Framing Global Climate Change in Newspapers, 2000-2015: A Five Nation Study


 Australia, New Zealand, India, United Kingdom, and United States newspapers from 2000-2015 were analyzed for global climate change (GCC) news framing. The combination of the time range, cross-national approach, and detail on framing is rare in research on GCC communication. This article reports on diversity across news frames, attends to some cross-national differences, and resituates the question of journalistic balancing in GCC reporting. Political framing along with environmental effects and science frames were dominant both cross-nationally and over time. Less prevalent were frames that more directly linked GCC to the economy, causal factors, and the lifeworld. Habermas’ legitimation crisis theory is used to bring attention to the significance of observable imbalance in reporting. Newspaper accounts situated GCC more as a crisis of the political sphere while deemphasizing economic or lifestyle causes and their significance for the private or public lifeworld. The findings are intended to inform social science studies of GCC as well as, potentially, informing journalistic practice in GCC and other environmental reporting.


Introduction 1
Media coverage of global climate change (GCC) was a heated topic through the 2 first two decades of the 2000s among academics, media watchers, journalists, and publics 3 with labels such as "inept", "tremendously challenging", "enjoyable", and "dangerous" 4 among others. For some media watchers: "[I]n New Zealand we just see lots of inept 5 media, and issues that complex just getting brushed to the side because no one 6 understands how to cover them properly…. There needs to be that climategate hook or 7 some personality conflict, drama… to make a climate change story work in New 8 Zealand" (Personal Interview, New Zealand Science Media Centre staff). For journalists: 9 "It's tremendously more challenging. I would put it at the absolute head of the list of 10 difficulty for journalists….
[T]his is on the short list of the hardest to write about 11 accurately and engagingly" (Personal Interview, US Journalist). For other journalists: "I 12 enjoy it the most because there are so many dimensions to it…. So you're talking about 13 energy technology, international law, high physics, just everything. It is a great topic" 14 (Personal Interview, UK Journalist). Some journalists faced other challenges. "…[I]t's 15 probably nastier. It's brought in a nastiness that I've never seen before covering politics. 16 There's a really well-organized anti-climate change lobby out there and they're quite 17 sinister and they all hook up on the internet and there are some real freaks out there… the 18 hate mail is worse than I've ever gotten. And there are academics in this country that get 19 death threats…. [People are] bombarding me with email and letters and stuff. It can get 20 quite nasty." (Personal Interview, Australian Journalist). Consistent with the Media 21 Centre quote above, academic studies of GCC journalism sometimes explicitly or 22 implicitly judged it based on its (in)accurate representation of "the science". Journalists 23 were either doing a good job of communicating the threat because they understood "the 24 science", or a poor job, perhaps influenced by or overly attending to conservative 25 commentators, fossil fuel interests, or skeptics. 26 Research reported here starts from a different standpoint, captured in these two 27 simple points: (1) GCC is complex and heterogeneous. Diversity in coverage is 28 necessary and should be expected.
(2) GCC action asks much of publics and their 29 political representatives, thus resistance should be anticipated. Under these conditions, as 30 Hulme (2009)  adopted Schnaiberg's (1980) "treadmill of production" theory as a starting point. The 44 basic idea is that acceleration of economic growth after World War II led to expanding 45 use of natural resources and generation of waste. Unless forced otherwise, capitalism 46 treated environmental fallout as an externality in economic decision-making in favor of 47 profits and competitive advantages. The state largely, but not exclusively, supported this 48 treadmill, assuming it generated economic stability. A treadmill of production was 49 accompanied by a complementary "treadmill of consumption" built over time through 50 advertising and marketing, improvements in the means of consumption, credit 51 accessibility, defensive consumption for social positioning (Curran 2017), and so on. 52 Labor, as well, adopted the growth model under the cultural influence of a work-and-53 spend cycle (Schor 1991) that equated capitalist growth with job security and 54 environmental regulation with job loss. 55 Within this research program capitalist growth is largely to blame for 56 environmental crises. Little can counter the structural forces of capitalism other than 57 environmental regulation forced through social movement activity. The media receives 58 minimal attention, other than how it is controlled by economic interests, perpetuates 59 capitalist ideology, and stimulates consumerism. Though focusing on economic rather 60 than environmental crises, Jürgen Habermas provides additional insight. For Habermas 61 ( 1975 [1973]), capitalist economic crises identified by Marx are conceptualized as system 62 crises where fewer possibilities for problem solving exist than are necessary for economic 63 stability (pg. 2). Crises in "liberal capitalism" of the 19 th century appeared as economic 64 steering problems. If unresolved they migrated to the public sphere where legitimacy of 65 the economic system may be questioned. In "organized capitalism" of the 20 th century, 66 the state intervened in the economy to handle unresolved economic crises, preventing 67 social upheaval. Economic crises were transferred to the state and citizens looked there 68 for solutions. If the state failed to meet expectations, legitimation crises of the state 69 emerged where authority and validity were challenged (Habermas 1975(Habermas [1973 The journalistic balancing norm --where views of mainstream and skeptic GCC 131 scientists are elicited for a "balanced" accountis critiqued in this literature for its 132 alleged undo attention to a minority of skeptics and generation of controversy. Boundary 133 work sometimes pervades the critique, where skeptics are situated outside a legitimate 134 GCC scientific community by emphasizing their insufficient credentials or representing 135 them as ideologically driven by fossil fuel interests (e.g., Oreskes and Conway 2010).

