Global Climate Change, Global Public Sphere?

Global Climate Change, Global Public Sphere

13 Global Climate Change, Global Public Sphere? Media Attention for Climate Change in 27 Countries Ana Ivanova, Andreas Schmidt and Mike S. Schäfer

Societal problems and political competences have transnationalized in recent years (see Knill, 2001; Wessels, 1997), as have technological possibilities and the economic incentives for transnational mass communication (see Löffelholz and Hepp, 2002: 15; Parks and Kumar, 2003). Consequently, the question of a transnationalizing of the public sphere (or spheres) has gained importance in communications (and in political science, e.g. Wessler et al., 2008; Koopmans and Statham, 2010b; Risse, 2010). This question is of particular relevance, in our view, to the case of anthropogenic climate change - a global phenomenon in its causes, effects and discussed solutions (e.g. Beck, 2007: 34). Therefore, we will examine whether a transnational public sphere is emerging around this issue. As '[v]irtually no other approach has the potential to bring communication studies further forward in the age of transnationalization' (Esser, 2013: 113), we will use a comparative approach for our study. Compared to other studies, we will analyze a rather large number of cases, 27 countries from all continents, going beyond the scope of many comparative analyses. 1

Conceptual framework and research question What exactly is a transnational public sphere, and how can it be identified empirically? For these questions no consensual answer exists. On the contrary, the relevant debate in communication sciences is diverse and conceptually broad. This makes it necessary to explicate what we mean by a transnational public sphere and which of its facets we intend to examine. 2 In order to do so, two conceptual specifications are necessary. First, we perceive a public sphere as a specific configuration of social communication. We do not take a normative perspective on public spheres (for an overview see Ferree et al., 2002a). Second, in accordance with the majority of the literature (cf. Tobler, 2010: 59f.; Wessler et al., 2008: 1), we do not 210

211

assume that a transnational public sphere has to be structurally identical to national public spheres. Therefore, it does not have to be a sphere of communication in which a unified, transnational demos (with a shared identity using shared media) controls supranational political institutions. To date, neither such a demos nor the respective mass media exist (see, for example, as early as Gerhards, 1993, 2000). Additionally, support for this understanding of transnational public spheres is rather rare nowadays (see Machill et al., 2006: 61). Instead, and in line with most scholars in this field, we assume that a transnational public sphere will most likely realize itself in the form of transnationalized national public spheres (see Gerhards, 1993: 12). This transnationalization of national public spheres involves different indicators (Wessler et al., 2008: 10): a) the extent to which national public spheres observe supranational governance institutions ('monitoring governance' or 'vertical' transnationalization, see Gerhards, 1993; Koopmans and Statham, 2010a: 36); b) structural similarities in the debates on issues discussed in the national public spheres, including transnationally synchronized attention to those issues ('discourse convergence', see Brüggemann et al., 2009: 400; Eder and Kantner, 2000; Schneider, 2006: 140); c) the exchange of topics and positions across borders ('discursive integration', see Brüggemann et al., 2009: 400; Risse, 2010: 157; Wimmel, 2004: esp. 11); d) the identification with a transnational community ('collective identification', see Kanter, 2006, 2009). In the case of this article, we focus on just one of these aspects: the transnational 'discourse convergence' (Wessler et al., 2008: 15). Specifically, we examine whether attention to a topic is similar in different national public spheres and whether it converges over time. We will use the amount of attention different mass media give to the issue over time as a measure for attention that the issue receives in the respective public spheres in general, because mass media are the 'master forum' of the public sphere (Ferree et al., 2002b: 10; see also Wessler et al., 2008; Koopmans and Statham, 2010b). Measuring issue attention is a simple indicator that is considered a part of, or at least a necessary condition for, transnational public spheres (Risse, 2010: 157; for criticism, see Wimmel, 2004: esp. 11). lt indicates that a topic is seen as relevant in different countries at the same time, and thus represents a first step in international communication. For this reason, as well as for scientific pragmatism, we decided on an analysis of media attention. Our comparative approach focuses on national public spheres as units of analysis - though, as will be shown later, this is only the first level. On the second level we compare transnational public spheres (which represent groups of national public spheres) as units of analysis.

