|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 A QUANTITATIVE CONTENT ANALYSIS AND TEXTUAL ANALYSIS OF COCACOLAS CRISIS RESPONSES TO A COLOMBIAN LABOR RIGHTS SITUATION By MARY ELIZABETH RUIZ A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DE GREE OF MASTER OF ARTS IN MASS COMMUNICATION UNIVERSITY OF FLORIDA 2011
2 2011 Mary Elizabeth Ruiz
3 To my parents, who always believed in me, even when I didnt believe in myself To the graduat e students who will come after me: keep going
4 ACKNOWLEDGMENTS My sincerest thanks to my thesis committee chair, Dr. JuanCarlos Molleda. His positive attitude and enthusiasm for life encouraged me to keep going even when my workload seemed impossible. I would also like to thank my committee members, Dr. Joanna Cleary and Dr. Michael Leslie. Both provided invaluable guidance throughout my academic career. I thank my father for always demanding my best and encouraging me to reach beyond my self imposed limitations. Without him I would not be in graduate school. I thank my mother for always being there to talk and for being a constant cheerleader. I would have gone insane without her. I thank my Abuelita for the countless prayers she has said on my behalf and for reminding me to include God in everything I do. I thank my Aunt Mary for always being the voice of truth and never letting me forget that my dreams will come true one day. Last but not least, I thank my friends for being good companions during thi s phase of our lives. I hope they always chase their dreams.
5 TABLE OF CONTENTS ACKNOWLEDGMENTS .................................................................................................. 4 page LIST OF TABLES ............................................................................................................ 7 LIST OF FIGURES .......................................................................................................... 8 ABSTRACT ..................................................................................................................... 9 CHAPTER 1 INTRODUCTION .................................................................................................... 10 Historical Context .................................................................................................... 15 Colombia: from Spanish Colony to War Torn Republic .................................... 15 Labor Unions: Unwelcome from the Start ......................................................... 18 2 LITERATURE REVIEW .......................................................................................... 22 The Crisis Management Process ............................................................................ 23 Prevent the Crisis ............................................................................................. 23 Assemble a Crisis Team ................................................................................... 24 Identify and Engage Your Audiences ............................................................... 25 Centralize Crisis Management Operations ....................................................... 26 Test and Retest the Crisis Plan ........................................................................ 26 Realizing a Crisis Exists ................................................................................... 26 Identification ............................................................................................... 27 Crisis Types ............................................................................................... 28 Containment and Resolution ............................................................................ 31 Situational Crisis Communication Theory ............................................................... 33 Journalistic Ethics ................................................................................................... 36 Research Questions and Hypotheses ..................................................................... 38 3 METHOD ................................................................................................................ 42 Sampling ................................................................................................................. 43 Content Analysis Coding Sheet Construction ......................................................... 46 Variables ................................................................................................................. 47 Conceptualizations ................................................................................................. 47 Measurements ........................................................................................................ 50 Intercoder Reliability ............................................................................................... 51 Data Analysis .......................................................................................................... 51
6 4 FINDINGS ............................................................................................................... 53 Research Question 1 .............................................................................................. 53 Research Question 2 .............................................................................................. 54 Research Question 3 .............................................................................................. 57 Research Question 4 .............................................................................................. 59 Research Question 5 .............................................................................................. 59 Additional Findings ................................................................................................. 60 Hypothesis 1 ........................................................................................................... 61 Hypothesis 2 ........................................................................................................... 61 Hypothesis 3 ........................................................................................................... 62 Hypothesis 4 ........................................................................................................... 62 Hypothesis 5 ........................................................................................................... 