136
Even journalists from the elite press have acknowledged the balancing problem from 137 earlier GCC reporting and mostly discontinued it despite the marketability of controversy 138 (Boykoff 2007; Schmid-Petri, Adam, Schmucki, & Häussler 2017). 139 However, GCC balancing might be conceptualized differently: as attending to and 140 balancing across a range of relevant dimensions, such as those depicted by the news 141 frames elaborated below. For the media to effectively serve as a 4 th estate, good balance 142 is necessary, enabling publics to see GCC as a multi-faceted set of problems and 143 opportunities. This variation would not just align to different publics and stakeholders, 144 but also a public to keep it broadly informed. Diversity across the range of issues moves 145 closer to, though inevitably short of, Habermas' goal of generating a public sphere. 146 This conceptualization leads to a hypothesis that the range of coverage and 147 balance increased over time. GCC infiltrates many social spaces including science, 148 environment, politics and policy, formal and informal education, social inequality, and so 149 on. As GCC evolved as a news issue, so might the range of stories as journalists became 150 more informed about novel aspects (as reflected in the UK journalist's interview account 151 above). The relevance of GCC across these social spaces is not linear, thus one would 152 not expect media coverage to have proceeded linearly. GCC scientific research and 153 political deliberations are as relevant to the public interest today as they were 20 years 154 ago, as are other dimensions like social inequality and public knowledge.

