Framing the climate issue : Patterns of participation and prognostic frames among climate summit protestors

Transcription

1 Framing the climate issue : Patterns of participation and prognostic frames among climate summit protestors Mattias Wahlström University of Gothenburg Department of Sociology and Work Science Box 720, Gothenburg Sweden mattias.wahlstrom@gu.se Magnus Wennerhag University of Gothenburg Department of Sociology and Work Science Box 720, Gothenburg Sweden magnus.wennerhag@gu.se Christopher Rootes University of Kent School of Social Policy, Sociology & Social Research, Cornwallis NE, University of Kent, CANTERBURY, Kent CT2 7NF UK C.A.Rootes@kent.ac.uk This is the preprint version of a manuscript accepted for publication in the journal Global Environmental Politics. Final version submitted

2 Framing the climate issue : Patterns of participation and prognostic frames among climate summit protestors Abstract Who protests at climate summits? We analyze responses to surveys of three large demonstrations in Copenhagen, Brussels and London, organized in connection with the 2009 UN Climate Change Conference (COP-15) in order to determine who demonstrated, and how and why the collective action frames employed by demonstrators varied. The demonstrations were products of the mobilization of broad coalitions of groups, and we find significant variation in the demonstrators prognostic framings, most notably a tension between system-critical and more individual action-oriented framings. A large proportion of demonstrators expressed affinity with the global justice movement (GJM), but we find little evidence of an emerging climate justice frame among rank-and-file protesters. The individual variations in framing may be attributed in part to differences between the mobilization contexts of the three demonstrations, the perspectives and values of individual participants, and the extent of their identification with the GJM. 2

3 Introduction Protests at and around climate summits are not new, but the patterns of participation in those protests have not been systematically investigated. The 2009 UN Climate Change Conference in Copenhagen centered around the 15th Conference of the Parties to the United Nations Framework Convention on Climate Change (COP-15). Attended by 190 national delegations, with 30,123 participants registered, 1 formal negotiations among political leaders were surrounded by meetings and seminars involving politicians, scientists and representatives of a wide range of NGOs. The summit also prompted extraordinary protest mobilizations, both in Copenhagen and elsewhere. Although, as many of the protesters expected, the summit ended without agreement on an effective regime to address global climate change, it provided an important opportunity for networking and mobilization with potentially enduring consequences for the development of a transnational climate movement. The events surrounding COP-15 provide a unique opportunity to explore the demographic characteristics and political goals of protesters mobilized around the most compelling transnational environmental issue of our time global climate change. By surveying different demonstrations related to a single transnational summit, the climate movement may be examined from the perspective of rank-and-file participants, which is an important complement to the images conveyed by the slogans and rhetoric of the social movement organizations (SMOs) engaged with climate change. Here we analyze data on participants in the three largest European climate demonstrations focused around COP-15: those in London and Brussels on 5 December 2009, the international day of action designed as a concerted display of demands for action by world leaders at COP- 15, and the Copenhagen march on 12 December, during COP-15 itself. 2 Our focus is on how the participants in these three large demonstrations understood the issue about which they were demonstrating. More specifically, what collective action framing did the participants employ concerning solutions to climate problems? In other words, we focus 1. Fisher 2010, The three demonstrations were surveyed as part of the international research program Caught in the act of protest Contextualizing contestation (hereafter CCC) (see van Stekelenburg et al. 2012) by its British, Belgian and Swedish teams, using the standardized method for surveying demonstrations developed within this research project. 3

4 on the so-called prognostic aspect of collective action framing. 3 If the climate movement had not yet found a unifying frame to bridge its various factions, 4 a strong candidate is climate justice with roots in the broader frame environmental justice 5 which has lately become increasingly salient, at least within parts of the movement. 6 Nevertheless, collective action framing of a campaign is seldom uniform among SMOs, and still less so among individual participants. 7 Considering the divisions within the climate movement, the broad range of organizations in protest coalitions, and participants different backgrounds, we expected to find significant variation among individual participants framings, despite the relative unity that some might expect to find in an internationally coordinated protest campaign. 8 We aim to illuminate coherence and variation within the (European) climate movement and the relative significance of factors that hypothetically explain differences in framing among participants in climate demonstrations. We focus on three distinct types of solutions to the climate crisis found in participants frames: action at the system level and through increased global justice (reflecting the emergent climate justice frame); individual action (either reformist or the life politics associated with new social movements); policy changes within existing political systems (reformist factions). 9 After briefly describing the three protests, we shall turn to our theoretical and methodological approaches and data analysis, focusing on: who demonstrates; how demonstrators frame solutions to the climate problem; and the relationship between demonstrators sociodemographic and organizational affiliations, and their framing of the climate issue. Three climate demonstrations The Copenhagen demonstration, largest of the numerous protests in Copenhagen during COP- 15, attracted between 50,000 and 100,000 people. It was organized by a coalition of environmental, religious, political, trade union, and solidarity organizations, embracing Benford and Snow Moser 2007, cf. Newell Capek 1993; Taylor Chatterton et al. forthcoming; Newell Cf. Benford Hadden Reitan and Gibson

