Contemporary Discourses of Green Political Economy: A Q Method Analysis

ABSTRACT For over two decades, the concept of sustainable development has been salient in political discourse. But its promise of reconciling economic development, social welfare, and environmental sustainability has proven rather elusive. In recent years, we've seen numerous competing concepts emerge in debates about sustainable economic development. While many advance ideas of a green economy and green growth, others talk about wellbeing, gross national happiness, inclusive wealth, harmony with nature, de-growth, steady-state economy, and buenvivir (living well). This rhetorical diversity shows that there is no single vision for reconciling environmental sustainability and economic development. But the varied terminology itself obscures actual points of agreement and disagreement. This article reports on a bilingual ‘Q study’ of international debates about sustainable economic development. It reveals that three discourses underpin these debates: Radical Transformationism; Cooperative Reformism; and Statist Progressivism. The article dissects these discourses and contextualizes their key points of contention in wider sustainability debates over the past two decades.


Introduction
The concept of 'sustainable development' has had a long career. Introduced by the Brundtland Commission in 1987, the term was quickly embraced in the global North and South by policy-makers and civil society alike. It has established a firm footing in national and international policy agendas, international negotiations, and political and popular discourse. Promising to reconcile environmental sustainability, social welfare, and economic development, the allure of this concept is easily understood. Yet, throughout its 30-year career, this promise has on the whole proven elusive. Certainly there have been development achievements to celebrate (e.g. since 2000 the proportion of people living in extreme poverty has halved, as has the proportion of people without access to safe drinking water). But, the global environment is hardly in a healthy state. Scientists now warn that four of nine planetary boundaries have been crossed as a result of human activity (Steffen et al., 2015). 'Sustainability' may well mean different things to different people, but the enduring capacity of the planet to support human life must be at the core of any sustainability discourse.
A burgeoning body of scholarship aims to identify and problematize the principles, values, and ideologies that underpin contemporary unsustainable economic systems. By bringing economic policy into the realm of political analysis, the field of 'green political economy' rejects the myth of economics as an objective and value-free science (Barry, 2012, p. 122-123). This scholarship instead reveals the diversity of options and pathways for pursuing social and economic development, and privileges those that would do so in an environmentally sustainable fashion. Much of this literature questions the feasibility and desirability of exponential economic growth and explores alternatives that allow for human development within planetary limits (e.g. Barry, 2012;Blauwhof, 2012;D'Alisa, Demaria, & Kallis, 2014;Demailly, Chancel, Waisman, & Guivarch 2013;Ferguson, 2015;Hayden, 2014;Tanuro, 2013;Vanhulst & Beling, 2014). But the question of how to pursue economic development under conditions of continuing environmental degradation is not a purely academic one. Rather it has been widely debated in political and public spheres, especially in the lead up to (and aftermath of) the Rio+20 Summit in 2012.
Perhaps owing to the evident gap between the promise and reality of sustainable development, debates about sustainable economic development are increasingly conducted using a wider range of language and concepts. This is the case in scholarly, political, and public spheres. There are those who talk of green economy and green growth, and others who promote wellbeing, gross national happiness, inclusive wealth, harmony with nature, de-growth, steady-state economy, and buenvivir (living well). There is good reason to question whether this new (and revived) rhetoric represents anything substantively different to the ideas advanced as 'sustainable development', or whether it is simply 'old wine in new bottles'. The aim of the study presented here is to determine the substance underlying these diverse rhetorical labels, and the extent to which they represent genuinely new ideas about how to reconcile economic development and environmental sustainability. 1 The animating assumption of the study was that beneath these labels lies a set of discourses about how the international community should pursue economic development under conditions of continuing environmental degradation. I refer to these as discourses of green political economy. 2 Identifying the contours and coordinates of contemporary discourses of green political economy allows scholars to identify how stakeholder debates overlap with scholarship on green political economy. Strengthening connections between scholarly and stakeholder debates could be mutually productive in terms of connecting political analysis with democratic spaces. Nuanced understandings of public and stakeholder debates are especially important for the many scholar-activists that populate the field of green political economy and aim to translate their scholarship into social and political change. 3 Recognizing points of agreement and disagreement in debates about sustainable economic development is also important for global environmental governance. The idea of 'sustainable development' gained considerable traction throughout the 1990s and 2000s largely due to the fact that it was sufficiently vague to accommodate a range of distinct and incompatible interpretations. But given that many indicators of environmental quality have continued to decline despite two decades of sustainable development policies, there is value in confronting tensions and inconsistencies within this idea. It is beyond the scope of this article to elaborate on a model of inclusive global environmental governance that draws in representatives of all known environmentalist discourses. But the analysis documented here can support precisely that endeavour. 4 The first step in including all relevant discourses in debate and decision-making is identifying what those discourses are. Human beings are reflexive creatures with a capacity to revise worldviews on the basis of changing circumstances and new information. It is therefore important to periodically review the discourses we assume to exist on any issue of political importance.