155
A second hypothesis is to expect variation in media coverage across nations as 156 they grappled with different environmental impacts, global political positions, scientific 157 contributions, adaptive requirements, and publics. Much overlap is expected as well, as 158 some issues were more globally conventional and wire services reached more than one 159 nation. But variation might be anticipated, for example, between India and the U.S. in 160 their focus on social inequality or technological development, or between Australia and 161 New Zealand in their attention to different causal factors. Media may have located a 162 GCC crisis in different places across nations based on national contributions to 163 greenhouse gas emissions, political efforts to avert or adapt, and impacts on or 164 engagement of its residents. 165 The basic research questions then are whether the media delivered balance across 166 a range of news frames over time and across nations, yet also exhibited some cross-167 national variation to account for nation-specific concerns. Or was there a detectable 168 imbalance in media coverage such that a GCC crisis was locatable within a particular 169 sphere such as the economic or political system or the lifeworld? To address these Newspaper articles were used due to accessibility and comparability over 16 years 177 across five nations. Headline search terms "global warming" or "climate change" were 178 used on LexisNexis (now Nexis Uni) to identify GCC articles in ten national newspapers 179 ("elite press"), two from each nation. Factiva filled small gaps in coverage. Headline-180 only searches inevitably excluded some GCC relevant articles, but effectively reduced the 181 sample to what readers would more readily identify as GCC relevant. Additional 182 sampling (e.g., one of every three chronological articles) was used for newspapers like 183 The Guardian where population size was overwhelming. Across the ten newspapers, 184 approximately 3500 articles were read by the author. Analysis was interpretive but 185 performed with care and consistency. Nonetheless, percentages below might vary 186 slightly across researchers, so very small differences across frames, nations, and time 187 should not be overemphasized. Using both Goffman (1974) and Entman (1993), frames 188 were identified by how articles were structured to promote interpretations/dimensions of 189 GCC for accomplishing purposes of defining problems, identifying causes, making moral 190 judgments, or suggesting remedies (see also Lindström & Marais 2012, Tuchman 1978. 191 Due to the large sample size, which prevented extensive rereading and reanalysis 192 that inductive approaches require, a mostly deductive approach was used starting with a 193 typology of common GCC news frames and "subframes" (subsets of more general news education. While not the central focus of this paper, "subframes" added more granularity 198 (e.g., the economic costs of GCC, economic costs of GCC policies, and economic 199 opportunities of mitigation or adaptation as subframes of the economics news frame). 200 Table 1a lists the basic news frames used in the analysis; table 1b lists frames and 201 subframes. Other framing typologies were certainly relevant, but this typology permitted 202 a generalizable analysis across 16 years and five nations. 203 The frame, rather than the article, was the unit of analysis. Articles could contain 204 more than one frame. In most cases, subframes rather than the more general frame were 205 coded (e.g., an article framing GCC as an issue of international political deliberation was 206 coded there rather than the "political" news frame). Each of these subframe codes could 207 later be collapsed into the larger general frame, which was done for this paper. 208 Newspaper content data were supplemented by personal interviews with journalists and 209 other media watchers in all five nations for further information and triangulation. 210 For example, a 1 April 2014 article in The Guardian titled "Frame climate change 211 as a food issue, experts say" used an IPCC report and interviews with World Bank and 212 Oxfam experts to emphasize how impacts of GCC on food production and talk about the 213 phenomenon might change people's beliefs and behavior regarding GCC (Goldenberg 214 2014). Two subframes were central to the article: "environmental impacts on 215 agriculture" and "public beliefs and behavior". Each was coded and then, for this paper, 216 collapsed into two general news frames of environmental effects and public 217 understanding, knowledge, and education. 218 Table 2 lists newspapers used for this paper. Some mix across ideological divides 219 and between mainstream and financial newspapers was included to detect any significant 220 framing differences, though this was not of central interest in the project. For example, in 221 Australia the politically liberal-leaning Sydney Morning Herald and Rupert Murdoch's 222 News Corp The Australian were used. In the U.K., The Guardian and the Financial 223 Times were used. Tables 3a-e depict the relative representation of news frames during three time-227 periods (2000-05, 2006-10, and 2011-15) in each of the ten newspapers. Data were 228 gathered by year but collapsed into 5-6 year intervals to peripheralize effects of any one 229 national or global event and to clarify change. Table 3 entries indicate the percentage  230 presence of each frame relative to the total number of detected frames within that 231 newspaper and time-period. In other words, percentages represent the relative emphasis 232 on each frame within overall GCC coverage. Focusing on relative representation rather 233 than total representation, in my view, better captured the likely impact on readership.

234
GCC competes with other newsworthy items for the limited "carrying capacity" of public 235 attention (Hilgartner & Bosk 1988). To the extent that publics read GCC articles, they 236 would have attended to the relative emphasis on one or another frame, rather than the 237 total amount of frame coverage a significant portion of which would have been ignored in 238 favor of other newsworthy issues.

239
Tables 3a-e show diverse frames in each newspaper across the three time-periods, 240 indicating that journalists addressed a range of GCC dimensions expected from effective 241 media coverage. Separate sub-frame data (not included here) accentuates that range. 242 That said, the political frame stands out as the most dominant across time-periods, 243 newspapers, and nations, especially in the US and UK. This result is consistent with 244 research showing overall coverage mirroring major national and international political 245 meetings (e.g., Boykoff 2011, Schmidt et al. 2013) and likely reflects the popularity of 246 policy topics. But its broader significance should not be ignored.

247
With a couple of exceptions (e.g., Financial Times) both the environmental 248 effects and science frames are next, alternating in dominance across newspapers and 249 nations. These two frames have a symbiotic relationship since scientific research is often 250 tapped to identify or explain environmental impacts. Methodologically, if an article 251 focused primarily on an environmental effect, it was coded there. If it focused on 252 scientific research or scientists, with environmental effects casually mentioned as an 253 outcome, it was coded as a science frame. Many articles included both frames.

