|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 NEWS MEDIA AND CRISIS LIFE CYCLE IN A TRANSNATIONAL ENVIRONMENT: A CASE STUDY OF THE FOXCONN By WEI LING YANG A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLME NT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2012
2 2012 Wei L ing Yang
3 To my parents and those who inspired my intellectual curiosity, research interests, and supported my determination s t hroughout my lifetime
4 ACKNOWLEDGMENTS I would like to express my thanks and appreciations to the following people, who have given me inspiration s support s and encouragement s They made this milestone of my life possible. Thanks to m y committee chai r, Dr. Juan Carlos Molleda who provided me amazing inspiration, useful knowledge valuable guidance, countless time and efforts throughout this thesis His encouragements gave me great courage and support to fulfill this task. I should also like to thank my two other committee members, Dr. Johanna Cleary and Dr. David Ostroff. Thanks to Johanna Cleary for her kindness, patience, and warmest encouragement for me to start this study. I learned from her the passion in research and attitude for exploring new k nowledge. Thanks to Dr. David Ostroff for his thoughtf ul advice and insightful suggestions. He showed me wider perspectives and greater possibility in developing this research topic. In addition, I would like to give my gratitude to all my professors in th e College of study in Gainesville and Italy. Moreover, I would like to thank Jody Hedge at the Graduate Division for her patience and thoughtful work. Further, I would like to thank Jessica Salter, who dedicated her time providing me suggestions of English. Thank her for her assistance in editing my whole thesis and as a friend to support my life in the United States. A word of gratitude to those who devoted their time and comforted me with their laughs during my hard time of writing thesis : They are too numerous to be named, but they included Kai her Wu, Angel Jang, Yin Hsuen Chen, Crystal Hsieh, Joanna Lin Wei jie Huang and many sweet friends at University of Flo rida I especially thank to my senior alumna, Hsiao Ying Liu, for her assistance and supp ortive instructions through my m
5 Most of all, I would like to thank my dearest father Men Yeh Yang, my mother Li Na Li, without her determination, I wo uld not have chance to study here, my brother Wei Chen Yang and my grandparents They light up my life for their ever lasting love, care and support. Finally I would also like to express my deepest gratitude to Mr. Chien Tsung Chen, for his support to my personal life, his faith on everything I did and thank him for sharing every frustration and happiness with me.
6 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 10 CHAPTER 1 INTRO DUCTION ................................ ................................ ................................ ...................... 12 Crisis Management in Transnational Environment ................................ ................................ 12 The 2010 Foxconn Employee Suicide Crisis ................................ ................................ ........ 14 2 LITERATURE REVIEW ................................ ................................ ................................ ........... 19 Cross National Conflict Shifting ................................ ................................ ............................ 19 Media Agenda Setting and Cr isis Communication ................................ ................................ 23 International Crisis Communication and Media ................................ ................................ ..... 26 Crisis communication: A Focusing on Crisis Life Cycle ................................ ....................... 31 World Media System ................................ ................................ ................................ .............. 34 Case Background ................................ ................................ ................................ .................... 40 Foxconn Employee Mistre atment Controversy ................................ ............................. 40 Evolution of the Foxconn Crisis ................................ ................................ .................... 41 Research Questions and Hypotheses ................................ ................................ ...................... 43 3 METHODOLOGY ................................ ................................ ................................ ..................... 48 Define Foxconn Crisis Stages ................................ ................................ ............................... 48 Sampling ................................ ................................ ................................ ................................ 49 Sample Selection ................................ ................................ ................................ ............. 49 Sample Profile ................................ ................................ ................................ ................. 50 Coding Sheet Construction ................................ ................................ ................................ ..... 52 Pretest and Inter coder Reliability ................................ ................................ .......................... 53 Data Analysis ................................ ................................ ................................ .......................... 54 4 FINDINGS ................................ ................................ ................................ ................................ .. 57 Research Question 1a ................................ ................................ ................................ ............. 57 Phase One: Signal Detection (Mitroff, 1996) ................................ ................................ .. 57 Phase T wo: P reparat ion and P revention ................................ ................................ .......... 58 Phase T hree: A cute and C ontainment ................................ ................................ ............. 58 Phase F our: L earning, D amage C ontainment, and R ecovery ................................ .......... 59
7 Phase F ive: R esolution, R ecovery, and L earning ................................ ............................ 60 Hypothesis 1a ................................ ................................ ................................ .......................... 60 Hypothe sis 1b: ................................ ................................ ................................ ........................ 61 Hypothesis 1c ................................ ................................ ................................ .......................... 62 Research Question 1b ................................ ................................ ................................ ............. 62 Length of N e ws A rticles ................................ ................................ ................................ .. 62 Number of N ews A rticles ................................ ................................ ................................ 63 Hypothesis 2a ................................ ................................ ................................ .......................... 63 Hypot hesis 2b ................................ ................................ ................................ ......................... 64 Hypothesis 2c ................................ ................................ ................................ .......................... 64 Research Question 2a ................................ ................................ ................................ ............. 65 Research Qu estion 2b ................................ ................................ ................................ ............. 66 Hypothesis 3 ................................ ................................ ................................ ........................... 67 Research Question 3 ................................ ................................ ................................ ............... 68 Research Que stion 4 ................................ ................................ ................................ ............... 68 Hypothesis 4 ................................ ................................ ................................ ........................... 69 Post Hoc Hypothesis ................................ ................................ ................................ .............. 70 5 DISCUSSION ................................ ................................ ................................ ............................. 79 Discussion and Implications of the Media Model for the Crisis Life Cycle .......................... 79 Discussion and Implications in Practices ................................ ................................ ............... 83 News Frequencies ................................ ................................ ................................ ............ 84 Themes Covered ................................ ................................ ................................ .............. 86 Frames Used ................................ ................................ ................................ .................... 87 Sources Cited ................................ ................................ ................................ ................... 88 Limitation and Future Research ................................ ................................ .............................. 89 APPENDIX CODING SHEET ................................ ................................ ................................ ...... 92 LIST OF REFERENCES ................................ ................................ ................................ ............... 95 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 101
8 LIST OF TABLES Table page 2 1 Timeline: the employee suicide crisis at Foxconn in 2010 ................................ ................ 47 3 1 The explanation of statistical tests for each hypothesis and research question .................... 55 4 1 Theme frequencies in each crisis stage ................................ ................................ ................. 71 4 2 Cross tabulation of news agencies and crisis stages ................................ ............................ 72 4 3 Themes covered in each news agency ................................ ................................ .................. 72 4 4 Correlations between news agency and number of themes ................................ .................. 73 4 5 Frames used by news agencies ................................ ................................ ............................. 73 4 6 Source frequencies in the news coverage of three news agencies ................................ ........ 73 4 7 Frames used in each crisis stage ................................ ................................ ........................... 74
9 LIST OF FIGURES Figure page 4 1 News frequency of three agencies by crisis st age. ................................ ............................... 75 4 2 The number of themes by stage. ................................ ................................ ........................... 75 4 3 News frequencies of Xinhua by crisis stage. ................................ ................................ ........ 75 4 5 News frequencies of AP by crisis stage ................................ ................................ ............... 76 4 6 The mean of article length of three news agencies ................................ .............................. 76 4 7 The number of article in three news agencies ................................ ................................ ..... 77 4 8 The number of themes covered in news articles from Xinhua ................................ ............ 77 4 9 The number of themes covered in news articles from CAN ................................ ............... 77 4 10 The number of themes covered in news articles from AP ................................ ................... 78 4 11 The number of themes covered in news agencies ................................ .............................. 78 5 1 News f requency of t hree a gencies by s tage ................................ ................................ ......... 91
10 Abstract of Thesis Present ed to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts in Mass Communication NEWS MEDIA AND CRISIS LIFE CYCLE IN A TRANSNATIONAL ENVIRONMENT: A CASE STUDY OF THE F OXCONN By Wei Ling Yang May 2012 Chair: Juan Carlos Molleda Major: Mass Communication Crises are unpredictable events. With the trend of globalization, the development of emergent communication technologies, and the diversity of the w orld media system have posed a challenge to transnational corporations (TNCs) in crisis management. In order to find guidance for this complicated issue, this study examines and analyzes a transnational crisis, the Foxconn ve content analysis of news stories generated from news agencies of three involved countries, this study examines the news coverage about Foxconn employee suicide outbreak, which was between January 23, 2010 and December 26, 2010. With a series of suicide attempts occurred within Foxconn factories, the company faced critics about labor right mistreatment and also affected its clients, such as Apple and Dell In this study, the printed newswire coverage of Foxconn crisis from Chinese Xinhua News Agency, Taiwanese Central News Agency, and Associated Press of the United States are analyzed to identify the characteristics of news content in terms of news frequency, themes covered, frames used, and sources cited. The study found that the news frequency of Foxconn crisis was affected by the crisis stages and the locations of news agencies. News media in China, Taiwan and the United States would cover different themes about Foxconn crisis and used different news frames as well as
11 the news sources. There is a significant difference found between the five crisis stages in terms of the theme number, source number, and the frames used.
12 CHAPTER 1 INTRODUCTION Crisis Management in Transnational Environment Crises are unpredictable events. If poorly managed, they viability, credibility, reputation and relationships with their stakeholders (Howell & Miller, 2006; Wang, 2005). Crisis communication management in the international environment, including prevention and response/reparation, is a complex task in public relations and communication management. Globalization, development of emergent communication technologies, and the diversity of the world media system, have posed a challenge to transnational corporations (TNCs) in managing cris es, issues and challenging situations in multiple home and host locations This study used a quantitative content analysis of news stories about Foxconn suicide outbreak, which was between January 23, 2010 and December 26, 2010, from different media syste ms to demonstrate how crisis shift from one to other locations and how newswire coverage differed through the media systems when covering a same transnational crisis. According to Molleda and Connolly Ahern (2002), transnational corporations (TNCs) decisi ons, actions, and operations influence not only domestic, but also global publics that include people where TNCs are headquartered and centrally operated. The blurred boundary between domestic and international business means crises are no longer restricte d by national borders (Molleda, 2011; Seymour, 2002). In order to study crisis management in an international context, researchers have developed a series of propositions regarding cross national conflict shifting (CNCS) that mainly states that a transnati onal crisis in a country may potentially shift to other places in the world (Kim & Molleda, 2005; Molleda & Quinn, 2004). Assisted by increasingly sophisticated communication technology, international and national news media play important roles in informi ng global and local public issues involving
13 transnational corporations (TNCs) the cross national c onflict shifting (CNCS) theoretical framework, international news media were considered as a variable that could affect transnational crises management. As one of the propositions that Kim and Molleda (2005) proposed, news media of host and home countries will frame the same crisis event differently. Molleda (2011) also presented CNCS propositions that were relevant to the preferences of international newswire coverage in a transnational crisis. These studies revealed that international news media are power ful tools for setting the agenda for publics and framing their opinion since TNCs also use mass media to communicate with their stakeholders during crises through controlled and uncontrolled corporate messages (Howell & Miller, 2006; Kim & Molleda, 2005; M olleda, 2011; Molleda & Connolly Ahern, 2002). emphasized media reports as one of the potentially determining factors in crisis planning and response models. He sug gested that news media in different countries have various highlights of news coverage and knowing these preferences will aid TNCs and public relations practitioners in predicting, anticipating, and addressing unique dimensions of crises subject to interna tional exposure (Freitag, 2002). Thus, this study seeks to identify the key elements that news stories privilege. Moreover, this study also aims to integrate these elements with crisis response strategies and to provide an effective framework for monitorin g and managing crisis communication at every stage in a transnational crisis. Although the preferences of news coverage in particular countries involved in a transnational crisis have become a favorite topic for analysis, there is little research that inte grated crisis response with international news coverage, which is an important step for
14 in China toys found (Molleda, 2011, p. 67). For integrating crisis response and media coverage in their study, Howell and Miller stage crisis life cycle model with news coverage to a framework by analyzing a domestic crisis of Australian based Ansett Airline to better predict the changes and media coverage d uring crisis in order to develop a series of efficient crises communication strategies. They found the different interests of Australian news media in each stage of signal detection, probing, containment, learning, and recovery (Howell & Miller, 2006). How ever, there is still a gap between domestic practice and managing crisis in a global context, which is more dynamic and complicated because of the differences of media structure and function, as well as cultural norms among countries. The purpose of this s tudy, therefore, is to bridge the gap between domestic and international practice as well as combining nature of news coverage with crisis life cycle in a crisis response mode. Also, this study hopes to add useful data of international media coverage in ex isting research by analyzing the news coverage of the 2010 Foxconn employee suicide crisis The 2010 Foxconn Employee Suicide Crisis This crisis started in China by Foxconn a TNC headquartered in Taiwan. It is one of the biggest electronic product manufacturers in China that has large b usinesses with transnational corporations (TNCs) such as Sony Dell Apple and Hewlett Packard Development Company (also known as HP) among others. Similar to other manufacturers who operate
15 their fac tories in China, the main comp eti tive advantage of Foxconn is low labor cost. In m anaging large amount of Chinese employees, Foxconn concentrated and somehow not humanitarian. According to the news of Agence France Press (2011), the company is alleged to treat employees as machines. Foxconn confronted the most terrible crisis regarding employee relations in a host country where it has most of its factories, C hina. The employee mistreatment crisis shifted from the host country, China, to its home country, Taiwan, immediately, and then to the United States since Apple was also involved as Foxconn & Scholars A gainst Corporate Misbehaviour [SACOM](2010), a Hong Kong based non government organization, 17 Foxconn workers attempted suicide from January to August, 2011, which resulted in 13 deaths with 4 injured. Along with the continuous news reports on the employ ee suicides, Foxconn was heavily criticized for its employee management by the news media, a variety of publics in China and Taiwan, activists, and scholars. H owever, Foxconn did not respon d appropriately at the beginning of the crisis. Its CEO Terry Gu o did not choose to face the media. Instead its management team in China put suicide prevention netting around employee dorms and asked employees to sign no suicide pledges. After the multiple suicide events, the Foxconn crisis impacted its clients, as w ell as well known global electronic product corporations such as Apple Sony and Dell The crisis involving employees of the (2010) report, people who buy the prod Apple was most involved in this crisis since it was the major customer. During the crisis life cycle, Apple was often associated with Foxconn in man y news stories since it is more recognizable.