This article reports on a bilingual 'Q study' of green political economy discourses, and is organized as follows. In the following section (Method), I introduce 'Q methodology', and document how this was used to identify the discourses underlying debates about sustainable economic development, specifically in the period surrounding the Rio+20 Summit. The study revealed three discourses (Radical Transformationism; Cooperative Reformism; and Statist Progressivism), which are detailed in the Results section. The Discussion then considers the relevance of these findings, and reflects on how the discursive terrain of green political economy has altered in recent years. I conclude with a summary of my findings and a reflection on the value of Q methodology for environmental policy analysis.

Method
By mapping debates about sustainability in discursive terms, I am following an established tradition in environmental policy and politics (e.g. Adger, Benjaminsen, Brown, & Svarstad, 2001;Barry & Proops, 1999;Connelly, 2007;Dryzek, 1997;Hajer, 1995;McManus, 1996). These studies have been influenced by diverse conceptualizations of discourse but perhaps the most widely cited definitions are those of Maarten Hajer and John Dryzek.
The definition one chooses has implications for how one conducts discourse analysis as these definitions focus the analyst's attention in slightly different ways. Hajer defines discourses as 'specific ensembles of ideas, concepts and categorization that are produced, reproduced and transformed in a particular set of practices and through which meaning is given to physical and social realities ' (1995, p. 44). Analysis informed by this definition focuses on the social practices through which discourses are produced. Dryzek defines discourse as a shared way of apprehending the world. Embedded in language, it enables those who subscribe to it to interpret bits of information and put them into coherent stories or accounts … . Each discourse rests on assumptions, judgements, and contentions that provide the basic terms for analysis, debates, agreements, and disagreements. (2013, pp. 9-10) This definition directs attention not to practices but rather to identifying how assumptions, values, and ideas fit together into coherent discourses that can coordinate action.
This study is informed by Dryzek's definition of discourse, which is compatible with the discourse analysis method of 'Q'. Developed by William Stephenson in the 1930s, Q methodology combines quantitative and qualitative techniques to access personal experiences, preferences, and beliefs. In short, Q is a method for discerning a set of discourses within the 'universe of subjective communicability surrounding any topic' (Brown & Good, 2013, p. 1149. It is designed to discover the finite range of viewpoints (or discourses) within the vast amount of communication about a particular issue (referred to as concourse); in this case, the issue of how to approach economic development under conditions of continuing environmental degradation. There is no one single way in which concepts like sustainable development, green economy, and green growth are understood; in fact, one of the key stumbling blocks during the Rio+20 negotiations was over the precise meaning of 'green economy'. There may also be numerous overlaps between understandings of existing and alternative economic paradigms. Q methodology allows the identification of such distinctions and overlaps, and thereby enables the identification of discrete perspectives or discourses.
A Q study comprises six steps: (1) define the research question; (2) design the Q-set; (3) select participants; (4) administer the Q sort; (5) conduct a factor analysis using specialized Q software; and (6) interpret the qualitative meaning of the factor structure. Each of these steps is explained below.