212 Ana Ivanova, Andreas Schmidt and Mike S. Schäfer

Global Climate Change, Global Public Sphere

Furthermore, we decided to focus on a single topic. The reason for this is that transnational public spheres have proved to be segmented, at least in the case of the (potential) European public sphere. Intensive transnationalization emerges not only for specific publics or particular events (for a summary, see Brüggemann et al., 2009), but also for the coverage of specific topics. Irene Neverla assumes that topics relating to the environment or other risk issues are particularly prone to transnationalized public debates because of their immediacy and range (Neverla, 2007). Additionally, the topic of climate change is used by several authors as a textbook case of transnationalization. For instance, Ulrich Beck describes the topic as a catalyst for an already existing 'global public sphere' (Beck, 2008: 81), while Elisabeth Eide et al. (2010), Peter Berglez (2008), and Ulrika Olausson (2009, 2010) all identify specific aspects of transnationalization relating to the topic. Hence, we study to what extent structural similarities in media attention to climate change can be identified in different countries and country groups, and how those similarities develop. Specifically, we ask:

Global public sphere

111

II

1. Does media attention to climate change show transnationally similar

levels and temporal synchronicity? 2. Does media attention in the countries under study converge over time?

1

Our study draws on insights and recent advances both in the field of research into transnational(izing) of public spheres and in comparative communication research in general. Additionally, the study goes beyond the state of the art in these fields on four points: 1. We aim for a large(r)-N comparative analysis: Esser (2013: 114) shows that

comparative communication research has moved from small studies comparing only two - or a limited number of- countries towards large-N, variable-oriented approaches, which allow for a greater degree of universalization of findings. We follow suit in providing an analysis of issue attention for climate change in 27 countries. 2. The geographic scope of our study goes beyond Western countries: In comparative analysis on various media and communication related questions, Western - and particularly North American and Western European countries are the most frequently analyzed. We will follow the request for a 'de-Westernization' in comparative communication research, which can be found in the literature (e.g. Pfetsch and Esser 2012: 39; Hanitzsch and Esser 2012: 512), and also include African, Asian and Latin American countries in our analyses. This is also motivated by the theoretical interest in distinguishing transnationalized public spheres of different scopes (Figure 13.1). Different scopes of transnationalization are conceivable and might involve, at first,

213

Figure 13.1

A selection of the conceptually distinct ranges of transnational public

spheres

only European countries ('European public sphere'); second, the addition of the USA and Canada ('transatlantic public sphere'; Wessler et al., 2008: 9); and third, all Western countries (the 'Western public sphere', which by broad consensus also includes Australia and New Zealand; see Inglehart, 1997; Toynbee, 1966; UNDP, 2007). Fourth, there is the possibility of transnational public spheres beyond the Western world - e.g. the emergence of a Latin American, African, or Asian public sphere (Brüggemann et al., 2009: 395), including sub-continental public spheres, for example in the Near East (e.g. Lynch, 2003). And finally, the emergence of a 'truly global public sphere' is conceivable (Wessler et al., 2008: 10). In contrast to these various conceptual possibilities, however, most studies focus on Western and European countries, especially Western Europe (see Brüggemann et al., 2009; Machill et al., 2006). 3 On the basis of the existing empirical evidence it is therefore not possible to judge whether convergences