16 After the ninth suicide, Apple suppliers to treat all workers with dignity and resp During the crisis, there were a large number of news stories covering the development of this event in China, Taiwan, and United States. However, the media environment in each country is different. This includes elements s uch as levels of press freedom that might influence news coverage. For media in China, they received instructions from the government to stop reporting Foxconn suicide s after the 12 th employee committed suicide. According to Freedom H ouse (2010), governme level s of press freedom, which means more restriction on news content. According to the 2010 report of Freedom House in press freedom section, Taiwan, China and United States have dif ferent rating s in press freedom, which considers the criteria of (1) legal environment that laws restrict the media, (2) political environment which means degree of political control over the content of news media, and (3) economic environment for the media such as media ownership. mass communication in each political system, news media in China, Taiwan and United States belong to these five concepts: (1) authoritarian, (2) western, (3) communist, (4) revolutionary, and (5) developmental. These differences in media system make it interesting in comparing news coverage from thes e three countries and treat media freedom as a variable in affecting news contents. This case study provides a chance for international public relations research to integrate the field of cross national conflict shifting (CNCS) crisis communication, globa l journalism, and
17 world communication system together. As a transnational crisis in which three countries were involved, Foxconn the researcher to (1) test and revise existing news media trend model of crisis communication in a global set ting. Different media environment in these three countries, China, Taiwan, and United States, provides a good opportunity to (2) examine the differences in news coverage among the three countries that may be affected by their media environment. As Freitag (2002) suggested, future studies should focus on smaller number of nations and take media structure and function into consideration when analyzing news coverage a systems and cultural syndromes might permit even more precise development of matrices to help addition to looking for the nature of international media covera ge during different stages in Apple and Chinese and Taiwanese governments) since the crisis involved many stakeholders. This increases complexity of the associations among the actors. Through this study, scholars and international public relations practitioners are able to better understand evolution of a transnational crisis an d how it interact with news media to guide decision making of crisis response strategies. In the literature review chapter, these purposes will be explained by incorporating theories of cross national conflict shifting (CNCS) crisis communication response agenda setting effects, and press freedom. This research will be accomplished by conducting a content analysis of newswire coverage related to Foxconn crisis in China. These analyses will examine printed news stories of three major international news ag encies in China, Taiwan, and the United States in order to apply the
18 model of crisis communication to a global context. The study is not only expected to shed light on international public relations research, but also expected to suggest a working framewor k for public relations and communication management practitioners in order to bridge the gap of academic research and practical implications.
19 CHAPTER 2 LITERATURE REVIEW Cross National Conflict Shifting As one of the consequences of the globalization tren d, business corporations are involved and affected by transnational incidents or challenging situations (Wakefield, 2001). For further research about these types of transnational incidents, international business scholars Welge and Holtbrgge introduced th e concept of cross national conflict shifting in 1998 and extended this concept later in their book, International Management (as cited in Molleda, 2002). Molleda and Connolly Ahern (2002) later borrowed this cross national conflict shifting concept from a n international strategic and business management discipline, after which they introduced it to the public relations field. According to Molleda and Connolly na they operate and e TNOs, the globalization trend is a double edged sword that enhances their business expansion to the global market while also posing unanticipated and complicated crises challenges. W ith an unparalleled force, the emerging communication technologies, such as satellite and Internet, local issues can easily shift across national borders. Molleda and Laskin (2009) explained, national conflict shifts involve a variety of publics a t various geographical levels, namely, host, home, and transnational publics (e.g., NGOs and activist groups, global media outlets, international news agencies, pan
20 To better understand and illustrate cross national conflict shifting (CNCS) in the global public relations arena, Molleda and Connolly Ahern (2002) conceptualized CNCS by providing a case study of a legal crisis that involving America Online Latin America (AOLA) in Brazil. The AOLA crisis shifted from Bra zil to the United States, and then to Europe, causing financial consequences for the organization in terms of its stock value. They further expanded their conceptualization of CNCS, stating that: There are organizational decisions, actions and operations t hat affect publics in one country and have an impact internationally. This impact seems to be greater at the home country of the organization or organizations involved, which could be explained by the relevance and proximity of organization for the home pu blics. Domestic conflicts are increasingly shifting worldwide because of the growth of international transactions, transportation and communication, especially information technology. (Molleda & Connolly Ahern, 2002, p. 4) Molleda and Quinn (2004) elabora ted this concept of CNCS with a focus on three major components: (1) the characteristics of the issue, (2) the ways in which a national conflicts reach transnational audiences, and (3) the parties involved or affected. They articulated 10 propositions of C NCS and some of them were tested and supported by Molleda, Connolly Ahern, and Quinn (2005) through a content analysis of international news coverage of the Lesotho bribery scandal: Proposition 1 : Cross national conflict shifting is mainly related to cor porate social performance issues and negative economic consequences of globalization. Proposition 2: The magnitude of a cross national conflict shifting will increase when it starts in an emergent or developed economy because of the greater pressure the t ransnational corporation will face in the host country and from the international activist community. Proposition 3: Conflicts that occur in developed nations have usually a shorter life and do not cross borders as often as conflicts that start in developi ng nations or emergent economies. Proposition 4: A greater number of involved parties will characterize a cross national conflict shift in which a developed nation transnational corporation is the principal participant of the crisis.
21 Proposition 5: A low er number of involved parties will characterize a cross national conflict shift in which a developing nation or emergent economy corporation is the principal participant of the conflict. Proposition 6: Transnational corporations that produce or commercial ize tangible, boycottable products are more likely to receive attention than those who produce and commercialize intangible services. Proposition 7: Transnational corporations headquartered in developed nations that produce or are part of a national confl ict outside their home country, will attract great attention from global NGOs, international regulatory bodies, national governments, organized citizen groups, and international news agencies and global media outlets. Proposition 8: The direct involvement of a transnational corporation in a cross national conflict shifting will produce greater consequences and demand a more comprehensive set of responses than a transnational corporation that is indirectly related to the issue. Proposition 9: National conf licts shift to the international arena when primarily global NGOs or media report on the situation to audiences or publics in different parts of the world. Proposition 10: National conflicts with a great human interest focus are likely to be shifted to th e international arena. (Molleda & Quinn, 2004, pp. 5 7) Kim and Molleda (2005) advanced CNCS theory by combining CNCS theory with crisis management and analyzing the Halliburton bribery probe case in Nigeria. Their study was based crisis communication strategies which are: attack, denial, excuse, justification, ingratiation, corrective action, and full apology. They also expanded their original CNCS theory in order to include political aspects in a dynamic global context. Kim and Mo lleda (2005) developed three additional propositions: Proposition 11: Although a transnational corporation that does not produce or commercialize tangible, boycottable products, if the CEO or top level management have a cohesive relationship with the home country government or another highly visible institution, it will draw more attention from home country media, international media, international NGOs and regulatory bodies, and the issue will have greater political repercussions and debates.
22 Proposition 12: Domestic or national conflicts are not only perceived differently by related parties in the home country and host country, but also framed differently by the host country and home country media. Proposition 13: National or domestic conflicts of trans national corporations are sometimes combined with other related conflicts or issues that negatively affect the reputation of transnational corporations in home and host countries and, therefore, require more complex responses and public relations strategie s (until the conflicts resolve). (Kim & Molleda, 2005, pp. 14 16) included into cross national conflict shifting (CNCS) at wh ich information and news travel has created a more informed citizenry and galvanized them Teflon crisis in China also suggests CNCS should be interpreted from three p erspectives which have taken media into consideration: the crisis managemen t performance of the involved transnational organizations (TNOs) the level of media interest in the involved issue, and the unique and complicated social and cultural context of th e involved country. To advance cross national conflict shifting (CNCS) theory within a more complex global context that includes media influence, Molleda (2011) examined news coverage of the 2007 in articulated new propositions about international media coverage in a CNCS for theory implications: Major transnational corporate crises will result in greater international newswire coverage, especially from the countries i nvolved in the conflicting situation. The news stories will be mainly filed from major cities of the countries involved where the conflicting situation has greater resonance to more active and affected stakeholders. International newswire agencies will f rame conflicts differently and will highlight particular aspects of a CNCS depending on where they are headquartered in relation to the
23 The international newswire coverage of a transnational crisis will pri vilege corporate and governmental responses and actions, which could indicate that newswire services would favor powerful voices or that these sources are habitually available for news reporters during the coverage of a cross national conflict shift. (Moll eda, 2011, pp. 67) Media, as described in these propositions, is considered when examining crisis management efforts of a transnational conflict; especially the news media. The concept that media has an effect on audience such as agenda setting and framing has gained prominence in contents, issues and frames during corporate crises and to the relationship between media and 1). According to Olson (2001), the theories of cultivation, agenda setting, and spiral of silence all may be considered powerful media effects. setting theory. By analyzing international news content and poll data from the Chicago Council on Foreign Relations, they found a clear relationship between perceptions toward those nations. The results of the study showed the more the news media covered a n ation, the more likely audiences were to think that nation was vitally important to ace issues on the news Media Agenda Setting and Crisis Communication Through emergent communication and media technologi es, an event that is worth Connolly Ahern (2002), internationa l news agencies (e.g., Associated Press, United Press
24 International, and Reuters) and global media (e.g., CNN International, BBC World Service) are considered as global news media that rapidly disseminates the incidents of which transnational businesses ar e involved. In the perspective of global public relations, communication media not only serve as an ahan, 2001). Although there setting, the majority of Researchers have conspicuously paid litt le attention to the effect of media on crisis management (Carroll & McCombs, 2003; Howell & Miller, 2006) The agenda setting effect of news media is defined as the ability to influence both the topic and the perception among audiences (McCombs & Reynolds, 2002). Carroll and McCombs (2003) drew the role of agenda setting in media out of its original domain (i.e., political issues and public affairs) and found it fit equally well in the world of business communication. They believed that the everyday selecti propositions on the effects of first and second level agenda setting on corporate reputations among t agenda setting are: P 1: The amount of news coverage that a firm receives in the news media is positively P 2: The amount of news cover age devoted to particular attributions of a firm is positively related to the proportion of the public who define the firm by those attributes. P3: The more positive that media coverage is for the particular attribute, the more positively will members of the public perceive that attribute. Conversely, the
25 more negative that media coverage is for a particular attribute, the more negatively will members of the public perceive that attribute. P4: The agenda of substantive and affective attributes associated with a firm in business news coverage, especially those attributes specifically linked with a firm, P5: Organized effort to communicate a corporate agenda will result in a significant degree of correspondence between the attribute agenda of the firm and news media. (pp. 39 42) Some research has attempted to integrate the effects of agenda setting in news media and crisis communication within an intern ational setting. Ha (1999) analyzed news coverage of the Clinton a global agenda that, as part of crisis response efforts, organizations should monitor and be aware of news coverage relevant to the crisis (Ha, 1999; cited in Freitag, 2002). Lim (2008) conducted an experiment that focused on how host consumers of automobiles may perceive the co rporate reputation of major transnational organizations (TNOs) with two types of crises (e.g., a massive product recall and a bribery scandal) through news media. She found that the potential customers of a host country form negative attitudes and negative behavioral intentions by reading news coverage of a TNO involved in a crisis (Lim, 2008). In other words, as Patterson (2004) said, 1 ). According to Lim (2008) and Wang (2005), this effect is harder to control once issues regarding cross national conflict shifting (CNCS) have come out and the transnational organization (TNO ) will need to be flexible and dynamic in an information transaction in order to have a better tran snational crisis defense system. Synnott and Mckie (1999) noted that when crisis shifts internationally, there are other potential factors would impact public perception of crisis and organization. They further illustrated that mic development
26 bears on perceived priorities of public relations research needs, and that geographic region and International Crisis Communication and Media rders. Worldwide audiences get the same interview (as cited in Molleda & Connolly Ahern, 2002, p. 3). With gradually sophisticated communication technology and increa sing public access toward information, media has become a powerful force that can make a national happening become a transnational or global event; and even impact the parties involved in their home, host, and transnational locations around the world (Moll eda & Connolly influence over a crisis has evolved with the digital revolution to instantaneous and exhaustive, o set public agenda and an important tool to crisis management (Howell & Miller, 2006; Patterson, 2004; Molleda & Connolly Ahern, 2002; Seeger, Sellnow, & Ul mer, 2001). Much research has been conducted for the purpose of analyzing the nature of news coverage and the angles taken by the media, which is of the Firest one tire recall crisis identified the elements of a crisis that could predict international media coverage and public reactions and the factors affecting the nature of coverage. He then proposed adjustments that are needed in a crisis response plan for an international environment. By analyzing international news coverage of the Firestone Bridgestone tire recall, Freitag (2002) concluded in his study what issues have covered and news editorial preferences of certain countries. He found, for example, the ev ent was prominently covered by Japanese news media, although its consumers were not significantly affected by the recall, as a result of the corporate
27 link between Bridgestone a Japanese based transnational corporation and Firestone. In contrast, the c overage in Denmark, India, Malaysia, South Korea, and Thailand was limited and only related to consumer assurances or business analyses (Freitag, 2001). Freitag (2001) also emphasized that factors such as cross cultural communication dynamics and mass medi a structure issues might affect news coverage. At the end of his study, Freitag (2001) called for more studies that focus on one national or geographic area and suggested that future research should consider the roles of cultural patterns, media structure, Teflon crisis in China from three perspectives: the crisis management performance, the level of media interest, and the complex of social and cultural context in where the crisis took place. She suggested a reversed cross national conflict shifting (CNCS) proposition that a conflict involving a t ransnational organization (TNO ) could shift from a home country to a host country through international media and cause greater impact in the h ost country (Wang, 2005). The study revealed that Chinese news coverage differed significantly from those covered in the United States through the elements such as event location, story focus, and primary problem attribution (Wang, 2005). In addition to th e studies that focus on the different way news media frame crises among countries, there are studies that discuss how cross national conflict shifting affects the perception of host customers toward transnational corporations involved in the conflict. Dive rgently from research showed that audiences exposed to the news story about massive product recall of a t ransnational organization (TNO ) resulted in a more negative attitude and behavioral intention than the news story about bribery scandal of the same company.
