Recognizing Carbon Forestry’s Uneven Geography: A Response to Purdon and the Structure-Agency Dichotomy That Never Was

Abstract

We here respond to the critique by Purdon of an article on carbon forestry that we published in this journal last year (Carton and Andersson). While we welcome critical engagements with our work, Purdon’s argument is wide of the mark and appears based largely on misconceptions regarding our theoretical entry point and empirical findings. Underlying this are fundamental disagreements about the nature of carbon forestry, structure-agency dynamics, and how to understand environmental interventions in the global South more broadly. We argue that we are unlikely to “find common ground” in our respective analyses of the Trees for Global Benefits project unless we share a common understanding of the unequal power relations and fundamental geographical unevenness within which carbon projects operate. Contrary to what Purdon argues, this position has nothing to do with ignoring local benefits, nor with denying the agency of the smallholder farmers who participate in the project. We see no contradiction between an analysis that does justice to the various structural conditions that frame carbon forest projects, and a recognition of local agency.

Keywords:

• Carbon forestry
• offsetting
• political ecology
• Uganda
• uneven development

Introduction

In a recent article for this journal (Carton and Andersson 2017), we present findings from our research on the “Trees for Global Benefits” (TFGB) project in Western Uganda, a community-based carbon forestry project that engages smallholder farmers in tree planting and sells the resulting carbon credits on the voluntary carbon market (Ecotrust 2018a, 2018b). Our contribution is part of a special issue on the “subsumption of nature” (cf. Boyd, Prudham, and Schurman 2001; Smith 2006; Boyd and Prudham 2017), an analytical framework that focuses on the different ways in which (different) natures are brought into capitalist production processes, and the challenges and opportunities that occur along the way (Carton, Jönsson, and Bustos 2017).

The main point advanced in our article is twofold. On the one hand, we highlight the various ways in which the TFGB project aligns local afforestation efforts with the requirements of the global carbon market, by promoting fast-growing indigenous trees and implementing a planting and management regime that aims to ensure the reliability and economic efficiency of the carbon offsets the project generates. We show that this process is significantly complicated by ecological factors, or what in line with recent academic debates can be termed nature’s “unruly” or “uncooperative” materiality (Bakker 2004; Bakker and Bridge 2006; Leach and Scoones 2015b). On the other hand, we demonstrate that the difficulties in guaranteeing a reliable “carbon commodity” are also profoundly social, in that project participants commonly do not fully understand their contract obligations, do not always follow specified planting and management practices and fail to meet their planting objectives. Our findings show that this is at least in part attributable to insufficient attention to local realities in project design, inadequate support and follow-up activities, as well as unrealistic assumptions by Ecotrust, the organization coordinating the project, on the pre-existing capacity, knowledge and/or willingness of participants to persist, over a period of years, with the specific management practices required, particularly when problems occur. We draw a parallel here with Marx’ (1977) discussions on the subsumption of labor, and suggest that in addition to ecological factors, the project’s difficulties in guaranteeing a reliable degree of carbon sequestration can be seen as an indication of smallholder farmers “unruly” involvement in the global carbon market, in other words, of the fact that they have considerable leverage over their own land use and farm management practices and are therefore “less subsumed” to capital than for example industrial laborers are. Taken together, our case study underlines the value of applying a subsumption framework to non-industrial sectors and capital accumulation processes, and highlights the need for research that integrates perspectives on the subsumption of nature and labor.

In a commentary on our article, Purdon (2018) criticizes our work on a number of grounds, including our choice for a case study approach and our supposed denial of the project’s benefits. Above all, he argues that we have analyzed the TFGB project through a “structural Marxist filter” that “den[ies] political agency” of smallholders by “treating noncapitalists as passive actors in the global economic system” (p. 3). While we welcome critical engagements with our work, Purdon’s main argument appears to hinge on a peculiar and untenable dichotomy between structure and agency and seems based more on preconceived ideas about Marxist analysis than an engagement with the gist of our argument. This provides a rather shaky foundation for finding the “common ground” that Purdon is calling for. We here refute his critique and underscore the value of a political ecological analysis of carbon forestry – a perspective that more accurately represents our approach than the ill-defined “structural Marxism” that Purdon maligns. The remainder of this text addresses the main arguments put forward in Purdon’s commentary, focusing mostly on the structure vs agency critique before briefly commenting on methodological considerations and empirical disagreements.

The Structure-agency Dichotomy That Never Was

Purdon’s most substantial critique revolves around our alleged “denial of rural smallholder political agency” (p. 6). He claims, for example, that our analysis denies “the possibility that farmers might have participated willingly—and some strategically—in the TFGB project” (p. 6) and that we “close the door on the possibility that smallholders might be interested in planting trees on their lands in exchange for carbon payments”. Purdon in this way paints our argument as saying that farmers were coerced into the TFGB scheme, which he sees as a logical consequence of our allegiance to “subsumption theory” which, as all “structural Marxist” analysis, “implies subordination and passivity” and therefore denies “the agency of smallholders” (pp. 6–7).