62 5 DISCUSS ION ......................................................................................................... 72 Summary of the CocaCola Crisis ........................................................................ 72 Summary of Crisis Types Used to Portray the CocaCola Case .......................... 72 Summary of CocaColas Crisis Responses ......................................................... 76 Denial as the Primary Response ...................................................................... 76 Addition al Responses ....................................................................................... 77 Reminder as CocaColas Primary Response................................................ 80 Application of Situational Crisis Communication Theory ......................................... 81 Evaluation of CocaColas Crisis Responses and Recommendations ................. 83 Summary of General Tone ...................................................................................... 84 Summary of the Word Count Tone Relationship .................................................... 85 Summary of Competing Voices .............................................................................. 86 Summary of Publication Date and Frequency ......................................................... 87 Implications for Public Relations and Communication Management ...................... 88 Limitations and Suggestions for Future Study ........................................................ 89 APPENDIX A CODING SHEET .................................................................................................... 91 B CODE BOOK .......................................................................................................... 93 LIST OF REFERENCES ............................................................................................... 96 BIOGRAPHICAL SKETCH .......................................................................................... 101
7 LIST OF TABLES Table page 3 1 LexisNexis search term results ........................................................................... 52 4 1 Frequency and significance of the crisis types in the source types .................... 70 4 2 Frequency and significance of the crisis responses in the source types ............ 70 4 3 Frequency and significance of corporate and critic voices in the source types .. 70 4 4 R elationship between tone and source types ..................................................... 71 4 5 Frequency and significance of tone in the source types ..................................... 71 4 6 Relationship between word count and tone ........................................................ 71 4 7 Group statistics for word count as it relates to the source types ......................... 71
8 LIST OF FIGURES Figure 4 1 Bar graph illustrating the crisis types that were present in the source types ...... 64 page 4 2 Bar graph illustrating the presence of crisis responses in the source types ....... 65 4 3 Bar graphs illustrating the presence o f the corporate voice and critic voice in the source types ................................................................................................. 66 4 4 Illustration of the relationship between tone and source types ........................... 67 4 5 Illustration of the relationship between tone and word count .............................. 68 4 6 Bar graph illustrating the source types publication date and how many communications were published on those days ................................................. 69
9 Abstract of Thesis Presented to the Graduate School of the Univers ity of Florida in Partial Fulfillment of the Requirements for the Master of the Arts in Mass Communication A QUANTITATIVE CONTENT ANALYSIS AND TEXTUAL ANALYSIS OF COCACOLAS CRISIS RESPONSES TO A COLOMBIAN LABOR RIGHTS SITUATION By Mary Elizabeth Ruiz May 2011 Chair: JuanCarlos Molleda Mass Communication The tremendous increase in globalization has fundamentally altered the way the world does business (Osterhammel & Petersson, 2003). The transfer of production and other processes fr om developed countries such as the U.S. to developing countries has transformed U.S. corporations into multinational corporations (Jones, 2005). The creation of multinational corporations has brought complex social and political problems. In the case of At lanta based multinational CocaCola, the issue of labor rights for foreign subcontracted laborers became a public battle when an employee was murdered on company premises in 1996. Coombs (2010b) Situational Crisis Communication Theory is used to evaluate how CocaColas crisis and crisis responses were framed in newspapers and official corporate communications. The presence of corporate and critic voices as well as tone is also analyzed.
10 CHAPTER 1 INTRODUCTION The tremendous increase in globalization has fundamentally altered the way the world does business (Osterhammel & Petersson, 2003). The process has made it easier than ever to transport goods and ideas, such as capitalism, all over the world. Capitalism is an economic ideology that emphasizes com petition, open markets, and profits ( Globalization, 2010). In pursuit of capitalisms ideals, U.S. based corporations have outsourced almost every element of the supply chain to developing countries. The transfer of production and other processes to these countries has transformed U.S. corporations into multinational corporations, or corporations that maintain a headquarters in one country but have branches in other countries (Jones, 2005). One such multinational corporation is The CocaCola Company In th e early 20th century, this beverage giant helped pioneer the globalization of U.S. based corporations (The Chronicle of CocaCola, 20062010). The corporation was founded in Atlanta, Georgia by Asa Candler, a businessman who purchased all of the rights to CocaCola before its inventor, Dr. John Stith Pemberton, died ( The Chronicle of CocaCola 20062010). The caramel colored, fizzy drink proved so popular that CocaCola expanded out of Georgia to Texas and was soon sold in every state. CocaCola o riginally was sold in soda shops but after the invention of bottling machinery, it was sold as a co nsumer product By about 1918 over 1,000 bottlers existed across the United States In the 1920s Robert Winship Woodruff, the president of Coca Cola create d the Foreign Department (now known as The CocaCola Export Corporation), which was meant to facilitate gl obal expansion.