28 Previous studies illustrated that media play an important role in setting the public agenda and shaping public opinion through the selection of news content. For the reason that the public would learn the important issues from the amount of coverage given to them in the news media, monitoring media coverage has been considered extremely crucial while managing a crisis (Coombs, 2006; Valentini & Romenti, 2011). Seeger, Sellnow, and Ulmer (2001) also stressed crisis (p. 162). In sever al research studies, the framing role of media during corporate crises, and the relationship between media and public opinion are again being emphasized with more focus on the international news media. In order to compare crisis issues, tones and frames pr oduced by national and international newspapers, Valentini and Romenti (2011) examined how Italian and differently by analyzing both Italian and international ne ws coverage. The authors indicated the when developing crisis communication strategies and sending messages to the media in the international environment (Val entini & Romenti, 2011). The study identified eight main crisis issues that were primarily covered by Italian and the international news media during the 2008 crisis. These issues are: positi relations, labor union relations, employee relations, and strikes/protests. Valentini and Romenti (2011) found that Italian and international news media would se lect different issues among these eight as well as the different tone and frame when covering the crisis. They also stated that media
29 systems and news practices have impact on choosing the tonality and associated frames among Italian and international news agencies, which proved the previous studies by Hallin and Mancini (2004) that the Italian news media are highly politicized with a preference for the conflict frame. in pro duct recall crisis in 2007 and suggested that transnational organizations (TNOs) involved in a transnational crisis should monitor and analyze the highest peak of news coverage concerning the global flow of information and international newswire. Through t his strategy, TNOs are able communications and corporate actions during crisis (Molleda, 2011). By examining the news coverage from China, Europe, and the Unit ed States, he indicated that newswire services would privilege different voices in the coverage of cross national conflict shifting (CNCS) (2011) study expanded his original 13 propositions of CNCS and articulated four more propositions regardin g to international news coverage for advancing the CNCS theory. One of frame conflicts differently and will highlight particular aspects of a CNCS depending on where 67). Furthermore, he also pointed out the preference of international newswires when covering a transnational crisis that: verage of a transnational crisis will privilege corporate and governmental responses and actions, which could indicate that newswire services would favor powerful voices or that these sources are habitually available for news reporters during the coverage of a cross national conflict
30 For developing better crisis management strategies and bridging the gap between academic research and real practice, many scholars integrated crisis response disciplines into their studies with an emphasis on me dia effects (e.g., Coombs, 2006; Freitag, 2001; Howell & Miller, 2006; Huang, 2006; Len Ros, 2010; Lim, 2008; Molleda & Connolly Ahern, 2002; Molleda, Connolly study on the Fi restone Bridgestone tire recall crisis, as aforementioned, discussed factors that may influence media coverage on a crisis, such as cultural dimensions, and media structure and function which determine crisis planning and response strategies. Seeger, Selln ow, and Ulmer (2001) also mentioned that, used as an effective crisis response strategy, organizations follow media closely so as to defend themselves against any false information that is reported. And if successful, the organization then has the ability Different from other research that focus on analyzing news content and examining the effects of the media on an audience during a crisis, Molleda (2011) found a link between issues that were raised by key terms presented in a newswire and the transnational organizations (TNOs) by monitoring the series of chronologically published headlines and lead paragraphs (p. 67). In addition to identify those key terms, he pointed out that these key terms may change during the chronological cycle of the cross national conflict shifting (CNCS) successful transnational crisis control program should be supported by continuous executive media training, the efficient coordination of responses and actions, and, particularly, effect ive
31 developing crisis response strategies in media perspective is necessary (p. 67). Crisis communication: A Focusing on Crisis Life Cycle According to Fearn Ba nks (2001), the public relations profession has been crisis driven since it was created in the early 20 th century. With the advancement of technology, key publics have grown up accompanied by various communication tools, such as television, radio, newspape r, books, magazines, and the internet, that keep providing its audience with an abundance of information. Therefore, because the public demands to know issues about corporations, concerning their financial condition, and communications they use to thwart t he problems, organizations are increasingly seeing the need for crisis communication (Fearn Banks, 2001). Fearn use verbal, visual, and/or written interaction between the org anizations and the public; often through the news media, and it is designed to minimize damage to the reputation of the organization. Coombs (1999) suggested in Ongoing crisis communication: planning, managing and responding the need for public practition ers to understand crisis life cycle in order to plan the different response for each stage. The crisis life from the marketing product life cycle model. The concept of a crisis that c ould progress was staged model that was applied initially to better four stages in the development of a crisis which are : (1) prodromal clues and warning signals that begin to emerge; (2) acute/crisis breakout an event or incident occurs followed by a negative outcome; (3) chronic the effects of the crisis linger as to limit the crisis progress; and
32 (4) resolution t he crisis is no longer a concern to stakeholder and the organization returns to normalcy (Coombs, 1999; Fearn Banks, 2001). Barton (1993), Mitroff (1996), and other scholars (e.g., Coombs, 1999; Sturges, 1994 ) staged model to a five staged model. Crisis expert Ian Mitroff (1994) divided crisis management into five phases which emphasizes more prescriptive functions: (1) signal detection organizations identify the warning sign for a new crisis, (2) probing and prevent ion organizations search known crisis risk factors and make plans to avoid the crisis through proactive strategies to reduce the potential of harm, (3) damage containment imit the duration of the crisis and try to prevent the crisis damage from spreading, (4) recovery organizational operation return to normal, and (5) learning organizations review, evaluate, and determine their crisis management effort as a lesson for t he next time (Coombs, 1999; Fearn Banks, 2001; Mitroff, 1987). Usually, the purpose of crisis communication is to offset the potential negative impact while organizations fail to respond to warning signs in the early stage (Coombs, 1999; Howell & Miller, 2 disseminate messages to public sped up less than a decade ago. Typically, the framework for crisis communication research makes the assumption that media coverage is important a s it could changes in content of media coverage during the crisis (Howell & Miller, 2006). According to Holtzhausen and Roberts (2009), there is a strong associat ion between crisis stage and media portrayals of individuals in the crisis. Len Rios (2010) evaluated the
33 sexual harassment by analyzing local news coverage. Fol lowing the suggestion of Brinson and Benoit (1996) that there are benefits in examining news coverage with crisis stages, she included crisis stage factors into her study. The study also determined how crisis stages affect the tone of local newspaper portr different strategies are used at different stages in the crisis as well as the themes selected by local ut the Howell and Miller (2006) believed that mass media coverage during a crisis can be fe cycle contains different themes in mass media coverage, and each theme exhibits attributes pertinent to amended crisis life cycle model that maps the news medi a content during each stage of crisis and they illustrated it by analyzing the local news of the safety crisis in Ansett an Australian based airline. Their crisis life cycle is comprised of five stages that integrated previous studies of crisis stages. ( 1) phase one which termed prodromal (Fink, 1986; Barton, 1993), signal detection (Mitroff, 1996), and detection (Fearn Banks, 1996), the model predicts that the issues that have the potential to become crises appear as small news stories in mass media; ( 2) phase two which termed preparation (Barton, 1993), probing (Mitroff, 1996) and prevention (Fearn Banks, 1996), the model suggests that those trigger themes in initial stages will continue to appear in mass media coverage and also suggests how to manag e those themes which are vital to organizations; (3) phase three which termed acute (Fink, 1986) and containment (Barton, 1993; Mitroff, 1996; Fearn prodrome will be th e focus of mass media content, and mass media coverage will evolve to
34 which termed chronic (Fink, 1986), learning (Barton, 1993; Gonzalez Herrero & Pratt, 1996), damage containment (Mitroff, 1996) and recovery (Fearn Banks, 1996), the model indicates that (Howell & Miller, 2006, p.6); (5) phase five which is termed resolution (Fink, 1986), recovery (Barton, 1993; Mitroff, 1996) and learning (Fearn summarize how the crisis occu rred, who and/or what was to blame or responsible for the crisis, relations practi tioners to develop more efficient crisis communication strategies to communicate with different stakeholders in each of these stages. However, whether this model can be applied to the international environment has not yet been tested. For managing cross na tional conflict shifting (CNCS) there is a need to consider factors such as the cultural syndrome, media structure, and its function in each country that are involved in the conflict, according to Freitag (2002). Valentini and Romenti (2011) also argued t specific features of the national media system and news practice before choosing crisis response news practices, and the choi ce of crisis news frames are necessary to gain insight into managing and communicating crises in different international contexts (Valentini & Romenti, 2011). World Media System rames to understand an event or issue by defining problems, diagnosing causes, making moral judgments, and suggesting remedies (Valentini & Romenti, 2011). Framed news coverage provides
35 audiences with a visible public expression of approval or disapproval of corporations and their behaviors (Carroll & McCombs, 2003; Valentini & Romenti, 2011). There are many internal factors in journalism that can affect a news frame; such as the communicator, the text, the receiver, and its culture. However, external fact ors, such as media system in countries where the media operates, also determine media framing (Entman, 1993; Valentini & Romenti, 2011). Several scholars have provided theoretical models defining and classifying dynamic world media systems (e.g., Hachten & Scotton, 2007; Hallin & Mancini, 2004; Siebert, Peterson, & Schramm, 1963). According to Hachten and Scotton (2007), political and cultural conflicts influence the e nature and role of journalism and mass communication are rooted in divergent political systems : early in 1956 by Siebert, Peterson, and Schramm. They believed media should ideally perform under certain political systems and social values. These values were then entitled as the authoritarian theory, the libertarian theory, the social responsibility theory, and the soviet comm well within modernity, it still accounts for an important point in the classification of world media systems; which had a tremendous influence on proceeding studies on the freedom of the press posed that dynamic world media systems are reflected in five
36 contemporary political concepts of the press: (a) authoritarian, (b) western, (c) communist, (d) revolutionary, and (e) developmental. g the four, indicates the press should always support and advance the policies of the government in power. According to to challenge, criticize, or in any way un authoritarian press can be found in some countries, such as Burma, Malaysia, Kenya, and Zimbabwe (Hatchen & Scotton, 2007). lassified as a of authoritarianism, according to Berry et al. (1995). The purpose of the press under the libertarian concept is to inform and entertain the audi ence, as well as discover the truth and supervise the government. This concept was rooted in the philosophical view that man is rational and has the ability to discern the truth. ligation of that the purpose of the media is not only to inform and entertain audiences, but also to ensure that all voices and views in the society are heard. F responsibility, the government has limited influence in media operations and regulations, which were demonstrated in the early practices of the United States. The soviet concept presents the role of media under the Soviet socialist system as a tool for serving the interests of the communist party. According to Schramm (1956), the Soviet press system is planned and completely party controlled with no permission of competing private
37 media. He further illustrated t hat news practices under this system are defined as providing te the Soviet Union, the Soviet communist concept has been questioned by scholars and received several critiques (e.g. Berry et al., 1995; Hatchen & Scotton, 20 07) regarding whether it can environment (Berry et al., p. 131). For press in Russia, (the former Soviet Union), it now can be described more as authoritarian (Hatchen & Scotton, 2007). For the media systems currently under control of the communist party, such as China, they no longer meet all the characteristics of the original Soviet communist theory, but are still considered as a different form of communist concept. As Freedom House (2011) described, the China Communist Party (CCP) keeps direct control over news media coverage through its Central Propaganda Dep artment (CPD) and The revolutionary theory shares a common trait with Western concept that suggests media systems operate outside of gover revolutionary press is a press of people who believe strongly that the government they live under commu nication tools, such as radio stations, the photocopying machine, the audiocassette and personalized media present challenges to centralized autocracies trying to control news and information.
38 As the revolutionary concept is viewed for short term implemen tation, developmental 2007). It holds the values that the press must be utilized by the government to assist the development of the society, the media should sup port government instead of challenging it, and information or news is the property of the state that must be used to further the national goal. This model also suggests that nations have sovereign right to control both foreign journalists and the flow of n ews across the borders (Hatchen & Scotton, 2007). Hallin and Mancini (2004) added new insights on the interaction between the media and the political structure. They further identified three models of media systems among Western tradition that the media is a medium of ideological expression and political mobiliza tion. The tendency for media to express partisans because of the high degree of political parallelism, the early development of press freedom with the tradition o f limits on state power, and the high degree of journalistic professionalism as well as a notion of commitment to public interest. The newspapers developed relative ly early with little state involvement. Within this model, the strong tradition of political neutrality affords the informational style of journalism to dominate. relat ively practical and flexible way to explain the difference between media systems nowadays. (absolute authoritarianism) at one end, to no control (pure libertariani
39 Although scholars (e.g., Hallin & Mancini, 2004; Hatchen, & Scotton, 2007) argued that globalization trends result in an uprising degree of homogenization among media systems, especially in commercial structures and the culture o f journalism; the trend also increases opportunities for conflicts between different concepts because of the increasing transnational communication. This literature review seeks to provide theoretical background to this research of cross national conflict shifting (CNCS) that aims to suggest a working model for efficiently managing the media during a crisis by testing the previous relationship model between news media and the crisis life cycle (Howell & Miller, 2006). Since globalization as a trend influen expanding business in the international arena, there is a raising imperative for transnational corporations to have ability to deal with corporate crises across national borders. Crises are unpredictable; therefore proactive and efficient managing strategies are needed to reduce the damage of crises on corporations. Studies and propositions of cross national conflict shifting (CNCS) provide a theoretical framework for managing crises in the international environment which is marked as more dynamic and complex. Based on the agenda setting and cycle model articulated a new perspective of integrating media coverage with crisis stages. The crisis life cycle model explains cri sis evolution with the nature of news coverage as a strategy of crisis response. The study derives insights from research in the macro context such as media systems, and the level of press freedom into cross national crisis management. Those perspectives n ot only reflect political systems in each country, but also present the diversity between nations and their news practices; which therefore can better inform crisis communication.
40 Case Background For better illustrate the relation between media coverage an d crisis life cycle of a cross national crisis, this research is going to use 2010 Foxconn employee suicide to map the news media content during each stage of the crisis and to test the model provided by Howell and Miller (2006) in international setting. The Foxconn crisis is well suitable for this research since it is a cross national conflict that shifted to three countries which have different media and political environments that author is especially interested in. Foxconn Employee Mistreatment Contr oversy In this sub section, the brief introduction of Foxconn Technology Group will be summarized for reader to understand the background of the corporation. The controversy of employee mistreatment that happened four years before the crisis hit in 2010 will also be described to put the study in context. Foxconn Technology Group is a subsidiary of the Taiwan based multinational corporation, Hon Hai Precision Industry Company, which is headquartered in Tucheng, Taiwan. The company focuses primarily on p roducts from mobile handsets to computers, including products as iPhone and iPad (Hoovers, 2011). According to Fortune magazine (2010) and Hoovers (2011) business report, the company is ranked 112 out of 500 corporations and owns large amount of contract s with many high profile consumer makers, including Apple Dell H P Motorola Nokia and Sony As the world leading electronics manufacturer, Foxconn operates its factories worldwide with primary installations in Taiwan and China, as well as builds production facilities in the United States and Europe. Similar to many multinational manufacturers that operate in China, Foxconn views low labor and land cost as factors that increase its competitive edge (Dean, 2007). With a strong workforce of 900,00 0 workers in China, the company aims to maximize its productivity that
41 makes employees to work overtimes ( SACOM, 2010). According to Dean and Tsai (2010) in numb er of overtime hours, and that its military style rigor and repetitive working conditions 17). Agence France Press (2011) also said that Foxconn is accused of treat ing employees as machines. The controversy of empl oyee mistreatment actually started early in 2006 when the Daily Mail in the United Kingdom alleged that Foxconn mistreated its employees with long work ). Apple responded to the Daily Mail Apple is committed to ensuring that working conditions in our supply chain are safe, workers are treated with respect and dignity, and manufacturing processes are environmentally responsible" (Evans, 2006). Evolution of the Foxconn Crisis Early in January 2010 Foxconn faced the disastrous event regarding employee mistreatment in its host country, China (Table 2 1 ). According to SACOM (2010), 17 Foxconn employees tried to commit suicide during January to August, 2010 resulted in 13 deaths and four injuries. T he suicides are heavily covered by news media in China and quickly spread overseas through international media. Large amount of related news are reported in its home country, Taiwan because of the issue proximity, and the United States since Apple and Del l also were involved as Foxconn mistreatment issue has transformed into an ethical concern for stakeholders. In the report of Hong Kong based labor activists, Student and Scholar s Against Corporate Misbehavior (SACOM) described that: Other than Foxconn, electronic brands like Apple Nokia, HP, Dell, Sony Sony Ericsson, and Motorola, which have placed orders with Foxconn, also bear indispensable
42 responsibility in the tragedies. Al l these brands are making huge profit at the cost of the workers. Likewise, we are consuming the blood and tears of the workers, a fact hidden from us by fancy advertisements. (2010, p. 4) There are even people who compared iPhone /iPad The eleventh suicide in May and media pressure forced the president of Foxconn Terry Guo, to face the media and to apologize while he gave a tour to a number of Taiwanese reporters in Foxconn Shengzhen base, according to the Bloomb erg Businessweek (Foreman, 2010). Meanwhile, the clients of Foxconn including Apple Dell and H P expressed the concern about the suicides and announced that they are in touch with Foxconn to investigate the work conditions. As quoted in the report o f The Wall Street Journal Apple saddened and upset by the recent suicides at Foxconn has taken this case seriously (Dean& Tsai, 2010). Unfortunately, the tragedy of Foxconn didn t stop after the apolog old male Foxconn worker jumped off the dorm to his death and another worker slit his own wrist trying to commit suicide (Foreman, 2010). Instead of developing efficient stra tegies to enhance internal employee relations and work conditions, Foxconn only put suicide prevention netting around the employee dorm, requested workers to sign the anti suicide pledge, and to attend an anti suicide rally. According to Ricadela (2011) i n Bloomberg, a group led by Apple Foxconn and met with Foxconn President Terry Gou in China and provided measures Foxconn adopted to prevent more death. Apple also supported its manufacturer by claiming that 2 ).