5 organizations from 67 countries. 10 The march, from central Copenhagen to the summit venue, proceeded peacefully, but in one section of the march a skirmish broke out between some demonstrators and police, precipitating the arrest of nearly 1,000 protesters, almost all of whom were released without charge. Other notable protests were mounted during the summit 11 but, because it is the broad climate movement in which we are most interested and whose manifestations we seek to compare, we surveyed the demonstration on 12 December. This was by far the largest and likely included most who participated in smaller protests. The Brussels march, which attracted 15,000 participants, was organized by the Climate Coalition, established in 2007 and comprising 70 organizations, including environmental, trade union (Socialist/social democratic and Christian), and youth organizations. One of the largest environmental demonstrations in Belgian history, the lightly policed march was peaceful and notable for the participation of prominent people such as Belgian Prince Laurent and his family. The London march on the same day attracted over 40,000 participants. The largest demonstration on an environmental issue in British history, the march was co-ordinated by the Stop Climate Chaos coalition 12 and supported by a wide range of organizations, from large international aid, development, human rights and environmental NGOs to small local groups, with more modest participation from trade unions and student groups. It too was peaceful and with a low-key, facilitative police presence. The organizers employed private security to marshal demonstrators and provided large quantities of printed placards but, whilst the large NGOs were especially prominent, there was also a large number of homemade placards with idiosyncratic slogans. As expected, the platforms and themes of the three demonstrations were very similar. In each case, organizers called for strong resolutions at COP-15 and urgent action, and all mentioned 10. Available online at accessed 27 July Respondents to our questionnaire named 19 different countries of residence, including Denmark (51.2 percent), Sweden (21.3 percent), Germany (8.3 percent), UK (3.8 percent), France (2.9 percent), Belgium (2.9 percent), and Norway (2.5 percent). In contrast, 99.6 percent of London respondents named the UK, and 99.1 percent of marchers in Brussels gave Belgium as their country of residence respectively. 11. Cf. Reitan Rootes 2011; Saunders

6 the need to support countries with fewer resources. However, in all three demonstrations individual participants slogans were more varied, which underlines the importance of our participant-oriented approach to movement framing. Collective action framing and transnational protest coalitions The concept of framing has been employed to illuminate how movements define problems (diagnostic framing), devise solutions to them (prognostic framing), and mobilize new followers for their cause (motivational framing). 13 These different aspects are often linked in various ways in the interpretive frameworks of movements, but here we are interested in prognostic framing as an analytically distinct dimension. Frames are frequently portrayed as the outcomes of strategic activities of critical communities and the leaderships of SMOs. 14 Examining the aggregated patterns of collective action framing among participants in demonstrations is a way of mapping the distribution of old and emergent collective action frames among grassroots activists whilst taking the latter seriously as active contributors to the identity and goals of a movement. Although movement leaders and intellectuals are principal disseminators of frames and interpreters of the meaning of collective actions, participants framings may differ, be more or less complex and thus may affect the development of the movement. Participants framing especially when expressed in less public contexts, as in an anonymous survey can also be expected to differ from public organizational framing by being less strategically adjusted to resonate with salient public framings. 15 In line with other research on activism around COP-15, we expected climate justice to be an emergent if not yet fully established master frame for climate protesters in Copenhagen, Brussels, and London. 16 The concept of master frame is commonly used to designate more general frames that serve to bridge the differences among diverse social movement groups and organizations during campaigns or entire cycles of protest. 17 Climate justice functions as 13. Benford and Snow 2000; Gamson 1992; Johnston and Noakes 2005; Snow et al Brulle and Benford 2012; Noakes and Johnston 2005; Rochon Cf. Kubal Chatterton et al. forthcoming; Fisher Snow and Benford 1992; Snow 2004:390. 6

