Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda

Geoforum xxx (2013) xxx–xxx

Contents lists available at SciVerse ScienceDirect

Geoforum journal homepage: www.elsevier.com/locate/geoforum

Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda Adrian Martin ⇑, Nicole Gross-Camp, Bereket Kebede, Shawn McGuire, Joseph Munyarukaza School of International Development, University of East Anglia, Norwich, UK

a r t i c l e

i n f o

Article history: Available online xxxx Keywords: Environmental justice Environmental conflict Payments for ecosystem services Opportunity costs Recognition Protected areas

a b s t r a c t Distribution and procedure, two core social justice concepts, are central concerns for the design and practice of payments for ecosystem services (PESs). This paper explores the relationship between local conceptions of justice and the more globally referenced justice principles embedded in the design of PES schemes. The importance of this is that perceptions of justness are powerful determinants of human behaviour and, consequently, many environmental conflicts arise from contested visions of what constitutes ‘just’ environmental management. With that in mind we propose that PES schemes built on conceptions of justice that broadly align with those of prospective service providers will be better received than those that do not. In order to explore differences in justice conceptions, we specify three commonly defined dimensions of environmental justice: distribution, procedure and recognition. We predict that there will be differences in the importance different actors place on these different dimensions of justice and also differences in how each particular dimension is conceived. We interview 80 randomly selected respondents from a PES case in Rwanda and relate their views about justice to the design of the PES. Our findings challenge the implicit universalism in many market-based conservation interventions: that imposed framings of justice will resonate with local ones. They also challenge the assumption that different dimensions of justice are always mutually supporting – the fallacy of the rising tide that lifts all boats. We also conclude that an environmental justice framing provides a fruitful new analytical approach for research into global forest conservation efforts. Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction: why justice? Our paper examines the practice of payments for ecosystem services (PESs) in relation to conceptions of environmental justice. It does this through a PES case study in the Nyungwe National Park, Rwanda (see Gross-Camp et al., 2012). The paper asserts that justice is an important, yet relatively under-researched aspect of PES interventions. In particular, we argue that there are important lessons to be learned from examining the notions of socialenvironmental justice evident in the design of PES schemes, and the notions of justice held by local actors who serve as hosts and partners for such schemes. We posit that aligning projects to local conceptions of justice is intrinsically good, but equally important is instrumental in achieving long-term conservation objectives. Environmental justice has become a powerful narrative that is increasingly employed in both research and advocacy around environmental management (Lejano et al., 2002). Whilst the environmental justice movement has roots in the toxic burden suffered by ethnic minorities in the United States (e.g. Bunyan and Mohai, 1992; Bullard, 1990, 1993), claims for environmental justice are ⇑ Corresponding author. E-mail address: [email protected] (A. Martin).

now found around the world, as demonstrated in collections of works edited by Schroeder et al. (2008) and Carmin and Agyeman (2011). Furthermore, environmental justice work now engages with scales of analysis beyond the state, owing to the increasingly global nature of political-economic systems and of environmental issues themselves (Sikor and Newell, this volume; see also volumes by Walker and Bulkeley (2006), Holifield et al. (2009), Peet et al. (2010) and Martin (2013)). The multitude of environmental justice movements and claims are particular and place-bound local struggles. However, they are also connected through the vocabularies employed, global treaties referenced, and commodity chains and networks of actors. Narratives of environmental justice are therefore both locally and globally constituted. We propose that research into environmental justice is particularly relevant to PES and to our focus on tropical forests. Our argument is, quite simply, that justice matters. People’s individual and shared perceptions of (in)justice begin to develop from an early age (Sen, 2009; Almas et al., 2010). These non-pecuniary motives are known to be important determinants of human behaviour, often outweighing concerns of personal financial gain (Fehr and Falk, 2002; Montada, 2002). This understanding of the importance of justice to human behaviour has stimulated considerable bodies of research applied to other areas, such as organisational theories

0016-7185/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.geoforum.2013.02.006

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

2

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

of labour relations (e.g. Moorman, 1991; Folger and Kanovsky (1989); Folger (1989); Whiteman, 2009). Yet relatively little attention has been paid to its role in the implementation of forest conservation. This is surprising given that conservation work has justice principles running through its veins. Whilst not always explicit, these principles can often be inferred from the underlying justification for conservation (such as concerns for ‘common heritage’ and future generations), the selection of different approaches to forest protection (such as strict preservation or Integrated Conservation and Development Projects) and in the detailed design of specific interventions, such as decisions to compensate people for harm suffered, or reward them for services rendered. Whilst it is useful to attend to the justice principles implicit in forest conservation policies and interventions, we also propose the need to explore these in relation to local conceptions of justice. For example, compensation schemes tend to imply particular ideas about what are just outcomes, but these ideas are not always shared by local people (Whiteman, 2009), who may believe it more just to mitigate the risk of harm in the first place than to be helped to live with that risk. Likewise, access and benefit sharing schemes may be conceived as ‘fair’ within international frameworks such as the Convention on Biological Diversity, but their framing of justice may be different to that of intended beneficiaries, who might prioritise alternative dimensions of justice such as recognition of their culture, right to self-determination, or traditional decision making procedures (see e.g. Robinson, 2010; Vermeylen and Walker, 2011). For these reasons, the study of justice in biodiversity conservation can contribute to our understanding of conservation in practice. We acknowledge that ‘justice’ poses considerable conceptual challenges, not least because of the practical (if not intellectual) impossibility of reaching consensus. With the exception of some utilitarians, most liberal political philosophers, as well as their critics, accept that there are multiple conceptions of what constitutes a ‘good’ empirical outcome. However, there is considerably more controversy over the possibility of universal conceptions of rights, including the fundamental right for individuals to pursue what they conceive to be good (Sandel, 1998). For the purposes of this paper we identify two levels of debate about the nature of justice that serve as our conceptual framework. Firstly, there is debate about what constitutes the relevant pillars or dimensions of environmental justice going beyond matters of material distribution, to include dimensions such as procedure, recognition or representation. Vincent (1998) and Dobson (2007) propose that one can largely focus on matters of distribution, as these are foundational; similarly, feminist justice theorists (Fraser, 2001, 2009; Marion Young, 2011) argue for the primacy of recognition and representation; whilst Crocker (2008) asserts that just procedure is foundational in the sense that it helps bring about just distribution. In this research, we follow Schlosberg (2004, 2007) in not ascribing primacy to any one dimension and instead characterising social justice as multi-dimensional (Walker, 2012). This more open framing of justice allows us to explore the use of different justice dimensions, and to maintain an open mind about the relationship between these dimensions. We suggest that it is unsafe to assume that one dimension necessarily underpins or supports another dimension, i.e. just procedure will lead to just distribution (Crocker, 2008), because the research to tell us that has not yet been done. Conservation and development interventions often focus on the distribution of costs and benefits, for example seeking to reduce or offset the costs of forest conservation to the poorest, or to increase their share of benefits from activities such as tourism. Distribution of benefits can occur through a range of interventions such as integrated conservation and development projects, revenue sharing, PES and compensation schemes. Whilst it is relatively common

