar, but also in northwest Vietnam (Baird 2010; Fox and Castella 2013; Dao 2015; McAllister 2015). This is illustrated by the recent rubber boom in Laos. According to official estimates, the area of rubber in Laos rose from only 900 ha in 2003 to 234,000 ha in 2010. Nearly 60 percent of the total rubber area in 2010 was in the Northern Region, while around 17 percent was in the Central Region and 23 percent in the Southern Region. In terms of modes of production, about 61 percent of the total rubber area in 2010 was reportedly in large estates held under land concessions, 13 percent was planted under contract farming systems and 26 percent was planted by independent smallholders (NAFRI 2011). Conces- sions held by Thai investors dominated in the Central Region, while Vietnamese-held con- cessions dominated in the Southern Region. The proliferation of concessions, particularly in southern Laos, has had major negative impacts on rural livelihoods as villagers with inse- cure tenure lose access to some or all of their farming and forest land (Baird 2010; Laun- garamsri 2012). The promise of local employment on these rubber estates is likely to be overtaken by the use of large numbers of Vietnamese workers to meet the skilled labour requirements of the companies. In Northern Laos, however, where Chinese businesses dominated, large concessions were less prevalent and the majority of the planted area was equally divided between inde- pendent smallholders and contracted smallholders. The government’s Land and Forest Allocation (LFA) programme was implemented in this region earlier and more extensively than in the south, thereby making household and village tenure somewhat more secure. In Luang Namtha Province, one of the main rubber-growing areas in the north, the Provincial Agriculture and Forestry Office (PAFO) reported a total planted area of just over 33,000 ha in 2015, of which 11 percent was in concessions, 49 percent was held by independent small- holders (some of whom can be considered ‘supported’ or ‘organised’), and 40 percent was under some form of contract farming arrangement (Vongvisouk and Dwyer 2016, 45). These two forms of smallholding are now considered in turn. The Journal of Peasant Studies 7 275 280 285 290 295 300 305 310 315 Hadyao, a smallholder swidden and rubber village in Northern Laos, has become the paradigmatic case of independent (i.e. non-contracted) smallholdings (Manivong and Cramb 2008a, 2008b, 2015; Vongvisouk and Dwyer 2016). Hadyao is a Hmong village in Namtha District of Luang Namtha Province, close to the district centre and near the main road to the Chinese border. In 1994, 14 Hmong households from Yunnan migrated to Hadyao, where they had relatives, and introduced rubber cultivation to the village, drawing on over 15 years’ experience working in a rubber collective in Yunnan. The village headman and local authorities went to Yunnan to explore the possibility of planting rubber. They made a proposal to the provincial authorities and received low-interest loans for rubber cultivation, with repayments deferred until after tapping had commenced. This institutional credit proved crucial. A survey of 95 farm-households in Hadyao was conducted in 2005 and several follow- up visits were made in subsequent years (Manivong and Cramb 2008a, 2008b, 2015). From 1994 to 1996, 341 ha of rubber were planted by individual smallholders, using seed sup- plied by Chinese traders, and in 2002 about 266 ha of rubber trees began to be tapped, making Hadyao the first rubber-producing village in Laos. Many villagers have since expanded their rubber holdings, using seed obtained from existing trees. A further 296 ha of rubber trees were planted during 2003–2005, so the area for swidden agriculture was substantially reduced. Village leaders undertook a survey in 2008 and found that every household had enough rubber trees to fully utilise their household labour force; hence, in that year the village declared ‘no more growing rubber’. By late 2011 there were about 650 ha of rubber planted in the village, of which 360 ha were being tapped. Every household in the village owned a rubber holding and none grew upland rice. Those with immature trees undertook tapping for others with mature plots for a 25 percent share of the output. Some also tapped rubber in other nearby villages. Since tapping began in 2002, rubber had become the major source of income in the village, with farmers benefiting from the boom in prices between 2002 and early 2011. The rubber was sold as ‘tub-lump’ (the coagulated latex from a bucket) to Chinese traders who came to buy at the village, usually once a month. Every month the village auth- orities contacted the buyers in Yunnan by phone, looking for those who offered the highest price. Despite the establishment of a rubber-processing factory in the province by the Sino- Lao Rubber Company in 2004, Hadyao farmers continued to sell their rubber to the Chinese traders as they received a higher price. The village had set up a rubber cooperative, initially to assist new rubber farmers and manage the allocation of land, but also to improve their marketing skills and bargaining power. The cooperative came to include units from sur- rounding villages. This stood them in good stead when prices fell sharply from 2011, enabling them to get a price advantage of around 15 percent over unorganised smallholders (Vongvisouk and Dwyer 2016). Bio-economic modelling of Hadyao smallholders based on conditions in 2005 showed that, while investment in rubber was very profitable, if the price of latex fell to around LAK 5500/kg, further planting would be uneconomic, and if the price fell below LAK 3500/kg, it would no longer be economic for households to tap their existing stands (Manivong and Cramb 2008a).4 This has been confirmed by a recent survey at a time when prices had fallen as low as LAK 4000/kg which found that, while investment had stalled and tapping had declined, farmers were still willing to tap 2–3 ha using family labour at this price, though not to hire labour (Vongvisouk and Dwyer 2016). The rural wage rate had 4USD 1.00 = LAK 10,500. 8 Rob Cramb et al. 320 325 330 335 340 345 350 355 360 doubled in the intervening 10 years but the lower opportunity cost of family labour, especially where no other land was available, made tapping sufficiently worthwhile to con- tinue on this reduced scale. Almost as many rubber farmers in northern Laos began as contracted smallholders. Contracts between Lao farmers and Chinese rubber investors were initially in the form of a ‘2 + 3’ agreement, with ‘2’ referring to the land and labour contributed by the farmer and ‘3’ referring to the capital, technology and marketing provided by the investor. Contracts were for 30–35 years, with options for extension or renegotiation. The revenue distribution was based on the type of agreement and the remoteness of the investment zone; normally 60–70 percent accrued to the farmer and 30–40 percent to the investor. Depending on the particular contract, farmers had a choice whether to sell rubber to the investor, but the investor still received the agreed share of revenue. Rubber was sold at the prevailing market price, whether to the investor or to another trader. The ‘2 + 3’ model of contract farming was promoted widely by the government as a way to provide smallholders with the necessary inputs and to maximise their share of revenue, as well as to ensure they retained their access to the land (NAFRI 2011). While contract farming had the obvious advantage of providing new rubber planters with access to capital and technology, there was a major issue with regard to farmers’ inability to contribute the agreed labour during the long gestation period before the rubber could be tapped. Hence, in many cases what started as a ‘2 + 3’ contract has become a ‘1 + 4’ contract inwhich farmers have ceased to provide unpaid labour, contributing only their land,while the investing company has taken over responsibility for management of the rubber plantation during the pre-tapping years (perhaps employing the landholder who has thereby gained a regular cash income while the rubber is maturing). The agreed revenue sharing has then been reversed, with a lower share for farmers (30–40 percent) and a higher share for compa- nies (60–70 percent). In some instances, the sharing has involved a partitioning of land and trees rather than revenue (Shi 2008; NAFRI 2011; Manivong and Cramb 2015; Vongvisouk and Dwyer 2016). Thus the ‘1 + 4’ model, when aggregated across many landholders, approximates to a joint venture or even an outright land concession, with smallholders effec- tively excluded (Shi 2008; McAllister 2015). In addition to the acquisition of land by companies throughmodified contractual arrange- ments, there has been a tendency for both independent and contracted smallholders to sell their rubber plots to wealthier individuals, often urban-based investors or government offi- cials (Shi 2008; Vongvisouk and Dwyer 2016). Some companies, such as the Yunnan RubberCompany, have also been buying rubber land, fromboth smallholders and other com- panies wanting to exit following the fall in prices. However, the demand for rubber land was at its peak during the 2009–2011 price spike, and interest has waned as prices have tumbled. While there is some evidence that poor or indebted householdswere selling their rubber plots, it was also the case that those with more rubber than they could tap with family labour were selling ‘surplus’ land to businesses with the capital to hire labour. The fall in prices and the demand from Chinese companies to lease land for other crops, especially bananas, had prompted some households to cut down (some or all of) their rubber trees.5 Hence, the ulti- mate effect on the distribution of land is difficult to disentangle. The collective organisation of Hadyao and someother villages to prevent both excess plantingwithin the village and the loss of land to outsiders is noteworthy in this context. 