Research question: The first step of the study was simply to define the research question as 'what are the different perspectives on how we should pursue economic development under conditions of continuing environmental degradation?' Q-set design: The Q-set is a set of statements presented to participants that are sorted into a matrix to reflect their view on a given issue. These statements should as closely as possible represent the totality of communication on this issue (referred to as the concourse). Researchers typically interview participants to compile these statements. Interviews were not used in this study to identify the concourse due to the fact that a considerable amount of published material was already available (in the form of online reports, blogs, and statements). To minimize the requested time commitment of my target participants, it was considered preferable to draw on this existing written material for this stage of the study. The most important aspect of compiling statements is to ensure that they reflect the tone and substance of public or stakeholder communication, rather than the voice or perspective of the researcher. Rigorous and extensive data collection and sampling processes were conducted to maximize the validity of the Q-set as reflective of wider stakeholder debates about sustainable economic development. The data collection strategy was aimed at identifying material published in the two years bracketing the Rio+20 Summit (20 June 2011to 20 June 2013, in English and Spanish. 5 This was a period during which sustainable economic development was widely debated (often in the context of anticipating or reflecting on the summit). Material was first sourced from the websites of relevant events and inquiries (the UN High Level Panel on the Post-2015 Development Agenda; the World Happiness Report 2013; the Interactive Dialogue of the General Assembly on Harmony with Nature, 2013); existing compilations of material were then consulted (UN-DESA's guide to recent publications on the green economy, green growth, and low-carbon development (UN-DESA, 2012); and the 'Why Green Economy?' database (Kenner n.d.); and finally a series of internet searches was conducted. 6 Relevant documents accumulated in the lead-up to the study complemented these purposively compiled documents, resulting in a total of 451 documents. I considered this data to be an adequate representation of communication about how to pursue sustainable economic development under conditions of continuing environmental degradation.
Distilling the substance and tone of this communication into a Q-set of 40-80 statements required close content analysis, which (for practical purposes) was precluded by the size of the data. 7 To produce a sample for content analysis and statement extraction, NVivo was used to classify all the documents. A classification system of four attributes was used (author region; author type; key concept; and relevance). 8 A sample of 147 documents was then selected primarily on attributes of relevance and key concept, and secondarily on attributes of author type and author region. 9 A close reading of these documents was then carried out to highlight statements that were representative of the sentiment and message of each. The resulting list of 270 statements was then categorized, and reduced to 48 statements through several rounds of synthesizing similar statements. 10 Participants: Non-random sampling techniques are used to select participants (the 'P-set') in Q methodology. Opportunity sampling and random sampling are generally inappropriate because most Q investigators are interested in determining the range of viewpoints of a specific community of people. Q methodology is often identified as an inversion of R methodology, and this has important implications for participant selection. Unlike in R methodologies, whereby the participants constitute the sample and the attributes constitute the variables, in Q methodology the statements constitute the sample and each participant becomes a variable. Randomly selecting participants would therefore be as absurd as randomly selecting variables in a traditional survey (Watts & Stenner, 2012).
The authors of my original Q-set data (i.e. the 451 documents) constituted a body of potential participants for this study. 11 In the first instance, only those individuals associated with the smaller sample (i.e. 147 documents) were contacted by email and invited to participate. Invitations were subsequently extended to individuals associated with the wider collection of documents. A total of 173 individuals were invited to participate; 40 accepted the invitation and completed the online Q sort (see below). 12 Participants included both English and Spanish speakers, from countries of the global North and global South, and included members of civil society, intergovernmental organizations, bloggers, and publicly engaged academics.
Administering the Q sort: Participants were directed to the online platform 'PoetQ', 13 which was adapted to function entirely in both English and Spanish. This programme took the participants through multiple sorting rounds in which they indicated their agreement, disagreement, or neutrality with each statement. The scale of opinion ranged from -4 (most disagree) to 4 (most agree). Responses were forced into the distribution formula shown in Table 1.
The online sorting process produces a unique matrix for each participant (shown in Table 2). Each cell represents a statement, as identified by its number. Participants were given an opportunity to rearrange their matrix to best reflect their subjective position in the green political economy debate (i.e. to best reflect the statements with which they felt most strong and indifferent about). In the final stage of the Q-sort, participants were asked to explain how they felt about the statements they had ranked at -4 and 4 (i.e. to explain why they most agreed and disagreed with these statements). This final step provides an opportunity for participants to comment on the issue and explain their positions in their own words.