to analyse the effectiveness of the resulting distribution (e.g. whether it is financially regressive or progressive), we take a rather different line of evaluation by asking whether the very focus on income distribution supports or contracts opportunities in other dimensions of justice. In other words we leave open the possibility that a materially progressive outcome may not be conceived as ‘just’ by actors who prioritise other dimensions of justice. Without evidence to the contrary, we should assume that a focus on one dimension of justice (such as distribution) will not always have positive impacts on other dimensions of justice. In our case study, we especially consider the question of whether a distribution-focused intervention can undermine justice claims for recognition, a possibility we infer from some previous studies of access and benefit sharing interventions (McAfee, 1999; Schroeder, 2008; Vermeylen and Walker, 2011). We also consider different claims to justice within a particular dimension: what constitutes just distribution, just procedure and just recognition? In particular, we consider some of the most established ways of defining ‘just distribution’, employing concepts of ‘need’ and ‘desert’. Again, our basic supposition is that principles employed in the design of forest conservation interventions will not necessarily correspond with principles that are prevalent within local host communities. For example, a compensation scheme may allocate its limited resources according to pro-poor principles of distribution (needs-based), whereas most local people might believe that all should receive the same (egalitarian). Thus we ask two main questions in this paper: (1) how do local conceptions of justice fit with those of conservation interventions in our case study location and (2) has the recent emphasis on improving the distribution of costs and benefits impacted on the justice dimension of recognition. In addition to these questions there is an over-arching ambition to see whether the employment of an environmental justice framework provides a useful way of gaining new understanding of conservation in practice.

2. Conceptualising environmental injustice 2.1. Justice as plural There is a strong tradition of seeking universal principles of justice which might serve as benchmarks for subsequent normative judgments. This is evident in some canonical works such as Plato’s (1974) attempt to articulate a virtuous life, Bentham’s (2009) utilitarian principle of achieving the greatest happiness for the greatest number, Kant’s (1998) argument that individual rights must come prior to identification of utility, and categorical imperative that we should act in ways that we would be content to become universal laws of action, or Rawls’s (1971) egalitarian principles of liberty and pro-poorness. Although the pursuit of universal justice ideals is a worthy one, we adopt a pragmatic conceptual approach that assumes that the pluralism of justice is intractable. As such we follow Sen’s (2009) argument for a more practical ‘idea of justice’. Sen illustrates justice pluralism through a fable about a flute that is claimed by different children. The first child claims the flute based on her expertise; she alone can play the flute. The second child is uniquely poor having no other toys to play with. The third child reveals that she actually made the flute. Readers can see where Sen is taking us, and we could contribute further claims, for example another child might uniquely have the storage conditions to preserve the flute for future generations. Our moral persuasion will likely dictate who the ‘winner’ is and will differ accordingly from individual to individual. Sen (p. 13) suggests a libertarian might favour child 3, an economic egalitarian child 2, and a utilitarian might be swayed by the greater happiness the flute would bring to child 1. Sen’s point is to show that there will

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

always be more than one rational argument for resource allocation and it may be impractical for us to think that we can agree. ‘There may not indeed exist any identifiable perfectly just social arrangement on which impartial agreement would emerge’ (p. 15). This potentially intractable pluralism can also be seen in a common forest conservation dilemma, such as the exclusion of poor people from a biodiversity hotspot. A Benthamite utilitarian might consider the utility arising from the common heritage to outweigh the loss of utility for the few. Other consequentialist thinkers might attach greater moral weight to the harm done to the few. Pareto optimality might have been secured if the excluded had been compensated. Rights-based (deontological) liberal thinkers might focus more on the fundamental rights of individuals and groups to be free to pursue their own conception of utility – lives that they consider worth living. Rawlsians might highlight the institutional failure in decision making procedure that has led to a betrayal of the ‘difference principle’ – that the inequality arising from this decision operates against those already disadvantaged. Feminist justice theorists might focus on the recognition injustice: that those outside of the dominant way of thinking about conservation are muted unless they assimilate into it. Moral philosophers could no doubt continue this list for some time, and that is before even broaching non-western traditions of justice thinking. Fortunately, our purpose here is not to weigh the relative merits of each, but only to see a little way through the lenses that they provide for us. Most importantly, the acknowledgement of plurality enables us to identify differences in justice conceptions between different groups of actors, not with a view to judgment, but rather to consider the potential significance of those differences, with implications for how we might practically engage with them in the pursuit of long-term conservation and social development goals. It is almost inevitable that such differences exist because organisations typically justify environment and development activities in normative terms and villagers often engage critically with this rhetoric, for example finding it stupid or offensive (von Benda Beckmann, 1989). We should add that pluralism is not about falling into a moral void. We might not agree on what constitutes ‘perfectly just social arrangements’ but we find it relatively easier to agree on patently unjust ones. Where we can agree the scope for ‘more just’ conservation, we simply assert that there are a number of potentially reasonable responses. This is itself a challenge to conservation because finding the time and resources to adequately deliberate possible responses can be hard to reconcile with the practical exigencies of project implementation. Furthermore, there are different and competing framings of what constitutes just deliberation and consensus-making (Martin and Rutagarama, 2012; Whiteman, 2009). 2.2. Justice as multi-dimensional The second component of our justice framing is multi-dimensionality. We follow Schlosberg (2004) in considering three dimensions of justice: distribution, procedure and recognition. Distributive justice concerns the distribution of goods and bads between different individuals and groups and has dominated environmental justice debates to date (Walker, 2012). This is perhaps not surprising given the materiality of environmental goods and bads, and the implications of inter- and intra-generational distribution for the wellbeing of current and future people. Procedural justice is largely concerned with the decision-making process, including who is included and on whose terms. Procedure is sometimes treated as a separate dimension of justice (e.g. Schlosberg, 2004) and other times subsumed under distribution (e.g. Vincent, 1998). In the latter case, procedure is seen to be an antecedent of distribution, i.e. good procedures facilitate just distribution, but are not just or unjust in their own right. We opt to treat procedure

3

as a separate dimension on the basis that participation in decisionmaking is not only a means to good decisions, but potentially also an end in itself, for example because it is a form of freedom (Outhwaite, 2009; Sen, 2009; Dryzek, 2000). We define procedural justice as the fairness of participation, as observed in government and other organisational procedures for determining distribution. It is an important pre-condition for distributional justice and therefore a cause of injustices, but it may not always equate to distributional justice. Justice as recognition is rather more difficult to define. Its place in justice thinking became prominent during the 1980s as a result of studying social movements and the terms upon which they construct justice claims. In this way, the shift to theories of justice as recognition arose from the moral grammar of contemporary social movements that centred on respect for difference and the struggle to avoid domination by others (Bohman, 2007). Recognition injustice or ‘malrecognition’ refers to lack of respect for cultural difference. In the US environmental justice movement, the main social fault-line was ethnicity, with malrecognition taking the form of environmental racism (Pulido, 2000). Malrecognition is embedded in cultural norms at large in society, in discourses that frame the ways in which problems and their solutions are conceived, and even in the structures of language itself. Within literature on environmental justice, recognition has largely been expressed in relation to the claims of indigenous and local peoples to cultural respect and self-determination (e.g. Escobar, 1998; Castree, 2004; Schlosberg, 2004; Vermeylen and Walker, 2011). For our purposes we consider any instances in which forms of discursive power suppress the right to alternative ways of thinking. This extends to framings of justice itself: malrecognition can occur where conservation organisations uncritically assume that they are the legitimate source of justice norms (Whiteman, 2009). Such framing can determine whose knowledge counts and which problems and conflicts are rendered visible or invisible (Escobar, 1998). The exercise of such discursive power (McAfee, 1999; Schroeder, 2008; Sullivan, 2009) might be quite subtle, for example where local actors have little choice but to sign up to PES schemes even where these don’t align with the real nature of the injustice that they feel. In our case study in Rwanda, we are interested in whether PES and its associated discourses conflict with cultural norms and traditions of local people and whether there is pressure on local people to assimilate to these imposed ways of thinking.