43 After a series of suicides in May, the Chinese government requested censorship on news related to Foxconn since the company, governmen t officials, and some citizens believed that it was the intensive news coverage that led to the continuous employee suicide (Tam, 2010; Chang, 2010; NowNews, 2010). VOA News report published on its Chinese website on May 27, 2010, mentioned that the Chines e government had put restrictions on local news coverage of Foxconn News, 2010; cited in Qi, 2010). Although the quick actions and strategies the Foxconn adopted were commended by Apple the Foxconn both share markets of Foxconn and Hon Hai the parent co rporation of Foxconn experienced a recession during the crisis. Protests of anti labor abus e and anti Apple products in China and Taiwan were also covered as consequence of a string of suicides. Apple was also criticized and considered to have social r esponsibility supervising its suppliers. The crisis received less attention among news media with the release of the report jointly produced by 20 universities in Hong Kong, Taiwan, and mainland China. The report called Foxconn (SA COM, 2010). In another perspective, Apple (2011) also explained the Foxconn suicide incidents in its Annual supplier report that the measures Foxconn applied immediately after suicides are effectively saving life. Research Questions and Hypotheses This study aims to investigate the relationship between cross national crisis life cycle and international newswire coverage based on the study of Howell and Miller (2006). The author is specifically interested in exploring issues frames and news sources selec ted to cover Foxconn crisis. The inspirations of relevant research on cross national conflict shifting (CNCS) and media
44 effects enable this study for constructing hypotheses and research questions. In order to plan efficient crisis communication strategie s, Howell and Miller (2006) took advantage of agenda setting effect of news media and developed a model to predict news coverage in each stage during a crisis. This study borrows their model of news coverage during each crisis stage for managing corporate crisis in an international environment. Based on their study, following are the research questions and hypotheses that this thesis purposes to answer and to support or reject. RQ1 a : How would the news coverage of Foxconn crisis case be explained by Howell and revision? Howell and Miller (2006) stated that the number of trigger theme can be distinguished across crisis life cycle. They analyzed the news coverage of t he Ansett Airline crisis and identified 10 trigger themes appear ing in news content during the crisis. Their research confirmed that the amount of trigger themes in each stage is consistent with the model. is study hypothesizes that: H1a: T he news coverage of Foxconn crisis in China will gain different levels of attention during five crisis stages in terms of the length and number of articles. H1b: T he news coverage of Foxconn crisis in Taiwan will gain d ifferent levels of attentions during five crisis stages in terms of the length and number of articles H1c: T he news coverage of Foxconn crisis in the United States will gain different levels of attentions during five crisis stages in terms of the length and number of articles RQ1b: Are there any differences between the levels of received attention in China, Taiwan and the United States in terms or the length and number of news articles? H2a: T he number of trigger themes/issues that covered in the news of Foxconn crisis in China will differ during five crisis stages. H2b: T he number of trigger themes/issues that covered in the news of Foxconn crisis in Taiwan will differ during five crisis stages.
45 H2c: T he number of trigger themes/issues that covered i n the news of Foxconn crisis in the United States will differ during five crisis stages. R Q 2 a : Which themes/issues about Foxconn crisis were covered by the press in China, the United States and Taiwan? RQ 2b: Are there any differences between the number of trigger themes/issues in China, Taiwan and the United States? H3: The themes/issues that covered in the news of Foxconn crisis will differ between China, the United States and Taiwan. The summary of previous research on cross national crises and news media suggested an interesting path for researchers to examine news contents, tones, and frames of a cross national crisis in different countries (Freitag, 2001; Molleda, 2010; Valentini & Romenti, 2011; Wang, ies revealed the difference of news contents regarding cross national crisis between China and the United States while Valentini and Romenti (2011) considered factors include news practices or media system when covering crisis in Italy. Semetko and Valkenb urg (2000) pointed out the five mostly used news frames : (1) attribution of responsibility, (2) conflict, (3) economic consequences, (4) human interest, and (5) morality. An and Grower (2009) borrowed these frames and used them in analyzing 2006 crisis eve nts covered by n ews papers in the United States. Attribution of responsibility frames focus on the direct or indirect causal relations between crisis and role of the organizations or individuals. Conflict frames depict the dissent among individuals, groups, or organizations in crisis. Economic frames stress the financial impact of the crisis while frames of human interest focus on the human and emotional aspects. Morality frames examine the behaviors and moral/ethical issues of individuals or organizations t hat considered are responsible for the crisis. Several studies (An & Gower, 2009; Cho & Gower, 2006; Qi, 2011; Valentini & Romenti, 2011) also applied these five frames as a tool measuring the difference of news coverage between countries. Qi (2011) compar ed the editorial freedom of major newspapers in China in reporting the 2010
46 Foxconn crisis by examining the frames and news sources those articles selected in each newspaper. She also suggested that newspapers with more editorial liberty would use greater variety of sources than official news media, which cited primarily official sources (Qi, 2011). As Molleda (2011) stated that international news agencies have different editorial policies regarding the selection of news sources. In his study which analyzi product recalls, Molleda (2011) found Xinhua cited and quoted Chinese sources more often than is a government run news agencies selected will be analyzed as elements reflect editorial liberty in different media environments. According to Freedom House (2011), press in China, Taiwan, and the United States enjoy different level of editorial freedom (China ranked as 184, Taiwan ranked as 48, and the United States ranked as 17 out of 194). Therefore, this study a ssumes and ask s : RQ3: Do the news coverage of China, Taiwan and the United States use different cri sis news frames when discussing Foxconn crisis? And if so, which one did they use respectively ? RQ4: Do the news coverage of China, Taiwan, and the United States use different news sources when covering Foxconn crisis? And if so, which one did they use r espectively ? H4: The news coverage of Foxconn crisis enjoyed different level of editorial freedom in China, Taiwan and the United States in terms of news sources and issues covered After providing a list of research questions and hypotheses this researc h is seeking for answers, the methodology of the quantitative content analysis of international newswire coverage will be explained in the next chapter. The methodology section will illustrate the sampling procedure, the statistical measurements of data an alysis and an appendix of research instruments to use to answer the stated research questions and hypothesis.
47 Table 2 1. Timeline: the employee suicide crisis at F oxconn in 2010 Date Event January 23, 2010 Ma Xiangqian, age 19, was found dead in fro nt of his dormitory. The police changed their determination of his death finally to suicide March 11, 2010 Li Hongliang, age 28, jumped from the 5 th floor of the dormitory and died. March 17, 2010 Tian Yu, an 18 year old female Foxconn employee, jumped from the 4 th floor of her dormitory at the Foxconn Longhus base in Shenzhen and was heavily injured. March 29, 2010 Liu Zhijun, age 23, jumped from the 14 th floor dormitory and died instantl y. April 6, 2010 Rao Shuqin, an age 18 female worker of Foxconn from the 7 th floor of her dorm. She survived but was also heavily wounded. April 7, 2010 Ning, an 18 year old female worker jumped from her dormitory and died. May 6, 2010 Xin Lu, age 24, jumped off the balcony of his dormitory in Shenzhen and died. May 11, 2010 Chenming Zhu, a female Foxconn assembly line worker, aged 24, was dead because she jumped off the 9 th floor of the apartment building that is near Foxconn plant. May 14, 2010 Chao Lian, a 21 year old male, jumped from the 7 th floor and died. The police found four knife wounds in the body and a bloody dagger. May 21, 2010 Gang Nan, a male employee of Foxconn age 21, jumped off the dorm building to his death. May 25, 2010 Hai Li, a 19 year old male Foxconn worker, committed suicide by jumping off the 5 th floor of his dorm in Shenzhen base. May 25, 2010 Foxconn suicide May 26, 2010 Terry Guo, Foxconn and apologized for the series of suicides May 26, 2010 Customers, including Apple Dell HP and Nokia express the concern about Foxconn employee suicide events. May 26, 2010 He, a 23 year old Foxconn worker jumped to his death from a building in the plant. Another worker slit his wrist attempting suicide. May 28, 2010 Chinese government requested censorship on Foxconn suicide news stories. June 8, 201 0 Foxconn confirmed to raise the wage of its workers up to 30 percent. June 29, 2010 Foxconn put anti suicide net around the employee dorms.
48 CHAPTER 3 METHODOLOGY This thesis investigates printed news wire coverage of Foxconn employee suicide cri sis in well suited to the study of communications and to answering the classic question of To answer the research questions and to test the hypotheses, this study collects electronic news articles of major news agencies from three of the most involved countries during the crisis: China, Taiwan, and the United States. Alth ough there are other types of news coverage, such as video and radio transcripts, the main focus of this research is on printed newswire coverage. Those printed news stories from news agencies were chosen for analysis instead of other types of media outlet s because of the global reach of newswire services and their important role in the daily activities of multinational media enterprises (McPhail, 2010; Molleda, 2011). News agencies are the major source of international news for regional, national, and loca l news media worldwide. Define Foxconn Crisis Stages In order to examine the difference of newswire coverage in each selected country with crisis life span, a time series identification of crisis stages was employed according to the definitions of each st age. According to scholars (e.g., Coombs, Fink, and Mitroff ) and studies (Co mbs, 2011; Mitroff, 1996; Howell & Miller, 2006) relevant to crisis life cycle, the Foxconn crisis can be divided into five phases regarding the timeline and the crisis characteri stics: Phase one : From January 23, 2010, when first death of Foxconn employee reported, to the end of February. Phase two : From March 2010, the second suicide in Foxconn factory marked as the beginning of prevention and probing stage, to the end of April.
49 Phase three : The crisis broke out at early May with the death of a Foxconn worker who jumped off the balcony of his dorm. This period ends at the end of May. Altogether six people committed suicides. Phase four : This stage is longer than o thers lasting from June to the end of September. In this Phase five : The establishment of labor union in Foxconn implied that the organization l earned its weakness from the crisis. In this period, organization reviewed its October to December, 2010. Sampling Sample Selection To answer the three research questions and test the four hypotheses, content analysis is conducted on the news coverage of the Foxconn crisis yielded by China, Taiwan, and the United States. For those news agencies, providing news contents and newswire services to other news organizations are their ma in business purpose. With foreign correspondents, news agencies generate and exchange news information regarding the events occurred abroad. The national and international news reached audiences through one or more of major international news agencies whet her in print or electronic form; therefore, to set the news agenda for other media outlets nationally and internationally (Golan, 2006; Sterling, 2003). Three of the influential news agencies with different national roots and editorial positions in the th ree involved countries were selected for analysis. Chinese Xinhua, Taiwanese Central News Agency (CNA), and U.S. Associated Press (AP) were chosen as different news sources reflected three media systems and level of press freedom. According to the studies of world media system (e.g., Hallin & Mancini, 2004; Hatchen & Scotton; Yin, 2008), the media in the United States is classified as Western concept with some characteristics from both revolutionary and developmental systems. Chinese media system is identif ied somewhere between communist
50 and developmental concepts while press in Taiwan enjoys the relatively freer environment among Asian countries and is recognized to be similar to the Western concept. The level of press freedom performed differently between these three countries: the press in United States is the freest among the three and the media in Taiwan are also free and ranked as 47 out of 196; while Chinese press marked as not free (Freedom House, 2011). es with different editorial orientations was to capture the nuance of the flow of global news, which goes hand in hand with the main claims Chinese language news from CNA, and English language news from AP and Xinhua. The author was going to use articles in their native language for better understanding of the editorial positions in each country selected and considering the agenda setting effect on national public. The therefore turned to LexisNexis to search for English version of news from Xinhua general news service. Sample Profile The news articles were generated from three major international and national news news articles containing the key word news digests of the week and Foxconn for analysis through three separate searches (Xinhua, N = 40; CNA, N = 90; AP, N = 32). China: Xinhua News Agency company. It is listed as one of the major global news agencies following Reuters, Associated Press, Agence France Press, Bloomberg, and Dow Jones. For the reason that Xinhua is tightly
51 con trolled by the Chinese government, the news agency is considered as a mouthpiece of the Chinese government. As McPhail (2010) mentioned, Xinhua serves as a gatekeeper of both Republic of China. The author chose Xinhua News Agency as one of the sample sources not only for its international prominence, but also for its position to represents the current news practice in China. According to Qi (2011), Chinese government has put ce nsorship on Foxconn crisis in covering the crisis. There are total of 67 news stories from Xinhua, already excluded articles ocusing on the crisis, such as Foxconn plan as well as the news digests of the week. Taiwan: Central News Agency (CNA) official news agency. It received part of its funding from government and serves as a role of non profit organization. Although CNA is not important as other international news agencies, for example, Associated Press, Reuters, and Xinhua, it still plays a significant role as foreign correspondent in Taiwan and reporting international news to news organ izations in Taiwan. The author gathered news coverage of Foxconn articles that already excluded articles not focusing on the crisis, as well as the news digests of the week and Foxconn business plan. The United States: Associated Press (AP). AP is viewed as the second largest global news agency, according to McPhail (2010). It gained prominence from its establishment in the mid 1800s for the purpose of news cooperation. Serves as a nonp rofit organization, AP represents a news center of news in the United States and a premier wire service corporation in North America. Its global video news, APTN, focuses on reporting events in different time zone
52 and provides video footage of important ne ws by satellite to news organizations worldwide. To collect news articles from AP, the author did a search on LexisNexis by entering key word news articles appeared a nd 33 of them are selected as sample of AP because they pertained to the studied crisis. As aforementioned, articles yielded from this search that have no relations with the crisis, belong to news digest and described Foxconn business plan, have already e xcluded. Coding Sheet Construction The coding sheet is designed to facilitate the content analysis of news covering the Foxcon n crisis in 2010 ( Appendix A ). Variables or questions in the coding sheet are developed based on the coding framework of Howell a nd Miller (2006), five common news frames indicated by Semetko and Valkenburg (2000) and coding elements from Qi (2011) and Wang (2005). The coding sheet consists of four major categories includ ing : basic information, main themes/issues focus, and news fr ames. The basic information category included elements such as the name of the news agency and the identification of phase number of the Foxconn crisis which defined by the author at the beginning of this chapter In this category, the coders will be aske d to count words of the news articles both in Chinese and English. According to online discussions on translation related forum, the ratio of Chinese characters and English words is approximately 1.5:1. The author transla ted ten Chinese news into English b y Google Translate and the ratios are about the same as mentioned. In order to compare the number of the words of news in Chinese and English, Chinese news stories will be translated into English by Google Translate only for coding the story length. Twen ty articles were selected by the researcher throughout the Foxconn crisis life cycle. The trigger themes that were identified by the researcher according to the list of issues that transnational corporations (TNCs) would face during a corporate crisis wer e proposed by
53 Molleda (2011). He pointed out that there will be one or a combination of those organizational and contextual issues: (1) employee treatment, (2) workplace safety, (3) c orporate decisions, policies, and performance, (4) living and natural eco systems, (5) community well being, (6) cultural clash, (7) strategic communications, (8) home and host government and legal systems, and (9) international relations. The 10 trigger themes/issues are developed to examine the difference a press focus in Chin a, Taiwan and the United States: Foxconn accusation, Foxconn environment in China, Apple market Impact, Apple supervising its suppli ers, Reactions of Foxconn Apple Dell and HP ), activist group, Foxconn Taiwanese government reactions toward crisis and the result of investigations. In the th ird category, news frames, variables were adopted from Qi (2011) and Hong (2007). In order to compare the five frames, Qi (2011) developed the sub frames, presented as questions on coding sheet, made each major frame mutually exclusive. If the news article has one or more than one sub frames, the certain frame is presented. The news sources coding elements are identified based on the researches of Qi (2011), Molleda (2005) and Wang(2005). Pretest and Inter coder Reliability The pretest was conducted to test the inter coder reliability for coding instrument. Ten percent of the news stories ( N = 1 6 ) were randomly selected from the news sample and coded by author and another graduate student separately. The consistency and inconsistency of the coding decision w ere assessed between the two coders. The inter coder reliability coefficient w as final coding sheet. A high level of the inter coder reliability coefficient (0. 875) was confirmed among the coders regarding coding decision.