7 a master frame in the sense that it fuses notions of climate protection with global justice. It thus bridges the concerns of groups associated with the global justice or alterglobalization movement with those more specifically motivated by concerns with climate change and, with increasing mutual recognition of the interconnections among their agendas, the climate justice frame has become more central since the late 2000s. 18 Environmental groups and issues were already present in the wave of protests brought to public attention in 1999 by demonstrations against the World Trade Organization in Seattle and summit meetings in years thereafter, but the justice framing then was primarily a critique of social injustices on a global scale and the democratic deficit of supranational bodies. 19 While new social movements of the 1960s and 1970s often promoted a form of life politics disconnected from traditional political institutions, the activists of the GJM have combined an emphasis on individuals own actions toward creating a more just world with concrete proposals to reform and democratize political institutions on global and national levels. 20 According to Pleyers, these frames of the GJM have also infused in the climate movement a focus on issues of global governance and social justice alongside individuals responsibility to promote alternative ways of life. 21 However, the broad climate movement is still divided in respect of its frames and discourses concerning the climate issue and what to do about it. 22 Bäckstrand and Löfbrand identified three general discourses of climate governance green governmentality (broadly emphasizing steering instruments connected to national and international governmental organizations), ecological modernization (envisaging free market and technological solutions), and civic environmentalism (stressing equity and broad participation in solutions) arguing that a reformist version of the latter was emerging as the dominant discourse, despite the lasting impact of ecological modernization in contemporary political 18. Pleyers 2010, ; Reitan and Gibson della Porta 2007; Wennerhag Sörbom and Wennerhag Pleyers 2010, Newell Following Johnson (2002), we regard discourses as comprehensive meaning structures governing what it is sensible to say and do, whereas frames refer to more specific aspects of a discourse, often used strategically. 7

8 practice. 23 In line with the reformist civic environmentalist discourse, Reitan and Gibson observe how the leading civil society network in the COP-15 protest campaign Climate Action Network (CAN) promoted regulation through fair, ambitious and binding targets and treaties, sustainable economic growth supported by technology and aid transfers to the South, and individual behavioral change. 24 Meanwhile the other major network, Climate Justice Now! (CJN), bridged frames of social, ecological, and gender justice 25 associated with the GJM expressed in a climate justice frame. In sum, in view of widespread criticism by the climate movement of the ecological modernization discourse, we focus specifically on comparing three important tendencies within the movement: framing solutions in terms of systemic changes involving global justice; policy changes within existing institutions; and individual behavioral changes. Hypotheses Variation in the ways participants in demonstrations frame the climate issue might be explained in terms of a variety of significant factors operating on different levels. The Copenhagen demonstration included people who were sufficiently motivated to travel from abroad, a category of protesters who in similar mobilizations have been demonstrably more radical and less satisfied with democracy in their home countries. 26 Demonstrations are also located in specific mobilizing contexts, and are shaped by local and event-specific circumstances. Therefore, both the location of the demonstration and the barriers that filter out certain kinds of non-domestic demonstrators but possibly attract others might be expected to be reflected in participants collective action framing. Since frames are proposed by social movement organizations (SMOs), the different organizational affiliations of participants might be presumed to entail dissimilarities in their framing of the problem. However, social movements can also be considered as networks of people and organizations held together by collective identities and, sometimes, shared ideologies. 27 Individuals framing of an issue might therefore be connected to their ideological stance as well as their identification with the movement (or rather, with one of the sometimes plural definitions of a specific movement). 23. Bäckstrand and Löfbrand Reitan and Gibson 2012, Ibid. 26. Bédoyan et al Diani

9 We retain as a point of reference the climate justice frame, which, as well as being the leading contender to function as a master frame for these demonstrations, is the one that previous research indicates to be most strongly associated with the radical left and groups rooted in the GJM. 28 This leads us to formulate the following five hypotheses: Hypothesis 1: Hypothesis 2: Hypothesis 3: Hypothesis 4: Hypothesis 5: Individuals framings will be affected by the mobilizing contexts of the demonstrations. Participants who have travelled from abroad to protest will articulate more radical frames, framing the climate issue in terms of system change and climate justice. Organizational affiliation will impact significantly on collective action framing, with active members of global justice and leftist political organizations having more system-critical perspectives (in opposition to the global neoliberal economy) on solutions to the climate problem. Protesters who place themselves on the left of the political spectrum will more often frame solutions to the climate problem in terms of system change or climate justice. Individuals who identify with the GJM will more often frame solutions to the climate issue in structural terms, or as a matter of climate justice. Method Sampling Respondents were sampled using the standardized method of the CCC research program. 29 At each demonstration, to minimize sampling bias, potential respondents were systematically selected according to a common protocol. All selected respondents were given a questionnaire with a reply-paid envelope, and every N th respondent was additionally interviewed face-toface, using a short, single-sheet schedule, to collect basic data about the demonstrators in order to facilitate control for non-response bias. Because non-response to the oral questions was very low, respondents interviewed could be regarded as broadly representative of the sampled population (for numbers of interviews, see Table 1). 28. Reitan and Gibson van Stekelenburg et al