3. Methods Our research methods were designed to explore local conceptions of injustice, namely those pertaining to distribution, procedure and recognition. We were interested in capturing the nuances of justice and their potential conflicts as perceived by the target communities based on our preceding speculation on recognition. Our study location included selected rural settlements surrounding the Nyungwe National Park in Rwanda, where we have established an experimental PES scheme (Map 1). Other social features of the location that are worthy of note are the relatively low levels of income (Government of Rwanda, 2007; Ansoms and McKay, 2010), high population densities (300+ people/km2; Masozera and Alavalpati, 2004) and recent designation of National Park status in 2004 leading to greater exclusion than under its previous designation as a forest reserve. The establishment of the park was largely based on its ecological importance as one of the largest remaining high altitude forests on the continent (Vedder et al., 1992). Furthermore the park houses a high level of biodiversity and is greatly valued for its hydrological services for the country and Congo–Nile watersheds.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

4

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

Map 1.

Household interviews took place in four cells (an administrative unit containing between 300 and 1130 households) that were randomly selected for inclusion in an experimental pilot PES scheme. Within each cell, 48 households were also randomly selected for a household livelihoods survey conducted in September and October 2009. We then stratified this sub-sample according to income, selecting the relatively 10 richest and 10 poorest households for semi-structured interviews (N = 78) that provide the main data source for this paper. These interviews were conducted in April and May 2011. The structured, deductive component of the interviews was influenced by the desire to test pre-determined conceptions of justice, and to relate this to the design of the PES scheme. We describe the PES scheme in more detail elsewhere (Gross-Camp et al., 2012) but in brief, local communities enter into contracts to communally provide an ecosystem service, namely biodiversity conservation gauged by the level of human impact including measures of hunting, bamboo and tree-cutting, the establishment of new trails and mining. Depending on the level of service delivery, variable financial payments are made, with a proportion going to cell-level for communal spending and the remainder distributed to household level. The proportions going to cell and households are decided through participatory planning events during annual contract negotiations and differ from cell to cell. The starting point for a discussion of distributive justice conceptions was a simple ranking exercise in which individual respondents put in order of preference four different statements about possible ways of distributing payments for ecosystem services: 1. Effort, 2. Opportunity Cost (as based on prior park use), 3. Need, 4. Egalitarian. The first three of these statements represented different forms of justice as ‘desert’. The first proposed distribution according to merit, based on the level of effort in support of deliv-

ering the service (see e.g. Vincent, 1998). The second proposed desert in terms of opportunity cost whereby those previously most dependent on park use be rewarded most. This statement reflects a commonly advocated approach to determining the level and distribution of payments within PES schemes (Wunder et al., 2008). The third proposed distribution should favour those most in need, the poorest – a distributive conception that is also a common way of framing equity in PES schemes (Pascual et al., 2010). The fourth option was not based on desert but on egalitarianism, proposing that all households receive the same. The ranking exercises were followed with qualitative probing of underlying reasons. Exploration of conceptions of procedural justice were initiated in a similar fashion, with statements reflecting different approaches to decision making with respect to the PES scheme: formal cell leaders, leaders following community consultation, consultation and voting, PES project team. We also attempted to look at the perceived relationship between distribution and procedure by eliciting responses to two brief versions of a story, one in which concerns about procedure dominated, and the other that was distribution-oriented. Conceptions of recognition could not be broached in quite the same way, because it is harder to identify clear and different conceptions of recognition that could serve as the basis for a ranking exercise. Instead, we began with a simple exercise in which respondents were asked to give yes/no responses to statements that we considered to be relevant recognition issues in this context, such as whether they hold different views about the park to those of the park authorities. Again, the initial exercise served as a ‘way in’ to more open discussion. These more deductive approaches to exploring pre-conceived ideas about justice were supplemented by more open, inductive approaches. First, we asked respondents to reflect on a forestrelated episode that they considered unjust. We then encouraged

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

5

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

(A)

(B)

Fig. 1. Respondent’s choice for method of payment distribution – first (A) and last (B) choices based on a selection of four statements.

discussion around these justice ‘stories’ to elicit the reasons why they considered it unjust. Second, we asked an open question about whether there was anything about the PES project that they considered unfair. Where possible, evidence from the justice survey is supported by evidence from other sources, including distributional preferences revealed through decisions about payment distributions and spending of the public component; observation of community interaction with park management authorities and local government officials; and informal follow up discussions with Rwandan cultural experts in November 2011 to explore the context in which justice conceptions have formed.

4. Results and discussion 4.1. Distributive justice Fig. 1 summarises rankings of statements about distribution. Respondents selected egalitarian (38%) as their first choice for payment distribution more often than expected by chance, followed by effort, need and opportunity costs (29%, 23% and 9%, respectively; Chi-square = 14.4, df = 3, p = 0.002). Similarly respondents’ least preferred method of payment distribution was opportunity costs (54%; Chi-square = 35.3, df = 3, p = <0.001). There were no significant differences in responses between wealth and gender classifications. The reasons specified for not liking distribution based on opportunity costs centred on concerns about perverse incentives, whereby it would be wrong to reward those who illicitly used park resources. For some, this was combined with concerns about whether such costs were genuinely borne, i.e. whether previous users had really stopped or reduced their forest use. Rejecting opportunity cost as a fair basis for payment determination seems to go against what is commonly held to be fair and efficient amongst PES theorists (Wunder, 2007; Jack et al., 2008). This reflects two particular contexts of this PES case: firstly, park use is illegal and therefore so is the ‘opportunity’ being considered; secondly, this is a community-based PES design and the monitoring of collective rather than individual performance incurs some potential for free-riding. Given that context, distribution according to opportunity cost would have the potential for the historically biggest illicit forest users to gain the biggest payments for foregone activity whilst not actually forgoing them. Prior to final design of the PES scheme, discussions with environmental managers showed considerable alignment with this community rejection of opportunity cost approaches. This was again due to the potential for a perverse and unjust outcome, with added concerns that it would send the wrong signals to locations not involved in the PES pilot.