254
Despite the centrality of economics in environmental decision-making, this frame 255 was less dominant than the above three. The Financial Times aside, economic framing 256 consisted of roughly 10% or less of total frames across the three time periods. An more. Subframe data indicate that more attention was given to "policy impacts on the 265 economy" in The Australian than the Sydney Morning Herald, but also on "economic 266 opportunities of mitigation policies" which is often found in environmentalists' 267 discourse. Economic framing was very limited in the two India newspapers. 268 Socio-cultural framings, captured in the analysis by the public 269 understanding/knowledge/education and social inequality frames, were less prevalent. Hindustan Times often depicted GCC educational events at a university or secondary 273 school in short, announcement type articles. The Guardian attended to public 274 knowledge, beliefs, and values and was more attuned to social science research on the 275 topic. The social inequality frame received less attention, even in the two India 276 newspapers where higher levels were expected. 277 Technological framings were few and focused on mitigation rather than 278 adaptation. This frame was coded if articles highlighted any technological intervention 279 whether large, complex, and system wide or smaller for home or personal use. The dominance of political framing should not be taken for granted, as it often is 295 under a linear science-to-policy progression. Some journalists shared this view. A New 296 Zealand journalist explained in a personal interview: "…there's a change that starts with 297 reality [and]… the state of the science…. It is public and the level of public 298 understanding of the science. Then there's flowing through to the political will to act, 299 then the actual policy outcomes of that debate and finally the impact that that has on the 300 economy and business and incentives." Other journalists indicated their newspapers 301 covered the gamut of GCC issues, but policy issues are "where the debate is," or "the 302 politics get the most coverage because that's where it starts and finishes" (Personal 303 Interview, Australian Journalist). Scholarly research on media representations also might 304 have been affected by this assumption as it gravitated to reasons for political inaction 305 (ineffective communication, fossil fuel industry manipulation, etc.) rather than focusing 306 on its construction as political. 307 An alternative assumption, however, is that GCC is at all times about science, 308 environment, politics, economics, social inequality, public values, and so on. Imbalance 309 across those frames in media accounts, then, raises questions and holds consequences. 310 For instance, the dominance of political framing was congruent with and likely partially 311 responsible for survey data indicating that the public viewed GCC as a problem for the 312 state, but simultaneously didn't trust their government's efforts nor wish to absorb costs 313 (Pidgeon 2012). This public positioning of the problem as political also created a 314 quandary for politicians, many desiring to pass strict regulations but recognizing the 315 likelihood of political pushback. For example, Australian Prime Ministers Kevin Rudd 316 and Julia Gillard paid this price following their government's imposition of a carbon tax. 317 The combination of economic impacts and causal factors framing was much less 318 prevalent than political framing, placing GCC mostly in the realm of the political system 319 rather than an outcome of economic activity. 320 From a theoretical standpoint, political frame dominance and limited attention to 321 causal factors and economy indicate that GCC was represented mostly as a crisis of state. 322 What could easily have been represented as an outcome of capitalism and its 323 externalization of environmental damage, or population growth, instead was transferred 324 to the state where policy actors appeared accountable. Explanations for this centering 325 cannot be placed in journalists' assumptions that publics were already well-informed 326 about causes and consequences due to clear survey data indicating the opposite. Also, 327 causes and economic impacts are complex, varied, newsworthy, and not at all obvious. 328 For example, when journalists did attend to causal factors, there was significant cross-329 national difference across subframes that reflected nation-specific differences. For 330 example, agriculture as a cause received attention in methane emitting New Zealand 331 while largely absent elsewhere. 332 Empirical research on the politicization of GCC and the media's attention to 333 political divergence, while accurate especially in the US, largely ignores the prior 334 questions of why and how it was represented as mostly political. Partly due to linear 335 science-to-politics thinking and partly due to journalists' interest in politics, GCC became 336 primarily a political problem. Once there, some mainstream journalists felt pressure to 337 adhere to the scientific consensus construction to avoid being perceived as supportive of 338 sceptics. In a personal interview, one young New Zealand journalist described the fear of 339 getting something scientifically wrong because of the angry letters it would yield. This 340 stance ironically may have moved GCC further away from the normal debate and 341 disagreement central to scientists' work. In the political sphere, as would be predicted, 342 GCC was subjected to the typical political ideological and power wielding maneuvering 343 that other issues with similar socio-cultural and economic implications experience. 344 The high percentages of science and environmental effects frames are consistent 345 with assumptions that science is the main source of GCC information and environmental 346 impacts (not socio-cultural) the main effects. These frames fit the science and 347 environment background of many elite press GCC journalists looking for novel and 348 relevant topics. While a dominant frame across nations, environmental effects subframes 349 varied with nation-specific factors. For example, fire and drought subframes were more 350 prevalent in Australia; agriculture and fisheries were more prevalent in India. There also 351 was cross-national consistency, with extreme weather events forming a common 352 environmental subframe across all five nations. Most science framing focused on new 353 reports or studies with much less attention on scientists and scant attention on science 354 funding and infrastructure (with some variation across nations). This emphasis 355 contrasted with some researchers' and media watchers' claims that scientific controversy 356 was overemphasized. It did receive attention in The Australian after about 2009, but in 357 general, stories were framed around research results and reports rather than the people 358 and resources responsible for them.