54 Data Analysis SPSS 17.0 was used to analyze the data collected through the content analysis of news articles selected as samples. Frequencies and descriptive statistics were run to examine the tabulations were tested to find the relationship between nominal variables. Multivariate Analyses of Variance (ANOVA) tests were conducted to analyze the potenti al associations between studied variables. ( Table. 3 1)
55 Table 3 1. T he explanation of statistical tests for each hypothesis and research question Hypothesis or Research Question Descriptive and/or Statistical Tests RQ1a How would the news coverage of Fo xconn (2006) model of crisis life cycle and media coverage over time? Dose the model need revision? Frequencies/Timeline H1a The news coverage of Foxconn crisis in China will gain different levels of atten tion during five crisis stages in terms of the length and number of articles. Frequencies/One Way ANOVA H1b The news coverage of Foxconn crisis in Taiwan will gain different levels of attention during five crisis stages in terms of the length and number of articles. Frequencies/ One Way ANOVA H1c The news coverage of Foxconn crisis in the United States will gain different levels of attention during five crisis stages in terms of the length and number of articles. Frequencies/ One Way ANOVA RQ1b Are the re any differences between the levels of received attention in China, Taiwan and the United States in terms or the length and number of news articles? One Way ANOVA H2a The number of trigger themes/issues that covered in the news of Foxconn crisis in Chi na will differ during five crisis stages. Frequencies/ One Way ANOVA H2b The number of trigger themes/issues that covered in the news of Foxconn crisis in Taiwan will differ during five crisis stages. Frequencies/ One Way ANOVA H2c The number of trigger themes/issues that covered in the news of Foxconn crisis in the United States will differ during five crisis stages. Frequencies/ One Way ANOVA RQ2a Which themes/issues about Foxconn crisis were covered by the press in China, the United States and Taiw an? Frequencies/ Cross tabulation and Chi Square/Correlation RQ2b Are there any differences between the number of trigger themes/issues in China, Taiwan and the United States? One Way ANOVA H3 The themes/issues that covered in the news of Foxconn crisi s will differ between China, the United States and Taiwan. Frequencies/Cross tabulation and Chi Square RQ3 Do the news coverage of China, Taiwan and the United States use different crisis news frames when discussing Foxconn crisis? And if so, which one d id they use respectively? Frequencies/Cross tabulation and Chi Square
56 Table 3 1. Continued Hypothesis or Research Question Descriptive and/or Statistical Tests RQ4 Do es the news coverage of China, Taiwan, and the United States use different news sou rces when covering the Foxconn crisis? And if so, which one did they use respectively? Frequencies/Cross tabulation and Chi Square H4 The news coverage of Fox co nn crisis enjoyed different level of editorial freedom in China, Taiwan, and the United State s in terms of news sources and issues covered. Frequency/ One Way ANOVA
57 CHAPTER 4 FINDINGS Research Question 1a According to the model provided by Howell and Miller (2006), the number of the trigger themes covered, the number of articles, and the length of news stories should differ during five stages. The model predicts th at the media coverage will peak in the second stage and will decrease throughout the rest of the crisis life cycle. They also noted that news media content has various patterns in different crisis stages, such as the number and the type of trigger themes. After examining the results of the Foxconn (2006) model is able to partially explain the media coverage of the Foxconn as a cross national conflict, the Foxconn crisis provides a diffe rent perspective than that of the Ansett Airline crisis, which was considered to have had an impact only within Australia. This study focuses on the impact of Foxconn Phase One: Signal Detection (Mitroff, 1996) The result of the Foxconn data shows that the number of articles ( N = 3) among three news agencies were very few in comparison to the other stages (Figure 5 1). Consistent with the model, news media paid little attention to the Foxconn employee sui cides with reporting a small h indicates that the news media simply reported the employee suicides as traditional or ordinary news stories. However, there are few trigger themes ( N = 5) covered in this stage, nor was these statistical significance through ANOVA found between the numbe r of trigger themes and stages. The themes covered by the
58 news media during this period show that they have not yet arrived but have the potential to becoming a crisis if badly managed. Phase T wo: P reparation and P revention In the stage two, the news stori es of the Foxconn N = 6). Instead of beginning to focus on one trigger theme, as the model states, the ne ws coverage of the Foxconn crisis explored new suicide related themes; for Apple as well as assigning responsibility (Table 5 1). Even though the number of arti cles did increase in the second stage because more suicides occurred in April, the coverage of the Foxconn the model; as it was considered to be less in comparison to the third and fourth stages. This indic ates that the news media paid more attention to Foxconn the second, as the model predicts. In the case of the Foxconn treated its suicides as ordinary social events until the number of s uicides soared in the next stage. Phase T hree: A cute and C ontainment This stage is considered a period that trigger issues brea k out and evolve into an actual crisis. Consistent with the model, the fiscal, physical, or emotional damage to the Foxconn v and its stakeholders was reported in the mass media during this very short and most intense stage. As the model predicts, specific trigger themes of the Foxconn media, and their coverage included extensive discussions of t he crisis. In addition to the Foxconn suicides theme (which appeared in 98% of the stories in this stage), the news media the series of suicides. Those trigger t hemes may serve as an important medium for audiences to
59 Foxconn Foxconn and the Chinese government to stop the continued suicide attempts among Foxconn workers. Although the model of Howell and Miller (2006) predicts that the largest amount of article related to will appear in the second stage, the peak of article amount of the Foxconn crisis is in the third stage (Figure 4 1). The high level of coverage reflects the intensity of this stage in the crisis life cycle. There were a total of six people who committed suicide during this stage, which is also the highest number through out the whole crisis life cycle. Phase F our: L earning, D amage C ontainment, and R ecovery In this stage, the attention of news media coverage and the number of news stories on the crisis event is decreasing. While the number of news stories is decreasing, th e number of trigger themes covered increased (Figure 4 (2006) model that the crisis will be revisited by news media while investigating the casual relations of the crisis and trigger themes as well trigger themes. In the Foxconn (Table 4 1). The n ews media also included a factory strike on Honda Vehicles in China (which is a Japanese car maker) into news stories of the Foxconn suicide crisis in order to bring up the s increased as well. Foxconn reactions toward crisis were also covered in news reports on themes such as salary increases, additional entertainment facilities in Foxconn plants, and changing management styles. However, the data shows that the issue of th labor
60 Phase F ive: R esolution, R ecovery, and L earning According to the model, mass media in the final stage focuses on the resolution of trigger themes and summarizes the crisis a s a caveat for the future. As the model predicts, the Foxconn 2). Although the news medi the issues to the labor right concern in China, which moved to the a latent status. The results of Foxconn reactions for resolving this crisis were examined as well as the response from Chinese government. Consistent with the model, the number and the length of news stories in this stage is smaller in comparison to the other four stages, which suggests that the news media gave less attention on the Foxconn impact of Foxconn Hypothesis 1a The first hypothesis predicts that the news coverage of the Foxconn different amounts of attention in each stage during the crisis life cycle in terms of article length and num ber. The content analysis evaluated 62 news stories that related to the crisis from Xinhua news agency in China. Although the relationships between news length and crisis stages do not have statistical significance ( df = 4/35, f = .517, p =.724) through th e performance of Analysis of Variance (ANOVA), the means of article length show that the first and the final stages tend to have the least amount of words (stage 1, M = 244; stage 2, M = 454; stage 3, M = 440.14; stage, 4 = 400.25: stage, 5, M =206.67). De spite the relationship between article length and the stage, this does not reflect the attention that the crisis draws from the news media. The difference between the number of articles in each of the five stages suggests that the news media in China
61 devot ed greater attention and interest to the Foxconn suicides issue in the third and fourth stages (Figure 4 3). The ANOVA between number of news stories and stages was unable to run due to the way the data was set up. Therefore, Cross tabulations with Chi S quare tests of news agencies and stages are used to further exploring the relationship between the number of articles from selected news agencies and each crisis stage. Significant association was found between these two variables ( X 2 (8, N = 162) = 29.877 p = .000). (Table 4 2) Hypothesis 1b The hypothesis is partially supp orted, which states that the news coverage of the Foxconn through examining the article length and number. A total of ninety four news stories from the Central News A gency were evaluated and ANOVA was used to find the associations between article length and crisis stages in Taiwan. This relationship was non significant (df = 4/85, f= 1.488, p = .213). However, the number of news stories reveals the difference in attent ion the Taiwanese media gave to the Foxconn crisis during its whole life cycle. The Central News Agency has the highest number of news stories in the third crisis stage which reflects the intensity of the actual crisis period. The article number in this s tage accounts for over half the amount of the reports for the Foxconn crisis from CNA. This shows that the CNA offered more extensive coverage and attention about the crisis in stage three (Figure 4 4). The result of Cross tabulation with Chi Square tests also partially supports the hypothesis that there are significant relationships found between news number and crisis stages ( X 2 (8, N = 162) = 29.877, p = .000). (Table 4 2)
62 Hypothesis 1c Hypothesis 1c is partially supported, which foretells the difference of attention from news media will exist between each crisis stage in terms of the article length and number. There are total of thirty two articles which were coded and analyzed by ANOVA to assess the associations between article length and crisis stage. Although there is no significance found between these two var iables (df = 3/28, f = .364, p = .779), the mean of article length in each stage indicates the difference of space the Associated Press gave to the Foxconn crisis (stage 2, M = 431; stage 3, M = 494.67; stage 4, M = 540.13; stage 5, M = 284). Concerning t he number of articles may reflect the interest of news media on Foxconn crisis, the result of frequency analysis shows that there is no news coverage in the first stage from AP. The highest number of stories occurred in the fourth stage, which indicates t hat AP has gave more attention to the Foxconn crisis at the stage four (Figure 4 5 ). The significant association also has found between number of articles from each news agency and crisis stages though Cross tabulation with Chi Square tests (X2 (8, N = 16 2) = 29.877, p = .000). (Table 4 2) Research Question 1b Length of N ews A rticles As shown in Fig ure 4 6, there is a clear difference between the length of article from each news agenc y The length of the articles is significantly larger in the AP than in Xinhua and CNA ( df = 2/159, f = 11.971, p = .000). Although the length of the news coverage might depend on the news resources the organizations have, the expectations of services from different target clients, t he larger length of the news coverage also suggest s that the news agency might devote greater space to the story; in other words, this phenom enon indicates that the news agency is considered more comprehensive and provides more details about the news events.
63 Number of N ews A rticles In terms of total published story frequency by each news agency selected, the CNA has the highest number of news s tories followed by Xinhua and AP, in that order (referred to Figure 4 7 for the number of article in three news agencies). The home news agency of Foxconn CNA, has more coverage about the crisis; whereas the news coverage in China was less than expected Although the Xinhua coverage was located where the crisis occurred, the Chinese government prohibited the news media from reporting the Foxconn suicides in May. According to CNA (2010), the Chinese government requested the news media in China to limit t when covering the Foxconn crisis. The different number of news articles published in each news agency has revea led the interest and attention that news media paid on the crisis event in certai n level. Hypothesis 2a The hypothesis predicts a difference between trigger themes/issues that are covere d in the news of the Foxconn types. As shown in Figure 4 8, the number of themes covered in the news stories peaked in the stage three which reflects the intensity of the actual cr isis stage. According to the model of Howell and Miller (2006), the news media will isolate one trigger theme as the main explanation of a crisis. Although there are still various themes cov ered by Xinhua, the results of c ross tabulation of trigger themes and stages show that there are only two themes that dominant in this isis stages. However, the ANOVA analysis reveals that there is no significant relation between number of themes in individual article and its crisis stages (df = 4/35, f = .327, p = .858).
64 Hypothesis 2b Hypothesis 2b forecasts that the relationship between the number of trigger themes covered by news agency will vary in each crisis stage. There is no signifi cance found between number of trigger themes in each news article and crisis stages in the analysis of ANOVA (df = 4/ 85, f = .433, p = .784). Even though it is not statistically significant, a small difference in the frequency of themes covered by CNA is visible between each stage. The Figure 5 9 presents the number of themes reported in the news stories from CNA in Taiwan, which reached its highest level in the stage three. The coverage of this stage tends to mirror the intensity of suicides occurring wi thin stage three, that lead the certain trigger themes to become a mature crisis. According to the model of Howell and Miller (2006), the news media content in this stage will focus on specific trigger themes as the crisis prodrome and will also have exten sive discussions about crisis. The coverage of the Foxconn explanations of the crisis. The Cross tabul ation with Chi Square test shows significant X 2 (4, N = 90) = 14.235, p = .007), which is the only theme that is increasingly covered from stage three to stage four. Hypo thesis 2c While examining the data results, there is no significance found between number of themes that are covered by each news story and crisis stage by performing the result of ANOVA analysis (df = 3/ 28, f = 2.128, p = .119). However, even though there is no statistical significance, the frequency of the number of themes covered in each stage has small visible differences.