Further, they were concerned that if community members considered this unjust, it would create dissatisfaction and conflict. Such thinking is prominent among park managers who see the management of their relationship with communities as a key task, and as a key responsibility for anybody introducing new conservation initiatives. It is clear then, that for park managers, finding alignment between community conceptions of justice and those of other conservation partners, is a significant concern. The opportunity cost approach may correspond with reasoned and seasoned economic thinking about distribution, but looked at through a justice lens, it appears unacceptable in this case. This is a key example of how we should not assume that dominant narratives of fairness will align with local ones. It contributes to emerging concerns over the appropriateness and implications of the opportunity cost approach (Gregersen et al., 2010; McAfee, 2012). Egalitarianism was the most popular first choice (38%) and also happens to be the model employed by the PES project. Respondents’ reasons for such a selection were variable. Firstly, people referred to notions of collective responsibility as well as a link between equal distribution of benefits and the potential strength of collective action: ‘if all people are supported they would put their forces together’. Others referred to notions of equality (‘all people are equal’), and that differentiating distribution by need is inappropriate due to the relative homogeneity of wealth: ‘there is little gap between rich and poor in this community’. In open questioning and respondents’ stories about injustice and unfairness, one of the strongest themes to emerge was the conception that injustice arises from not being treated equally.1 For example, being fined for collecting grasses from the forest is widely considered fair, but not where others are not fined for doing the same thing, or where the fine is higher than for others. Similarly, the unfair aspects about the PES include the fact that some, not others gain wages for particular administrative functions and not all have land assets necessary to access benefits such as tree seedling distribution. One of the more prominent themes that people connect to equality is the need to avoid conflict (‘those who receive less will be frustrated’) which we also noted as a primary concern of park managers. We will pick up on this theme later, i.e. that ‘the right thing to do’ is often

1 The term Ubutabera is the closest single Kinyarwanda word to ‘justice’ but refers specifically to the process of law and judiciary. The terms of most relevance to our interviews are Ikintu kidafite uwo kibangamiye, which we understand as ‘Something that isn’t detrimental to anyone’ and Ikintu kitabera, understood as something that does not favour particular persons or groups. We claim no linguistic expertise but note – perhaps unsurprisingly – that the language that frames articulation of justice (just as avoiding detriment, justice as relational) seems to resonate with many of the views discussed later in the paper.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

6

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

considered in terms of the avoidance of conflict and insecurity in Rwandan society. 4.2. Procedural The most preferred decision-making procedure for allocation of community funds was decisions by leaders in consultation with community members. This was selected as first choice significantly more often than expected by chance (50%; Chi-square = 36.3, df = 3, p = 6.61e 08) followed by the community by itself, then PES project staff and then elected cell leaders alone (29%, 17% and 4%, respectively). The last choice was also significant, with leaders operating alone being the least preferred procedure followed by the community, PES project staff, and leaders consulting with the community (49%, 27%, 21% and 4%, respectively; Chisquare = 32.3, df = 3, p = 4.62e 07). Neither the most nor least preferred methods differed according to wealth classification (p = 0.3 and p = 0.82, Fisher’s exact test, FET). Significantly more women than men selected leaders and the community as their least preferred method (46%; p = 0.001, FET). In addition, men selected leaders significantly more than women as their least preferred choice (63%). This difference appears to arise because there are more women who do not feel empowered to contribute to decision-making and who therefore value less the idea of community involvement. By contrast, men are often more confident in their capacity to improve upon decisions taken by leaders alone. There were no significant gender differences for the most preferred choice. Overall, only 4% of respondents placed ‘leaders’ as their first choice procedure, and 49% placed it last. One key reason was the perceived inequities in previous cases of benefit distribution, including petty corruption. Furthermore, cell level leaders are seen as subservient to their line managers and to higher level plans and therefore unresponsive to their own constituencies. Rwanda operates a form of decentralisation that involves district mayors entering into performance agreements with the president’s office. They are judged on attainment of received performance targets and these targets pass down the local government chain from district, to sector, to cell. Ultimately then, there is a system of upwards accountability in which decisions about development priorities taken by cell administrations reproduce and report against priorities that are centrally determined. There is little downward accountability and local government priorities can easily differ to those expressed by their constituencies. ‘Leaders alone take unfair decisions that give advantages to rich people. Leaders report to their higher authorities so they take decisions that will please their superiors and not necessarily the community’. In Rwanda, as elsewhere, project and policy design does not take a frictionless passage into community level practice. Field realities are inevitably mediated and transformed by local government bureaucracies with their own agendas. In some cases, the pursuit of higher agendas has actually involved a removal of PES monies that were designated for distribution to households, in order for the cell to meet its targets for school construction and, most commonly, health insurance coverage. Many households due a 9000 Rwandan Franc (RWF) payment for example, would find that 1000 RWF was taken from this to pay their Credit Mutuelle health insurance – no bad thing in many respects but an unjust act in the eyes of many who it happened to. ‘Leaders do not use our money for what it is meant to. For example, now the project thinks that we have received 9000 RWF but it’s not true. The money does not reach the community as promised’.

The most popular procedure for decision-making, involving consultation between community and leaders, was justified through reference to fairness of outcomes, legitimacy and empowerment. Thus many regard participatory procedures as a means to better decisions and as instrumental to distributional justice, as illustrated by the assertion that ‘if the community works with leaders, there is little chance of the money being misused’. But such consultation is also seen as securing legitimacy and buy-in (for example, ‘if people are involved in the decision, the decision will be followed whole-heartedly’) and, for a small minority of respondents, as inherently good due to its emancipatory potential (‘if the community does make decision, they would be empowered and that is good’). The significant differences between men’s and women’s preferences for decision-making by leaders ties in with men expressing greater distrust of leaders than women. For example, one male stated that ‘leaders may misuse the money for other things or remove some amounts of money for unknown reasons’. This contrasts with the female response that, ‘leaders work for the people and they are trusted by the people. Their decisions are more likely to be followed by the community’. Several female respondents also expressed specific concerns over community involvement (‘the community is not able to command themselves. They are like bees without their queen’ and ‘if it is up to the community, a lot of conflict can arise’) and greater preference for leaders making decisions (‘the leaders are more educated than the community’ and ‘a person of low rank [the community] cannot do things before passing a person of higher rank [a leader]’). We explored the relationship between distributive and procedural dimensions in more detail by eliciting responses to two versions of a story about a king distributing cows among villagers, one in which distributional principles are foremost, the other in which procedure is foremost. The procedure-oriented scenario involved an inclusive process of decision-making, but leading to no cow being allocated to the respondent. The distribution-oriented scenario did not involve the respondent in the decision-making process, but resulted in her receiving a cow. Significantly more people (69.2%, n = 54) preferred the distribution-oriented (noninclusive) procedure if it meant that they received the benefit (preference for end rather than means; Chi-squared: 11.54, df = 1, p-value < 0.001). In probing respondents’ preferences we learned that people were fixated on the fact that the woman in the scenario was poor and often made statements that placed considerable emphasis on this fact. For example, one respondent said, ‘‘A woman received a cow from the leaders and she needed it, that’s all that counts’’. Procedure was considered important by some respondents, but not where it was perceived to trade-off with the opportunity to be a beneficiary. In an effort to further distinguish preference for distribution versus procedural aspects we introduced a second scenario that more neutrally depicted money distribution without consultation versus possible money distribution with consultation. This scenario was done in only two of the four participating cells with respondents selecting the version emphasising procedure significantly more often than expected by chance (63%, n = 24 of 38, Chi-square = 4, df = 1, p-value = 0.0455). Many of the respondents considered that an inclusive procedure would be desirable even if they did not receive benefits as a result: ‘If things are done in transparency and if I would not be selected as a consequence, I could understand and accept that’, for example where there are others who are more poor and therefore more deserving. Others did not see their own involvement as serving their own ends alone ‘When I am involved in the process I can vote for the people that really deserve the money’. Perhaps most interestingly, the case for participation once again employed references to conflict and insecurity, with several respondents highlighting