5A report in 2015 indicated that farmers in Luang Namtha had cut down 500–600 ha of rubber trees (Rubber prices may decrease in Luang Namtha 2015). The Journal of Peasant Studies 9 365 370 375 380 385 390 395 400 405 Summarising in terms of the three research questions: (1) the agro-economic character- istics of rubber are generally favourable to smallholders, except for the long gestation period with no income, requiring some form of financial support; (2) there are profitable opportunities for agribusiness firms to support smallholder production with inputs, credit, technical advice and marketing; (3) rubber planting can adversely affect access to land for subsistence production and attract outside investors wanting to buy up established hold- ings, unless there are strong village institutions capable of managing these tendencies. Smallholder oil palm in Sarawak, Malaysia Oil palm (Elaeis guineensis) has been the most extensively planted Southeast Asian cash crop in recent decades (De Koninck and Rousseau 2012). The oil palm grows best in con- tinuously wet, equatorial conditions; hence, it is planted primarily in a zone from southern Thailand, the Malay Peninsula and Sumatra in the west, through the southern Philippines (Palawan andMindanao), Borneo and Sulawesi, to New Guinea in the east (though attempts are being made to extend these limits into Mainland Southeast Asia) (Cramb and Curry 2012). With the surge in demand for vegetable oil, driven primarily by population and econ- omic growth in the large economies of South and East Asia, as well as the more recent growth in demand for biofuels, world palm oil production increased from five million tons in 1980 to 61 million tons in 2015, an annual growth rate of seven percent.6 While Malaysia was the dominant player up to the mid-1990s, Indonesia has accelerated its expan- sion from that point, overtaking Malaysia in terms of planted area in 1997 and surging to over 10 million ha by 2014, compared with Malaysia’s total of just over five million ha (though much of this expansion in Indonesia is due to Malaysian and Singaporean invest- ment). Between them the two countries account for 84 percent of the world’s palm oil pro- duction. Each country exports about USD 18 billion of palm oil and palm oil products – in the case of Malaysia, including almost USD 3 billion in oleochemicals (Cramb and McCarthy 2016). The oil palm industry in Southeast Asia has many examples of land grabbing and is dominated by large plantations (Cramb 2013a), though there is a sizeable and growing smallholder sector in Indonesia, now accounting for 40 percent of the planted area, and in Thailand small- and medium-sized holdings have always predominated. The industry has also experimented with almost all the intermediate arrangements listed in Table 2. Especially in Malaysia and Thailand, but less so in Indonesia, some smallholders have been supported by state-provided or state-subsidised services such as the provision of plant- ing materials, fertiliser and technical advice. Smallholders in Indonesia are now frequently organised into cooperatives or farmer groups as a way of interacting with state services or private plantations. The major approach used in Indonesia has been the NES scheme, with the nucleus estate occupying 30 percent of the land and outgrowers occupying 70 percent (Zen, Barlow, and Gondowarsito 2005). In this arrangement the agribusiness company pro- vides inputs and technical support to the surrounding smallholders who in turn sell their fruit to the company’s mill. In contrast, in Malaysia the major approach has been the managed smallholder schemes of FELDA, which opened up forested land for settlers, and FELCRA and similar state-based agencies, which developed land on behalf of existing smallholders. Initially, the smallholders in these schemes looked after their own plots, but increasingly the schemes are centrally managed as plantations, with the landowners 6Compound growth rate computed from time series data in FAOSTAT (2014). 10 Rob Cramb et al. 410 415 420 425 430 435 440 445 450 receiving a share of the profits. Joint ventures between smallholders and private plantation companies have been initiated in both Malaysia and Indonesia. These involve the pooling of smallholder land from the outset so it can be managed entirely as a plantation, with the smallholders receiving shares and dividends according to the area of land contributed. However, these schemes have proved to be very disadvantageous to the participants, who have no control over the management of the estate and receive little or no return for their land (Cramb 2013b). There are no important economies of scale in oil palm production that would suggest oil palm is inherently unsuited to smallholders (Barlow 1986). A diversified rural household with other farm and non-farm activities can cope with the up-front costs of oil palm devel- opment. Where governments, plantation companies, traders or farmer organisations are able to provide capital to smallholders in the form of credit or grants for seedlings and fertiliser, the extra returns easily justify the initial outlay. Once the necessary skills are acquired (sometimes through family members who have worked on estates), household labour can just as easily undertake the tasks of planting, crop maintenance and harvesting as an estate labour force, and without the need for costly supervision. Thus, assisted or otherwise favourably situated smallholders can in fact obtain comparable yields to private or state-run plantations at significantly less cost. The major economies of scale with respect to oil palm emerge in processing. The har- vested fresh fruit bunches (ffb) need to be transported within 48 hours to a local mill for crushing to produce crude palm oil. While small-scale processing can be viable, a standard commercial mill in Malaysia or Indonesia capable of handling 60 tons of fruit per hour can process up to 300,000 tons of fruit per year at full capacity. This would require 15,000 ha of oil palm producing 20 tons of ffb per ha (The Palm Oil Mill 2011). A mill can still run prof- itably at less than full capacity, and many do, but to undertake the large investment that a commercial mill entails requires assurance that there will be a large productive area (at least 5–10,000 ha)AQ19 ¶ supplying the mill. These economies of scale provide an incentive for planta- tions to maximise their direct control over production. In both Indonesia and Malaysia, many investors have argued that models which give ‘unreliable’ smallholders responsibility for production entail significant financial risks. By avoiding dependence on smallholders and directly controlling the largest area of production possible, plantations argue they can boost profits by maximising and stabilising the daily input of high-quality fruit to their mills. In large part this logic underpins the shift away from the NES models of the New Order period in Indonesia (McCarthy, Gillespie, and Zen 2012a). The same motiv- ation has been behind the push to replace smallholdings with shareholdings in both govern- ment-managed and joint-venture schemes in Malaysia. However, there is no inherent reason why the supply area for a mill should not be some combination of estates and smallholders, or even a purely smallholder landscape, provided the necessary transport infrastructure is in place (Shiel et al. 1999AQ5 ¶ ; Rival and Levang 2014). In essence, this is a coordination problem that could be solved internally by a large private company or by a public agency or through private (or government-brokered) contracting between a processor and smallholders, whether acting individually or as a marketing cooperative. The state of Sarawak in Malaysia has been an important example of state policies biasing the oil palm boom towards large-scale private estate development (Cramb 2011). However, the area of smallholder oil palm in Sarawak has increased dramatically in the past decade, from 9000 ha in 2001 to 96,000 ha in 2013, a growth rate of 20 percent. This compares with negligible growth in managed smallholdings over the same period and growth of around 10 percent in joint venture schemes. The growth in smallholdings has occurred with little government assistance compared to the concerted political The Journal of Peasant Studies 11 455 460 465 470 475 480 485 490 495 campaign in support of the joint venture approach – indeed, it has occurred in the face of active discouragement (Cramb 2011; Ngidang 2002; Majid Cooke, Ngidang, and Selamat 2006). Moreover, the expansion has occurred without the agrarian differentiation among smallholders that has been observed in the Indonesian context (McCarthy 2010). We now examine how smallholders in Sarawak have inserted themselves into the rapidly expanding landscape of large-scale plantation development, based on a case study of a cluster of Iban longhouse-communities that have been among the pioneers of oil palm planting in Miri Division in northern Sarawak (Cramb and Sujang 2011, 2013).7 The case study is based on a questionnaire survey of 72 Iban farm-households con- ducted in 2009. The oil palm planters in the survey were essentially independent small- holders who had taken on the activity with little or no assistance from government agencies or contractual obligations to private traders and financiers. Though there was a small number of landholders with 10–30 ha of oil palm who had started earlier and now made use of hired labour, by far the majority had planted under 8 ha, typically 2–4 ha, and operated with labour provided by husband and wife (who also engaged in other farm and non-farm pursuits). Less than one in five households had not planted oil palm, most of whom lacked non-farm sources of cash flow to provide the start-up capital and/ or did not have suitable land. Thus, although there were different strata in terms of the area planted with oil palm, this did not reflect a process of agrarian differentiation in the sense that wealthy households were acquiring land from poorer households. Land remained relatively abundant in this former swidden area and the strata were more the product of initial differences in wealth and status within the longhouse and differences in ‘human capital’ affecting access to off-farm sources of cash flow, rather than a necessary conse- quence of the oil palm boom. The evidence from other surveys, anecdotal evidence and casual observation elsewhere in Sarawak suggest that the survey respondents were fairly typical of indigenous oil palm smallholders in the state. Many of the smallholders surveyed began planting oil palm desultorily and experimen- tally in the 1980s, mainly to confirm their claim to the land, which was under pressure from large-scale plantation development. But once they started selling their fruit, oil palm became their major on-farm source of income – ‘strategic agriculture’ had become a genuine ‘livelihood strategy’. Their oil palm smallholdings provided them with steady employment and cash income throughout the year, without them being subject to the direc- tions of plantation managers or having onerous financial obligations to suppliers of inputs and credit. This strategy was buttressed, on farm, by low-input swamp rice cultivation for subsistence and, off farm, by urban wage and salary employment, and in a few cases self- employment in rural transportation and marketing. The survey confirmed that smallholder oil palm is a livelihood strategy that generates relatively high returns to household labour and capital, even though yields per unit of land are lower than on plantations. However, this is consistent with the circumstances of most farm-households in Sarawak, where farming land is still relatively abundant and underutilised but family labour is increasingly scarce and costly, given the opportunities available for urban-based employment and residence. Capital has also been scarce and costly, with limited availability of private credit for smallholder farm development and restricted access to the subsidised capital resources channelled through government schemes for smallholders. This constraint is being progressively lifted as increased 7The Iban constitute the largest ethnic group in Sarawak and are numerically dominant in the oil palm zone of Miri District. 