10 Table 1: Basic data on the three demonstration surveys Brussels Copenhagen London Oral interviews conducted Mail-back questionnaires distributed Response rate mail-back questionnaire (%) 40% 31% 40% To test for non-response bias, we compared the answers to certain questions between those respondents who were interviewed face-to-face and returned the self-completed questionnaire, and those who were interviewed but did not return it. There were no significant differences between these two groups in respect of gender, year of birth, university education, level of interest in politics or whether the respondent was a member of the organization staging the demonstration. Operationalization of the dependent variable Most frame analytic studies of social movements employ a qualitative approach; quantitative survey studies of framing are still very rare. 30 A secondary aim of our investigation is therefore to demonstrate how open survey questions may be used to map the distribution of protest participants framings. Since we focus on climate protesters prognostic framing, we used the question: What should be done to address this issue [climate change/global warming]? Two of the authors were involved in the coding of each response, to ensure the reliability of the result. These codes are inductive in the sense that we did not begin with a preconceived set of theoretically motivated codes, but instead added codes when necessary to encompass the various positions expressed by respondents. Because many respondents answered by listing several things, we gave each code a separate dummy variable. Answers that expressed uncertainty or were too vague to code are treated as missing. The initial coding produced agreement about 91 percent of cases; remaining discordances were then carefully 31 negotiated. We assigned the following codes: 30. See however, e.g., Marullo et. al. 1996, and the quantitative analysis of speech acts by McCammon Apart from the codes described below, a small number of respondents also mentioned protest activities and reducing world population as solutions to climate problems. Another 8.8 % mentioned an unspecified reduction of CO2. 10

11 System change may designate the classic socialist conception of abandoning the capitalist economy. We also include those asserting that the economy/society should be radically changed in other ways, e.g., to a steady state or de-growth economy, including responses ranging from Consider a transition from our economic model towards a more sustainable model (Brussels demonstrator) to getting rid of capitalism (London demonstrator). Global justice includes only explicit mentions of this or synonymous terms, as well as references to the relationship between richer and poorer countries. Arguably, this implies a system change, but we kept this separate code in order to single out more explicit references to a global justice frame. Below we collapse this and the System change code for analytic purposes. Change individual behavior / raise awareness. A relatively broad category, this includes statements focusing on general changes of individuals and their behavior, e.g.: We know how we can t go on, so let us change our way of life (Copenhagen demonstrator). Also included are those who explicitly mention information and education ( raising awareness ), as well as changes in values. Reducing consumerism (without mentioning any kind of political regulation) is included, since it is assumed to be related to lifestyle. This frame can be based in everything from liberal ideology seeing individuals as atomized actors (and being skeptical towards collective action) to the life politics attitude towards collective action found in the new social movements, stressing the everyday role of individuals in reproducing and/or counteracting structural injustices. Technological change / investments reflects views that climate problems should be wholly or partially solved through technological (and scientific) development. Includes statements that investments should be made in technology, and statements about belief in the power of expertise to solve climate problems: leave it to experts politicians can t fix it (Copenhagen demonstrator). Legislation / policy change includes political measures, such as changes in legislation and taxation. Legislation, as not enough people are prompted by information to do the 11

12 right thing especially big business (London demonstrator). International agreements are included, as well as more or less specified notions of regulation and limitation, and those mentioning changes obviously based on political measures, such as increased public transport. Change industry / production: industry/production ought to change in ways not explicitly involving technological development, instead implying actions such as downsizing, shifting to existing climate-friendly alternatives, and simply following regulations, e.g.: replace fossil fuels by solar, wind, wave, tidal and nuclear energy rapidly! (London demonstrator). As distinct from system change, change industry/production does not necessitate drastic changes in the economic system at the overall level of ownership and societal planning. Our dependent variables focus on the distinction between variables indicating individual action-oriented frames of the climate issue ( change individual behavior ), system-oriented perspectives including those indicating notions of global justice ( system change, and global justice ), and the framing of solutions in terms of legislation and policy change (which is perhaps closest to the type of outcome one might have expected from COP-15). To operationalize a system-critical global justice perspective, we created the combined variable system change and/or global justice. Operationalization of the independent variables Organizational affiliation was operationalized by responses to a battery of questions about whether the respondent had been an active member of different types of organizations during the previous 12 months. The item measuring membership in political organizations was disaggregated since it is reasonable to assume that membership in green or red (Socialist/Social Democrat, Left-Socialist, and Communist) parties and organizations has a different impact on framing the climate issue than membership of liberal or right-wing parties or organizations. As a drastically simplified indicator of individuals ideological orientation, we used respondents self-placement on an 11-point scale from left (0) to right (10). We used dummies representing far left (0 2 on this scale), center-left (3 5), and right (6 10). Affiliation 12