11 Colombia was one of the countries that was targeted for expansion during this period (The Chronicle of CocaCola, 20062010). Aroun d the time CocaCola began operations in Colombia, three other U.S. based corporations, United Fruit, the Tropical Oil Company, and the Shell Oil Company, were expanding their Colombian business ventures (Chomsky, 2008; Bucheli, 2005). The influence of t hese corporations helped shape Colombian labor policies which heavily favored multinational corporations. For example, multinational corporations were allowed to form contract relationships with their Colombian workers instead of formally employing them. T his arrangement prevented the workers from demandi ng specific rights, such as an 8 hour workday as these rights were only guaranteed to formal employees. Contracting employees was advantageous to the multinationals because they could control the terms of employment. The lack of worker power led to frustration and compelled them to turn to politics for a solution. When political solutions failed the workers, many of them aligned with subversive groups who were sympathetic to their plight (Bucheli, 2005). Li ttle has apparently changed since the early 20th century as Colombian CocaCola workers allegedly have many of the same problems as their predecessors. In 2001, a lawsuit was filed in Miami, Florida by the Colombian labor union, Sinaltrainal. The lawsuit alleged that CocaCola bottlers Panamerican Beverages and Bebidas y Alimentos hired right wing paramilitaries to threaten and violently intimidate the unions members. The facts that prompted this lawsuit occurred on December 5, 1996 in Carepa, Colombia ( Rayner, 2004).
12 On this day Sinaltrainals chief negotiator a gatekeeper at the Bebidas y Alimentos bottling facility, named Isdro Segundo Gil, was gunned down by two paramilitaries on motorcycles ( Rayner 2004). The murder was followed by an attempt to ki dnap another leader and the bombing of Sinaltrainals local office. Sinaltrainal believes the paramilitaries targeted them at the behest of bottlers because workers had requested an increase in wages and benefits. In addition, workers allege they saw the p aramilitaries communicating with Panamerican/Bebidas y Alimentos bosses in the days leading up to December 5. About a week after this tragedy, paramilitaries allegedly stormed into the bottling plant and forced all of the union members to type up letters o f resignation from the union. After this mass exodus of union members, the Carepa chapter of Sinaltrainal ceased to exist ( Rayner 2004). The alleged reason behind the violence directed at Sinaltrainal was its keenness to negotiate more favorable terms of employment with CocaColas Colombian bottlers (Rayner, 2004). Like the banana and oil workers of the 1920s, many of Sinaltrainals members were contract laborers who lacked the rights afforded to formal employees. The formation of Sinaltrainal gave the w orkers an opportunity to try and change their circumstances, but these attempts were not met with favor by management. This is because l ab or unions in Colombia attract violent attention from right wing paramilitaries as labor unions are seen as part of the left wing guerillas, which are the right wings mortal enemy (Chomsky, 2008). According to the International Labor Rights Forum (2009), Colombia is the most dangerous country in the world for union leaders. Since the 1980s about 4,000 labor union members have been murdered. To put this figure into perspective, 101 labor
13 unionists were murdered worldwide in 2009, with 48 of these unionists coming from Colombia (International Trade Union Confederation, 2010). The dangerous environment surrounding Colombian labor unionists coupled with the timing of Gils death (he was spearheading a campaign to increase wages and benefits for Coc a Cola workers) indicate he was probably murdered because of his union activity. In spite of this connection, there was insuffici ent proof to definitively link CocaCola to Gils death and the intimidation of Sinaltrainal. The lack of legal evidence did not protect CocaCola from receiving a lot of negative publicity, which threatened its corporate reputation. The Campaign to Stop Killer Coke, an activist group dedicated to mobilizing U.S. citizens against CocaCola presented the corporation with its biggest challenge because the group focused its efforts on persuading U.S. college s tudents to pressure their universities into ending contracts