12 Rob Cramb et al. 500 505 510 515 520 525 530 535 540 income from oil palm provides the working capital to purchase farm inputs and expand the planted area without resort to credit, but the level of fertiliser use is still low, suggesting that capital remains limiting. Even with low inputs of labour and capital, and hence yields of around 12 tons ffb per ha (compared with 20–25 tons/ha on plantations), smallholders with about 3 ha of mature palms were able to earn around MYR 7000 per year, providing a return to family labour of over MYR 40 per day when prices were low (around MYR 300/ton in 2009), rising to over MYR 120 per day with higher prices (around MYR 600/ton in 2011).8 Nevertheless, small- holder returns to labour and capital could be improved, particularly by greater use of ferti- liser to increase yields. This suggests a need for greater use of well-targeted support measures enabling the acquisition of good-quality planting material, the use of higher levels of recommended fertilisers, and improvement in infrastructure. An approach pioneered by the Malaysian Palm Oil Board (MPOB) in Miri District looks promising. This involves the establishment of smallholder buying groups (Kelom- pok Baja Terkumpul, KBT) to improve access to fertiliser and credit. The concept is that a commercial agent supplies fertiliser on credit to the participating farmers, who agree to deliver their fruit to a specified mill, which in turn agrees to deduct the cost of the fer- tiliser and repay the supplier. MPOB’s role is to oversee the arrangement and to supply good-quality planting material and technical advice to the farmer group. The uptake of this scheme in pilot sites has been rapid and the results in terms of increased yields and incomes impressive. A typical outcome is for fertiliser use to increase fourfold to 800 kg/ha and yields to 20 tons/ha, generating a return to family labour of MYR 570/ day (at MYR 600/ton) – a more than fourfold increase over the returns cited in the pre- vious paragraph. This approach not only alleviates the constraint on capital, ensuring greater use of fertiliser, but also provides access to good-quality planting material and technical advice, and facilitates savings in the transportation and assembly of fruit for processing, while leaving the overall development and management of the farm in the hands of the smallholders. A further area of improvement would be in the upgrading of farm access roads to lower the cost of harvesting and hauling fruit, and to assist those whose land is currently not sufficiently accessible for them to consider planting. Thus, as with the rubber case above, well-targeted programmes for supported or organ- ised smallholders have the potential to spread the benefits of the oil palm boom more widely and equitably. Returning briefly to the research questions: (1) the agro-economic characteristics of oil palm are even more favourable to smallholders, with a shorter investment period and lower labour requirements, though to avoid being locked in to low yields, access to high-yielding planting material and credit for fertiliser is important; (2) once an oil palm industry is estab- lished, there are profitable opportunities for input suppliers and palm oil mills to support smallholder production with credit arrangements, infrastructure, transportation and market- ing; (3) where land is relatively abundant and livelihoods diversified, oil palm planting can benefit the majority of smallholders in accessible villages without resulting in significant differentiation, but where land is scarce and livelihoods more vulnerable, poorer farmers can experience adverse incorporation and be caught in the backwash of oil palm development. 8USD 1.00 =MYR 4.03.These returns to labour are well above rural wage rates of up to MYR 30–35 per day. The Journal of Peasant Studies 13 545 550 555 560 565 570 575 580 585 Smallholder cassava in Tbong Khmum province, Cambodia Cassava (Manihot esculenta) is a small perennial shrub that produces a starchy root. The crop is typically cultivated on an annual basis and has a diverse range of uses, including direct consumption as a food crop, livestock feed, starch production and biofuel production. The recent rapid spread of commercial cassava cultivation has been one of the major land- use changes in Southeast Asia, where over 3.6 million ha are now cultivated (De Koninck and Rousseau 2012; Lefroy 2014). Cassava production began increasing in Thailand in the 1970s, and then in Vietnam and Indonesia, to meet new market opportunities, particularly for livestock feed in Europe. Since 2000, changing trade policies and rising incomes have seen the market for cassava products become increasingly focused on Asia, particularly China. Southeast Asia now contributes over 95 percent of global cassava exports, with Thailand and Vietnam accounting for the majority of both dried cassava chips and cassava starch. Cassava is subject to a range of production risks, including drought and emerging pests and diseases, and is particularly affected by price fluctuations in a range of global commod- ity markets that compete for land (such as maize, rubber and sugarcane) and/or substitute in final product markets (such as maize, oil, sugar and potato starch) (Newby 2016). Never- theless, cassava production is considered a suitable activity for resource-poor farmers, given its low demand for inputs, its ability to grow on marginal land and its relatively quick turnaround. This makes it potentially important for local livelihood development in marginal upland communities. However, such smallholders need support from public- or private-sector actors to access suitable varieties, other inputs and technical advice (e.g. with regard to pests and disease control, and soil conservation measures). In Thailand, the state has funded research, extension, the free distribution of improved planting materials, and financial support for smallholders, paid for in part by a levy on exports (as for rubber). This has permitted the growth of a prosperous smallholder industry. Cassava production in Vietnam has also become an important livelihood activity for upland smallholder farmers, but here there has been limited state support. However, there are now almost 100 large starch factories operating throughout the country, as well as hundreds of small-scale, family-operated starch extraction enterprises. Cassava starch factories do not have the option of obtaining large land concessions within Vietnam, hence they have an interest in developing formal and informal relationships with traders and smallholders to secure a reliable supply of good-quality feedstock. The perishable nature of the crop means that coordination of delivery is important to maintain quality and processing effi- ciency. Many starch factories have been providing planting material (stakes), credit and extension services to farmers within their supply zones in the absence of a public extension system. The incentive to do so depends on the company’s ability to capture the benefits from increasing the productivity of farmers. In situations where there are competing fac- tories or alternative value chains, hence an increased risk of side-selling, there is less incen- tive to provide inputs and technical advice. However, other inducements and sanctions are often given to traders and farmers to maintain loyalty. One factory in the Central Highlands of Vietnam pursues a number of business models within its supply zone. Previously, formal contracts were signed with a large number of farmers, which was a complicated and inefficient process (Wandschneider 2008). Now the company has formal contracts with traders to deliver roots according to a schedule, and these traders develop a network of farmers, often using informal contractual arrange- ments. Often these key traders are advanced money from the factory which they supply to the farmers in their network. These specialist traders have an incentive to increase the 14 Rob Cramb et al. 590 595 600 605 610 615 620 625 630 productivity of their farmers and to monitor quality and coordinate delivery. In other cases, particularly with the Ede ethnic minority, farmer groups have been established, with a farmer leader as the main intermediary between the factory and the farmers, providing inputs and coordinating the harvest and supply of roots to the factory. Given the limited opportunity for further expansion in the main producing countries (in fact, the planted area is declining in Thailand and Indonesia, and the central government in Vietnam would like the planted area to be reduced), the industry has turned to Cambodia, Laos and more recently Myanmar to meet the growing regional demand. However, unlike the first wave of expansion in Thailand, Vietnam and Indonesia, which was dominated by smallholders supplying traders and processors, the current growth in Cambodia, Laos and Myanmar has been driven by domestic and foreign companies receiving large concessions to establish both plantations and processing facilities, with smallholders delivering varying amounts of feedstock under different market linkages (Zola 2008; Wright 2009; Manoram et al. 2011). While the expansion of estates could perhaps be explained by the ‘pioneer industry’ argument (i.e. the need to ensure an initial supply area for a new processing plant), it is a cause for concern that this expansion is occurring in countries with ‘weak land governance’ where large land concessions for other boom crops such as rubber and sugarcane have displaced smallholder livelihoods. Nevertheless, thousands of smallholder farmers are also growing cassava in these newer cassava-producing countries, as indepen- dent producers or under various contractual arrangements.9 The most prominent newcomer has been Cambodia, where the area has increased 15 times in the past 10 years. An ongoing study of smallholder cassava producers includes a case study of Dambae, one of six districts in the newly established province of Tbong Khmum, previously part of Kampong Cham in the country’s east. Unlike newer cassava areas in the west where there has been an influx of migrant farmers, Dambae