Factor analysis: Dedicated Q methodology software, PQ Method, was used to perform correlation and factor analysis of the 40 completed Q-sorts. A three-factor solution was ultimately accepted through an iterative process of factor extraction, rotation, and interpretation. While a set of statistical criteria can ultimately determine whether a solution is accurate and valid, Q methodological factor analysis is a deeply interpretive and qualitative process. It is therefore important that considerations of statistical significance do not preclude qualitative judgements about appropriate factor solutions. The scope for interpretation is by no means boundless; the data will only reveal relationships among the variables that actually exist, it is then the role of the researcher to decide which relationships make most substantive sense. The most commonly used indicator of a statistically sound factor solution is the Eigenvalue (EV) (i.e. the sum of its squared factor loadings). Generally only factors with an EV of more than 1.00 are considered significant. Additionally, a factor is only considered viable if it has at least two significant factor loadings (i.e. at least two participants are highly correlated with that factor and no other). For this study, a significant factor loading was calculated at 0.37 (at the level of p<.01).
Initial extractions of five and four factors were rejected on the basis of insufficient significant loadings. Although the Eigenvalues in both cases were over 1.00, the fourth and fifth factors each had only one significant loading. Three factors were subsequently extracted using the Centroid method and Varimax rotation. Although this solution captured 54% of the variance (i.e. the full meaning and variability within the data), and produced factors with appropriate EVs and significant loadings, it too was rejected because Factors 2 and 3 were quite highly correlated (.48). Correlation at this level is generally taken as a sign that too many factors have been extracted or that the two correlated factors are 'alternative manifestations of a single viewpoint' (Watts & Stenner 2012, p. 141). Extracting two factors produced a plausible solution (plausible both statistically and substantively). Nevertheless this extraction offered visual evidence to suggest that a three-factor solution would be preferable to a two-factor solution, as revealed in Figure 1. This figure shows the correlations of all the Q sorts with Factors 1 and 2. Factor 1 accounts for a particularly large proportion of the variables. This in itself is unsurprising. Yet, two fairly distinct clusters of participants are evident within this Factor, suggesting there may be substantively significant differences within the participants loading on Factor 1. To determine whether these clusters were different in any meaningful way, the Factor 1 loadings were isolated and subjected to an additional Centroid factor extraction and Varimax rotation. This revealed that these clusters largely diverged on the issue of valuing nature in economic terms and the importance of pursuing new measurements of progress. This points to the importance of qualitative judgement in determining a correct factor solution. Familiarity with green political economy debates informed my judgement to take this distinction seriously by pushing the Q-sort clusters onto separate factors. This was achieved by extracting three factors (Centroid method and Varimax rotation) and subsequently hand rotating Factors 1 and 3 by -8°. This pushed the lower right-hand cluster onto Factor 3, while keeping the upper left-hand cluster on Factor 1. The resulting solution was statistically sound: 55% of the variance is explained by these factors; factor correlations range from -.0975 to .4296; Factors 1, 2, and 3 have significant EVs of 13, 7, and 1, respectively. 14 McKeown and Thomas argue that 'it does not understate the case unjustly to stipulate that all that … factor analysis does is lend statistical clarity to the behavioural order implicit in the correlation matrix by virtue of similarly (or dissimilarly) performed Q sorts' (2013, p. 52). The statistical actions performed in Q methodology support us in interpreting the discrete subjective views on any given issue. Ultimately the plausibility of any solution rests on the coherence and insight provided by the substantive account of each factor. In the following section, I aim to demonstrate that this factor solution is substantively sound by providing a narrative account of each factor. First though, it is important to clarify how these narratives were drawn from the factorized data. Just as each Q-sort produces its own correlation matrix, so does each factor produce a composite correlation matrix. Van Exel and de Graaf explain: '(t)he composite Q sort of a factor represents how a hypothetical respondent with a 100% loading on that factor would have ordered all the statements of the Q-set' (2005, p. 9). Table 3 below shows the position (or value) of each of the statements in the composite correlation matrix of each factor. As explained earlier, a value of -4 signifies strong disagreement, while a value of 4 signifies strong agreement. The substance of each factor was interpreted on the basis of four considerations, each of which can be seen in Table 3: a The statements valued at -4 and 4; b The statements valued higher in each factor than any other factor; c The statements valued lower in each factor than any other factor; d The significance of each statement's Z-score. 15

Results