with the corporation (Rogers, 20042011). Two of the most high profile universities to ban CocaCola were the University of Michigan and New York University (Rogers, 20042011). The loss of these contracts combined with the negative press i n U.S. media compelled CocaCola to mount an aggressive public relations campaign in the U.S. to counter the effects of the allegations. Part of the campa ign included official communications and statements made in U.S. newspapers. These communications we re analyzed through a quantitative content analysis to discern the types of crisis responses Coca Cola employed in its campaign. A textual analysis supplement ed the quantitati ve content analysis as it allowed for more in depth insights to be drawn. Both m ethods were informed by Coombs Situational
14 Crisis Communication Theory (SCCT, 2010 b ) which posits that the type of crisis experienced should be identified before a crisis response is selected (Coombs, 2007a). Official Coca Cola communications and newspaper articles were chosen as the observations because the former represent the official corpor ate voice and the latter often publish these statements. At the peak of the crisis in 2006, newspapers were still one of the primary ways that Americans got their news so studying relevant newspaper articles allow ed for a more complete a nalysis of the crisis responses (Pew Research Center, 2006). In addition, journalists consistently show a conservative bias in their news coverage so they were likely to reiterate the crisis responses Coca Cola published on its official website (Herman, 2000). Their interest in maintaining business relationships with major advertisers such as CocaCola also likely affected the published content (Altschull, 1995). Due to these facts, a review of newspaper articles would provide insight into whether or not Coca Colas corporate voice was noticeable in U.S. newspapers and if the messages presented in the newspapers were consistent with the messages on the CocaCola website. Studying C oca Colas public relations crisis has relevance for several reasons. As previously mentioned, Colombia has more labor unionist s die each year than any o ther country in the world. Additionally, because CocaCola is a prominent U.S. corporation it represents all U.S. citizens wherever it does business. The corporation must represent them positively and adhere to the highest ethical standards, even beyond U.S. borders. Lastly, the crisis Coca Cola experienced is an example of cross national conflict shifti ng, a new phenomenon that is becoming increasingly prevalent as a result of rapid globalization. Cross national conflict shifting describes a conflict that
15 involves a transnational organization whose headquarters is located in one country (home), but it d oes business in another (host) (Molleda, Connolly Ahern, & Quinn, 2005, p. 99). The conflict typically erupts in the host country, and its effects are present both in the host country and in the corporations home country. Therefore, this type of conflict includes host, home, and transnational publics (e.g., NGOs and activist groups, global media outlets, shareholders), (Molleda, Connolly Ahern, & Quinn, 2005, p. 89). Because multiple audiences have interest in cross national conflict shifts, public rel ations professionals must generate a crisis response that addresses each groups concerns without sending an inconsistent message. Oftentimes this task can be very challenging so an analysis of how one multinational, in this case CocaCola, handled such a problem will help other corporations navigate similar conflicts in the future. Before previous research is addressed, a more detailed explanation of Colombian political and labor history is merited to help the reader understand why labor unionists are constantly threatened and violently attacked. Historical Context Colombia: f rom Spanish Colony to War Torn Republic Simn Bolvar liberated Colombia from Spanish rule in 1819 (Simons, 2004). About 30 years after gaining its freedom, the Liberal and Conservat ive political parties were formed and are in existence today. Since these parties inception, they have engaged in violent power struggles as each seeks to control Colombia. The first major spate of violence occurred from 18991902 and was known as the War of a Thousand Days. Tens of thousands of people were slaughtered in a dark time, which seemed to foreshadow Colombias present situation (Safford & Palacios, 2002).