359
Economic framing had a significant but not overwhelming presence, even in the 360 US where policymakers placed environmental problems and solutions in cost-benefit 361 frameworks. Subframe data indicate some variability across nations on the important 362 distinction between economic impacts of GCC or costs of GCC policies. The latter 363 assume domestic policy of some sort, so this subframe was more prevalent in Australia 364 and New Zealand where mitigation policies were passed. Another notable point is that 365 economic frames almost exclusively focused on mitigation rather than adaptation. 366 A few exceptions aside, less attention was given to public dimensions and civil 367 society activity than to politics, economy, science, and environmental impacts. Even less 368 attention was given to social inequality, which many social scientists consider key to 369 GCC and GCC policy. Exceptions included The Guardian and Hindustan Times. As 370 predicted, social inequality frames were more prevalent in the two India newspapers, but 371 less than expected given India's global position as a low per capita emitter and 372 vulnerability to impacts. 373 Minimal attention to public and social inequality frames placed GCC largely 374 outside the lifeworld. The GCC crisis remained highly centered in the state without 375 transfer to the lifeworld where public and private blame and responsibility might become 376 more salient. Some newspaper accounts emphasized, and occasionally deplored, the lack 377 of public scientific knowledge or the public's perceptions about GCC and GCC policy, 378 but minimally associated lifeworld activity as cause or solution. At best, the occasional 379 article included a "what you can do" list to reduce emissions. Social scientists also have 380 lamented a lack of public knowledge and concern about GCC. Research emphasizes how 381 communication can become more effective, perhaps by using more emotional language 382 or visual imagery, explaining scientific uncertainty better, providing means for action, 383 and so on. What is largely missing from this research is the prior point that GCC has 384 been minimally framed as a public or private lifeworld problem. Why should publics feel 385 responsible? Due to the lack of lifeworld attention in newspaper accounts, coupled with 386 what GCC solutions will require of us individually and collectively, we should not be 387 surprised with Norgaard's (2006aNorgaard's ( , 2006b) findings of implicatory denial among 388 Norwegians who otherwise tend to be environmentally informed and conscientious. That 389 denial is likely in other nations as well and its causes may extend beyond the 390 overwhelming and numbing nature of the problem. 391 In conclusion, several main points are worthy of further emphasis and 392 development. One is the meaning of media balancing in GCC reporting. Rather than 393 scientific consensus/skeptic balance, a different and in my view more productive 394 approach is locating it across diverse frames, such as those in the typology considered 395 here. This approach is simple but underutilized in the literature, even though it is 396 consistent with some journalists' and editors' thinking about GCC reporting. "If you're 397 trying to judge what we do as a …whole …on… climate change, then I think we… cover 398 the landscape pretty well. If you're judging an individual story, then yes, it's going to 399 have boundaries around it" (Personal Interview, U.S. Journalist). From this standpoint, 400 the data indicate the elite press in these five nations did address the range of frames 401 examined here. All frames appeared from time to time. Friedman (2015) called attention 402 to "environmental mainstreaming" in the media --the tendency to spread environmental 403 topics across politics, business, science, health, lifestyle, and technology sections rather 404 than maintaining a separate beat. This practice has been criticized due to its associated 405 decline in journalists with environmental (and science) specialties. While there is 406 agreement that the latter is a problem, there also may be significant benefits to balancing 407 GCC and other environmental issues across those sections even when non-specialists 408 write the stories. GCC is also about these other topics and need integration with them.

409
The results also indicate imbalance. Political, environmental impacts, science, 410 and economic framings received more attention than causal factors, public 411 knowledge/values/education, social inequality, and technological solutions framings. 412