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

the potential insecurities arising from procedures that are not public and visible: ‘when the process was transparent2 it is more secure. When you receive something without knowing the process, later you can have problems. For example, I know a man in another cell who unduly received a cow from the government’s Girinka program. Afterwards the Executive Secretary started asking him for bribes and they ended up in prison.’ Just decision-making is here related to personal security and contrasted with procedures that put people at risk of personal harm. In Rwanda, the specific risk of imprisonment is a common concern due to the very high imprisonment rate since the 1994 genocide (Tersakian, 2011; Walmsley, 2011). Furthermore, the use of the local Gacaca court system involves local identification of those accused as genocidaires, and a prevailing fear that any rumour of wrongdoing can have severe consequences. Where preference for procedure is expressed over distribution, this is often therefore rooted in practical concerns about consequences and in this respect, it is hard to disentangle the justice conceptions that we (outsiders) are looking for, from pragmatic day to day attention to household security. 4.3. Recognition We now consider whether a focus on distributional outcomes, which is common in conservation revenue sharing and PES schemes, serves to crowd out concerns about the recognition dimension of justice. In particular, we were interested in whether local people felt coerced into abandoning locally significant cultural practices or into assimilating to more dominant ways of thinking about environmental management. We opened discussion of recognition through a series of statements to which respondents were initially asked to agree or disagree with. Some of these statements were intended to establish the importance of recognition to participant communities, whereas others were seeking to establish whether the payment scheme was seen to contribute to recognition injustice. Perhaps surprisingly, the survey results found that lack of recognition is not expressed as an injustice by local people. This is not the same as finding that failures of recognition don’t exist and it is important to try to understand why this justice dimension is not articulated. Part of the explanation may be that rural Rwandans exercise considerable discretion in responding to questions, even with researchers who have been working with them for 2 years or more (as we had). As has been said, choices are influenced by a desire to pursue what is more secure, or least risky, and in doing so it is sometimes safest to assume that researchers are inseparable from authority. Thus, when asked whether ‘Our views about access to the park are different from those of RDB’, it is perhaps not surprising that 60% disagreed. Men were more likely to agree than women, 29.3% and 18.9%, respectively, and the rich more likely to agree than the poor, 34.3% and 16.3%. In other words, the relatively more powerful (male, wealthy) were more likely to admit to views that differ to those of authority. For the minority who say that community views do differ to those of authority, the main argument put forward is that they know that some people still want to use the forest and that it is only the fear of punishment that stops them. For the majority who say that community views are the same as those of the park authority, a wider range of arguments are put forward. First, they refer to the benefits that they get from a protected forest: ‘peoples 2

We use transparent to translate the Kinyarwanda phrase ‘‘Ikintu gikorewe mu mucyo’’ which translates more literally as ‘‘something that is done in the light’’.

7

views are the same as RDBs because they know the park’s benefits’. Second, they cite the increased knowledge about conservation: ‘now people know what conservation is and they agree with RDB, except some thieves’. Third, they cite the availability of substitute resources: ‘because the things we used to get from the forest we can now get outside’. Fourth they refer to a long process of alignment with park authorities: ‘even before RDB started a strong law people had already stopped using the forest’. Fifth, they refer to legal instruments: quite simply ‘it is the law’. The significance of rule abeyance will be discussed below, but for now we note that a rule established by authority tends to be considered a shared view (at least, in responses given to researchers). Respondents predominantly refused to agree or disagree with the statement that ‘RDB needs to recognise our views of the park’ (73%, N = 78) suggesting the controversial nature of such a statement. Of the 22 respondents that did answer, the majority (16% or 73%) strongly disputed the statement. For example it was stated that ‘Recognising our views would allow some people to destroy the forest’, that ‘it is impossible! What we are doing in the park is not good – we are harming the forest’ and that ‘peoples’ views should not be considered because they are not right’. A large majority (77%) agreed that ‘The rules of the park prevent us from passing on our traditions to our children’, with little difference observed across wealth or gender. However, in nearly all cases those who agree that traditions are being lost consider this positively, not negatively in terms of recognition failure. Respondents employed terms that represent positive aspects of modernity for them, such as ‘development’, ‘new technology’, ‘improvement’, ‘moving forward’, ‘a modern life’, ‘plastic’, ‘modern hospitals and medicines’. The lack of expressed cultural attachment to forest-dependent activities and traditional ways was very striking and contrasts with experiences reported elsewhere (Chapin, 2004). We sought additional help with understanding this finding, through informal interviews with three persons identified for their knowledge of Rwandan culture – a university professor, a senator in parliament, and an NGO director.3 The main suggested explanation for the lack of cultural attachment to forests is the long history of enforced change, initially through colonialism and its subservient evangelical missions, and then through political regimes with embedded discourses of the benefits of modernising change. Under colonial rule, traditional institutions became associated with repression due to Belgian rule through existing power structures of kings and princes who asserted their sovereignty through control of cows and land. More recently, traditional relationships of clan and ethnicity became associated with war and genocide. The history of violence therefore permeates the political discourse of modernisation and the rejection of traditional ways. This history of violence is also considered to be linked to what appears to outside observers as extreme abeyance. Social fragmentation along traditional fault-lines has been linked to challenges to authority and to disastrous outcomes. This is a factor in the authoritarian nature of post-colonial rulers, justifying hard emphasis on homogeneity and on obedience. In short, there is a very powerful political discourse linking the success of the Rwandan state to cultural unity, in which the traditional is fragmented and the modern is unified. A result is perhaps a sense of identity that is still imbued with the importance of cows, land and now money (Ingelaere, 2010), but which is progressively divorced from traditional social relations and material artefacts (de Lame, 2005). Whilst political modernisation helps to understand the loss of traditional cultural attachments to forest landscapes, the particular histories of exclusion from those forests is also important. In Nyungwe, forms 3 We are also grateful to an anonymous reviewer for questioning our more political explanation and highlighting one based more on the longevity and location of the protected areas.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