16 The War of a Thousand Days was followed by a period of Conservative rule in which the agri culture industry experienced tremendous prosperity (Lehrer, n.d.). Decades of Conservative rule yielded to Liberal control in 1930, and significant reforms were made. One of the most important reforms included the authorization of labor unions to organize. During this time frame, disgruntled workers and other unhappy members of society formed the Communist Party and the National Unity of the Revolutionary Left (Lehrer, n.d.). In 1947, the Liberal Party was defeated by the Conservatives who immediately soug ht to contain labor unions, which they viewed as adverse to capitalist interests (Lehrer, n.d.). Their efforts were not entirely effective, so many business owners took an anti union stance by organizing militia groups meant to intimidate labor union members. In 1948, Colombia experienced its second wave of political violence when the Liberal candidate for the presidency, Jorge Elicer Gaitn, was assassinated. The death of Gaitn prompted a major riot in Bogota, which marked the start of a decade long arme d conflict between Liberals and Conservatives known as La Violencia (Safford & Palacios, 2002; Simons, 2004). During La Violencia many citizens sought cover in the dense rainforest where they could establish their own governments within Colombia (Lehrer, n.d.). These unofficial governments were supported by rich Liberals who also funded armies to defend these governments from conservative right wing groups who wanted to take control of liberal territories. In the aftermath of La Violencia, a military dictatorship was installed, and it would remain in power until the Sitges Agreement was ratified in 1957. This bipartisan agreement evenly split government positions between the two main parties, the Liberals
17 and the Conservatives. This move angered members and supporters of other political parties and prompted them to form a rebel group called the Southern Bloc, which would one day become the Fuerzas Armadas Revolucionarias de Colombia (Spanish acronym FARC) [Revolutionary Armed Forces of Colombia] which contin ues to operate today (Simons, 2004; Chomsky, 2008). The formation of governments by the people and their creation of rebel groups ignited a forceful response from the official Colombian government who believed these groups were conspiring with Communist Cuba (Lehrer, n.d., Simons, 2004). The response was known as Plan LAZO, and it was supported by the U.S. government. Far from quelling the proliferation of rebel groups, Plan LAZO actually served as the catalyst for the formation of the Ejrcito de la Liberacin Nacional (ELN) and Movimiento19 (M 19). Over the ensuing years these rebel groups gained more followers and capital from sympathizers in other countries, allowing them to strengthen. Eventually these groups would turn to drug trafficking, kidnapping, threats, and murder to intimidate anyone who challenged their authority (Lehrer, n.d.). The ELN, FARC, and Autodefensas Unidas de Colombia (AUC) represent the three largest rebel groups currently operating in Colombia (Garamone, 2003). These groups have hundreds of millions of dollars to finance operations against the government which make them a formidable enemy for the government to combat (Garamone, 2003, 6). Recent gains by the Colombian government over the rebels have given Colombians hope that an end to this lengthy conflict will materialize (BBC, 2010). However these groups have infiltrated virtually every industry in Colombia, including multinational subsidiaries, which makes eradicating them a challenge.
18 Labor Unions : Unwelcome f rom the Start According to Chomsky (2008), U.S. multinational corporations have been in Colombia since the early 1900s. Their extended existence in this country has given them the opportunity to influence labor policy during times when the government was in transition or weakened due to social unrest (Simons, 2004). The industries that have the longest history of labor struggles with U.S. corporations are the banana fruit and oil industries, but in reality all other industries receiving foreign investment have had simi lar experiences (Chomsky, 2008). Understanding the relationship between multinational corporations, the Colombian government and the U.S. government and how they conspired to disband unions is fundamental to understanding why the violence against Sinaltrai nal members occurred and why no one has been brought to justice. Scholars and historians such as Bucheli (2005) and Simons (2004) note that Bostons United Fruit Company pioneered the Colombian banana industry. The produce corporation ventured to Colombia in order to take advantage of the cheap labor and poor working conditions. United Fruit initially began its Colombian operations in the Caribbean town of Santa Marta. In 1928, banana workers in this town made the decision to strike against United Fruit, w hich ended with the murder of hundreds if not thousands of banana workers at the hands of the Colombian army. Subsequent generations of unionists have never forgotten this tragedy, and its memory serves both as a feared outcome and motivator to continue th eir fight against what is perceived as U.S. imperialism (Bucheli, 2005; Safford & Palacios, 2002; Simons, 2004). Seeking more fertile land, United Fruit set its eyes on a fertile area of the Department of Antioquia known as Urab (Bucheli, 2005). The probl em with this area was that it already had inhabitants, mainly poor descendents of African slaves.