has a long- established population, most of which was locally born, with some migrants from other dis- tricts within the province. In addition to lowland rice, grown mainly for family consump- tion, farmers in Dambae have been planting cash crops since the early 1990s, following de- collectivisation, the restoration of peace and stability, and improvement in infrastructure. Cassava is now the major commercial crop, accounting for 15,940 ha in 2014 or 54 percent of total agricultural land. Other commercial crops include rubber, cashew, pepper and vegetable crops. Farmers also raise livestock for income, food and manure. Proximity to Soung, a major market centre, and to the border with Vietnam makes it easy to transport agricultural produce and obtain inputs. There is no government extension or support for cassava producers, and there are very few local starch factories. Most traders are linked to starch factories in Vietnam, which are looking to expand their supply zone and obtain out-of-season feedstock to keep their factories running. While there is a flow of planting material and advice from these processors and traders to Cambodian smallholders, the link is more tenuous than within Vietnam itself. Two major cassava-growing villages were selected for study – Pr Sreleu and Bangheur Khleng. The former had close proximity and a good connecting road to Soung, while the latter was more remote from markets and had a higher incidence of poverty. Interviews were conducted with 39 households in Pr Sreleu and 51 households in Bangheur 9For example, in Myanmar there are more than 15,000 ha of smallholder cassava grown in the Ayeyarwady Region supporting over 250 small-scale starch processors. By comparison, in Kachin State the Yuzana Company has a 100,000-ha cassava plantation (with 16,000 ha currently established) (Nyi Nyi 2015). The Journal of Peasant Studies 15 635 640 645 650 655 660 665 670 675 Khleng. Most households owned 2–3 ha of farming land, while around 20 percent owned 5–10 ha. A minority worked as daily-paid agricultural labourers within the village. Some landless poor had moved out of these villages to other nearby provinces such as Kratie and Prea Vihear where there was still land that could be cleared and occupied, as there was little room for expansion in either village. Upon implementation of decollectivisation in 1989, each family received 0.5–1.5 ha depending on the number of household members. Farmers slowly converted the village for- estlands for cash crop cultivation. They switched among multiple crops such as sesame, soy bean and cassava as prices fluctuated. Production was mainly for the local market, with some exported to Vietnam. Increasing demand for cassava tubers from starch factories in Vietnam has driven the cassava boom since the early 2000s. A high-yielding Thai variety (KU50) was introduced by traders to Pr Sreleu in 2000 and to Bangheur Khleng in 2004. Cassava stakes were imported from Vietnam and retained by farmers for sub- sequent plantings. The price of cassava roots has been fluctuating but the demand from traders has continued to increase. The introduction of commercial cassava led to abandon- ment of existing cash crops such as soybean and sesame. Farmers preferred cassava because it was much easier to plant and had higher returns and better market prospects, the village was accessible to traders, and newly cleared forest land was suitable for cassava production. The village studies found three implications of cassava planting for rural livelihoods. First, the conversion of land to cassava production led to loss of access to forest for food, timber and resins. Most people stopped raising cattle because they were replaced by tractors and there was very little forest available for grazing. In Bangheur Khleng, an increasing proportion of forest land had been cleared for soybean production. On adoption of commercial cassava farming, the rate of forest loss increased. In 20 years of cash crop production, the village forests disappeared. Second, commercial crop production caused rising land prices and attracted urban- based investors to buy land for cash crops. Land ownership is increasingly concentrated in medium and large holdings. Some farmers sold their land to urban investors or better- off households in the village as a means to survive when faced with price or yield down- turns or other financial shocks (i.e. distress sales). Others took the opportunity of increasing land prices to sell their land so as to invest in other, less labour-demanding livelihood activi- ties (i.e. strategic sales). Some sold their land to outsiders, fearing that their land would be grabbed due to the lack of a proper title. The third implication is that the cassava boom has led to the movement of some farm households out of the community. Findings from both villages confirm that push factors (many of them interrelated) included an initial lack of productive land, growing household numbers, distress sales of land, failure in commercial crop production, and the rising price of land locally. The main pull factor was the availability of forest land in Kratie, Rattanakiri and Mondulkiri provinces to the north that could be cleared for farming. Hence many of those caught up in the backwash of the cassava boom in Dambae were contributing to the spread of the boom to more remote and less-densely populated areas, where minority groups were less prepared for the transformations this involved (Mahanty and Milne 2006AQ6 ¶ ). In terms of the three research questions: (1) the agro-economic characteristics of cassava are favourable to smallholders, giving a quick return with low requirements for labour and inputs and flexibility in crop choice, though there are significant yield and price risks that can adversely affect smallholders who go into significant debt to plant the crop; (2) starch processors and traders have a strong incentive to support smallholder production within their supply areas by providing suitable planting material, credit and technical advice, provided the risk of side-selling can be minimised; (3) where rural 16 Rob Cramb et al. 680 685 690 695 700 705 710 715 720 poverty is high and land tenure insecure, rising land prices and indebtedness can induce poorer farmers to sell their land and migrate, thereby shifting the locus of cassava expansion to frontier zones in a cascading process of differentiation. Smallholder teak in Luang Prabang province, Laos Teak (Tectona grandis), a tall, deciduous, tropical hardwood, occurs naturally in a discon- tinuous distribution across India, Myanmar and Thailand, extending into northern Laos (Midgely et al. 2007, 2015).10 It has also become naturalised in Java, following its introduc- tion several centuries ago. Steadily increasing global demand for tropical hardwoods and depletion of native teak forests have created an incentive to establish teak plantations in these and other tropical countries, both as smallholdings (as in Thailand) and as large- scale plantations (as in Brazil). Teak requires little labour or other inputs. Apart from suit- able environmental conditions, productivity depends on selection of good planting material, appropriate spacing, periodic thinning and pruning, and harvesting to maximise the market value, which varies in increments with size class. The total planted area of teak is estimated to be about 30 million ha in 36 countries (Midgley et al. 2015). The Floresteca Group is the world’s largest privately owned teak pro- ducer, with 24,000 ha of planted teak forest in Brazil, producing 200,000 m3 of round timber annually. The Floresteca website frankly identifies the desiderata of plantation com- panies throughout the world: ‘The group benefits from ideal conditions in Brazil due to the availability of arable low priced land, sufficient rainfall, high soil quality and optimum climate conditions. Moreover Brazil offers access to an economically favourable labour force…’.11 However, as with other plantation crops, Midgely et al. (2007) report a general shift from large-scale teak plantations to small-scale lots incorporated in diversified farming systems, whether as independent holdings or as outgrowers of a vertically inte- grated concern. In Laos, there are no large-scale teak plantations. Rather, smallholder teak farms have been identified by government planners as a potentially valuable component of upland farming systems, enabling subsistence swidden farmers to move into commercial agricul- ture and out of poverty. Teak planting commenced with government support more than 50 years ago in Luang Prabang Province in northern Laos, but a rapid expansion has occurred in the last 20 years, with over 10,000 ha of smallholder teak now established in the province (Midgely et al. 2007). This crop provides high returns, especially in relation to current household incomes, and has good long-term market prospects. Most of the teak is sold to traders who supply local sawmills, but only five percent of the sawn timber remains in the region while 95 percent is transferred through Vientiane-based traders to Thailand, Vietnam and China (Mohns and Laity 2010AQ7 ¶ ). Local agribusiness actors purchase, harvest, transport and process teak logs, acquiring whatever they can, usually in a buyer’s market. Improved management and marketing practices have the potential to increase substantially the returns to teak-growing households (Dieters et al. 2014), but these local agribusiness actors have no real incentive to transmit these improved practices to smallholders (compared with the case of cassava processors considered above). 10Teak was also introduced to Vietnam in the early twentieth century. It was almost eradicated in the French and American wars, but has been revived in recent decades. 