The statistical analysis and interpretation described in the preceding section produced the following set of discourses of green political economy. 16 Factor 1: Radical Transformationism Factor 1 reflects a post-growth vision of a sustainable economy that is very different to the status quo. From a Radical Transformationist perspective, we need to accept that environmental sustainability is completely incompatible with continuing economic growth; we cannot expect to reduce pollution and preserve ecosystems while simultaneously growing the economy (36). Even if it were possible to reconcile economic growth and environmental protection, it would be desirable to rethink our commitment to growth-based economies because growth in itself can be problematic (8,11). As one respondent argued '(i)t's harmful and is part of what has caused us to be so spectacularly unsustainable in the first place' (p. 25). Another respondent couldn't imagine 'how strong economic growth could deliver a just and sustainable global economy given the current impacts of growth policies, including on the thousands of communities affected by industrial-scale mining, oil, and agricultural projects' (p. 37) Often further growth just results in greater inequality, which is undesirable from both social and environmental perspectives (11,2). To mitigate the existence and effects of inequality, governments ought to be pursuing redistributive policies such as agrarian reform, fiscal transfers, progressive  taxation, and other public spending (2). Capitalism is certainly not the only viable economic system; in fact, it 'absolutely not viable' and is actually only 'a very new invention in the history of humanity' (p. 37; p. 12). Therefore, it would be a mistake to limit ourselves to trying to make an unsustainable system sustainable (26). Lessons can be drawn from historical and contemporary experiences where communities have lived well under noncapitalist economies (p. 12; p. 30). In general terms, an alternative to our existing market-based economies lies in strengthening economic relations based on cooperation and sharing (12). A sharing-based economy could be a sustainable economy (12), especially if it relies on renewable energies rather than fossil fuels; this is possible in developed and developing countries. We should not assume that low-income countries will need to continue to use fossil fuels for the foreseeable future to reduce poverty and promote development (31). Insofar as market-based economic relations persist, these should become much more regionalized and localized. Re-localization, one respondent stressed, is 'the only sustainable way forward … . Having lengthy just-in-time supply chains is a recipe for disaster-they are brittle and rely on cheap fuel ' (p. 12). 'In the context of peak oil, re-localisation is an imperative' (p. 25); and this may require resisting pressure to reduce barriers to international trade and foreign investment (21, 39). Small-scale economies are inherently better for both people and the environment (39). They allow people to have greater control over the environmental resources on which their wellbeing depends, and this is especially important for poor and marginalized communities (1). A radical shift is needed because despite several decades of talking about sustainable development, the global economy is certainly not becoming greener (48). Existing market-based environmental solutions are part of the problem not part of the solution. The idea of protecting nature by pricing it is both dangerous and undesirable; 'it leads to the commodification of the natural world and legitimizes the privatization of essential common natural resources … . This attitude ignores the necessity for the conservation, restoration and protection of essential life supporting ecosystems ' (p. 15). For another respondent, this approach to environmental protection should be rejected because it is 'a dangerous extension of prioritizing economic growth ' (p. 11). In short, protecting nature is too important to be left to the whims of the market (18; 16). This perspective is also associated with a fairly high level of scepticism about the existing sustainability initiatives of businesses, governments, and international institutions (43).

Factor 2: Cooperative Reformism
This vision of a sustainable economy stresses the importance of cooperating to sustainably reform the economic system that we have (40). Cooperative Reformism differs from Radical Transformationism most significantly on the assumption that sustainability, capitalism, and economic growth are compatible. Cooperative Reformism strongly rejects the suggestion that we need to abandon capitalism and growth-based economies in order to live sustainably; such ideas are considered unnecessary and certainly unrealistic (36,9,45,23,27). As one respondent explained, '(t)here is no causality between profit and degradation. The economic system has simply not considered its consequences before, but is now beginning to. There will be innovation that satisfies new sustainability goals as well as economic growth' (p. 7). In short, 'environmental sustainability and economic growth are compatible' (p. 32). Radical ideas like reducing the standard working week or shifting to a steady-state economy were equally dismissed (5,33). Instead, we need to recognize the sensible and realistic opportunities that are available to us to ensure win-win outcomes. In contrast to the negative view of economic growth associated with Radical Transformationism, in the Cooperative Reformism discourse economic growth is important for improving lives and should be pursued (10,11). Indeed, economic growth 'generates wealth and wellbeing' (p. 32). Continued improvements in technology and efficiency will allow us to increase GDP, profits, and jobs while still reducing pollution and preserving ecosystems (36). Greening the economy does not have to threaten either workers or companies: '(i)t is being demonstrated that "green" jobs are on the increase. Fighting climate change and other environmental problems requires qualified people. At the same time, sustainable companies are more efficient in the long-term' (p. 31). We should be looking to redirect existing fossil fuelbased subsidies towards investments in new clean technologies (2). Energy-intensive sectors in countries with ambitious environmental policies should not be compensated or protected (15). At the same time, however, we must be 'practical' and 'realistic' (p. 2) and not assume that we can immediately break our dependence on fossil fuels (30). Moving towards a more sustainable economic system requires a concerted and collaborative effort in which everyone contributes to reducing their impact on the earth (40). Such '(c)ollaboration … is vital for sustainability to be acceptable for all affected ' (p. 27). Everyone consumes, often on an unsustainable scale, so everyone has a