8

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

of exclusion had progressed since the 1930s, such that genuine community control does not exist in living memory. Respondents’ offer a number of more concrete reasons for why they don’t feel loss for their traditional ways of using forest resources, and why they don’t see this as a failure to recognise their traditions. Firstly, they express the essential materiality of natural resources through a willingness to accept substitutes as equally good or better. Of 60 respondents who offered reasons for accepting the loss of tradition, 31 explicitly referred to substitutes. For example, losing the ability to make traditional mortars was acceptable because ‘it made people start thinking of planting their own native trees’, losing access to herbal medicines was acceptable because ‘now we have modern hospitals’, and not being able to eat bushmeat was ok because ‘[we can] raise our own livestock’. Secondly, respondents pointed out that hunting and gathering activities are not missed because they are not really economically viable, e.g. the returns from hunting are meagre set against the effort, and young people don’t choose to take up weaving when they can work for money. Thirdly, a subset of responses (11 of 60) draws on the kind of modernisation narratives that have formed part of Rwandan political discourse around modernity: ‘we cannot remain living in old ways’, ‘traditions like hunting. . . don’t fit in our country’s vision’. Fourthly, in terms of cultural significance, agricultural land trumps forest in Rwanda. Set against the small returns from hunting and gathering, agriculture is the mainstay of livelihood security with local people increasingly linking agricultural security to forest regulatory ecosystem services, namely local rainfall. This is likely in part because such narratives are deployed by state and NGO conservation agencies, but also because they are supported by personal observations. For example, some have visited other parts of Rwanda without forests and see that such areas are more arid, others say that they see the rain clouds coming from the direction of the forest. Whilst forest extractive activities may command some residual cultural value, it seems that most respondents identify more strongly with a recognition of agrarian identity and livelihood. Thus it is perhaps not surprising to hear responses such as ‘mining and snaring can’t be passed to the young generations...because if the forest is destroyed there would be no rain here’. Land is, as it were, the line that should not be crossed, and those who actually lost land to the park, remain vocally aggrieved: to have land taken away is to be dominated in a way that is unjust, whereas to have hunting and gathering access denied is generally not. This discussion suggests that in this particular context, ‘recognition’ has become constructed in terms of agrarian livelihoods and related identities, whereas more spiritual and traditional aspects of recognition that are often observed in Latin American studies of indigenous people (Whiteman, 2009; Escobar, 1999) have been eroded over a century or more for very particular political and

geographical reasons. This demonstrates the importance of empirical case work for understanding meanings of environmental justice. It is also important for what it tells us about the relationship between local conceptions of justice and those encountered in more globalised views of justice embedded in designs for PES and REDD+. Rwanda might in fact be relatively more receptive to PES approaches than many other places, owing to its alignment with environment as a predominantly material and distributive domain. Of course, there remain a small minority who for economic or other reasons continue to hunt, mine, fell trees to remove honey and so on. Given that the PES project provides incentives to the community as a whole to reduce forest use, it seems likely that this minority will find their activities increasingly suppressed by the majority. Indeed the structure of the incentive system explicitly requires supporters of the PES to monitor those who are non-compliant. Seventy-two percent of respondents agreed with the statement that ‘Because of financial rewards for not using the forest, poachers no longer have any respect in this cell’. Seventy-one percent agreed that ‘Those most dependent on the forest have less respect from the community than before the project began’. The denigration of a group of forest users is a common issue for community-based conservation. Whilst rarely stated in such terms, an underlying intent is to construct collective claims about environmental justice that criminalise certain social groups in the eyes of others in the community (Martin and Lemon, 2001). Some forthright views were expressed about this, such as ‘poachers are saboteurs and thieves’, that hunting is ‘a shameful act’, and even that ‘I wish it was up to us to punish poachers, we’d give them a heavier punishment than those few months that they spend in prison’. Of those who felt there was a loss of respect for illicit park users, about half (27 of 56) indicated that the PES had contributed to this: ‘before [the PES] people thought that poachers don’t have a choice but now because of the rewards they regard them as thieves’. ‘before using the forest was respected because someone who had been mining in the park would buy you a beer and you would see mining in the park as normal but now because of the payments they are seen as saboteurs . . .and desperate because they run the risk of being shot in the park by rangers’. On a more positive note, some respondents claimed that the PES has helped people to give up hunting, thereby gaining them more respect: ‘Before the project poor people used to depend on the forest but now they are being supported these same individuals have stopped using the forest and have more respect’

Fig. 2. Respondent’s choice for method of procedure for distribution of the communal payment – first (A) and last (B) choices based on a selection of four statements.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

If PES is to be used more widely as a conservation tool in support of protected areas, this line between PES as undermining respect for more forest dependent people, and PES as emancipatory, will require further consideration and may be one direction for a more just conservation. But this line between coercion and opportunity will be interpreted differently by different actors, dependent up personal situation and justice conceptions. The reality is that the transition towards viewing illicit park use as wrong has been there for much longer than our PES which is the latest in a number of conservation interventions aimed to steer communities towards that way of thinking. ‘people that used the forest before the time of Habyarimana were well respected but now that has greatly changed and these individuals are much disrespected. . . even more so because of the project. ‘lack of respect is not because of the project. It has been that way for a long time.’ We remain uncertain of the detailed history of this transformation, but we are sure that in the past there was considerable respect for hunters and this has now changed dramatically. This long-run receptivity to conservation ideas needs to be considered in the context of wider changes to Rwanda’s economy, in which labour can increasingly be allocated to higher return activities than hunting and gathering, and in the context of cultural transition towards the ‘modern’ that continues to link identity to land more than forests. We also assume that this transformation is related to exposure to norms of environmental justice from different stakeholders at different scales, through international donors and NGOs whose justice norms are informed by ‘global’ referents such as biodiversity, ecosystem services, opportunity costs, informed consent and future generations. 5. Conclusions Exploring local perceptions of a PES scheme through an environmental justice lens has yielded some intriguing results, including some tensions between norms that are dominant locally, and those that communities encounter through the ‘global’, in the form of globalised ethics around biodiversity conservation and mainstream economic thinking about fair (and efficient) alignment of costs and benefits. Our findings about distributional justice can be boiled down to three main observations. First, that the majority of respondents reject the dominant neoclassical economics approach to determining the distribution of payments for this PES scheme. The rejection of ‘opportunity costs’ was partly due to concerns about perverse incentives – recent history has seen a change in status of forest users such as hunters, from respected community members to people whose actions are not just, and who do not deserve compensation for ceasing them. The all-important context for this is Nyungwe’s legal designation as a National Park, and repeated reinforcement of the illegality of park use from park authorities and NGOs. This is a potentially significant finding if one follows the following key strand of PES thinking. First, Ferraro (2001) and others have presented PES as an efficient alternative to largely unsuccessful Integrated Conservation and Development Projects around parks; second, Wunder (2012) and other economists who work on PES design are insistent that efficiency requires payments to be linked to opportunity costs; thirdly, we find that this might not be appropriate around protected areas. We might conclude either that PES is least likely to work in those very locations where first it was advocated, or that we need to rethink the orthodoxy that opportunity costs is always the best way to determine PES payment distribution.