Global Climate Change, Global Public Sphere 215

214 Ana Ivanova, Andreas Schmidt and Mike S. Schäfer

indicate a European, transatlantic, Western or global public sphere (see Wessler et al., 2008: 10, 192): a question we also aim to investigate. 3. In addition to an extended geographical scope, we employ a longitudinal perspective: With respect to the time frame of our study, we investigate media attention between 1996 and 2010, thus analyzing an extended period of coverage. This addresses a need of the research field: Esser and Hanitsch (2012: 5; see also Esser 2013: 115) state that a combination of spatial (cross-country) with a longitudinal (cross-temporal) dimension would represent an important advance in comparative communication research that would provide the possibility of tracing the results of cross-country comparisons over time. Furthermore, every analysis of transnationalization - which is a process, after all - demands a longitudinal perspective (Wessler et al., 2008: 9). However, most existing studies only measure similarities in public debates at one point in time (see Brüggemann et al., 2009: 401). For example, some studies are able to show that certain topics or events are discussed at the same time (de Vreese et al., 2001; Norris, 2000: esp. 190; Tobler, 2002; van de Steeg, 2006), but in most cases they cannot deliver more than 'snapshots of a longer process' (Wessler et al., 2008: 9). Nevertheless, it is not clear whether those studies indicate a sustainable transnationalization or transitory phenomena (Wessler et al., 2008: 22). In contrast, longitudinal analysis is rare, and such studies as exist do not paint a coherent picture. Juan Diez Medrano (2003), for example, shows a convergence in media coverage of European integration in Spain, Germany and France. Steffen Schneider reveals 'largely parallel national trajectories' (Schneider, 2006: 140) for the media debate on genetically modified food, without significant convergence. Wessler et al. find no clear trends of convergence in their study of military interventions between 1991 and 2003 (Wessler et al., 2008: 105), while Cathleen Kantner identifies several phases of convergence and divergence for the same topic in a different country sample (Kantner, 2009: 170). Confronted with these limited and partly contradictory findings, we deem it sensible for our work to expand on the descriptive evidence for long-term accounts (see Wessler et al., 2008: 192). 4. We analyze phenomena lying beyond the nation state: As Esser (2013: 118) demonstrates, the transnationalization of communication can be analyzed from a 'glocalization' and a 'deterritorialization' perspective. The former focuses on how transnational phenomena are being dealt with at the national level. The latter questions the assumption that the national level is meaningful for analysis, and emphasizes phenomena beyond the nation state that (may) have emerged due to globalization. We adopt Esser's (2013: 120) suggestion 'to marry the "glocalization" and "deterritorialization" sub-approaches and integrate their logic into comparative designs'. We describe the level of attention to a global issue

of national public spheres and analyze communication phenomena beyond the nation state: transnational public spheres.

Data and methods We selected countries representing geographically different regions as well as Western (i.e. developed) and non-Western countries for analysis. 4 27 countries were selected in total (Table 13.1).

Table 13.1 Overview of analyzed countries and newspapers Lead media index*

Country

Newspaper

Algeria Australia Brazil

EI Watan The Australian Folha de Sao Paulo

4 4 4

Brunei Canada China France Germany India Indonesia Ire land Israel Jordan Malaysia Mexico

Borneo Bulletin Toronto Star People's Daily Le Figaro Süddeutsche Zeitung The Hindu Jakarta Post Irish Times Jerusalem Post The Star New Straits Times Reforma

3 4 4 4 4 4 4 4 4 2 4 4

Namibia Netherlands New Zealand Papua New Guinea Russia Singapore South Africa Spain Thailand UK USA Yemen Total

The Namibian De Volkskrant New Zealand Herald PNG Post Courier Izvestia Straits Times Sunday Times EI Pais Bangkok Post The Times New York Times Yemen Times

Region

N (articles) 549 13,906 3,617

4 4 4 4

Northern Africa Oceania Latin America/ Caribbean South-Eastern Asia Northern America Eastern Asia Western Europe Western Europe Southern Asia South-Eastern Asia Northern Europe Western Asia Western Asia South-Eastern Asia Latin America/ Caribbean Southern Africa Western Europe Oceania Oceania

4 4 3 4 4 4 4 3

Eastern Europe South-Eastern Asia Southern Africa Sothern Europe South-Eastern Asia Northern Europe Northern America Western Asia