11Floresteca Group website, http://www.globalforestry.com/floresteca-group-overview (accessed 1 May 2015).AQ20 ¶ The Journal of Peasant Studies 17 725 730 735 740 745 750 755 760 765 Smallholder teak planting in Luang Prabang has been strongly influenced by govern- ment land policy. The Land Law 1997 and Forestry Law 1996 provide the current frame- work for the eradication of swidden agriculture and the encouragement of sedentary farming, including the establishment of tree plantations. The state authorises individuals and households to use agricultural land in accordance with a village-based allocation plan and objectives – the LFA process mentioned above in relation to rubber. The area allo- cated includes land for food crops and up to 3 ha of ‘degraded forest land’, mainly land pre- viously used for swidden agriculture, where tree-planting activities are meant to be focused. If a household does not utilise the land it has been allocated within three years it may be returned to the village committee for redistribution to other farmers (Ducourtieux et al. 2005AQ8 ¶ ). As a result, many farmers planted teak on the plots they did not currently need for food crops to retain them for the future (Kolmert 2001; Midgely et al. 2007). This dis- advantaged households that were slow to realise the implication of the LFA process, as well as those arriving later in the village (either as part of a resettlement programme or through voluntary migration) who had a reduced pool of available land to claim. There was a brief boom in teak planting in Luang Prabang in 1988 and another larger one in 1996, after which the rate of planting slowed. Nevertheless, the total area has increased from 500 ha to over 10,000 ha in the past 20 years (Kolmert 2001; Midgely et al. 2007). Researchers have identified a variety of reasons why farmers have planted teak (Hansen, Sodorak, and Savathvong 1997; Kolmert 2001; Midgely et al. 2007): . depletion of wood supply from natural forest and the emergence of a market for rela- tively young teak timber (15–20 years old); . relatively secure private land tenure since the 1990s; . the permanent settlement pattern adopted by most villages; . expansion of the road system; . land allocation schemes that gave additional land for production of perennials; . promotion of tree planting by private investors through financial support, the pro- duction of stumps, and information dissemination; . promotion and extension by government agencies. Thus, while teak was planted by independent, village-based smallholders, the role of public agencies and private investors in providing planting material and technical support was significant. There is considerable diversity both within and between teak-growing villages in Luang Prabang. In November 2009 a survey of 127 households in five teak-growing villages was carried out in four districts of Luang Prabang Province to explore the differences within and between villages in teak planting and management (Newby et al. 2012). The five villages were selected to highlight differences in proximity to Luang Prabang City, ethnicity, reset- tlement history, population density, and other land-use opportunities. Four-fifths of the sur- veyed households had planted teak, averaging 1330 trees (1.4 ha) per household. The distribution of plantation size was positively skewed, so the average was inflated by a small number of larger plantations. About 20 percent of households surveyed had never planted teak and 40 percent had planted less than 1000 trees (1.1 ha). The largest 10 percent of plantations ranged from 3000 trees (3.2 ha) to 20,000 trees (just over 20 ha). However, the sampling was not able to capture absentee landholders (discussed below) whose holdings reportedly represented up to 50 percent of the planted teak area in the more accessible parts of Luang Prabang. 18 Rob Cramb et al. 770 775 780 785 790 795 800 805 810 The pattern of planting reported in the household survey showed a few years of exten- sive planting during the 1990s, and a major expansion in the 2000s. The data suggest that this more recent expansion was not primarily strategic, to take advantage of the LFA process, but was related to factors such as the implementation of restrictions on swidden agriculture, the new ability to transfer land, and the improvement in road infrastructure and market opportunities. Furthermore, seeing early planters reap the benefits as the trees reached harvestable age encouraged other households to plant. The survey data indicate that it was the better-off households that had settled earlier, had access to paddy land, had achieved higher education, had off-farm sources of income, were not reliant on swidden cultivation, and were less dependent on cattle that had adopted teak and planted more trees. Those households without teak typically had fewer assets and lived in houses constructed of temporary materials. This relative lack of assets was also related to other factors, notably the household’s relocation history (mainly affecting Hmong households resettled from the highlands). Some of the variation in assets can also be explained by village-level factors, such as the absence of electricity or a mobile phone signal in one village, or a recent fire in another that had destroyed the majority of the houses. The causal relationship between teak holdings and household assets was not straightfor- ward. Did teak households have more assets because they had planted and benefited from teak, or were they able to establish teak plantations because they had more resources to begin with? Analysis of the survey data showed that the asset levels of those households that had planted but not yet harvested teak were higher than those of households that had never planted teak, suggesting that initial household wealth was an enabling factor. However, early planters who had already harvested some teak tended to have greater wealth than other planters who had not yet harvested any trees, suggesting that teak planting also had a positive economic impact. Given the low rate of rice self-sufficiency in the region, households with little land that adopted teak faced a food shortage problem until the trees could be harvested. For this reason, Hansen, Sodorak, and Savathvong (1997) suggested that teak was primarily suitable for wealthier farmers, businessmen and government employees. These authors concluded that one of the main motivations for other upland farmers to plant teak was the possibility of selling young (1–3 year-old) plantations to investors. Furthermore, they supposed that, since plantations were predominantly established on flatter land next to roads, farmers had lost much of the best agricultural land for the production of other cash crops. According to Kolmert (2001), the selling of land recently planted with teak had been occurring even before farmers had the required certificates and resulted in many farmers not having enough land on which to grow food. During the 2009 interviews, respondents were asked if they had purchased or sold land in the recent past. Overall, 36 percent of households had purchased land and 21 percent had sold land. The data suggest that better-off households in the survey were purchasing both established teak blocks and fallow land that they later converted to teak. Given that the survey did not include absentee landowners, the full extent of land transactions was difficult to quantify. Anecdotal evidence, however, suggests that the scale of land transactions is increasing, with investors looking for land on which to plant both teak and rubber. The money from land sales was reported to be used for school fees and weddings, and to meet healthcare expenses. For households that depended on swidden agriculture for their livelihoods, the role that teak played in their livelihood strategy had changed little since the 1950s. Many of these households reported that they borrowed land for upland rice production on the condition that after harvest they established teak for the owner. The next year they would have to The Journal of Peasant Studies 19 815 820 825 830 835 840 845 850 855 find a new piece of land to grow their crops. However, because the area of teak had expanded, these households reported that they now have to travel farther afield, often walking to neighbouring villages and into more remote and steeper country. Formal and informal sources of credit were available to farmers in some villages. The incidence of borrowing was higher among non-teak households and those with smaller plantations, suggesting that teak plantations (like rubber plantations) substitute for bor- rowed capital, providing a ‘bank’ to fund household needs as they arise by selling a number of trees (regardless of whether they had attained optimal girth). This also suggests that indebtedness was a motivator for selling young teak plantations, which had come to imply selling the land underneath. Therefore, without access to alternative productive land or income sources, maintaining ownership of teak land will continue to be difficult for many households. As permanent land titles are progressively issued, permitting the land to be transacted legally, more households may sell some of their titled land in order to survive. Research on teak agroforestry systems that provide short- and medium-term cash flow (e.g. under-planting with rattan, intercropping with paper mulberry) may enable a more gradual transition to teak systems for households with little land. However, farmers with alternative livelihood activities and absentee landowners are unlikely to adopt these more diversified and labour-using systems. For these households, the establishment of teak planta- tions not only represents a source of future income andwealth, but also provides amethod for maintaining access to land beyond the area they can physically cultivate. Thus, even with increased research and extension efforts, the boom in teak planting is accelerating processes of agrarian differentiation, with a small group of better-off farmers and urban-based outsiders capturing the majority of the benefits, while those with greatest dependence on swidden agri- culture are actually made worse off through declining access to land and capital. Returning to our research questions: (1) the agro-economic characteristics of teak make it an attractive crop for smallholders as it can be easily established and requires little or no household labour to grow and harvest; hence, it can be used to secure claims to ‘surplus’ land while accumulating wealth for emergency needs and/or long-term investments; (2) governments and investors have an incentive to supply good-quality planting material and technical advice to smallholders to ‘stabilise’ land use and expand the supply of teak logs for millers and traders; (3) because teak planting locks up land without requiring sig- nificant labour input it favours better-off smallholders and outside investors, while creating a backwash effect on poorer households who are more likely to sell their teak stands early and push farther into marginal land to meet their subsistence needs. Discussion A wide range of factors have been apparent in considering the potential of the four com- modities to improve smallholder livelihoods, summarised in Tables 3–5. It is not possible to review each factor here. However, the influence of these factors on inclusive smallholder conjunctures can be illustrated by comparing the four case studies in terms of the three research questions. What are the agro-economic factors favouring or obstructing smallholder modes of commodity production relative to large-scale production entities? Each of the four com- modities considered has agro-economic attributes that advantage and disadvantage small- holders (Table 3). For example, rubber has a low requirement for major land development, being readily established