responsibility to reduce their individual impact (47). In the context of governance, this means that governments, business, and civil society should collaborate to envisage and implement policies for a sustainable economy (40). This will not happen automatically or immediately, but instead needs to be promoted over time: '(t)here exists a lack of citizen consciousness in relation to environmental problems. We need more education and awareness-building' (p. 31); 'generation by generation we must build a more self-conscious society' (p. 2). While recognizing our common responsibilities, we must keep in mind that the costs and benefits of transitioning to a sustainable global economy will not automatically be distributed evenly. We need to explore opportunities for fair burden sharing. The interests of poor people and low-income countries ought to be safeguarded in this transition to ensure that they are not negatively impacted (14). In 'equal societies', one respondent observed, 'it is easier to ask everyone to bear the "burden" of sustainability' (p. 4). There is no reason for developing countries to continue to rely on fossil fuels for their economic development (31), but they will need assistance to shift course. In this respect, greater technology transfer from industrialized countries to developing countries is important, and indeed 'obligatory' (p. 27) (19). Furthermore, given that inequality undermines sustainability, we ought to be pursuing a range of redistributive policies to mitigate inequalities and ensure that everyone can enjoy the benefits of a sustainable economy (2). In contrast to Radical Transformationism, Cooperative Reformism considered it unhelpful and unnecessary to be suspicious of governments' efforts to promote ideas like the 'green economy' (43,22). We shouldn't assume that the 'green economy is a "western imposition", it is a concept that makes sense for all countries … because all are affected by natural resource scarcity and environmental impacts' (p. 32). Such initiatives are genuinely needed to move us closer to a sustainable global economy.

Factor 3: Statist Progressivism
Factor 3 presents a vision of a sustainable economy based on the pursuit of wellbeing and happiness rather than gross domestic product (6). 'GDP', one respondent stressed, 'is an indicator that is increasingly partial and obsolete' (p. 38). We should not assume that the concept of happiness is too subjective and problematic to pursue as a policy goal; it can and should become the main measurement of progress. This is desirable from both an environmental and a social perspective (7). As one respondent summed up: '(y)ou get what you measure and we are measuring progress the wrong way' (p. 8). Wealthy countries in particular need to move away from a system in which economic growth is pursued as an end in itself (10). Continued economic growth is not necessary for transitioning to a fair and sustainable global economy (8). This critical questioning of economic growth places Statist Progressivism at a middle point between Radical Transformationalism and Cooperative Reformism. Statist Progressivism doesn't share either the wholehearted rejection or support for economic growth, but rather questions the desirability of an economic system oriented so exclusively towards economic growth. The state is seen to have a fundamental role to play in moving society towards a new green economic order that pursues wellbeing over growth. In short, one respondent argued, '(i)t is the responsibility of governments to formulate environmentally friendly policies' (p. 26). This will involve eliminating subsidies for environmentally damaging activities and use these funds to invest in new clean technologies (29); reducing the standard working week to share the benefits of employment more widely and reduce consumption (5); and potentially taking responsibility for moving workers out of unsustainable jobs and assisting them in moving into fair and sustainable work (4). This new wellbeing economy may still be based on some form of capitalism (27), a system that 'rewards innovation, which is necessary when dealing with global sustainability issues' (p. 9). This moderate position on capitalism distinguishes Statist Progressivism from Radical Transformationalism. It is evident that the model of capitalism supported by Statist Progressivism would involve much more state intervention and steering than dominant forms of liberal market capitalism. 17 Such intervention will need to include redistributive policies because existing levels of inequality undermine sustainability (2). The state can and must play a central role in making our economies sustainable. It is therefore unhelpful and unnecessary to be sceptical about existing governmental sustainability initiatives like 'green economy' plans (43), environmental standards (22), the valuing of 'natural capital' (17), and the efforts of leading businesses to integrate sustainability into their corporate culture and decision-making (24). Indeed, we can already see 'lots of good examples of how the green economy is benefiting the poorest' (p. 9). The rejection of green economy scepticism is what most sets Statist Progressivism apart from Radical Transformationalism. A further significant difference between the two concerns the issue of valuing nature. From the perspective of this factor, we should certainly be cautious about commoditizing nature and opening nature up to trade (17), but we should not assume that putting a price on nature is inherently dangerous and undesirable (18). 'We should not confuse value and price ' (p. 8). Nature is intrinsically valuable, but it is also economically valuable. This value is evident, for example, in the various services provided by wetlands and forests, as well as the income generated from nature-based tourism. To recognize this economic value does not detract from nature's intrinsic value. Recognizing the economic value of nature facilitates good policy-making, and this does not necessarily have to entail market mechanisms like pricing and trading. In some cases it may be appropriate to develop market mechanisms, but '(c)are needs to be taken to assess when (this) is a suitable approach' (p. 17).