This would have been convincing were it backed up with citations from our text that illustrate these claims. But Purdon can do no such thing of course, because they are not there. Rather, his critique is based on stereotypical generalizations of Marxist analysis (which he insists on calling “structural Marxism”, despite the latter being a specific stream within Marxist thinking, mostly associated with the work of Althusser in the 1960s and 1970s) and misconceptions about the structure/agency debate more generally. By equating Marxist theory with a denial of agency, Purdon seems to suggest that any analysis of the uneven, structural power relations inherent in market interactions (beyond the community-level inequalities that he himself mentions) somehow automatically deprives people of their ability to influence and alter their everyday realities. By doing this, Purdon sets up a rigid dichotomy between structure and agency that is entirely untenable, and that is far-removed from our argument and the critical literature we draw from. Framing structural analysis as antagonistic to concerns for political agency flies in the face of past and contemporary (and often Marxist-inspired) scholarship on class, gender and race, on uneven geographical development, on governmentality, in short, on the uneven social relations and positions of power that permeate everyday life and thereby shape and condition the opportunities and abilities of individuals to act. As this rich literature surely shows, there are a multitude of ways in which one can understand the interdependent and dialectical relations between structure and agency, without a concern for one having to constitute a denial of the other.

Clearly however, Purdon does not allow such nuance to stand in the way of his blanket disapproval of Marxist theory. He charges that “subsumption itself implies subordination and passivity” (p. 6) but this is demonstrably false. A careful reading of our argument shows that agency is at the center of not just our own analysis, but also of Boyd, Prudham, and Schurman (2001)’s original theorization of the subsumption of nature. One of the main points of the subsumption thesis, as outlined in both our original article and in the special issue editorial, is that nature (and labor) is not easily or willingly subsumed to the requirements of capital, but that it provides barriers, obstacles and unintended consequences, occurrences that affect and influence the process of commodification. In other words, the subsumption of nature/labor is a tendency, an intention deriving from capitalist dynamics, but an intention that is necessarily and inevitably incomplete exactly because of the irreducibility of labor and nature to capitalist processes (cf. Polanyi 2001). Due to its autonomous character, nature inevitably ends up resisting or otherwise influencing the way in which it enters into capitalist production processes. Hence, our analysis shows that selected tree species grow less well than intended, and are subject to challenges such as droughts, diseases and infestations. Similarly, local participants are recruited into the project on the conditions of the project organizers/the carbon market, but their agency is evident in their choice to join (based on the expectation that they will benefit from doing so), in how well they live up to their contract requirements, to what extent they adhere to management guidelines, etc. These are contingent and contextual processes that in turn relate to people’s situated expectations of the project and how well they feel these are being met. Evidently then, the agency of project participants as well as the unruly materiality of nature is fundamental to our argument and any understanding of the project’s ability to generate carbon offsets.

Unlike what Purdon asserts, also, we nowhere claim that “rational economic behavior among smallholder farmers in sub-Saharan Africa can only be imposed from outside” (Purdon 2018, p. 9), that participants are “isolated from the market economy” (p. 7) or that they are coerced into the project. Our argument is, and has been throughout, specifically related to the carbon offset market as a distinctive kind of market that provides particular opportunities, risks and relationships, and that places specific demands on smallholder farmers. Similarly, we suggest that participants are not fully integrated into wage–labor relations and that, together with the particular requirements posed by the offset market, this has repercussions for the project’s ability to produce reliable carbon credits. Far from arguing that smallholders are somehow incapable of rational market behavior, our point is rather that it is simplistic – and problematic – to reduce participants’ land management choices to a question of rational market behavior. It is meaningless to discuss participants’ motivations and choices without also paying attention to the various structural conditions that constrain their ability to join the project and take care of their trees in the long-term, or indeed their real opportunities to make informed decisions about participation, which require contracts, payment plans and management instructions that are locally appropriate (see e.g. Fisher 2012; Andersson and Carton 2017).