19 Paramilitaries and others in positions of power used deadly force to evict a great deal of these poverty stricken people from the lands, which eventually made their way into United Fruits control. Once the land was available for conversion into banana farms, United Fruit sent workers to start sowing the seeds and relocated its business from the coastal town of Santa Marta to the interior city of Urab (Buchel i, 2005). Workers on the banana plantations were forced to endure subhuman living conditions and work appallingly long days under United Fruit (Chomsky, 2008). The workers turned to left wing groups to help them fight United Fruits unfair labor practice s. The fusion of banana workers and the left resulted in the formation of a labor union known as Sintrabanano. Almost immediately, Sintrabanano joined ranks with the Fedeta and Trade Union Confederation of Colombian Workers (CSTC) unions. United Fruit res ponded by firing all unionists, and the Colombian government had them imprisoned because they jeopardized foreign investment (Chomsky, 2008). These actions would be supplemented by death threats and murders in the years to come. Colombian workers for U.S. oil corporations went through a strikingly similar struggle over labor rights with their employers; however, the oil industry and the unions it spawned began their fight a few years before United Fruit even set foot on Colombian soil (Chomsky, 2008). The t wo main U.S. oil corporations who refined and exported oil were the Tropical Oil Company and the Shell Oil Company, and they began operating in 1922. The following year Unin Obrera, the precursor to todays oil union, known as Unin Sindical Obrera (Spani sh acronym USO), was founded. In the first decade of the oil unions existence it waged two strikes against Tropical. The unionists received aid
20 and moral support from their fellow citizens which helped the strikes to continue (Chomsky, 2008). To suppress the laborers uprisings Tropical and other multinational corporations experiencing similar difficulties pressured the U.S. government to encourage the Colombian government to take action against unionists (Chomsky, 2008). The main way the U.S. government would pressure the Colombian government into complying with the multinational corporations desires was by threatening to eliminate much needed capital. Colombia was essentially dependent on foreign investment to run both the oil and banana industries. The se threats prompted the Colombian government to respond by sending police to physically stop unionists from striking. These encounters frequently resulted in the deaths of the unionists. Colombian oil workers saw the Colombian government as the United Stat es lapdog and, like the banana workers, turned to the left wing insurgent groups for help (Chomsky, 2008). Many workers which collaborated with insurgent groups did so in their spare time so as not to arouse suspicion. In a documentary aired on PBS calle d Colo mbia: The Coca Cola controversy two U.S. filmmakers traveled to Colombia to try and v erify the facts put forth in Sinaltrainals lawsuit against Coca Cola (Lapan & Harris, n.d.) During one of the segment s the filmmakers int erviewed the president of a banana union, Osvaldo Cuadrado. He openly admitted that he used to moonlight as a leftist rebel on the weekends while working as a unionized banana worker during the week (Lapan & Harris, n.d.). The double life these unionists led made it difficult to discern which workers were true pacifists interested in negotiating with their employer, and those who were
21 ultimately trying to take down the multinational corporations. The employers seemed to take the safe bet and assume all unionists were insurgents, which in their eyes justified retaliation against the workers. The complicated political and social problems in the background of CocaColas publicity crisis made formulating a crisis response all the more difficult as the corporation needed to show sensitivity towards the situation. The literature that follows will describe the ideal way to execute an effective crisis response and present Coombs (2010b) Situational Crisis Communication Theory, which will aid in the assessment of Coca Colas response.