in swidden fields; once tapping begins there is year-round employment for family labour, which can include men and women, young 20 Rob Cramb et al. 860 865 870 875 880 885 890 895 900 and old; there is little yield risk; and the product is storable and transportable in small quan- tities, giving market flexibility. However, there is a long time to first harvest, making it dif- ficult for smallholders to sustain family labour unless they also have other livelihood activities or there is some form of external support. Similar sets of trade-offs can be ident- ified in Table 3 for the other commodities. However, given strategic interventions in the establishment phase (e.g. provision of good-quality planting material and planting grants or concessional loans), smallholder production proved viable for a majority of households in all four cases, showing that arguments for large-scale agriculture cannot be based on commodity-specific factors alone. Given that smallholder production of the four commodities was viable and expanding in the case studies, what were the incentives for agribusiness firms to contribute to smallholder commodity production through roles other than direct farm management? As enumerated in Table 1, agribusiness firms and/or state agencies can provide a variety of inputs and services to commodity value chains, singly or in various combinations. The case studies showed a range of complementarities between smallholders, agribusiness firms and government agencies, specific to each context, summarised in Table 4. For example, in the cassava case, while farmers manage the entire production process, there is a strong mutualistic link between smallholders and starch processors, mediated by traders, such that planting material, credit and information are frequently sourced from the processors, who benefit from the increased supply of better-quality product. Even where the relationship is mono- psonistic, farmers have flexibility in their crop choice and need to be treated fairly by traders if they are to remain loyal suppliers. The case studies also shed light on the question whether smallholder commodity pro- duction can be broadly inclusive in the face of powerful tendencies towards agrarian differ- entiation and the impersonal operation of land markets. The agro-economic attributes of the commodities and the complementarities with agribusiness firms, discussed above, com- bined to allow smallholder production to be a profitable undertaking, increasing household incomes and assets, both tangible (e.g. land values, house quality, vehicles) and intangible (e.g. knowledge, skills). Moreover, adoption of these cash crops has been widespread, with most or all households in the case-study villages benefiting from planting the crop in Table 3. Agro-economic characteristics of four boom crops. Characteristic Rubber Oil palm Cassava Teak Need for major land development L–M L–M L L Investment up front for production M–H M–H L–M L–M Benefit of using improved planting material H H H M Time to first harvest 6–7 years 3 years < 1 year 10+ years Year-round labour requirements H H M L Susceptibility to mechanised production L L–M M–H L Yield risk L L H L Susceptibility to mechanised harvesting L L–M M M Harvesting frequency 1–2 days 3–4 weeks Annual Episodic Storability of harvested product M L L H Selling frequency Monthly 3–4 weeks Annual Episodic Price risk H M H M Coordination between harvesting and processing M H H L Percentage of processed product to raw material H H H H Investment in first-stage processing L–M M–H L–H L–M Source: Adapted from Byerlee (2013, Table 2.1). Note: L = Low; M =Moderate; H = High. The Journal of Peasant Studies 21 905 910 915 920 925 930 935 940 945 Table 4. Contributions of actors in four smallholder value chains.AQ22 ¶ Function Rubber (Luang Namtha, Laos) Oil palm (Miri, Malaysia) Cassava (Dambae, Cambodia) Teak (Luang Prabang, Laos) Land development Infrastructure already in place; land cleared by smallholders Infrastructure already in place; land cleared by smallholders Infrastructure already in place; land cleared by smallholders Infrastructure already in place; land cleared by smallholders Land settlement Land and Forest Allocation managed by village Customary tenure remains in force Land ownership recognised by village Land and Forest Allocation managed by village Production inputs Smallholders purchase seed and inputs Smallholders purchase inputs with cash; agency provides some planting material Smallholders purchase planting material and inputs Smallholders purchase planting material from government and private nurseries Knowledge Farmers learn from experienced farmers and traders Farmers learn from government agencies and other farmers Farmers learn from processors Farmers learn from government agencies Finance Government loans, repayable on maturity of crop Government planting grants; credit for fertiliser deducted by mill Credit from processors, moneylenders Purchase of standing crop Labour management Family and local labour Family and local labour Family and local labour Family and local labour Field operations Farm household Farm household Farm household Farm household Harvesting Family and local labour Family and local labour Family and local labour; harvested by buyer Harvested by buyer Acquisition Trader comes to farm to purchase tub-lump rubber Smallholder delivers fruit direct to mill or via local collector Trader/processor comes to farm to purchase, perhaps harvest, and transport tubers Trader/miller comes to farm to cut and transport logs Processing Large-scale rubber mills (in China) Large-scale palm oil mills Mostly large-scale mills (small- scale mills in Myanmar) Small- to medium-scale local sawmills Marketing Traders to mills in China Mills to refineries to exporters Starch processors to traders and end-users Sawmills to exporters Facilitation Government loans; village negotiates best price, arranges for traders to come Agency facilitates link between smallholders, fertiliser suppliers, and mills Processors/traders supply planting material, inputs, advice, scheduling N/A 22 R ob C ram b et al. 950 955 960 965 970 975 980 985 990 question. Nevertheless, in each case there was an emerging inequality in the area planted, and in some cases in the overall distribution of land owned. The key factors contributing to this differentiation are listed in Table 5. However, there was no simple causal relation between a given factor and an observed outcome. Rather, it was the conjuncture of multiple factors that influenced the degree of inequality that emerged. For example, in the oil palm case in Miri, though there was relatively abundant land (including widespread access to paddy land), no indebtedness, alternative sources of income, and no land transactions, there was a positively skewed distribution of oil palm holdings. But this was largely related to initial differences in land and labour resources, enabling some households to plant more oil palm, without impinging on the ability of other households to follow suit. This, then, was a case of ‘non-impoverishing differen- tiation’ in which all or most households were progressing but some faster than others. In the rubber case, farmers were more constrained. However, the early rubber planters in Hadyao were able to find enough land to grow rice, within and outside the village, and the later planters were able to work as tappers while their own holdings matured. Though some inequality in land and income emerged, largely related to the time of planting, the village leaders took a strong role in ensuring everyone had a rubber holding, limiting the total area planted, prohibiting land sales to outsiders, and promoting a cooperative approach to mar- keting. The outcome was ‘minimal differentiation’. However, contract farmers were much more vulnerable to loss of land due to their inability to fund their own labour during the establishment phase under ‘2 + 3’ contracts and the perverse twist by which renegotiated contracts often involved splitting not just the output but the land with the investor. In addition, purchases of rubber land by wealthy outsiders were resulting in a redistribution, not just within the village but between urban and rural landowners (as also in the teak case in Luang Prabang). Table 5. Factors potentially affecting agrarian differentiation in the four case studies. Factor Rubber (Luang Namtha, Laos) Oil palm (Miri, Malaysia) Cassava (Dambae, Cambodia) Teak (Luang Prabang, Laos) Initial wealth differentiation Low Moderate Low Moderate Land frontier Closed Closed Open Closed Land abundance Moderate High Moderate Moderate Tenure security High Moderate–high Moderate–high High Access to rice land Limited Widespread Widespread Limited Extent of participation in cash crop production High High Moderate–high Moderate–high Ease of absentee management Low–moderate Low Low High Income risk due to yield and/or price fluctuation High Low–moderate High Low Dependence on credit High (but not onerous) Low High Low Alternative sources of livelihood Limited Widespread Moderate Moderate Strength of community institutions Moderate–high Moderate–high Low Moderate The Journal of Peasant Studies 23 995 1000 1005 1010 1015 1020 1025 1030 1035 The cassava case study in Dambae shows a less well-insulated situation. Though farmers started on a fairly equal footing with land reform in the 1990s and had maintained their subsistence base, the expansion in commercial crops, especially cassava, combined with steady growth in population, had eaten into their available forest land, limiting forest-based livelihood pursuits such as cattle grazing and closing the frontier for new households. In addition, the high levels of indebtedness and the risks associated with cassava production meant that some households were forced into distress sales of land. More successful farmers were ready to buy up this land and enlarge their holdings, while outside investors were also attracted by the rising land values and the opportunity to profit from cassava production. Hence some households became landless or near-landless, and some left the village to occupy land in more remote provinces. Village institutions were not able to address this process of ‘impoverishing differentiation’, for example by prevent- ing absentee land purchases. Conclusion Land grabs in Southeast Asia are driven by many motivations, but the underlying dynamic is to cash in on favourable agricultural commodity markets, increasingly driven by demand from within Asia. Past and present crop booms have seen both small- and large-scale pro- duction entities emerge. While large-scale operations may have economic advantages, par- ticularly in a pioneering setting, it is clear that once an industry is underway and infrastructure is in place, there is a strong smallholder