496 2,497 273 6,787 1,542 9,946 8,676 112 95,472

590 7,773 2,575 4,112 6,894 5,710 2,492 6,151 742 101 1,757 4,061 801 2,652 4,961 838

Note: *The leading media index was constructed from the sum of four dichotomous variables (ls t~e news~ap~r

(a) national, _(b) ~haracterized by a high circulation, (c) published daily, and (d) of quahty. ). A value of 3 md1cates that not all four criteria are met (Brunei: circulation; South Afnca and Yemen: daily publication; Jordan: both). h1g~

Global C/imate Change, Global Pub/ic Sphere

216 Ana lvanova, Andreas Schmidt and Mike S. Schäfer

We then investigated media attention for climate change by analyzing coverage in leading national newspapers. These are publications whose opinion leadership is societally acknowledged (Wilke, 1999: 302) and based, for example, on their circulation, reputation or guality of journalism. Print media were selected since they offer a simpler means of methodically collecting and analyzing data, with respect to the geographic and temporal reach of the study. However, in some countries, print media may have a smaller influence than, for example, television. Using reputable sources (e.g. Hans-Bredow Institut, 2009), one leading print publication was selected for each country that: (a) is preferably published daily, (b) has a universal and national coverage, (c) has a large circulation, and (d) has high journalistic standards ('guality newspapers'). The basic population of this study was defined as all articles from these newspapers that explicitly mention climate change (even without climate as the main focus of the article). A reference to climate change exists when: (a) the keyword 'climate' appears in connection with words indicating change (i.e. change, development, warming, cooling); (b) the article includes words synonymous with climate change, such as 'greenhouse effect' or 'global warming'; or (c) when a global change of temperature is discussed. These conditions were operationalized by broadly-defined search strings 5 which were subseguently employed for full-text searches in electronic databases. In doing so, we had to restrict our search to coverage from 1996 onwards, i.e. from the year in which most newspapers in guestion were made available electronically. All articles containing positive search hits were downloaded. Non-relevant articles were later eliminated by employing corpus linguistic technigues, checking the sample for duplicate copies of articles, and elaborate relevance checks. After cleaning the data, more than 95,000 articles proved relevant from the 27 countries under study (see Table 13.1). Additionally, we recorded the absolute number of articles published in each newspaper. The numbers of articles referring to climate change were related to the absolute number in order to calculate coverage of climate change as a proportion of the absolute number of articles, by month in percent. This procedure established the 'functional eguivalence' of our data, which is a necessary condition for cross-country as well as longitudinal comparisons (Esser, 2010, 2013; Kolb, 2012).

Transnational public spheres on the topic of climate change? Results Figure 13.2 shows media attention to climate change in all 27 analyzed countries, and displays both similarities and differences. In all countries, coverage of climate change remains at a low level early on, but increases gradually and then more strongly later on (for similar findings regarding some of the countries see M. Boykoff, 2010: 22; Carvalho and Burgess,

_L,--_

_...._.,,,,,,..,

('"""'""""""""""""''

„ ........

00J

„„„„i

""'China

i,

:OlO

M

lm

~cno

J~

JyJ WJQ

t·

1

1

J

i

i~

·~

T. """~ Jordan

I ~J r. 'j"""""

'" Namibia

1 Refonna

1 The Namibian

T

W'~

·i ,„r-R--. I us_sra--~!i '"" 5mgapore . i:vest>Ja , j Strait:oTimc$

S'll. PNG Courier

: );X

t~

l!'.1%

....

20tll"

l "i" Thailand WIQ

·M 2003

* . .._ ..,.,,u

ll!%

.?00.l

2'01'1

; '"" United Kingdom

rrn::s.mpoHa\I

111%

1"

...„Jr\ 1 \

J

lhellmes

1

1•m

(

.

1

t?..16-

,„

S1.nd~v11on1i~

.t.JJ ,,,.L---„-,,""' T

~--- /

l~~ß

NttWY