65 Figure 4 10 indicates that the number of themes covered in each stage peaked in both stages three and four in AP. There is no coverage about the Foxconn crisis from the AP in the consider the crisis as important to its audience until the second stage. This may be explained by the Foxconn Apple resonance with the American public. Compared to the other three stages, stage five has the least number of themes covered in the news articles from AP. Research Question 2a Five main crisis themes were primarily covered by the news media in China, Taiwan and the United States during whole crisis life cycle, and t issues by appearance frequency are: (1) e mployee suicide issue (85%), (2) Foxconn reactions (50%), and (3) Chinese government reactions (47.5%). For the Taiwanese CNA, the frequencies of three prominent themes are: (1) employee suicide issue (96.7%), (2) Foxconn reactions (48.9%), and (3) Foxco nn labor mistreatment (41.4%), whereas for the AP in the United States, the top three themes were: (1) Employee suicide issue (84.4%), (2) Foxconn reactions (71.9%), (3) Foxconn labor mistreatment (65.6%) and Labor rights concern in China (65.6%). ( Tabl e 4 3) The Cross tabulation with Chi Square test was performed to assess the relationships between the themes covered and the three news agencies selected. Significant differences were labor mist X 2 (2, N =162) = 12.108, p X 2 (2, N =162) = 7.237, p
66 X 2 (2, N =162) = 23.984, p ( X 2 (2, N =162) = 14.106, p Apple X 2 (2, N =162) = 7.743, p X 2 (2, N =162) = 6.332, p X 2 (2, N =162) = 40.625, p X 2 (2, N =162) = 10.3 01, p = .006) that covered during crisis life cycle. As shown in Table 4 Apple than the news stories from CNA and Xinhua. The data shows that the AP tends to associate Apple and companies who consume Foxconn reflect the proximity of Apple and Foxconn for the maj crisis, whereas the AP covered it for over half of its news articles. Among the three news agencies, Xinhua of China gave much more its attention to concerned about fiscal damage gained less coverage in China in terms of the frequency; themes such as and Ap ple resonance of the Foxconn crisis with its home country resulted in more coverage of the Research Question 2b In order to understand how news media in the China, Taiwan and the United States covering the Foxconn crisis in terms of the number of the themes, ANOVA analysis was performed to assess the relationship between number of trigger themes and three news agencies. This relationship was found significant (df = 2/ 159, f = 25.038, p = .000). The Figure 4 11
67 indicates that the AP has covered more themes of Foxconn crisis in each news stories, whose Mean is 4.63, followed by CAN (M = 2.94) and Xinhua (M = 2.70). A correlation with the two tailed test was performed to further explore the relationship between number of the themes covered and news agencies. Moderate negative correlation w as found between number of articles generated from each news agency and the number of themes appeared in an articles ( r = .423, p < .000). The correlation indicates that, in a moderate level, the more themes included in a news story, the lesser the n umber of articles reported ( Table 4 4). Hypothesis 3 The hypothesis three is partially supported, which predicts the di fference of themes covered between the news media of three countries involved in the Foxconn crisis. The Cross tabulation with Chi Square teats were used to analyze the possible relationships foretold by this hypothesis. Significances were found between n ews agencies between eight themes, which are: Apple differences can also be found through frequency analysis. (Table 4 3 ) The top three themes that are covered by AP are: (1) Employee suicide issue (84.4%), (2) Foxconn react ions (71.9%), (3) Foxconn labor mistreatment (65.6%) and Labor rights concern frequ ency. In Xinhua, the first two themes are approximately the same as covered in the AP and
68 nearly half of the news articles from Xinhua. In addition to the differences of the top three themes that are covered in each news agency, there is a significant difference between three news while there is no coverage in Xinhua. The AP als o had many more news stories including the themes that were related to the Foxconn Apple appeared only in the news co verage from CNA. There is only AP associated the labor strike issue of Honda with Foxconn employee suicide crisis. Research Question 3 In terms of the crisis frames used when discussing the Foxconn crisis, the result of Cross tabulation analysis shows a significant diffe rence between news agencies ( Table 4 5). A significant associat ion was found between frames used and news agencies in the result of Cross tabulation with Chi Square tests (X2 (8, N = 162) = 32.038, p = .000). In AP, the major frame d by crisis. Wi thin three news agencies, Research Question 4 Cross tabulation results reveal that the news coverage of AP, CNA, and Xinhua differed significantly in terms of sources adopted. Chi Square tests were performed with cr oss tabulation analysis to help evaluate the relationships between sources cited in each of the three news media.
69 Apple three newswire services. In add ition, the significance was also found between the number of sources used and news agencies selected (df = 2/ 161, f = 7.196, p = .001). According to the result of ANOVA analysis, the AP tends to include more news sources in a news article while Xinhua has the least (Table 4 6). As shown in Table 5 Foxconn Apple were never adopted in the news article from China. Furthermore, the AP and CNA cited the additional sources from Honda, a Japanese company which has the similar labor issue as Foxconn and other Taiwanese companies operate in China. Hypothesis 4 This hypothesis forecasts that the level of editorial freedom differed between news agencies of different countries and may reflect on the news sources used, as well as the themes covered. The hypothesis is supported as the clear significances shown in the relationships between number of themes covered (df = 2/ 161, f = 25.036, p = .000) and number of sources
70 used (df = 2/ 161, f = 7.196, p = .001) and three news agencies. AP tended to include various sources and different themes in the news coverage of Foxconn crisis while the Xinhua covers less themes and news sources (News sources: A P= 11, CNA= 12, Xinhua= 6). For the news sources used, Xinhua used only six sources to report Foxconn articles. News sources such Apple cited in the articles of Xinhua news agency. Post Hoc Hypothesis For further exploration into the relationships between data of Foxconn crisis, an additional Cross tabulation analysis with Chi Square test was conducted. The relationship between ne ws frames used and crisis stages was found highly significant (X2 (16, N = 162) = 43.820, p = .000). According to the Table 4 7, when covering a cross national conflict shifting (CNCS) as Foxconn 3.3%), and was primarily used to cover the crisis in the second stage, while stage three focuses on the
71 Table 4 1. Theme frequencies in each crisis stage Themes covered Crisis Stages* Stage1 Stage2 Stage3 Stage4 Stage5 Frequency Percentage Frequency Percentage F requency Percentage Frequency Percentage Frequency Percentage Foxconn labor mistreatment 1 33.3% 3 50% 38 44.18% 24 40.6% 2 25% Employee suicide 3 100% 6 100% 84 98% 50 84.7% 5 62.5% Labor right concern in China 0 0.0% 1 16.6% 21 24% 25 42.3% 2 25% F inancial performance 0 0.0% 0 0.0% 8 9.3% 19 32.2% 3 37.5% Foxconn and Apple business impact 0 0.0% 0 0.0% 4 4.6% 4 6.7% 1 12.5% Apple issue 0 0.0% 2 33.3% 3 3.4% 4 6.7% 0 0.0% Reactions of Foxconn 0 0.0% 0 0.0% 6 6.9% 4 6.7% 0 0.0% Activist group 0 0.0% 0 0.0% 10 11.6% 17 28.8% 0 0.0% Reaction of Taiwanese company in China 0 0.0% 0 0.0% 4 4.6 1 1.6% 0 0.0% Foxconn reactions 3 100% 3 50% 37 43% 40 67.7% 4 50% Chinese government reactions 1 33.3% 0 0.0% 32 37.2% 11 18.6% 2 25 % Taiwanese government reactions 0 0.0% 0 0.0% 4 4.6% 3 5% 0 0.0% Investigation or results of labor mistreatment 1 33.3% 1 0.0% 18 20.9% 4 6.7% 0 0.0% Other 0 0.0% 0 0.0% 0 0.0% 3 5% 0 0.0%
72 Table 4 2. Cross tabulation of news agencies a nd crisis stages Crisis Stage Stage1 Stage2 Stage3 Stage4 Stage5 Total News Agency AP Count 0 1 6 24 1 32 % within News Agency .0% 3.1% 18.8% 75.0% 3.1% 100% % within Crisis Stage .0% 16.7% 7.0% 40.7% 12.5% 19.8% % of Total .0% .6% 3.7% 14.8% .6% 19.8% CNA Count 1 4 58 23 4 90 % within News Agency 1.1% 4.4% 64.4% 25.6% 4.4% 100% % within Crisis Stage 33.3% 66.7% 67.4% 39.0% 50.0% 55.6% % of Total .6% 2.5% 35.8% 14.2% 2.5% 55.6% Xinhua Count 2 1 22 12 3 40 % within News Agency 5. 0% 2.5% 55.0% 30.0% 7.5% 100% % within Crisis Stage 66.7% 16.7% 25.6% 20.3% 37.5% 24.7% % of Total 1.2% .6% 13.6% 7.4% 1.9% 24.7% Total Count 3 6 86 59 8 162 % within News Agency 1.9% 3.7% 53.1% 36.4% 4.9% 100% % within Crisis Stage 100% 100% 100% 100% 100% 100% % of Total 1.9% 3.7% 53.1% 36.4% 4.9% 100% Table 4 3. Themes covered in each news agency Themes covered News Agencies* AP CNA Xinhua Frequency Percentage Frequency Percentage Frequency Percentage Foxconn labor mistreatment 21 65.6% 37 41.1% 10 25.0% Employee suicide issue 27 84.4% 87 96.7% 34 85.0% Labor right concern in China 21 65.5% 18 20.0% 10 25.0% Financial or business performance 13 40.6% 14 15.6% 3 7.5% Foxconn and Apple business impact 2 6.3% 6 6.7% 1 2.5% A pple (supplier issue) 5 15.6% 3 3.3% 1 2.5% Reactions of Foxconn clients 5 15.6% 4 4.4% 1 2.5% Activist group 17 53.1% 10 11.1% 0 0.0% Reaction of Taiwanese company in China 0 0.0% 4 4.4% 1 2.5% Foxconn reactions 23 71.9% 44 48.9% 20 50.0% Chinese government reactions 9 28.1% 18 20.0% 19 47.5% Taiwanese government reactions 0 0.0% 7 7.8% 0 0.0% Investigation or results of labor mistreatment 3 9.4% 13 14.4% 8 20.0% Other 3 9.4% 0 0.0% 0 0.0%
73 Table 4 4. Correlations between news agency and number of themes News Agency Number of Themes News Agency Pearson Correlation 1 .423** Sig. (2 tailed) 0 N 162 162 Number of Themes Pearson Correlation .423** 1 Sig. (2 tailed) 0 N 162 162 **. Correlation is significant a t the 0.01 level (2 tailed). Table 4 5. Frames used by news agencies Frames News Agencies AP CNA Xinhua Frequency Percentage Frequency Percentage Frequency Percentage Conflict frame 7 21.90% 7 7.80% 3 7.50% Human interest frame 0 0.00% 13 14.40% 13 32.50% Economic consequence frame 12 37.50% 19 21.10% 7 17.50% Responsibility frame 6 18.80% 32 35.60% 3 7.50% Moral frame 7 21.90% 19 21.10% 14 35.00% Total 32 100% 90 100% 40 100% Table 4 6. Source frequencies in the news coverage of three news agencies News Sources Used News Agencies* AP CNA Xinhua Frequency Percentage Frequency Percentage Frequency Percentage Foxconn 21 65.6% 43 47.8% 23 57.5% Chinese government officials 8 25.0% 18 20.0% 21 52.5% Taiwanese government officials 0 0.0 % 7 7.8% 0 0.0% Foxconn workers, their families and friends 5 15.6% 11 12.2% 6 15.0% Activist groups 15 46.9% 9 10.0% 0 0.0% Researchers 9 28.1% 18 20.0% 8 20.0% Apple and other clients 5 15.6% 5 5.6% 0 0.0% Other media 10 31.3% 40 44.4% 1 2.5% Communities/Residents 2 6.3% 5 5.6% 2 5.0% Financial agencies 1 3.1% 3 3.3% 0 0.0% Internet user and unidentified 1 3.1% 10 11.1% 0 0.0% Others 4 12.5% 4 4.4% 0 0.0%
74 Table 4 7. Frames used in each crisis stage Article Frame Total Conflict f rame Human interest frame Economic consequence frame Responsibility frame Moral frame Crisis Stage stage1 Count 1 1 0 1 0 3 % within Crisis Stage 33.3% 33.3% .0% 33.3% .0% 100.0% stage2 Count 0 3 0 1 2 6 % within Crisis Stage .0% 50.0% .0% 16.7% 3 3.3% 100.0% stage3 Count 10 15 9 33 19 86 % within Crisis Stage 11.6% 17.4% 10.5% 38.4% 22.1% 100.0% stage4 Count 6 6 25 5 17 59 % within Crisis Stage 10.2% 10.2% 42.4% 8.5% 28.8% 100.0% stage5 Count 0 1 4 1 2 8 % within Crisis Stage .0% 12. 5% 50.0% 12.5% 25.0% 100.0% Total Count 17 26 38 41 40 162 % within Crisis Stage 10.5% 16.0% 23.5% 25.3% 24.7% 100.0% ( X 2 (16, N = 162) = 43.820, p = .000)
75 Figure 4 1. News frequency of three agencies by crisi s stage Figure 4 2. The number of themes by stage Figure 4 3. News frequencies of Xinhua by crisis stage
76 Figure 4 4. News frequency of CNA by crisis stage Figure 4 5. News frequencies of AP by cr isis stage Figure 4 6. The mean of article length of three news agencies
77 Figure 4 7. The number of article in three news agencies Figure 4 8. The number of themes covered in news articles from Xinhua Figure 4 9. The n umber of themes covered in news articles from CAN
78 Figure 4 10. The number of themes covered in news articles from AP Figure 4 11. The number of themes covered in news agencies
79 CHAPTER 5 DISCUSSION Discussion and Implications of the Media Model for the Crisis Life Cycle Because the research was based on and aimed to advance the model of relations between media coverage and crisis life cycle, the Foxconn crisis is indicative for testing the model in an international environment. Researchers (Freitag, 2002; Molleda, 2011; Molleda & Quinn, 2004; Valentini & Romenti, 2011) have noted in previous studies that it is important to consider the differences between media systems and cultural contexts. Following the suggestions of those stu dies (Freitag, 2002; Molleda, 2011; Molleda & Quinn, 2004; Valentini & Romenti, 2011), this study focused on the role of international news agencies for they are influential vehicles and channels that move corporate crises from one place to another in the globe. From the inter media agenda setting perspective, international news agencies, such as AP, AFP, CNA, Reuters and Xinhua have largely covered cross national crises and may have an impact on domestic news reports regarding the crisis (Golan, 2006; Moll eda, 2011). The Howell and Miller (2006) media coverage and crisis life cycle model was the foundation of this study. As it predicts, the differences between news frequencies number of themes covered, and frames and sources used exist between crisis stage s. Although there are some differences, the model has explained and guided the evolution of media coverage in the Foxconn crisis. The content analysis result revealed that the news coverage of the Foxconn suicides peaked in the third stage while the mode l suggests that the highest level of news frequency should appear in the second phase. Cross national crisis management should take these various factors into account, including the news practices in the countries involved. As shown in Figure 5 1, the peak s of news coverage are different between AP, and CNA and Xinhua. In AP, news frequency reached the highest level in the fourth phase; while the peak of
80 news coverage in CNA and Xinhua appeared in stage three. The United States is the least affected country by the Foxconn crisis among the three According to Freitag (2002), media content of the issue will differ through the news editorial preferences in each country. For the Associated Press, the themes related to Apple other American companies and its ec onomic or fiscal consequences as these are the issues that have the closest proximity to the news agencies coverage of the Foxconn crisis until those themes appeare d and moved into focus in news media, which is the fourth stage of the crisis. The theme numbers and the types of the issues covered are also emphasized in this model. As researchers (Howell & Miller, 2006; Mitroff, 1996; Seymour & Moore, 2000) suggested, the key to managing a crisis is to control the message. Stephens, Malone and Bailey (2005) warned that once the organization lacks message management, it relinquishes its control of communication to the media, and the media will dominate the rest of the cr isis stages; setting agendas according to their aims and perspectives. Pearson and Clair (1998) also stated that if organizations fail or choose not to respond to trigger themes at the beginning of a crisis, news media may cover rumors instead. On the othe r hand, if the organizations recognize and can properly implement strategies to those trigger themes, the negative impact can be reduced. In the Foxconn the few themes asso ciated with the specific prodrome were indeed focused on by news media. Foxconn N = 6) at the third stage lead the coverage from simply reporting suicides to labor mistreatment. For example, according to the Central News Agency
81 China and made an apology for the series of suicides. The reaction s and strategies Foxconn adopted to manage the crisis also caught media attention during this period. As reported, the issue that Foxconn suicide brought up extensive discussions in news media as one of the controversial strategies implemented by Foxconn Central News Agency (2010) also covered a rumor which stated that Foxconn invited monks to do an exorcism on its company plants because of the suicides. In the stage four, the charges, government inquiries, litigation, and responses from affected stakeholders (Howell & Miller, 2006, p. 6). The theme types of Foxconn crisis in stage four are consisten t with the workers, such as salary increases. In stage five, the coverage of business performance and bu siness plans means the operation of Foxconn is back to normal and the trigger themes are resolved or the resolution is partially achieved through management actions. The news media summarized the Foxconn crisis as a series of incidents caused by labor mi streatment and extended the issue to the labor right concern in China, which moved into a latent status as suggested in the model. Howell and Miller (2006) noted that if the organization is unable to resolve the themes that evolve to the crisis, then those themes will move into a latent status that may be reactivated by another event and brought back to the focus in the news media. Once the trigger themes are back into prominence, they will have potential to create a new crisis cycle and cause damage to the organizations again. When developing a crisis management plan, proactive media management is required with clear and concise information released in time to respond to the requests of the media in order to stop the rumors and maintain control of message in crisis stages. Furthermore, since the media