The dichotomy that Purdon constructs between smallholders either participating voluntarily in TFGB or being coerced to do so, as he mistakenly faults us for, needs to be seen in this context. It is not because we describe participants as being disciplined into the requirements of the carbon market that we are talking of a coercive relationship, let alone one that involves “the purposeful exploitation of farmers” (Purdon 2018, p. 6) by Ecotrust. Purdon reduces our analysis to an accusation of deliberate malpractices on behalf of the project organizers, while what we are actually saying is that in order to meet their planting objectives, participants need to conform to a set of practices and requirements that are stipulated by the demands of a global carbon market and that they therefore have little possibility of influencing. Our use of “disciplining” thus refers to a Foucauldian understanding of how power is exercised in various everyday, subtle and primarily non-coercive ways, as expressed for example in discussions on environmentality (Agrawal 2005) and the “will to govern” (Murray 2007). The question we are raising, essentially, is on whose conditions landscapes in the project area are being altered, and what the implications are when, as our analysis shows, the creation of “carbon farmers” turns out to be less easy than expected. The fact that people willingly (indeed, eagerly) choose to join projects like TFGB does not take away the fundamental inequalities that underpin the choices they make or their ability to (not) participate in the first place. That participants are “free” to join the project in no way diminishes the various unfreedoms that this is necessarily accompanied with. Purdon’s own interviews suggest as much, when he describes the limited bargaining power people had in drawing up their contracts:

We couldn’t bargain for things we didn’t know” one farmer explained. Another stated “The organization came as if it is going to give people help. So people were really ready to get help. There was no chance to bargain when someone is coming to help. (Purdon 2018, p. 5)

To us this points to the importance of recognizing the inevitably unequal relationship that is created when credit buyers from the global North engage in purportedly free and equal contracts with poor smallholder farmers. It shows the kind of concerns that arise when projects operate in the gray zone between development intervention and market exchange, where participants are approached as simultaneously in need of aid and as equal partners on the market. Consider in this context also Fisher et al.’s (2018) analysis of the TFGB project, which shows that most participants in fact have very little idea about the kind of market that they are involved in, essentially undermining the idea that they have given their informed consent to the project (cf. Corbera and Martin 2015). All of this suggests that Purdon’s opposition between voluntary versus coercive participation is an all too simplistic framework for capturing the complex dynamics at play. The point of pursuing “structural” analyses of the TFGB project, then, is not to deny agency but to conceive of it as historically situated, as necessarily operating within multiscalar socioeconomic and political relations that are by definition uneven and therefore have implications for how to understand the dynamics and outcomes in projects like TFGB.

Methodological Choices

Purdon also criticizes our case study method and argues that his own approach, based on comparative design that weighed the project’s benefits and shortcomings against each other, and combines qualitative and quantitative methods to allow statistical analysis, gives a more accurate picture of the TFGB project. But this is comparing apples and oranges. Surely, the research design and methodological approach that is chosen is relative to the specific questions one sets out to answer. We are researchers interested in examining specific aspects of the TFGB project, not consultants contracted to provide a cost-benefit evaluation of the project. We never set out to make a full assessment of the project, nor do we agree that any study of the project (or any project) should aspire to do so. As our introduction and analytical framework clearly indicates, the main objective of our article is to examine the opportunities and obstacles generated in the creation of forest-based carbon offsets itself, analyze how these manifest themselves, and attempt to theorize this. While our results are specific to the places we visited and the people we talked to, they also raise broader questions about uneven dynamics, risks and implementation obstacles in carbon forestry. Our research approach in this sense is an explanatory one, for which case study research is a widely accepted method (Gomm, Hammersley, and Foster 2000; Yin 2003). By adopting this approach we connect to a large body of literature, specifically in political ecology, that uses case study analysis to unpack the dynamics, promises and challenges that characterize environmental interventions in the global South (see for example, Wittman and Caron 2009; Osborne 2011, 2015; Shapiro-Garza 2013; Cavanagh and Benjaminsen 2014; Bryant, Dabhi, and Böhm 2015; Hendrickson and Corbera 2015; Leach and Scoones 2015a; Osborne and Shapiro-Garza 2018). Our article thereby intervenes in a different debate than the one Purdon seemingly wants us to engage with. To criticize us for not adopting a comparative approach to our analysis is simply to disagree with our focus and research questions.

On Empirical Disagreements

The examples Purdon gives of where his own research contradicts ours are unconvincing. He disputes our finding that many participants are not fully aware of project conditions and contract details with anecdotal evidence on the role of one of the local project coordinators in helping to develop “each smallholder’s household land-use management plans (“Plan Vivo”)” (p. 4). But while Purdon in this way argues that Ecotrust (by way of the local coordinator) has undertaken efforts to inform farmers about the project details, our interviews show that participants are commonly uncertain about, for instance, the planting, spacing and management requirements, as well as their expected economic compensation. We do not see a contradiction between these two findings. Being informed about the project conditions, and actually understanding them, are two fundamentally different things, particularly when the guidelines change along the way, as seems to have been the case. Rather than disproving our argument, this merely suggests that farmers are not fully aware of their contracts despite the organization’s efforts. If anything this reinforces rather than challenges our point on the inherent complexities that characterize community-based carbon forestry.