13 with the GJM was measured by asking whether respondents consider themselves to be part of the global justice movement. Data analysis In the descriptive analysis, we cluster the data on the basis of the three demonstrations. To test our hypotheses we used the standard statistical software package SPSS to conduct two separate binary logistic regressions for each of the three dependent variables. For each dependent variable, the protest locations were included as dummies in the first model, as well as a dummy for non-domestic demonstrators and a number of individual control variables. In the second model, we also included a number of variables related to hypotheses 3-5 to see to what extent the impact of the local mobilization context and being a non-domestic demonstrator was reduced. Such reductions, in turn, would arguably indicate that differences in framing between the three demonstrations could be in part attributable to differences with respect to the introduced variables. 32 Results Descriptives independent variables The socio-demographic characteristics of participants in the three demonstrations are consistent with previous research about who participates both in demonstrations 33 and in the new social movements from the 1960s through the recent mobilizations of the GJM. 34 Participants were quite equally divided between men and women, were mostly natives of their countries of residence, and 70 percent were either students or had previously studied at university (Table 2). Three-quarters of protesters identified themselves as middle class; since so few identified themselves as either upper class or lower class, in the regressions we collapsed these in the dummy other class. There were nevertheless some differences between participants in the three demonstrations: protesters in Copenhagen were significantly younger; fewer Brussels protesters had a university education; London protesters were more highly educated and included a higher proportion of women. 32. Aneshensel Van Aelst and Walgrave E.g., Wennerhag 2008,

14 Table 2. Socio-demographic characteristics of climate summit protesters (percent and mean values) Brussels Copenhagen London Total Cramer s V / Eta Gender Women *** Age Mean age *** Median age *** Ethnicity Born in country other than country of residence Education University degree / study at university * *** Subjective class Middle class ** Working class * Lower class n.s. Upper class n.s. None n.s. Count Comment: Measure of significance for gender, ethnicity, education, and subjective class: Cramer s V; for age we used Eta. * = 5%, ** = 1%, and *** = 0.1% significance. n.s. = not significant. When we consider the demonstrators activism in different types of SMOs during the year preceding the event (Table 3), one in four was active in an environmental organization, while between 10 and 15 percent respectively were active in a political party, church or religious organization, humanitarian or charitable organization, or an organization working on global justice, Third World, or peace issues. On average, 9 percent of protesters were active in a trade union. Overall, the single largest category of organization in which protesters were 14

15 active was environmental SMOs, but, overall, no single type of SMO accounted for more than one-fourth of protesters. Thus, these three demonstrations shared the pluralism that has often been ascribed to earlier mobilizations of the GJM della Porta 2007; Sörbom and Wennerhag

16 Table 3. Organizational/demonstration experience and political orientation of climate summit protesters (percent and mean values) Brussels Copenhagen London Total Cramer s V / Eta Active in organization (during last 12 months) Environmental organization * Green party ** Red party *** Global justice, Third world or peace organization ** Church or religious organization *** Charity or humanitarian organization *** Trade union ** Not active in any organizations (of the types *** mentioned above) Not member in any organizations (of those mentioned above) n.s. Global Justice Movement identification Identify with the GJM * Do not identify *** Left Right self-placement Mean (Left=0, Right=10) * Count Comment: Measures of significance for organizational affiliation, demonstration experience, and identification with the GJM: Cramer s V; for Left-Right placement Eta was used. * = 5%, ** = 1%, and *** = 0.1% significance. n.s. = not significant. There are some notable differences between the patterns of organizational activities of participants in each of the three demonstrations. Most conspicuous is the much greater participation in church / religious groups of London demonstrators, who were also significantly more likely to have been active in charitable, humanitarian, global justice, Third World or peace organizations, somewhat less likely to have been active in a trade union, and markedly less likely to have been active in no organizations at all. Thus, to a great extent, participants in the London demonstration appear to have been activists affiliated to the organizations that contributed to the Stop Climate Chaos coalition, the principal organizer of 16