9

Second, we found egalitarian distribution to be the most popular. We understand this partly in terms of perceived economic homogeneity (‘we are all poor’) and partly in terms of the aforementioned disrespect for illicit forest users and associated rejection of weighting payments to offset opportunity costs. However, we also have to understand this within the Rwandan political context, which we have characterised as promoting cultural homogeneity and that appears to contribute to an ethic of treating people the same: for many, injustice was articulated as relational, in terms of treating people differently. Thirdly, in practice, principles of distribution are quite varied. On the one hand, centrally controlled resources are often allocated according to principles of need, for example providing livestock to the poorest, or construction materials to those with the greatest need. On the other hand, three of the four cells participating in the PES decided at public meetings to only allocate a minority of funds (10–30%) to such public spending and to distribute the remainder equally among all households. Findings about procedural justice can also be boiled down to three issues. Firstly, the least just procedure is perceived to be allowing local (cell level) government officials to take unilateral decisions. In some cases this is based on assertions of corruption, in others it is because local government officials are seen as puppets for higher level government, making decisions based on understanding of the needs of sector and district governments rather than the needs of citizens. From a project point of view there is little choice but to work with local government, and indeed many advantages to doing so (Gross-Camp et al., 2012), but clearly this does not constitute a straightforward ‘community-based’ governance arrangement. Secondly, most respondents wanted more inclusive procedures and transparency. Transparency in this translation from Kinyarwanda refers to decisions that are in the public domain (‘in the light’), as opposed to decisions taken in private (‘the dark’). The reasons proposed for this confirm some wellversed normative and instrumental benefits of participation (Funtowicz and Ravetz, 1993; Horlick-Jones and Sime, 2004). But we also find a particularly strong emphasis on public deliberation and decision-making protecting against corruption, accusation, and trouble with authorities. More generally we find that conceptions of justice are strongly aligned with reducing risk of harm, notably punishment by authorities. Thirdly, our attempt to analyse procedure and distribution as distinct may not have connected with how respondents conceive decisions about resource distribution. Our findings suggest that the value of inclusion and transparency have limits when traded off against expected benefits. Not surprisingly, what people really want is win–win scenarios wherein participatory decisions deliver distribution that is favourable to them. So whilst 79% of respondents’ first preferences were for community decision-making or community consultation (Fig. 2), a majority would also forsake this to receive a benefit they would not otherwise get. Again, one of the key issues arising is the links to local government structures, not only due to their priorities for resource allocation but also due to decision-making procedures that are not liked. Our expectation when setting out on this research was that we would find evidence of ways in which a focus on distributional issues (an import of globalising environmental justice norms – Martin et al., 2013) was in conflict with locally important norms of cultural recognition. We have not observed this, perhaps partly due to methodological difficulties of researching this dimension, but certainly not exclusively so: we have worked with enough local people in Rwanda over the last 10 years to be confident that there is no strong spiritual attachment to forests that would reveal itself to more ethnographic methods. In trying to understand why justice claims are not overtly linked to cultural attachments, we considered the particular context of Rwanda, including the colonial and post-colonial histories of subjugation, political discourses of

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

10

A. Martin et al. / Geoforum xxx (2013) xxx–xxx

modernity and cultural unity, and the affirmation of agrarian (cow and land) ways of life over forest based ones. The key conclusion here is not so much about the details of this specific case, but that context matters, and that empirical casework on local and global conceptions of justice are important. The Rwandan case has global dimensions to be sure, including narratives about forest conservation and ecosystem services, but also unique characteristics that would ideally feature in planning for interventions such as PES and REDD+. We finish with two general conclusions. First and foremost, we confirm that it is naïve to think that the global justice referents embedded in PES, REDD+ and other global responses will always be aligned with those of local people. There is considerable variation in views among local people and we also find considerable variance from very common ‘global’ distribution principles. The implications of this are really quite profound, especially if Whiteman (2009) is right in arguing that many if not most conservation conflicts with indigenous people arise over differences in perceptions of justice. Given that it is not practicable for conservation actors to always understand local justice conceptions in advance, the pragmatic implication is that interventions need to be adaptively managed, with the sensitivity to learn and the flexibility to respond. Secondly, we think this study justifies the utility of a cross-scale environmental justice framing by providing a new analytical approach for understanding local receptivity to forest conservation interventions, and in particular providing a more nuanced way of interrogating the relationship between equity and effectiveness. We clearly have much to learn about the limitations of our own framing and methods, including our inevitable starting point in logics of justice that we are more familiar with. This includes the analytical cut we employ to distinguish three dimensions of justice, and to separate out conceptions of justice from, say, practical views about what actions are safest given the institutional context. Acknowledgements We gratefully acknowledge funding from the European Research Council through the project ‘Reconciling Biodiversity and Development through Direct Payments for Conservation’ (ReDirect). We are also thankful to the Rwanda Development Board – Tourism and Conservation, especially those at the Nyungwe National Park who made this research possible. Students and graduates of the National University of Rwanda biology department – Erik Ndayishimiye, Laure Rurangwa and Fabrice Ndayisenga, provided invaluable support with fieldwork and plenty of wise advice. Nozomi Kawarazuka provided thoughtful assistance with data analysis. The authors’ claim full responsibility for the interpretation of data. References Almas, I., Cappelen, A., Sorensen, E., Tungodden, B., 2010. Fairness and the development of inequality acceptance. Science 328, 1176–1178. Ansoms, A., McKay, A., 2010. A quantitative analysis of poverty and livelihood profiles: the case of rural Rwanda. Food Policy 35, 584–598. Bentham, J., 2009. An Introduction to the Principles of Morals and Legislation. Clarendon Press, Oxford. Bohman, J., 2007. Beyond distributive justice and struggles for recognition: freedom, democracy and critical theory. European Journal of Political Theory 6, 267–276. Bullard, R., 1990. Dumping in Dixie: Race, Class and Environmental Quality Boulder. Westview Press, San Francisco, Oxford. Bullard, R. (Ed.), 1993. Confronting Environmental Racism: Voices from the Grassroots. South End Press, Boston, MA. Bunyan, B., Mohai, P. (Eds.), 1992. Race and the Incidence of Environmental Hazards. Westview Press, Boulder, CO. Carmin, J., Agyeman, J., 2011. Environmental Inequalities Beyond Borders: Local Perspectives on Global Injustices. MIT Press.