Discussion
What this study has revealed are the meanings behind the terminology used in debates about economic development and environmental sustainability. The rise and renewal of terms such as green economy, green growth, inclusive wealth, harmony with nature, etc. reminds us that there is no single way of problematizing the relationship between economic development and environmental sustainability. But the varied terminology itself obscures actual points of agreement and disagreement. Q methodology can reveal agreement and disagreement in stakeholder debate that is otherwise obscured by rhetoric. Of course there are nuances in individual positions and narratives that have not been captured in this analysis. What Q-based discourse analysis does is identify how groups of people align with different ensembles of ideas and beliefs. Just as the members of a political party will disagree in some respects, they do agree on a basic vision and set of fundamental principles. 'Discourse coalitions' (Hajer, 1995) similarly articulate broadly shared visions, values, and principles. What is shared is ultimately more important than what is disputed. But discourses are fluid rather than fixed. This makes periodic discourse analysis important whether our motives derive from intellectual curiosity or a desire to enhance political representation. My aim in this section is therefore to compare the contemporary discursive terrain of green political economy with earlier times. A comprehensive review of existing analyses of environmental discourses is beyond the scope of this article. 18 Instead I focus on two influential examples: Dryzek (1997Dryzek ( , 2005Dryzek ( , 2013 and Hajer (1995). Dryzek described four basic environmental discourses: 'problem solving', 'sustainability', 'survivalism', and 'green radicalism', which were then dissected to reveal nine specific discourses including Promethean discourse, administrative rationalism, sustainable development, ecological modernization, and green consciousness. Hajer charted the rise of 'ecological modernisation' and showed how it conquered more radical discourses in the struggle to define 'sustainable development'. These more radical discourses were exemplified by the reports Limits to Growth, Blueprint for Survival, and Small is Beautiful published in the 1970s. Some 20 years have passed since the initial publication of Dryzek's and Hajer's analyses. Given the increasingly diverse language deployed in debates about sustainable economic development, it is worth considering whether the substance of debates has also altered. To some degree it is a case of 'plus ça change, plus c'est la même chose'. But there are some novel features in contemporary discourses. Cooperative Reformism is clearly a close relation of ecological modernization. Both discourses are focused on making the capitalist economic system less resource-and waste-intensive through the close cooperation of government and industry. Yet, a salient feature that emerged in the analysis above was a concurrent concern with burden sharing, which is absent from Dryzek's and Hajer's characterization of ecological modernization. This emerged as a concern for ensuring that everyone 'does their bit'-all citizens, businesses, and nations-but in such a way that supports the less well-off and protects their interests in the transition to more sustainable economies.
Radical Transformationism has evidently grown out of the radical environmentalist positions articulated in Blueprint for Survival and Small is Beautiful, which called for radical political and economic restructuring, including decentralization, self-sufficiency, and bottom-up self-governance (Hajer 1995, p. 84-85). Radical Transformationism is also closely aligned with Dryzek's radical green discourses, which he distinguished along two lines: one focusing on changing consciousness (green consciousness) and the other on changing political, economic, and social structures and practices (green politics) (2005, p. 181). With its focus on structural change, re-localization, and redistribution, Radical Transformationism closely resembles green politics.