Moreover, it should be noted that Bitereko, the project region that Purdon examined, was the site of TFGB’s pilot activities. Our findings from two surrounding sub-counties, and from the experiences of farmers in Bitereko who joined the project more recently, suggest that Ecotrust invested more resources in training and project guidance at the initial phase of the project (hence, in the pilot project region) than it did in recent years. The possibility of investigating local experiences of the project over time was precisely one of the motivations for focusing on this particular region. Also, while in our article we do not mention the coordinator that Purdon refers to, we did in fact interview her, but we of course did not rely on her exclusively for information about the project.

Finally, while it is true that we do not provide an analysis of the project’s benefits (for the methodological reasons mentioned above), we clearly mention alleged benefits and state (in the conclusion) that our argument in no way denies the project’s potential value to local participants. There is nothing in our text that contradicts the idea that participants gain something from their involvement in the project, which at a minimum they do in the form of payments (if and when they arrive), forest products and, ultimately, timber revenues. Indeed, participants commonly mention that they appreciate the manifold uses they derive from their trees. It is mainly the payments they have concerns about, because these come with long delays or not at all, are lower than anticipated, or are seen to insufficiently compensate for expended labor and the difficulties encountered in actual tree management. Our primary attention to these issues should not be seen as a denial of tangible benefits to participants. Nor does the existence of benefits somehow invalidate our conclusions, as Purdon appears to suggest. It is not because TFGB is advantageous for participants in some ways that their other concerns are rendered less relevant, or that it becomes less necessary to critically examine how, for whom, and on whose conditions the project is being implemented.

Conclusion

Our analysis of the TFGB project is situated within a long tradition of Marxist-inspired scholarship in political ecology that aims to unveil and critically examine the power relations with which human–environment relations are infused (Peet, Robbins, and Watts 2011; Robbins 2012). These perspectives have proven instrumental in studying carbon forestry projects, helping to contextualize the claims made on their behalf and make sense of the various tensions and trade-offs that are often identified (Leach and Scoones 2015a). Political ecology does this by providing a set of theoretical tools that acknowledges the variety of social, economic, political, ecological and cultural conditions under which projects unfold. It provides a necessary counterweight to the often simplistic ideas that underpin environmental interventions in the Global South, highlighting how local ecological changes and interactions are embedded in global networks of power and, vice-versa, how external interventions by necessity are articulated through complex interactions with existing landscapes, socioeconomic conditions, cultural practices and power structures. This theoretical commitment involves paying attention to the highly variegated and complex ways in which structural inequalities and power relations intertwine with, constrain, and are altered by the agency of individuals, communities, institutions, corporations, etc. in specific places.

Reflecting on this wider literature, this means that we indeed see carbon forestry as inherently constituted by tensions and trade-offs. We believe that the collective body of literature on projects such as TFGB convincingly unsettles the idea that these somehow constitute easy or cheap ways to reduce emissions. In our view – and this was not the focus of our article, but given the predominant focus that Purdon puts on it, it is worth spelling out – this is expressed in the unequal conditions that limit participants’ opportunities to negotiate the terms of their contracts or to influence the design of the project, the kind of trees to be grown, or the way in which plantations are to be managed. It is manifested also in their limited access to carbon credit buyers or project organizers when they run into challenges or seek to raise concerns about unclear, delayed or lacking payments. Clearly, of course, project design matters, and compared to other projects TFGB is arguably one of the more beneficial ones. But the fact that TFGB delivers benefits in no way takes away the relevance of highlighting the unequal conditions the project has to work with, or analyzing how these affect the distribution of risks and opportunities between the different actors involved.

Conflating this research approach with a denial of agency and an accusation of deliberate malpractice, as Purdon does, is deeply misguided. Our project is not to deny agency but to problematize predominant liberal notions of it that seem oblivious to the manifold political and socioeconomic relations that individual agency is constrained by. To the extent that Purdon’s critique reflects a commitment to such liberal ideas – and it certainly seems so to us – our differences are not just theoretical and methodological, but ideological. While it is worthwhile to try and tease out our disagreements through additional research, then, it seems unlikely that we, from our side, will find common ground with an approach that sees no value in analyzing the uneven geography of carbon forestry and dismisses concerns for structural power relations out of hand.

Acknowledgments

We would like to thank the two reviewers who provided constructive comments on an earlier version of this text. All remaining shortcomings are our responsibility alone.

Additional information

Funding

We acknowledge funding from the Swedish Society for Anthropology and Geography (SSAG) and the LUCID research school at Lund University, who both provided travel funding for our fieldwork in Uganda in 2015. Wim Carton’s research is funded through a mobility grant from Formas, dnr. 2016-00892.