22 CHAPTER 2 LITERATURE REVIEW A good reputation is one of a corporations most important assets because consumers associate it with the corporations integrity (Coombs, 2010a; Lerbinger, 1997). Sometimes the corporate reputation is called into question when it encounters a crisis. The word crisis has many different definitions because scholars have been unable to reach a consensus over the terms meaning. However, it is generally agreed that all crises have three basic elements: suddenness, uncertainty, and time compression (Lerbinger, 1997). For the purposes of this study Seeger, Sellnow, and Ulmers (1998) definition will be used. They defined a crisis as a specific, unexpected, and nonroutine event or series of events that create high levels of uncertainty and threaten or are perceived to threaten an organizations highpriority goals (Seeger, Sellnow & Ulmer, 1998, p. 233). The term crisis is often used interchangeably with the term disaster (Coombs, 2010b) Some scholars believe these phenomena are th e same and should be studied together while other scholars believe the two are separate. Those who distinguish the concepts view crises as m an made occurrences that ar e preventable and disasters as natural events that are merely containable (Mitroff & Anag nos, 2001). For example, if a bank executive embezzles money from his corporation and causes bank customers to sue the bank, this is a manmade occurrence and hence a crisis. If a tornado hits a major city and destroys all of the buildings, this is a natur al event or a disaster. Concentrating on the cause of a crisis is one way to distinguish a crisis from a disaster, but it is not the only way. Evaluating which groups are affected by the event is also telling.
23 According to Ronald Perry (2007), crises affec t specific groups and disasters affect whole societies. This difference is notable and important because the crisis communication strategy develops around the affected group. During a crisis, communicators must tailor their message to address specific stak eholders (FearnBanks, 2007, Harvard business essentials, 2004). In contrast, during a disaster the target audience is the entire society, obligating communicators to create a message that transcends all demographics (Perry, 2007). It can be argued then th at because crisis responses require more precision, it is essential for a situation to be classified as either a crisis or a disaster before the management response continues. Ample literature exists that substantiates both sides of this argument, but elaborating upon this discussion is out side of the scope of this study Nevertheless, the researcher finds it necessary to mention the controversy surrounding these two concepts. The Crisis Management Process Crisis management should begin long before a crisi s strike s (Regester & Larkin, 2008; Zaremba, 2010). As with most problems, prevention should be the first goal. Crisis experts maintain that all corporations should have a crisis management plan in place before a crisis strikes The steps involved in each plan vary depending on the author, but the main steps include preventing the crisis, assembling a crisis team, identifying and engaging the target audience, centralizing crisis management operations, and testing the crisis management plan (Mitroff, 2004). Prevent the C risis The best way to prevent a crisis is to pinpoint the potential crises that could befall a corporation (Coombs, 2010a; FearnBanks, 2007). To do so a corporation should highlight the areas where a crisis is most likely to occur. Preemptiv ely identifying
24 potential crisis situations provides countless benefits, yet many managers admit they do not take this step. It is not enough to just acknowledge potential sources of crises, an organization must also identify which crises are preventable a nd take steps to ensure these crises do not occur (Harvard business essentials, 2004). Assemble a Crisis Team It is crucial to designate a crisis team that will come together and formulate a crisis response when a crisis strikes ( Coombs, 2010a). Planning out this team will ensure that the crisis response runs more smoothly. The crisis team members should be carefully chosen because they will control how the crisis response is implemented. Crisis experts recommend that a crisis team consist of at least a se nior operating officer a director of communications, the head of the affected division and a spokesperson (Lerbinger, 1997). Some experts believe legal counsel should be kept off of the crisis team because lawyers are by nature risk adverse and will advi se limited communication to avoid future liabilities. Experts that are of this mindset still find it necessary to keep legal counsel informed about the crisis teams plans because corporations should always be weary of lawsuits (Regester & Larkin, 2008). Other experts think the crisis team should include a lawyer in order to ensure that l egal risks are kept to a minimum. While limiting the information that comes from an organization during a crisis is a smart move from a legal perspective, it is a potentially devastating move from the communications perspective. Communication professional s know that the best way to protect the corporate reputation is to be as open as possible with the public ( FearnBanks, 2007) As the legal and communication fields are nat urally at odds in this regar d,