17 the event, and probably influenced by the earlier experience of the Make Poverty History coalition, 36 which in 2005 drew together many of the same strands to demand action on Third World poverty. The Brussels demonstrators, by contrast, were much more likely to be active in a trade union or Green party, which reflects the specificity of the Belgian mobilization, where the socialist trade union ABVV (Algemeen Belgisch Vakverbond) had a prominent role, and a mobilizing context in which Green parties were relatively prominent and where only two weeks previously the Flemish government conceded to Greenpeace's campaign and promised to build no new coal-burning power stations. 37 We also asked whether respondents were members of (as distinct from being active in) any of the organizations staging the demonstration. The majority (53 percent) were members of one or more relevant organizations, but the patterns of membership varied between demonstrations. The organizations of which more than 5 percent of Brussels demonstrators were members were Greenpeace (25 percent), Natuurpunt/Natagora (15 percent), the Christian trade union ACV-CSC (11 percent), WWF (7 percent), Oxfam (7 percent), and the green party Ecolo (6 percent). The corresponding figures for Copenhagen were Greenpeace (7 percent), Friends of the Earth (6 percent), and the Danish red-green party Enhedslisten (5 percent). For the London demonstration, the corresponding figures were Co-op (16 percent), Friends of the Earth (16 percent), Christian Aid (11 percent), Oxfam (9 percent), the Catholic development aid organization CAFOD (7 percent), the Royal Society for the Protection of Birds (7 percent), and Greenpeace (5 percent). Although it was, in all three demonstrations, environmental groups in which participants were most likely to be active, almost half of demonstrators in both Copenhagen and Brussels were not active in any organization. Thus, even though the climate marches organized around the Copenhagen summit can be considered part of the same transnational mobilization, different constellations of SMOs, their activists and supporters were mobilized in each city. Asked whether they considered themselves to be part of the global justice movement, overall, 47.1 percent of participants in the three demonstrations said yes, while only 22.5 percent said no, with 30.4 percent answering not sure/don t know. However, although organization members more often identified with the GJM, there is considerable variation by 36. Rootes and Saunders accessed

18 type of SMO. Of those active in an environmental organization, 58.1 percent considered themselves part of the GJM. Among active members of other types of organizations, identifiers varied between 54.4 percent (green parties) and 80.0 percent (global justice, third world or peace organizations). Of the large minority of protesters (41.1 percent) who said that they had not been active in any organization during the previous twelve months, only 33.5 percent identified with the GJM. If the London marchers were more likely than those in Copenhagen or Brussels to identify with the GJM, it appears to be because in London a higher proportion of marchers had been active in organizations closely associated with that movement. In line with previous evidence about the political composition of the GJM, 38 most participants (82.2 percent overall) in the three climate marches placed themselves on the left of the political spectrum, and only 5.5 percent placed themselves right of center. There was remarkably little difference between demonstrations; Copenhagen demonstrators placed themselves only slightly further to the left. Descriptives the framings of climate protesters The solution to the climate crisis most frequently mentioned by demonstrators in Brussels was changing individual attitudes and behavior (Table 4). Copenhagen and London demonstrators, however, rated legislation and policy change highest, followed by changing individual behavior. A significantly higher proportion of Copenhagen protesters than in Brussels or London mentioned global justice, system change and technological change / investment frames. 38. della Porta 2007; Wennerhag

19 Table 4. Climate summit protesters framing of how to solve or counteract the climate crisis (percent) Brussels Copenhagen London Total Cramer s V Change individual behavior *** Legislation / policy change n.s. Change industry / production ** Technological change / investments *** System change *** Global justice *** Count Comment: Measure of significance: Cramer s V. * = 5%, ** = 1%, and *** = 0.1% significance. n.s. = not significant. Explaining climate protesters framing The regressions for each of the three dependent variables (Tables 5 7) were conducted in two steps, first including only dummies for location (with Brussels as reference category) and the control variables, only in a second step adding the other focal explanatory variables. Brussels demonstrators stand out as significantly more likely to use the frame change behavior than those in London and Copenhagen (Table 5). This is explained neither by demonstrators active membership in different categories of organizations, nor by their identification with the GJM, and only marginally by their left-right self-placement. Among the individual control variables, both gender and age have a significant effect; women and people aged were more likely to use the change behavior frame, as were those who admitted to no class identification. 19

20 Table 5. Binary logistic regression of predictors for use of the change behavior frame. Model 1A Model 1B Control variables and location Accumulated model Variable B S.E. Exp(B) B S.E. Exp(B) Control variables Age (<30 years = ref.) years ** * years Gender: Woman ** *** Non-domestic demonstrator Ethnicity: Born in other country University degree / study at university Subjective class (middle class = ref.) Working class Other class No class identification ** ** Location (Brussels = ref.) Copenhagen * London ** ** Active in organization Environmental org Global justice org Green party Red party Trade union Charity or Humanitarian org Church or religious org Left Right self-placement (far left = ref.) Centre-left ** Right No L R position Identification with GJM Constant * * Observations Nagelkerke s pseudo R p<10%, * p<5%, ** p<1%, and *** p<0.1% Copenhagen protesters were more than twice as likely as those in Brussels to use the global justice and system change frames (Table 6). We attribute the fact that the odds of the Copenhagen protesters are slightly reduced in model 2B, and that the London protesters are no longer significantly different from those in Brussels, to the inclusion of three (at least weakly) significant variables in the model: left-right self placement, which means that part of the difference between Copenhagen protesters and those in other places is the slightly more 20