impulse to engage in commodity pro- duction, such that the share of planted area and output attributable to smallholders increases over time. In some policy settings, a successful industry can be developed from the outset entirely by small- and medium-holders, such as with all four commodities in Thailand. Hence the agro-economic attributes of the commodities associated with land grabbing do not in themselves warrant a policy emphasis on large-scale land investments. This under- scores the point made at the outset that it is primarily politics rather than economics that underlies land grabbing. Where the politics restricts the opportunity for land grabbing and facilitates the expan- sion of smallholder commodity production, the case studies underscore that smallholder success depends on crucial contributions to value chains by private- and public-sector actors. These include upstream contributions, notably the provision of quality planting material, production inputs, technical knowledge and finance, and downstream contri- butions, especially transportation, processing and marketing. Public agencies committed to smallholder development can broker innovative arrangements between farmers and agri- businesses that ensure all parties benefit, such as the fertiliser buying groups (KBT) for oil palm smallholders in Sarawak. Even without such public-sector facilitation, specialised processors dependent on smallholder supplies of feedstock often have an incentive to provide technical and financial support to smallholders in their catchment, as with cassava starch factories in Vietnam (though not extending to the opportunistic cross- border trade with Cambodian producers). To avoid the high transaction costs of dealing with many smallholders, intermediaries in the value chain become important to coordinate activities, including traders and farmer groups. There is little doubt that widespread smallholder engagement with agricultural com- modity chains is associated with more inclusive patterns of rural development than are large-scale land concessions that typically restrict and displace traditional rural livelihoods. However, what Tania Li (2011) calls ‘everyday processes of accumulation and disposses- sion’ can undermine the effectiveness of smallholder-oriented policies in reducing rural 24 Rob Cramb et al. 1040 1045 1050 1055 1060 1065 1070 1075 1080 poverty, leading instead to ‘small-scale land grabs’.AQ21 ¶ All four case studies presented here showed this tendency towards growing inequality among smallholders. Nevertheless, depending on the contextual factors in play, the emerging differentiation can be either impoverishing (as with teak in Laos and cassava in Cambodia) or non-impoverishing (as with oil palm in Sarawak). One important factor is the initial inequality and vulnerability of poorer households in the community. Another is the incentive and opportunity for outside investors to acquire increasingly valuable land from smallholders, many of whom find they end up without their foothold in the village economy and have to resort to selling their labour or migrating. However, strong village institutions can enable all households to benefit from a profitable cash crop, limit the extent of internal differentiation, prevent a land grab by outside investors, and engage on a cooperative basis with the market. The pioneering rubber village in Laos is a rare example of this degree of foresight and self- organisation. However, secure land tenure and smallholder-oriented policies, as pursued in Thailand and (in the past) in Malaysia and Indonesia, can go a long way to ensuring the benefits of engaging in commodity chains are widely spread. Thus, our analysis leads us to agree with Byerlee, who concludes that historical experience has shown the importance of providing a level playing field for small- holders. Where support services have been put in place, including research, extension, land- tenure security, and finance, a vibrant smallholder sector has eventually emerged to dominate the industry. This has not only alleviated land conflicts, but also promoted inclusive rural devel- opment. (Byerlee 2013, 39) There are clearly alternatives to land grabs that enable small- and medium-scale family farms to benefit from cash crops, but these are radically contingent on the conjuncture of favourable land policy and institutions, robust farmer organisations, and well-targeted support from state agencies and agribusinesses. Acknowledgements We are grateful to Niels Fold, the editors and two anonymous reviewers for helpful comments. Pre- vious versions of this paper were presented at the Conference on Land Grabbing: Perspectives from East and Southeast Asia, 5–6 June 2015, Chiang Mai University; at the Southeast Asian Studies Sym- posium, 14–16 April 2016, University of Oxford; and at a seminar in the Department of Geography, University of Copenhagen, 31 May 2016. We would like to thank participants in those conferences for their feedback. Disclosure statementAQ9 ¶ No potential conflict of interest was reported by the authors. Funding Funding for the fieldwork reported here was provided by the Australian Research CouncilAQ10 ¶ and the Australian Centre for International Agricultural ResearchAQ11 ¶ through several projects and scholarships. References Allen, D.W., and D. Lueck. 1998. The nature of the farm. The Journal of Law and Economics 41, no. 2: 343–386AQ12 ¶ . Baird, I.G. 2010. Land, rubber and people: Rapid agrarian changes and responses in Southern Laos. Journal of Lao Studies 1: 1–47. The Journal of Peasant Studies 25 1085 1090 1095 1100 1105 1110 1115 1120 1125 Baird, I.G., and J. Fox. 2015. How land concessions affect places elsewhere: Teleconnections and large-scale plantations in southern Laos and northeastern Cambodia. Paper presented to Conference on Land Grabbing, Conflict and Agrarian-Environmental Transformations: Perspectives from East and Southeast Asia, June 5-6, Chiang Mai University. Barlow, C. 1978. The natural rubber industry: Its development, technology, and economy in Malaysia. Kuala Lumpur and New York: Oxford University Press. Barlow, C. 1986. Oil palm as a smallholder crop. Palm Oil Research Institute of Malaysia (PORIM), Ooccasional Paper No. 21. Kuala Lumpur: PORIM. Barlow, C., and S. Jayasuriya. 1984. Problems of investment for technological advance: The case of Indonesian rubber smallholders. Journal of Agricultural Economics 35: 85–95. Barlow, C., and T. Tomich. 1991. Indonesian agricultural development: The awkward case of small- holder tree crops. Bulletin of Indonesian Economic Studies 27, no. 3: 29–53. Bernstein, H. 2010. Class dynamics of agrarian change. Halifax: Fernwood. Borras, S. M. Jr., R. Hall, I. Scoones, B.White, andW.Wolford. 2011. Towards a better understanding of global land grabbing: An editorial introduction. Journal of Peasant Studies 38, no. 2: 209–216. Byerlee, D. 2013. Are we learning from history? In The Global farms race: Land grabs, agricultural investment, and the scramble for food security, ed. M. Kugelman and S.L. Levenstein, 21–44. Washington, DC: Island Press. Byerlee, D. 2014. The fall and rise again of plantations in tropical Asia: History repeated? Land 3: 574–597. Byerlee, D., and S. Haggblade. 2014. African food systems to 2030: Towards inclusive business models for provisioning Africa. In Frontiers in food policy: Perspectives on sub-Saharan Africa, ed. R.L. Naylor and W.P. Falcon, 509–533. Centre on Food Security and the Environment, Stanford University.AQ13 ¶ Byerlee, D., D. Kyaw, U.S. Thein, and L.S. Kham. 2014. Agribusiness models for inclusive growth in myanmar:Diagnosis andways forward.MSU International DevelopmentWorking Paper 133. East Lansing: Department of Agricultural, Food, and Resource Economics, Michigan State University. Byerlee, D., W.A. Masters, and D. Robinson. 2015. From land grab to land development: The past and potential of private investment in frontier agriculture. Paper prepared for the 2015 World Bank Conference on Land and Poverty, Washington DC, March 23-27. Chambon, B., Ruf, F., Kongmanee, C., and Angthong, S. 2016. Can the cocoa cycle model explain the continuous growth of the rubber (Hevea brasiliensis) sector for more than a century in Thailand? Journal of Rural Studies 44: 187–197. Chayanov, A.V. [1966] ¶ 1925 ¶ . The theory of peasant economy, ed. Daniel Thorner, Basile Kerblay, and R.E.F. Smith. Homewood: Richard Irwin (American Economic Association). Cramb, R.A. 2011. Reinventing dualism: Policy narratives and modes of oil palm expansion in Sarawak, Malaysia. Journal of Development Studies 47: 274–293. Cramb, R.A. 2013a. A Malaysian land grab? The political economy of large-scale oil palm develop- ment in sarawak. LDPI Working Paper 50. The Hague: Land Deal Politics Initiative. Cramb, R.A. 2013b. Palmed off: Incentive problems in joint-venture schemes for oil palm develop- ment on customary land in Sarawak, Malaysia. World Development 43: 84-99. Cramb, R.A., and G.N. Curry. 2012. Oil palm and rural livelihoods in the Asia-Pacific region: An overview. Asia Pacific Viewpoint 53, no. 3: 223–239. Cramb, R.A., and J.F. McCarthy, eds. 2016. The oil palm complex: Smallholders, agribusiness and the state in Indonesia and Malaysia. Singapore: NUS Press. Cramb, R.A., and P.S. Sujang. 2011. ‘Shifting ground’: Renegotiating land rights and rural liveli- hoods in Sarawak, Malaysia. Asia Pacific Viewpoint 52, no. 2: 136–147. Cramb, R.A., and P.S. Sujang. 2013. The mouse deer and the crocodile: Oil palm smallholders and livelihood strategies in Sarawak, Malaysia. Journal of Peasant Studies 40, no. 1-2: 129–154. Dallinger, J. 2011. Oil palm development in Thailand: Economic, social and environmental consider- ations. Oil Palm Expansion in South East Asia: Trends and Implications for Local Communities and Indigenous Peoples. FPP and SawitWatch. Dao, N. 2015. Rubber plantations in the Northwest: Rethinking the concept of land grabs in Vietnam. The Journal of Peasant Studies 42, no. 2: 347–369. Deininger, K. 2011. Challenges posed by the new wave of farmland investment. Journal of Peasant Studies 38, no. 2: 217–247. De Koninck, R., and J-F. Rousseau. 2012. Gambling with the land: The contemporary evolution of Southeast Asian agriculture. Singapore: NUS Press. 26 Rob Cramb et al. 1130 1135 1140 1145 1150 1155 1160 1165 1170 Dieters, M., Newby, J., Cramb, R., Sexton, G., McNamara, S., Johnson, M., Sakanphet, S., Sodarak, H., Kikeo, Savathvong, S., Midgley, D. 2014. Enhancing on-farm incomes through improved sil- vicultural management of Teak in Luang Prabang Province of Lao PDR. Final Report FST/2004/ 057. Canberra: Australian Centre for International Agricultural Research (ACIAR). Eastwood, R., M. Lipton, and A. Newell. 