82 have the powerful ability to communicate and spread news expeditiously while a crisis happens, effective response and good relations with the news media can prevent a crisis from developing. Scholars (Howell & Miller, 2006; Valentini & Romenti, 2011) indicated that the public will learn the importance of events from the amount of coverage in the news media. The selection of news frames in the information provided by news media is also influential in formatting advancing the model, this study may suggest that there is a trend in the frames chosen by international news media groups that should be integrated into the crisis media model. In the first and second stages, the news media simply describe the trigger themes that have potential to stories during these two stages. The organizations should be careful to manage any trigger themes with appropriate and timely responses to the events covered in the new content. For stage three, where the crisis actually crisis. This suggests the media focus has moved from simply reporting of crisis events to seeking culpability of creating the crisis. In this stage, the o rganization should also respon d to assigning the r esponsibility and be cautious when applying the strategies to resolve the problems since any reactions of related actors will be under scrutiny. The news media will focus on the use of the d fifth stages. When these themes are covered, the news media tend to turn their attention to the fiscal damage and business impact of the crisis. Additionally, any investigation of the crisis and any corporate reactions will be reported. In the Foxconn e xample, news articles covered the results of strategies adopted by Foxconn in stage three, such as the protection nets installed a round the dormitories, the rais ing
83 of salaries and the change management style. The organizations should still keep the messa ge released concisely and avoid the issues and messages that will cause the damages to organization be revisited in the future stories. The main purpose of the media model of the crisis life cycle is to clearly point out a pattern or trend in the news medi a as they report crises related, and to provide as much information as possible for prediction, such as the amount of news coverage, the type of themes, the frames and the sources the news media would use. With this, the PR practitioners and crisis managem ent teams are able to quickly develop and implement effective strategies to control the message and maintain control of media content during crisis. Discussion and Implications in Practices According to the results of the data of Foxconn s crisis, this st udy may suggest some implications and actions for practitioners based on the media model. In stage 1 and stage 2, the crisis management team should start to monitor the trigger themes that appeared in the news stories and follow the trend of news coverage since those trigger theme may have potential to become a crisis The actions respond to those trigger themes are also essential ; otherwise, the news media may cover rumors to fi ll out the spaces. As the model predicts the news media will start to focus o n few trigger themes as main explanation for the crisis in stage 3. The crisis management team should treat specific trigger theme carefully in this stage, as well as respond to the media request to maintain good relationship with news media and control of message. Therefore, a proactive media management may be needed with clear an d concise information released. In stage 4, although the real crisis stage ends in the stage 3, the crisis management team should still follow up the events, activities and mass m edia coverage of crisis to date. Keeping record of crisis events and coverage may assist organization to take advantage of previous successful management and to compensate the failures. In stage5, the practitioners should
84 prevent trigger themes move to lat ent status which might have potential to damage the organization in other crisis life cycle. Concerning of the relationship between cross national conflicts and frequent global information exchange, transnational corporations (TNCs) and organizations that are involved in the crisis should take the news practices and the media environment of different world locations into account for projecting flexible and proactive actions. As described in the literature review chapter, the three international news agenci es selected are suitable for an analysis on differences between various media environments in terms of the level of press freedom. According to the that order; their ranks being 184, 48, and 17. This ranking also reflects the news coverage of the Foxconn crisis in terms of the coverage frequency, themes covered, frames used and sources cited. News Frequencies Through this study, the trend o f news frequency provides guidance for practitioners when to release important information and adopt reactions toward the crisis. The trend of news frequency clear points out the peak of the media attention of the crisis and offers an opportunity for organ ization to release the message in a timely manner Close examination of the news articles published by three news agencies in China, Taiwan and the United States, the Central News Agency from Taiwan has the highest coverage of the Foxconn crisis ( N CNA= 9 0> N XH= 40> N AP= 32). There are three reasons that explain the coverage frequency. Firstly, the CNA and Foxconn headquarter s share a home country. As suggested in the cross national conflict shifting (CNCS) propositions, Molleda (2011) noted that the home n ews agencies, which have more resonance with the TNC involved, will offer more extensive coverage about the crisis. Compared to Taiwan and China, the United States is neither
85 the home country of Foxconn nor the location where the crisis occurred. The AP o nly covered the themes and issues that have some internal relation, including the issues related to American transnational corporations (TNCs) ; as well as the economic impact, which shows that the AP treated the Foxconn crisis from a global perspective. T he amount of coverage from the Xinhua News Agency greatly differs from the CNA even though China is the place where the crisis actually occurred. Although it was not reported in the Xinhua News Agency, the news stories from the CNA and the United States no ted that the Chinese government has put censorship on news coverage of the Foxconn crisis in May, 2010 and requested the Chinese media to use only 2010, the Chinese government prohibited the media from covering the Foxconn suicides because they regarded the media as one of the proponents that encouraged the continued suicide attempts. Although Yin (2008) noted that the Chinese media are not exactly the same as the o riginal Soviet press model, which lacks any exposs of negative news, and despite that some of the Chinese media allow negative view points by approval from the government, some sensitive be eliminated from news content. As Freedom House (2011) reported, the media environment in China remained highly repressive in 2010 for some sensitive issues like the imprisonment of the Nobel Prize winner, Liu Xiaobo. which can also curbed coverage related to public health, environmental Foxco nn
86 Themes Covered Even covering the same event, the news content will somehow differ in preferences of news agency. In the news from Xinhua, the number of themes covered is less than those included in the AP and the CNA. Despite the fast economic growth and considerable reforms in China, the Chinese social context is still consider ed under the tight control of the Chinese Communist Party. Any event considered that will affect social stability is censored. In additional to the concerning forms o (Wang, 2005, p. 84). Therefore, the news stories from Differing from practice closer to the social responsibility model. As Hatchen and Scotton (2007) stated, med ia Apple ons of Foxconn more exposure in the news from the AP than the CNA and Xinhua. News coverage from the AP often associate Foxconn with Apple Apple Appl e is more recognizable for audiences in the United States. The CNA covered various themes during the five crisis stages. It is the only news agency that covered the reactions from the Taiwanese government, which indicates its proximity with the Taiwanese public. According to the Freedom House report in 2005, Taiwa n has one of the
87 Freedom, 2006, 5). Taiwanese media are sensat ional and eager to expose scandals of business corporations, government corruptions and private life of people of high recognition. The themes ing among the most competitive media environments in Asia, the Taiwanese media tends to cover news with or without grounded sources. For example, while reporting the reactions toward the crisis from Foxconn only Taiwanese media included the Frames Used Similar to the choice of themes, the selection of news frames also differ between news agencies. The AP has more concern about financial impact, which is why they chose themes more crisis. For CNA, it preferred og that is likely to expose things that would have been hidden from public view without investigations by the media. By determining the responsibility and blame of the suicides, the Taiwanese news media took over polices and government officials responsibi lity by investigating and making judgments. The result of this study found that the Xinhua News Agency as the official news proximity of suicides, the media in China te nd not to cover the sensitive issues that would lead to conflicts endangering the social stability. Plus, in this case study, the Chinese government limited the news reports of the Foxconn crisis. Even the order of censorship or limitations by the news media in China, a self censoring atmosphere with invisible regulations affected the selection of news frames when covering the Foxocnn crisis.
88 Sources Cited For crisis management team, to know the news sources that preferred by journalists is important. The preferred news sources can serve as another channel for practitioners to release the information which has benefits to the organization. The use of sources also appeared to be different between three international news agencies. In the Foxco nn news source, Xinhua adopted the official news sources from the Chinese government more often; whereas the AP used more news sources from activist groups and the CNA included more sources from other media. As for the number of sources used, the AP and the CNA used a variety Apple and other clients of Foxconn different editorial policies of each news agency (Molleda, 2011). While the other two news agencies are independent of using more news sources from Chinese gove rnments because it is a states commissioned news agency. Freeman (2000) explained that in a tightly controlled media environment, the priority of the media is to maintain the stability of society, investigative reporting is discouraged and they are encoura ged to rely on official sources. Extremely different from the way Xinhua did in China when covering the Foxconn crisis, the CNA has twice the number of news sources. Among the es. According to Yin (2008), while the Taiwanese press enjoys the freest media environment, some of them are considered to be the source of social chaos that is often accused of making reports of rumors and covering news that is not true. If the crisis ma nagement practitioners are familiar with the media practice in different media systems, they can adjust information that are going to release in different location and to
89 avoid extra management costs for unnecessary mistakes. Knowing the dynamic nature of media practices in different social and cultural settings helps us to find a clear pattern for news coverage about an event, especially important for cross national crisis management. Limitation and Future Research This research aims to assist both academi c studies and practitioners to better understand the trends in news media when covering the crisis throughout its life cycle. While a crisis process is viewed as one continual improvement, understanding and tracking media content change during crisis stage s are vital to control the mass mediated message. This study illustrates and characterizes the content of media coverage from the aspect of cross national crisis management. To advance the media model developed by Howell and Miller (2006), the research add ed different components into the original model, such as news frames and sources used. This revised media model of the crisis life cycle provides guidance for when to adopt appropriate strategies as well as the predictions of news media practices to a cris is event. However, there are some limitations for the research. The media content analysis was limited to the selected news samples retrieved from online databases. Besides, the criteria those databases used for collecting the news may affect the profile of news samples. Thus the result based on these news coverage may not simulate to border media coverage. At the same time, the amount numbers of news articles related to the crisis are also confined by the news agencies selected. Different crises types, th e locations involved and the organizational behaviors of each news agencies/newspaper may well influence the coverage of the crisis. Like a double sided sword in research, these limitations can also act as suggestions for future study in a similar area. In order to refine the media model, these key elements of news practices in local news agencies should be further examined and studied.
90 Furthermore, the inter media agenda setting theory has received much attention for explaining its applications to internat ional news coverage to domestic news practices. The question of the gap between coverage from international news agencies and local newspapers in terms of crisis issue selection is still a field needing further exploring. The future research should include possible work of case studies with different sampling methods and samples to test and amend the media model expanded by this research. Testing different samples such as major newspapers is recommended to further improve the result of this study. Since the research has already considered the differences between media systems as a variable that may affect the outcome of media coverage of a crisis, additional research with a focus on detailed preferences or choice of themes, frames and sources of each media e nvironment are also needed from the perspective of the crisis. Additionally, the model should also be tested for the different types of cross national crises. For example, there might be a difference in the news practices between the types of mass product recall, a boycott crisis related to human interests or a cultural clash. The differences between the amount of news, themes, frames and sources used for the various types of cross national crises should be further studied. Last but not least, the applicat ion of this media model should be tested integrated with the crisis management strategies throughout the crisis life cycle. The combination of crisis responses from involved organizations takes a huge step toward to the practical application. The model can be improved by examining the real crisis response situations. After all, the original idea about developing this model was to better understand the news media practices in a crisis and enable crisis management teams to make efficient and well prepared cri sis management plans. Thus, it is important to do a case study of the best practices of crisis management. This may provide efficient and practical guidelines for this complicated and dynamic global issue.
91 Figure 5 1. News f requency of t hre e a gencies by s tage
92 APPENDIX CODING SHEET Column Variable label Value Labels Column Record 1 3 Identification Number 101 300 4 Coder Identification 1 = Coder 1 2 = Coder 2 5 Blank 6 News Agency 1 = AP 2 = CNA 3 = Xinhua 7 Crisis Stage 1 1 = Stag e 1 2 = Stage 2 3 = Stage 3 4 = Stage 4 5 = Stage 5 8 11 Length Words 2 0000 5000 12 Blank 13 Foxconn Labor Mistreatment 0 = No 1 = Yes 14 Employee suicide issue 0 = No 1 = Yes 15 Labor right concern in China 0 = No 1 = Yes 16 Financial or bu siness Performance 0 = No 1 = Yes 17 Foxconn or Apple business market Impact 0 = No 1 = Yes 18 Apple issue) 0 = No 1 = Yes 19 Reactions of Foxconn 0 = No 1 = Yes 20 Activist Group 0 = No 1 = Yes 21 Reacti on of Taiwanese Company in China 0 = No 1 = Yes 1 The coder should follow this instruction according to the stages defined in literature review: Stage 1: January 23, 2010 to Februa ry, 28, 2010 Stage 2: March 1, 2010 to April, 30, 2010 Stage 3: May 1, 2010 to May 31, 2010 Stage 4: June 1, 2010 to September 30 Stage 5: October 1, 2010 to December, 2010 2 The number of the words in an article can be calculated both in English and Chine se
93 22 Foxconn Reactions 0 = No 1 = Yes 23 Chinese Government Reactions 0 = No 1 = Yes 24 Taiwanese Government Reactions 0 = No 1 = Yes 25 Investigation or results of labor mistreatment 0 = No 1 = Yes 26 Other 0 = No 1 = Yes 27 28 Number of Themes 01 98 29 Blank 30 Conflict Frame 3 0 = No 1 = Yes 31 Human Interest Frame 4 0 = No 1 = Yes 32 Economic Consequence Frame 5 0 = No 1 = Yes 33 Responsibility Frame 6 0 = N o 3 to conflict frame. Does the news story emphasize the conflict between Foxconn and its workers? 4 n event, issue or If the is categorized as human interest frame : Does the news st ory tell stories of a specific human or humans? Does the news story have an emotional angle to present to issue in an effort to personalize or dramatize the news event, by using words that express feelings such as a sense of loss and suffering? 5 The econ If the answers of the following question have one or is categorized as e conomic consequences frame : Does the story mention the compensation Foxconn gave to the workers who died or were injured at its factories or to their families? Does the story mention the financial impacts on Foxconn ? 6 The responsibility frame attribu uman interest frame: Does the news article suggest that the Chinese government is responsible for deaths at Foxconn ? Does the news article suggest that management of Foxconn is responsible for deaths at Foxconn ? Does the news article suggest that the victims themselves are responsible for deaths at Foxconn ?