21 leftist complexion of the former; active membership in organizations, where only active in Red party has a weakly significant effect (p<10 percent); and identification with the GJM. Female demonstrators were significantly less likely than men to want to change the system or mention the importance of global justice in relation to solutions to the climate problem. The opposite is true of non-domestic demonstrators, a correlation which appears only marginally attributable to leftist self-placement and identification with the GJM. 21

22 Table 6. Binary logistic regression of predictors for use of the system change and/or global justice frame. Model 2A Control variables and location Model 2B Accumulated model Variable B S.E. Exp(B) B S.E. Exp(B) Control variables Age (<30 years = ref.) years years Gender: Woman ** ** Non-domestic demonstrator ** * Ethnicity: Born in other country University degree / study at university Subjective class (middle class = ref.) Working class Other class No class identification Location (Brussels = ref.) Copenhagen *** ** London Active in organization Environmental org Global justice org Green party Red party Trade union Charity or Humanitarian org Church or religious org Left Right self-placement (far left = ref.) Centre-left Right * No L R position Identification with GJM * Constant *** *** Observations Nagelkerke s pseudo R p<10%, * p<5%, ** p<1%, and *** p<0.1% The last regression (Table 7), regarding the frame stressing legislation / policy change, shows that left right self-placement matters only marginally. However, London demonstrators were more likely than those in Brussels to frame climate problems as matters that should be addressed by political decisions and agreements. Of individual control variables, age matters most, with demonstrators aged 50 or older less likely than younger demonstrators to frame solutions as a task for politicians. 22

23 Table 7. Binary logistic regression of predictors for use of the legislation / policy change frame. Model 3A Model 3B Control variables and location Accumulated model Variable B S.E. Exp(B) B S.E. Exp(B) Control variables Age (<30 years = ref.) years years *** *** Gender: Woman Non-domestic demonstrator Ethnicity: Born in other country University degree / study at university Subjective class (middle class = ref.) Working class Other class * * No class identification Location (Brussels = ref.) Copenhagen London * Active in organization Environmental org Global justice org Green party Red party Trade union Charity or Humanitarian org Church or religious org Left Right self-placement (far left = ref.) Centre-left Right No L R position Identification with GJM Constant Observations Nagelkerke s pseudo R p<10%, * p<5%, ** p<1%, and *** p<0.1% In sum, the regressions provide general support for our first hypothesis that the local mobilizing context matters for collective action framing. Most prominently, the Copenhagen context has an independent positive effect on demonstrators use of global justice or system change in their prognostic framing. Our second hypothesis is also confirmed: demonstrators who traveled from abroad to protest were more likely to use system critical/global justice frames, even when controlling for all the variables in the model. We 23

24 found little support for hypothesis 3 that being active in different types of organizations has an independent effect on collective action framing. However, in accordance with our fourth hypothesis, left-right self-placement had a significant effect; the main cleavage is between farleft and center-left protesters, the latter being more likely to embrace a change behavior frame and less likely to propose to change the system or promote global justice. Supporting our fifth hypothesis, those identifying with the GJM were more likely to employ the system change / global justice frame. Discussion Dana Fisher suggests that COP-15 marked the emergence of the climate justice movement. 39 Around half the protesters we surveyed considered themselves to be part of the GJM, which confirms that the mobilizing networks of the climate campaigns substantially overlap with those of the GJM. 40 The different types of political demands made by climate protesters include life politics frames stressing the contribution of individual actions to social change, demands for concrete policy changes directed towards existing political institutions, and more radical demands for changes in the economic system or the distribution of wealth and power between the global South and industrialized countries. Even though climate protesters who consider themselves part of the GJM were relatively more likely to frame the climate issue in more radical terms, demanding changes on the system level, others who identified with the GJM saw climate change as something that has to be solved through individual actions and policy changes within existing political institutions. However, on the whole, global justice-related frames were hardly salient. In contrast to what might have been expected of a protest dominated by slogans of global justice and climate justice, relatively few protesters named solutions to climate change that involve system change or global justice (although significantly more did so in Copenhagen than in Brussels or London). Instead, among the protesters, the dominant prognostic framing was that climate change must be dealt with by changing the attitudes and behavior of individual citizens and / or by legislation and / or policy change. Among participants in the 2009 climate demonstrations, there are significant differences in framing between the three protest events. Thus the mobilizing context and embeddedness of 39. Fisher 2010, Reitan and Gibson 2012,