2010. Farm size. In Handbook of agricultural economics, vol. 4, ed. R. Evenson and P. Pingali, 3323–3397. Amsterdam: North-Holland. Fairbairn, M. 2014. ‘Like gold with yield’: Evolving intersections between farmland and finance. The Journal of Peasant Studies 41, no. 5: 777–795. FAOSTAT. 2014. Statistics division, food and agriculture organisation of the United Nations. http:// faostat.fao.org/ (accessed August, 2014). Fox, J., and J-C. Castella. 2013. Expansion of rubber (Hevea brasiliensis) in Mainland Southeast Asia: what are the prospects for smallholders? Journal of Peasant Studies 40, no. 1: 155–170. Hall, D. 2011. Land grabs, land control, and Southeast Asian crop booms. Journal of Peasant Studies 38, no. 4: 837–857. Hall, D., P. Hirsch, and T.M. Li. 2011. Powers of exclusion: Land dilemmas in Southeast Asia. Singapore: NUS Press. Hansen, P.K., H. Sodorak, and S. Savathvong. 1997. Teak production by shifting cultivators in Northern Lao PDR. Paper to a workshop on Indigenous strategies for intensification of shifting agriculture in Southeast Asia, Bogor, Indonesia, June 23-27. Cornell University and the International Centre for Research in Agroforestry, Bogor. Hayami, Y. 1998. The peasant in economic modernisation. In International agricultural development. 3rd ed, ed. C.K. Eicher and J.M. Staatz, 300–315. Baltimore and London: Johns Hopkins University Press. Hayami, Y. 2010. Plantation agriculture. In Handbook of agricultural economics, vol. 4, ed. R. Evenson and P. Pingali, 3305–3322. Amsterdam: North-Holland. Kolmert, A., 2001. Teak in Northern Laos. Minor Field Studies No. 175. Swedish University of Agricultural Sciences, Uppsala. Laungaramsri, P. 2012. Frontier capitalism and the expansion of rubber plantations in southern Laos. Journal of Southeast Asian Studies 43, no. 3: 463-477. Lefroy, R. 2014. Non-rice crops in rice-based farming systems in Mainland Southeast Asia. In Trajectories of Rice-Based Farming Systems in Mainland Southeast Asia, ed. R.A. Cramb, 111–138. ACIAR Monographs No. 177. Canberra: Australian Centre for International Agricultural Research (ACIAR). Li, T.M. 2011. Centering labor in the land grab debate. Journal of Peasant Studies 38, no. 2: 281–298. McAllister, K. 2015. Rubber, rights and resistance: The evolution of local struggles against a Chinese rubber concession in Northern Laos. The Journal of Peasant Studies 42, no. 3-4: 817–837. McCarthy, J.F. 2010. Processes of inclusion and adverse incorporation: Oil palm and agrarian change in Sumatra, Indonesia. Journal of Peasant Studies 37, no. 4: 821–850. McCarthy, J.F., and R.A. Cramb. 2009. Policy narratives, landholder engagement, and oil palm expansion on the Malaysian and Indonesian frontiers. Geographical Journal 175: 112–123AQ14 ¶ . McCarthy, J.F., P. Gillespie, and Z. Zen. 2012. Swimming upstream: Local Indonesian production networks in “globalized” palm oil production. World Development 40, no. 3: 555–569. Mahanty, S., and S. Milne. 2016. Anatomy of a boom: Cassava as a ‘gateway’ crop in Cambodia’s north eastern borderland. Asia Pacific Viewpoint 57, no. 2: 180–193AQ15 ¶ . Majid Cooke, F., D. Ngidang, and N. Selamat. 2006. Learning by doing: Social transformation of small-holder oil palm economies of Sabah and Sarawak, Malaysia. Report submitted to UNESCO, Participatory Programme Research Project No. 27213203 MAL. Manivong, V., and R.A. Cramb. 2008a. Economics of smallholder rubber expansion in northern Laos. Agroforestry Systems 74: 113-125. Manivong, V., and R.A. Cramb. 2008b. The adoption of smallholder rubber production by shifting cultivators in Northern Laos. In Smallholder tree growing for rural development and environ- mental services, ch. 5, ed. D.J. Snelder and R.D. Lasco. Dordrecht and London: Springer.AQ16 ¶ Manivong,V., andR.A.Cramb.2015. Impacts of smallholder rubber on shifting cultivation and rural live- lihoods inNorthernLaos. In Shifting cultivationand environmental change: Indigenous people, agri- culture and forest conservation, ed. M.F. Cairns, 826–840. London and New York: Routledge. Manoram, K., D. Hall, X. Lu, S. Katima, M.T. Medialdia, S. Siharath, and P. Srisuphan. 2011. Cross- border contract farming arrangements: Variations and implications in the Lao people’s demo- cratic republic. Manila: Asian Development Bank. The Journal of Peasant Studies 27 1175 1180 1185 1190 1195 1200 1205 1210 1215 Midgely, S., M. Blyth, K. Mounlamai, D. Midgley, and A. Brown. 2007. Towards improving profit- ability of teak in integrated smallholder farming systems in Northern Laos. ACIAR Technical Reports 64. Canberra: Australian Centre for International Agricultural Research (ACIAR). Midgley, S., R.T. Somaiya, P.R. Stevens, A. Brown, Nguyen Duc Kien, and R. Laity. 2015. Planted teak: Global production and markets, with reference to Solomon Islands. ACIAR Technical Reports No. 85. Canberra: Australian Centre for International Agricultural Research. NAFRI. 2011. Rubber research in Lao PDR. Vientiane: National Agriculture and Forestry Research Institute. Netting, R.M. 1993. Smallholders, householders: Farm families and the ecology of intensive, sustain- able agriculture. Stanford: Stanford University Press. Newby, J.C. 2016. Cassava in Asia: Exposing the drivers and trajectories of the hidden ingredient in global supply chains. World Congress on Roots and Tubers, January 18-22. Nanning China. http://www.gcp21.org/wcrtc/ppt/PS05-JNewby.pdf. Newby, J.C., R.A. Cramb, S. Sakanphet, and S. McNamara. 2012. Smallholder teak and agrarian change in northern Laos. Small-Scale Forestry 11: 27–46. Ngidang, D. 2002. Contradictions in land development schemes: The case of joint ventures in Sarawak, Malaysia. Asia Pacific Viewpoint 43: 157–180. Nyi Nyi. 2015. Future prospects for cassava starch industrialization in Myanmar. Presentation given at 4th Starch World Conference, 29-30 January 2015, Inya Lake Hotel, Yangon. Pletcher, J. 1991. Regulation with growth: The political economy of palm oil in Malaysia. World Development 19, no. 6: 623–636. Rival, A., and P. Levang. 2014. Palms of controversies: Oil palm and development challenges. Bogor: Centre for International Forestry Research. Rubber prices may decrease in Luang Namtha. 2015. Vientiane Times, 23 April. Sheil, D., A. Casson, E. Meijaard, M. van Noordwijk, J. Gaskell, J. Sunderland-Groves, K. Wertz, and M. Kanninen. 2009. The impacts and opportunities of oil palm in Southeast Asia: What do we know and what do we need to know? Bogor: Centre for International Forestry ResearchAQ17 ¶ . Shi, W. 2008. Rubber Boom in Luang Namtha: A transitional perspective. Rural Development in Mountainous Areas of Northern Laos. Vientiane: Lao-German Program. Stiglitz, J. E. 1986. The new development economics. World Development 14 (2): 257-265. Sutton, K. 2001. Agribusiness on a grand scale – FELDA’s Sahabat complex in East Malaysia. Singapore Journal of Tropical Geography 22: 90-105. The Palm Oil Mill. 2011. April 5. http://palmoilmill.co.nz/index.php?option=com_content&view= article&id=57:management-of-a-palm-oil-mill&catid=62:mill-mangement&Itemid=71 (accessed September 12, 2013). Van Vliet, J.A., Schut, A.G.T., Reidsma, P., Descheemaeker, K., Slingerland, M., Vande Ven, G.W.J., and Giller, K.E. 2015. De-mystifying family farming: Features, diversity and trends across the globe. Global Food Security 5: 11-18. Vongvisouk, T., and Dwyer, M. 2016. Falling Rubber prices in Northern Laos: Local responses and policy options. Draft Report prepared for Helvetas. Wandschneider, T. 2008. The role of contract farming in the development of a competitive and sus- tainable cassava sub-sector in Vietnam. Paper presented at the 8th Asian Cassava Research Workshop, Vientiane, October 20-24. Wright, S., 2009. Agriculture in transition: The impact of agricultural commercialisation on livelihoods and food access in the Lao PDR. Vientiane: UN World Food Programme Laos (draft report). Zen, Z., C. Barlow, and R. Gondowarsito. 2005. Oil palm in Indonesian socio-economic improve- ment: A review of options. Working Papers in Trade and Development No. 2005/11. Canberra: Division of Economics, Research School of Pacific and Asian Studies, Australian National University. Zola, A., 2008. A preliminary assessment of contract farming arrangements and plantations in the agriculture and natural resources sector of Southern Lao PDR. Vientiane: ADB. Rob Cramb is professor of Agricultural Development in the School of Agriculture and Food Sciences at the University of Queensland, Brisbane, Australia. His research focuses on agricultural develop- ment and agrarian change in Southeast Asia. He recently co-edited The oil palm complex: small- holders, agribusiness and the state in Indonesia and Malaysia (NUS Press, 2016) with John 28 Rob Cramb et al. 1220 1225 1230 1235 1240 1245 1250 1255 1260 F. McCarthy. He is currently involved in research on cassava value chains in Southeast Asia. Email: r.cramb@uq.edu.au Vongpaphane Manivong is deputy director of the Agriculture and Forestry Policy Research Centre in the National Agricultural and Forestry Research Institute (NAFRI), Vientiane, Lao PDR. He has pub- lished on rice intensification, livelihood diversification, migration and rubber smallholding in Laos and is currently involved in research on mechanisation of lowland rice production and improvements to smallholder cassava production in Laos. Email: v.manivong@gmail.com Jonathan C. Newby is an agricultural and natural resource economist with the International Center for Tropical Agriculture (CIAT) based in Vientiane, Lao PDR, and a research fellow in the School of Agriculture and Food Sciences, University of Queensland, Brisbane, Australia. He has published on conservation farming in the Philippines, rice intensification in Laos and Cambodia, and small- holder teak production in Laos. He is currently coordinating a comparative project on smallholder cassava production and marketing in Indonesia, Vietnam, Laos, Cambodia, and Myanmar. Email: jnewby@cgiar.org Kem Sothorn is a research officer with the Cambodian Development Resources Institute (CDRI) and a PhD scholar in the School of Agriculture and Food Sciences at the University of Queensland, Bris- bane, Australia. He has conducted research on various aspects of agricultural policy in Cambodia and is currently studying the implications of the expansion of cassava production in Cambodia for rural livelihoods and agrarian change. Email: sothornkem@gmail.com Patrick S. Sibat is an agricultural and environmental scientist with a private rural development con- sultancy in Kuching, Sarawak, Malaysia. Previously he was a soil scientist with the Sarawak Depart- ment of Agriculture. He has conducted research on social and environmental impacts of oil palm development and resettlement of rural communities, and has published on shifting cultivation and smallholder oil palm in Sarawak. Email: pssrsb@gmail.com The Journal of Peasant Studies 29 1265 1270 1275 1280 1285 1290 1295 1300 1305