94 1 = Yes 34 Moral Frame 7 0 = No 1 = Yes 35 Blank 36 Foxconn 0 = No 1 = Yes 37 Chinese Government Officials 0 = No 1 = Yes 38 Taiwanese Government Officials 0 = No 1 = Yes 39 Foxconn Workers, Their families and Friends 0 = No 1 = Yes 40 A ctivist Groups 0 = No 1 = Yes 41 Researchers 0 = No 1 = Yes 42 Apple and Other Clients 0 = No 1 = Yes 43 Other Media 0 = No 1 = Yes 44 Communities/Residents 0 = No 1 = Yes 45 Financial Agencies 0 = No 1 = Yes 46 Internet User and Unidentified 0 = No 1 = Yes 47 Others 0 = No 1 = Yes 7 n the article is categorized as moral frame: Does the news article raise the questions of how the Chinese government should behave? Does the news article raise the questions of how the management of Foxconn should behave? Does the news article raise th e questions of how Foxconn Does th e news article raise the questions of how the Apple should behave? Does the news article raise the questions of how society should behave?
95 LIST OF REFERENCES Agence France Press. (2011). Foxconn workers treated like 'machines': labor group Retrieved September 20, 2011, from Taipei Times: http://www.taipeitimes.com/News/biz/archives/2011/05/04/2003502331 An, S. K., & Gower, K. K. (2009). How D o the N ews M edia F rame C rises? A C ontent A nalysis of C risis N ews C overage. Public Relations Revi ew 35(2), pp. 107 112. Apple ( 2011 ). Supplier Responsibility Retrieved October 11, 2010, from Apple Inc. : http://www.apple.com/supplierresponsibility/ Babbie, E. R. (2010). The Practice of Social Research (12 ed.). Belmont, Calif: Wadsworth Cengage. Barton, L. (1993) Crisis in O rgani z ations: Managing and C ommunicating in the H eat of C haos. C incinnati, Ohio: South Western Publishing Company. BBC. (2010, Jun 2). News Technology Retrieved October 24, 2011, from BBC: http://www.bbc.co.uk/news/10212604 Berry, W. E., Braman, S., Christians, C., Guback, T. G., Helle, S. J., Liebovich, L. W., et al. (1995). Last Right: Revisiong Four Theories of Press. Urbana: University of Illinois Press. Brinson, S. L., & Benoit, W. L. (1996) I mage R epair S trategies in the B reast I mplant C risis. Communication Quarterly, 44(1), pp. 29 41. Carroll, C. E., & McCombs, M. (2003). Agenda S etting E ffects of B usiness N ews on the P I mage and O pinions a bout M ajor C orporations. Corporate Reputation Review 16 ( 1), pp. 26 36. Central News Agency (2010, May, 13). Rumor Says Foxconn Calls on Monks to Stem Suicides. Central News Agency Retrieved October 26, 2010, from CNA Database Central News Agency (2010, May, 26).Employees Against to Sign the Anti suicide Pledge. Central News Agency Retrieved October 26, 2010, from CNA Database Central News Agency (2010, May, 27).The 12 th Suicide Shocked Chinese Society. Central News Agency Retrieved from CNA Database Chang, S. (2010, May 29). China Government Request s Censorship On Foxconn News Retrieved 10 10, 2011, from MICgadget: http://micgadget.com/4418/china government requests censorship on foxconn news/
96 Cho, S. H., & Gower, K. K. (2006). Framing E ffect on P R esponse to C risis: Human I nterest F rame an d C risis T ype I nfluencing R esponsibility and B lame. Public Relations Review 32(4), pp. 420 442. Coombs, W. T. (1995). Choosing the right words: The development of guidelines for the response strategies. M anagement Communication Quarterly 8, 447 476. Coombs, W. T. (1999 ). Ongoing crisis communication: Planning, managing, and responding Thousand Oaks, CA: SAGE Coombs, W. T. (2006). Crisis M anagement: A C ommunicative A pproach. In C. H. Bota n, & V. Hazleton (Eds.), Public R elations T heory (pp. 171 197). Mahwah, NJ: Lawrence Erlbaum Associates Cu rtin, P. A., & Gaither, T. K. ( 2007). International public relations: Negotiating culture, identity, and power. Thousand Oaks, C A: Sage Publications Daily Mail. (2006, Aug 18). The stark reality of iPod's Chinese factories Retrieved September 25, 2011, from MailOnline: http://www.dailymail.co.uk/news/article 401234/The stark reality iPods Chinese factories.html Dean, J. (2007, Aug 11). The Forbidden City of Terry Gou Retrieved October 15, 2011, from The Wall Street Journal: http://online.wsj.com/public/article/SB118677584137994489.html?mod=blog Dean, J., & Tsai, T. I. (2010, May 27). Suicides Spark Inquiries Retrieved October 21, 201 1, from The Wall Street Journal: http://online.wsj.com/article/SB10001424052748704026204575267603576594936.html Entman, R. B. (1993). Framing: Toward C larification of a F ractured P aradigm. Journal of Communication 43, pp. 51 58. Evans, J. (2006, Jun 14) Apple responds to iPod factory claims Retrieved October 2, 2011, from MacWorld: http://www.macworld.co.uk/mac/news/?newsid=14935 Fearn Bank K (200 1). Crisis Communication: A Review of Some Best Practices In R. L. Heath, Handbook of public relations ( pp. 26 9 289 ). Thousand Oaks: Sage Publications. Fink, S. (1986). Crisis M anagement: Planning for the I nevitable New York, N.Y: American Management Association. Foreman, W. (2010, May 27). Latest Foxconn suicide comes after visit by boss Retrieved Octob er 17, 2011, from Bloomberg Businessweek: http://www.businessweek.com/ap/financialnews/D9FUVFTO0.htm
97 Fortune Magazine. (2010). Global 500 Retrieved October 20, 2011, from CNN Money: http://money.cnn.com/magazines/fortune/global500/2010/snapshots/11204.ht ml Freedom House. (2011). Freedom of the Press 2011. Retrieved October 27, 2010, from http://www.seguridadcondemocracia.org/administrador_de_carpetas/biblioteca_virtual/pd f/Freedom_of_the_press_2011.pdf Freeman, L A (2000) Closing the S hop: Informati on C artels and Japan's M ass M edia. Princeton, NJ: Princeton University Press Freitag, A. (2002). International media coverage of the Firestone tyre recall. Journal of Communication Management 6 (3), pp. 239 25. Golan, G. (2006). Inter Media Agenda Set ting and Global News Coverage. Journalism Studies, 7 (2), pp. 323 333. Gonzalez Herrero, A., & Pratt, C. B. (1996). An I ntegrated S ymmetrical M odel for C risis c ommunications M anagement. Journal of Public Relations Research 8(2), pp. 79 105. Ha, S. (1999) News Coverage of the Clinton Lewinsky Scandal: Intermedia Influence at an International Level. (Unpublished master s thesis). University of Texas, Austin. Hallahan K (200 1). Strayegic Media Planning: Toward an Integrated Public Relations Media Model In R. L. Heath, Handbook of public relations (pp. 461 478 ). Thousand Oaks: Sage Publications. Hallin, D. C., & Mancini, P. (2004). Comparing M edia S ystems Three M odels of M edia and P olitics. Cambridge: Cambridge University Press Hatchen, W., & Scotton J. (2007). Changing Ideologies of Press Control. In W. Hatchen, The World News Prism: Global Information in a Satellite Age (7 ed.). Malden: Blackwell Publishing. Holsti, O. (1969). Content analysis for the social science s and humanities, Reading, MA: A ddision Wesley. Holtzhausen, D. R ., & Roberts, G. F. (2009). An I nvesti gation into the R ole of I mage R epa ir T heory in S trategic C onflict M anagement. Journal of Public Relations Research 2 1 (2), pp. 165 186 Hong T. (2007). Information Control in Time o f Crisis: The Framing of SARS in China Based Newspapers and Internet Sources paper presented to the 2007 Annual Meeting of the International Communication Association Hoovers. (2011, Oct 20). Foxconn Technology Co., Ltd. Profile. Austin, TX.
98 Howell, G., & Miller, R. (2006). How the relationship between the crisis life cycle and mass media content can better inform crisis communication. PRism 4 (1), pp. 1 14. Huang, Y. H. (2006). Crisis Situations, Communication Strategies, and Media Coverage. Communic ation Research 33 (3), pp. 180 205. Kim, R. J., & Molleda, J. C. (2005). Cross national conflict shifting and crisis management: an 8th International Public Relations Research Conference. Miami, F L. Len Ros, M. E. (2010). Image Repair Strategies, Local News Portrayals and Crisis Stage: A Case S tudy of Duke University's Lacrosse Team Crisis. International Journal of Strategic Communication 4 pp. 267 287. Lim, H. J. (2008). The influence of a cross national conflict shifting on a transnational corporation's host customers. University of Florida, Gainesville. McCombs, M. E., & Reynolds, A. (2002). News Influence on Our Pictures in the World. In J. Bryant, & D. Zillmann, Media Effects: Advances in Theory and Research (2 ed., pp. 1 16). Hillsdale, NJ: Lawrence Erlbaum. McPhail, T. L. (2010). Global communication : theories, stakeholders, and trends. Malden, MA: Wiley Blackwell. Mitroff, I. I. (1994). Crisis Manageent and Environmentalism: A Natu ral Fit. California Managment Review 36 (2), pp. 101 113. Mitroff, I. I. (1996 ). Essential G uide to M anaging C orporate C risis: A S tep by step G uide. New York: Oxford University Press. Mitroff, I. I., Shr ivastava, P., & Udwadia, W. A. (1987). Eff ectiv e C risis M anagement. The Academy of Management Executive, 1(3) pp. 283 292. Molleda, J. C. (2010). Cross N a tional Conflict Shifting: A Transnational Crisis Perspective in Global Public Relations. In R. L. Heath, The SAGE Handbook of Public Relations (pp 679 688). Thousand Oaks, Calif.: Sage Publications. Molleda, J. C. (2011). Advancing the Theory of Cross National Conflict Shifting; A Case Discussion and Quantitative Content Analysis of a Transnational Crisis' Newswire Coverage. International Journal of Strategic Communication 5 pp. 49 70 Molleda, J. C., & Connolly Ahern, C. (2002). Cross national conflict shifting: Conceptualization and expansion in an international public relations context,. Miami, Florida, USA: the Association for Education in Journalism and Communication.
99 Molleda, J. C. Connolly Ahern, C., & Quinn, C. (2005). Cross national conflict shifting: Expanding a theory of global public relations management through quantitative content analysis. Journalism Studies 6 (1), pp. 87 102. Molleda, J. C., & Laskin, A. (2010). C oordination and control of g lobal public relations to manage cross national conflict shifts: A multidisciplinary perspective for research and practice. In G. J. Golan, T. J. Johnson, & W. Wanta (Eds.), International media communication in a global age (pp. 319 344). New York, NY: Routledge. Molleda, J. C., & Quinn, C. (2004). Cross national conflict shifting: a global public relations dynamic. Public Relations Review 30 pp. 1 9. NowNews. (2010, May 28). Retrieved November 2, 2011, from NowNews: http://www.nownews.com/2010/05/28/91 2608481.htm Oliver, C. (2010, Aug 31). Foxconn Shares Tumble After Results Announcement Retrieved September 28, 2011, from Market Watch: http://articles.marketwatch.com/2010 08 31/mark ets/30704161_1_net loss shares morgan stanley Olson B (200 1). Media Effect Research for Public Relations Practitioners In R. L. Heath, Handbook of public relations (pp. 26 9 289 ). Thousand Oaks: Sage Publications. Patterson, B. (2004, December). A Cris is Media Relations Primer. TACTICS 13 Peng, H. Y. (2010, May 25). TVBS News Retrieved 10 23, 2011, from TVBS: http://www.tvbs.com.tw/news/news_list.asp?no=arieslu20100525152821 Pearson, C. M., & Clair, J. A. (1998). Reframing C risis M anagement. Academ y of Management Review 23(1), pp. 59 77. Press freedom compared. (2006). The China Post online. Retrieved October 20 20 10 from: http://www.chinapost.com.tw/editorial/detail.asp?onNews=&GRP=i&id=84782 Qi L (20 11 ). Tragedy Reporting in China: A Compar ative Study of Framing between Semi independent Business Newspapers and Official News Media. Point Park University, Pittsburgh. Ricadela, A. (2011, Feb 14). Apple's Cook Visited Foxconn in 2010 Suicide Response Retrieved October 19, 2011, from Bloomberg: http://www.bloomberg.com/news/2011 02 14/apple report says cook visited foxconn in 2010 suicide response.html SACOM. (2010). Workers as Machines: Military Management in Foxconn. Hong Kong: SACOM.
100 Seeger M W., Sellnow, T. L., & Ulmer, R. R. (200 1). Med ia Effect Research for Public Relations Practitioners In R. L. Heath, Handbook of public relations (pp. 26 9 289 ). Thousand Oaks: Sage Publications. Seymour, M. (2002). Does the media cause crises? Frontline 24 (3), pp. 22 23. Siebert, F., Peterson, T. & Schramm, W. (1963). Four Theories of the Press: The Authoritarian, Libertarian, Social Responsibility and Soviet Communist Concepts of What the Press Should Be and Do. Urbana: University of Illinois Press.
101 BIOGRAPHICAL SKETCH Wei Ling Yang was born in Kaohs iung, Taiwan. She bacame a B achelor of Arts in p olitical s cience with specialization in i nternational r elations from the best university in Taiwan, National Taiwan University. With international background, s he joined the graduate program of i nte rnational c ommunication in the College of Journalism and Communication at the University of Florida in fall 2010. In 2012, s he graduated and received her m aster s d egree in m ass c ommunication in the spring of 2012. Her research interests are internat ional public relations, international communications, media systems, crisis communication and international advertisement.