25 the protest event in specific protest networks influences which of the frames present within a transnational movement are locally predominant. In London and Copenhagen, the climate justice frame appears to have been more prominent in the organizers rhetoric than in Brussels, which may have contributed to the less frequent climate justice framings among Brussels demonstrators. Nevertheless, although we cannot rule out the effect of different constellations of organizations in the protest coalitions, the marginal effects of active participation in different types of organizations lend no support to that conjecture. More plausibly, differences may be an artifact of the informal mobilizing networks of the coalitions that staged the demonstrations. The Copenhagen demonstration, held during the conference itself, was a deliberate attempt by an ad hoc organizing committee to mobilize in the streets the largest possible number of those who had come from across the globe to demand a global agreement to effectively address climate change. It attracted a relatively large number of leftist transnational protesters, and the vicinity of the critical community of the alternative climate summit (Klimaforum09) probably increased the salience of system-critical and global justice frames among Copenhagen protesters. The surprise is that it did so only rather modestly. The London demonstration preceded the conference and was an attempt to draw attention to it by a formal campaign coalition formed in 2005 and dominated by well-established national aid, development and humanitarian as well as environmental NGOs. The partners in Stop Climate Chaos, the demonstration s principal organizer and a lineal descendent of Make Poverty History, made considerable efforts to mobilize their members. The Brussels demonstration was organized by established SMOs, NGOs and trade unions, and was legitimized by the participation of members of the royal family; mainstream environmental organizations were significantly more prominent than in London or Copenhagen, which is very likely to have affected the mobilization pattern. Our analysis underlines that rank-and-file participants in a movement do not necessarily frame political solutions to a problem as do its movement intellectuals, or according to the official stances of the SMOs to which they belong. When studying mobilizations claiming to express the will of the many, there is a particular need to illuminate differences between the official framings of SMOs and those of activists, as well as participants who have no 25

26 organizational affiliation. Above all, however, close attention should be paid to the local and recent historical contexts of particular demonstrations, even where they are embedded in transnational mobilizations against a common transnational target, and to the character and organizational efforts of the groups and / or campaign coalitions that organize those demonstrations. 41 On this evidence, announcements of the birth of a transnational climate justice movement appear premature. At best, if it is a movement, it is one better embedded among organizations and movement intellectuals than among rank-and-file activists. 41. Research on environmental protest in western Europe during the 1990s (Rootes 2003) reached similar conclusions. 26

27 References Aneshensel, Carol S Theory-Based Data Analysis for the Social Sciences. Thousand Oaks: Pine Forge Press. Bédoyan, Isabelle, Peter Van Aelst, and Stefaan Walgrave (2004) Limitations and Possibilities of Transnational Mobilization: The Case of EU-summit Protesters in Brussels, Mobilization 9(1): Benford, Robert D. (1993) Frame Disputes within the Nuclear Disarmament Movement. Social Forces 71(3): Benford, Robert D., and David A. Snow Framing Processes and Social Movements: An Overview and Assessment. Annual Review of Sociology 26: Brulle, Robert J., and Robert D. Benford From Game Protection to Wildlife Management: Frame Shifts, Organizational Development, and Field Practices. Rural Sociology 77(1): Bäckstrand, Karin, and Eva Löfbrand Climate Governance Beyond 2012: Competing Discourses of Green Governmentality, Ecological Modernization and Civic Environmentalism. In The Social Construction of Climate Change: Power, Knowledge, Norms, Discourses, edited by Mary E Pettenger, Aldershot: Ashgate. Capek, Stella M The Environmental Justice Frame: A Conceptual Discussion and an Application. Social Problems 40(1): Chatterton, Paul, David Featherstone, and Paul Routledge. Forthcoming. Articulating Climate Justice in Copenhagen: Antagonism, the Commons and Solidarity. Antipode DOI: /j x. della Porta, Donatella The Global Justice Movement: Cross-National and Transnational Perspectives. Boulder: Paradigm Publishers. Diani, Mario The Concept of Social Movement. The Sociological Review 40(1): Fisher, Dana R COP-15 in Copenhagen: How the Merging of Movements Left Civil Society Out in the Cold. Global Environmental Politics 10(2): Hadden, Jennifer Divided Advocacy Networks: Conflict and Competition in Global Climate Change Politics. Working paper, University of Maryland, MD. Johnston, Hank Verification and Proof in Frame and Discourse Analysis. In Methods of Social Movement Research, edited by Bert Klandermans and Suzanne Staggenborg, Minneapolis: University of Minnesota Press. 27