Castree, N., 2004. Differential geographies: place, indigenous rights and ‘local’ resources. Political Geography 23, 133–167. Chapin, M., 2004. A challenge to conservationists. World Watch Magazine (November/December), 17–31. Crocker, D., 2008. Ethics of Global Development: Agency, Capability and Deliberative Democracy. Cambridge University Press, Cambridge. de Lame, D., 2005. A Hill Among a Thousand. Transformations and Ruptures in Rural Rwanda. The University of Wisconsin Press, Madison/Tervuren. Dobson, A., 2007. Social justice and environmental sustainability: ne’er the twain shall meet? In: Agyeman, J., Bullard, R., Evans, B. (Eds.), Just Sustainabilities: Development in an Unequal World. MIT Press, Cambridge, MA, pp. 83–95. Dryzek, J., 2000. Deliberative Democracy and Beyond: Liberals, Critics, Contestations. Oxford University Press, Oxford. Escobar, A., 1998. Whose knowledge, whose nature? Biodiversity conservation and the political ecology of social movements. Journal of Political Ecology 5, 53– 82. Fehr, E., Falk, A., 2002. Psychological foundations of incentives. European Economic Review 46, 687–724. Ferraro, P., 2001. Global habitat protection: limitations of development interventions and a role for conservation performance payments. Conservation Biology 15, 1990–2000. Folger, R., 1989. Effects of procedural and distributional justice on reactions to pay raise decisions. Academy of Management Journal 32, 115–130. Folger, R., Konovsky, M., 1989. Effects of procedural and distributive justice on reactions to pay raise decisions. Academy of Management Journal 32, 115– 130. Fraser, N., 2001. Recognition without ethics? Theory, Culture and Society 18, 21–42. Fraser, N., 2009. Scales of Justice: Reimagining Political Space in a Globalizing World. Columbia University Press, New York. Funtowicz, S., Ravetz, J., 1993. Science for the post-normal age. Futures (September), 739–755. Government of Rwanda, 2007. Economic Development and Poverty Reduction Strategy 2008–2012, Preliminary Draft June 2007. Ministry of Finance and Economic Planning, Kigali. Gregersen, H., El Lakany, H., Karsenty, A., White, A., 2010. Does the opportunity cost approach indicate the real cost of REDD+. Rights and Resources Initiative, Washington, DC. Gross-Camp, N., Martin, A., McGuire, S., Kebede, B., Munyarukaza, J., 2012. Payments for ecosystem services in an African protected area: exploring issues of legitimacy, fairness, equity and effectiveness. Oryx 46, 24–33. Holifield, R., Porter, M., Walker, G. (Eds.), 2009. Spaces of Environmental Justice: Frameworks for Critical Engagement. Special Issue of Antipode 41(4). Horlick-Jones, T., Sime, J., 2004. Living on the border: knowledge, risk and transdisciplinarity. Futures 36, 441–456. Ingelaere, B., 2010. Do we understand life after Genocide? Center and periphery in the construction of knowledge in postgenocide Rwanda. African Studies Review 53, 41–59. Jack, K., Kousky, C., Sims, K., 2008. Designing payments for ecosystem services: lessons from previous experience with incentive-based mechanisms. Proceedings of the National Academy of Sciences 105, 9465–9470. Kant, I., 1998. Groundwork of the Metaphysics of Moral. Cambridge University Press, Cambridge. Lejano, R., Piazzo, B., Houston, D., 2002. Rationality as social justice and the spatialdistributional analysis of risk. Environment and Planning C: Government and Policy 20, 871–888. Marion Young, I., 2011. Responsibility for Justice. Oxford University Press, Oxford. Martin, A., Lemon, M., 2001. Challenges for participatory institutions: the case of village forest committees in Karnataka, South India. Society and Natural Resources 14, 585–597. Martin, A., Rutagarama, E., 2012. Just deliberation: can communicative rationality support socially just environmental conservation in rural Africa? Journal of Rural Studies 28, 189–198. Martin, A., 2013. Global Environmental Justice, in Practice: Introduction, Geographical Journal, In Press. Martin, A., McGuire, S., Sullivan, S., 2013. Global environmental justice and biodiversity conservation, Geographical Journal, In Press. Masozera, M., Alavalpati, J., 2004. Forest dependency and its implications for protected areas management: a case study from the Nyungwe Forest Reserve, Rwanda. Scandinavian Journal of Forest Research 19, 1–8. McAfee, K., 1999. Selling nature to save it? Biodiversity and the rise of green developmentalism. Environment and Planning D: Society and Space 17, 133– 154. McAfee, K., 2012. The contradictory logic of global ecosystem services markets. Development and Change 43, 105–131. Montada, L., 2002. The justice Motive in Everyday Life. Cambridge University Press, Cambridge. Moorman, R., 1991. Relationship between organisational justice and organisational citizenship behaviours: do fairness perceptions influence employee citizenships? Journal of Applied Psychology 76, 845–855. Outhwaite, W., 2009. Habermas: A Critical Introduction, second ed. Stanford University Press. Pascual, U., Muradian, R., Rodríguez, L., Duraiappah, A., 2010. Exploring the links between equity and efficiency in payments for environmental services: a conceptual approach. Ecological Economics 69, 1237–1244. Peet, R., Robbins, P., Watts, M. (Eds.), 2010. Global Political Ecology. Routledge, London and New York.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006

A. Martin et al. / Geoforum xxx (2013) xxx–xxx Plato, 1974. The Republic, second ed. Penguin Books, Harmondsworth. Pulido, L., 2000. Rethinking environmental racism: white privilege and urban development in Southern California. Annals of the Association of American Geographers 90, 12–40. Rawls, J., 1971. A Theory of Justice. Harvard University Press, Cambridge, MA. Robinson, D., 2010. Locating biopiracy: geographically and culturally situated knowledges. Environment and Planning A 42, 38–56. Sandel, M., 1998. Liberalism and the Limits of Justice, second ed. Cambridge University Press, Cambridge. Schlosberg, D., 2004. Reconceiving environmental justice: global movements and political theories. Environmental Politics 13, 517–540. Schlosberg, D., 2007. Defining Environmental Justice: Theories, Movements and Nature. Oxford University Press, Oxford. Schroeder, R., St. Martin, K., Wilson, B., Sen, D., 2008. Third world environmental justice. Special Issue of Society and Natural Resources 21 (7). Schroeder, R., 2008. Environmental Justice and the Market: the politics of sharing wildlife revenues in Tanzania. Society and Natural Resources 21, 583– 596. Sen, A., 2009. The Idea of Justice. Penguin Books, Harmondsworth. Sullivan, S., 2009. Green capitalism, and the cultural poverty of constructing nature as service provider. Radical Anthropology 3, 18–27. Tersakian, C., 2011. ‘‘All Rwandans are afraid of being arrested one day’’: prisoners past, present and future. In: Straus, S., Waldorf, L. (Eds.), Remaking Rwanda: State Building and Human Rights After Mass Violence. University of Wisconsin Press, Madison, pp. 210–220.

11

Vedder, A., Hall, J., Harcourt, A., Monfort, A., Wilson, R., 1992. Burundi and Rwanda. In: Sayer, A., Harcourt, C., Collins, N. (Eds.), The Conservation Atlas of Tropical Forests: Africa. IUCN and Simon and Schuster, New York, pp. 102–109. Vermeylen, S., Walker, G., 2011. Environmental Justice, Values and Biological Diversity: the San and the Hoodia Benefit Sharing Agreement. In: Environmental inequalities beyond borders: local perspectives on global injustices. MIT Press, Cambridge, Mass., 105–128. Vincent, A., 1998. Is environmental justice a misnomer? In: Boucher, D., Kelly, P. (Eds.), Social Justice: From Hume to Walzer. Routledge, London, pp. 120–140. Von Benda Beckmann, F., 1989. Scape-goat and magic charm: law in development theory and practice. Journal of Legal Pluralism 28, 129–148. Walker, G., Bulkeley, H. (Eds.), 2006. Geographies of environmental justice. Special issue of Geoforum 37. Walker, G., 2012. Environmental Justice: concepts, evidence and politics. Abingdon, Routledge. Walmsley, R., 2011. World Prison Population List, ninth ed. University of Essex: International Centre for Prison Studies. Whiteman, G., 2009. All my relations: understanding perceptions of justice between companies and indigenous peoples. Organisation Studies 30, 101–120. Wunder, S., 2007. The efficiency of payments for environmental services in tropical conservation. Conservation Biology 21, 48–58. Wunder, S., 2012. Of PES and other animals. Oryx 46, 1–2. Wunder, S., Engel, E., Pagiola, S., 2008. Taking stock A comparative analysis of payments for environmental services programs in developed and developing countries. Ecological Economics 65, 834–852.

Please cite this article in press as: Martin, A., et al. Whose environmental justice? Exploring local and global perspectives in a payments for ecosystem services scheme in Rwanda. Geoforum (2013), http://dx.doi.org/10.1016/j.geoforum.2013.02.006