Statist Progressivism has inherited many elements from Limits to Growth, particularly its critique of exponential economic growth (Hajer 1995, p. 79-83). 19 This earlier report's faith in top-down management is also reflected to some degree in this contemporary discourse's emphasis on state agency, authority, and responsibility. Novel discursive aspects emerge in the emphasis on wellbeing and happiness, which are much more salient in environmentalist debates now than in earlier years. GDP has long been a focus of green critique, but it is only in recent years that this critique has moved closer towards the policy mainstream from the radical margins. This is evident, for instance, in the proliferation of new measurements of progress (Happy Planet Index, Better Life Index, etc.); initiatives such as the French Government's Commission on the Measurement of Economic Performance and Social Progress (headed by Joseph Stiglitz), and the UN's commission into 'Broader Measures of Progress'.
The most notably novel aspects of Radical Transformationism and Statist Progressivism compared to earlier radical discourses is the contentious debate about using monetary valuation to conserve nature. The values underpinning rejection to valuation may be found in the older radical positions but the salience of this issue has clearly increased in the past two decades. While payment-for-ecosystem-services initiatives were in place in the 1990s, it is only in the past few years that environmental accounting and ecosystem valuation has been widely institutionalized at the international level. 20 Questions of whether and how to value nature in monetary terms are therefore provoking stronger debate (e.g. Kenner 2014), and this is an important distinguishing element in Radical Transformationism and Statist Progressivism.

Conclusion
The question of how to pursue economic development given persistent environmental degradation cannot be treated purely as a technical or objective matter. There are diverse visions for a sustainable economy, and these encompass different social and political values. The assumptions associated with different discourses may be contested or invalidated, but not in such a way that will reveal one true and legitimate discourse. It has been beyond the scope of this article to elaborate on the normative implications of this discursive diversity. Here it must suffice to suggest that identifying the range of contemporary discourses of green political economy is a necessary step towards improving the representation and inclusiveness of debate, decision-making, and governance. This study has shown that the substance of this diversity cannot be gleaned simply by listening out for different language and concepts. Arguments might be advanced under the banner of many different terms such as green economy, green growth, wellbeing, gross national happiness, inclusive wealth, harmony with nature, de-growth, steady-state economy, and buenvivir.
Q methodology is a useful method for pushing beyond such terminology to map the actual positions in these debates. As with all research methods, Q does have limitations. Perhaps the principal weakness is that rich and nuanced arguments have to be reduced to rather short statements to facilitate the Q-sort. Lengthy statements with multiple clauses are impractical and undesirable. Participants can become frustrated by the lack of nuance and the inability to qualify their agreement or disagreement with certain statements. 21 The final stage of the Qsort offers participants the opportunity to explain in their own words how they feel about their strongest statements, which mitigates this problem. Traditionally Q-sorts have been conducted in-person, which has the advantage of allowing the participant to talk through their decisions with the researcher. This approach has proven fruitful for localized sustainability debates (e.g. Barry & Proops 1999;Lo forthcoming), but it is impractical for studies in which the focus is on international debates. Maximizing the geographical diversity of participants will generally necessitate carrying out the Q sort online.
My initial assumption that diversity of sustainability terminology obscures the precise points of agreement and disagreement was supported by the study presented in this article. Three distinct perspectives were distilled by conducting a bilingual Q-study based on international communication surrounding the Rio+20 summit in 2012. Radical Transformationism reflects a post-growth and post-capitalist vision in which economic relations become more localized, and in which there is no place for putting a monetary price on nature. Cooperative Reformism sees capitalism and growth-based economies as potentially compatible with a sustainable environment. Fundamental social and economic change is rejected in favour of sensible and cooperative changes that promise win-win outcomes while protecting the needs of the poorest. Finally, Statist Progressivism acknowledges the limitations of growth-based economies and holds a vision of a sustainable economy based on the pursuit of wellbeing and happiness rather than gross domestic product. The state has a central role to play in redirecting society in this way, and ensuring that nature is valued carefully to capture its economic and intrinsic qualities. Adopting broad concepts like 'sustainable development' allow us to paper over these evident tensions and disagreements. But given that many indicators of environmental quality have continued to decline over three decades of sustainable development policy, a more fruitful albeit challenging task would be to openly reflect on which of these policies and practices the planet can genuinely sustain in the decades ahead.