25 some crisis experts think it is best to avoid including a lawyer on the crisis team in order to prevent indecisiveness amongst team members The crisis teams most important member is arguably the spokesperson because he or s he is the most visible. This person represents the corporation delivers the strategic message and answers the medias questions (FearnBanks, 2007; Harvard business essentials, 2004). The person who is selecte d for this position must be articulate, comfor table with answering difficult questions, and be credible. Some experts in crisis management advise against usin g the head of public relations as the spokesperson because he or she could be perceived as less credible due to the nature of his or her position These people argue for using a prominent executive as the spokesperson, usually the Chief Executive Office (CEO) ( Coombs, 2010a). Identify and Engage Your Audiences Another integral step is to establish who the key external stakeholders are and when they should be addressed ( Regester & Larkin, 2008, Zaremba, 2010). The external stakeholders are the persons who are affected by the corporations crisis and who stand to change their opinion of the corporation depending on how the crisis is managed. Failing to communicate with external stakeholders may result in frustration and loss of credibility amongst the stakeholders, wh ich threatens the corporate reputation. As one of its most important assets, a reputation must be protected at all reasonable costs ( Co ombs, 2010a). While external communication is a pivotal aspect of crisis management, crisis experts strongly suggest that management implement an internal communications plan that will keep employees informed of the latest developments ( FearnBanks, 2007). The rationale behind this suggestion is that employees act as additional informal
26 spokesper sons for the corporation, making it essential that they are well informed. Family, friends, and acquaintances will likely question the employee about the crisis, es pecially if the crisis is particularly salacious, and management must ensure that they spread the desired message ( Lerbinger, 1997). Centralize Crisis Management Operations In addition to putting forth a central message, a corporation must designate a cent ral location for the crisis team to meet and execute its strategies. The crisis headquarters could be either externally or internally located. The decision depends on space and resource availability as well as preference (Zaremba, 2010). Test and Retest th e Crisis Plan Once a crisis management plan has been developed, it should be tested in a mock crisis situation Doing so will allow everyone who is part of the crisis team feel more comfortable with their roles and will uncover any weaknesses within the pl an (Zarem b a, 2010). Preparing for a crisis can help reduce its negative consequences although it can never eliminate the possibility of having a crisis In the Coca Cola crisis these precrisis steps cannot be applied because the situation had already reached a crisis level. The crisis management steps that are applicable to the case study are identifying the type of crisis, containing the crisis, and solving it (Coombs, 1999; FearnBanks, 2007; Harvard business essentials, 2004). Realizing a Crisis Exists Knowing when a crisis exists may not be as obvious as one would believe. In some instances, a corporation will trivialize an event because it fails to view the situation from an external perspective. Crisis experts maintain that crises exist as a matter of
27 perception. What this means is that if the external stakeholders believe a crisis exists or even think that a crisis exists, then a crisis does exist no matter what a corporations management perceives (Coombs, 2010a). Stakeholders opinions matter the most because their opinions uphold the corporate reputation; therefore, they should defer to stakeholders. Most crisis experts assert that all crises give off warning signals management must be trained to recognize. As an objective list of signs does not exist, crisis experts advocate periodically reviewing other departments reports or media sources to see if there is a potential for a problem. Exampl es of internal reports include financial audits, safety/accident records, and a record of customer complaints (Coombs, 1999; FearnBanks, 2007). The exact moment Coca Cola realized it was experiencing a crisis does not need to be determined as it does not affect the analysis of its crisis response. All that matters is that a crisis existed and that the corp oration recognized it as such. Given the public outcry and boycotts referenced in CocaColas official communications, it will be assumed that CocaColas stakeholders viewed the situation to be discussed as a crisis. Because Coca Cola responded to thes e reactions by refuting the accusations and providing evidence to the contrary, it will be assumed that Coca Cola recognized its stakeholders concerns and treated the allegations as a crisis. Identification Once management has realized a crisis exists, t he crisis management process moves to the identification stage. Identifying the type of crisis a corporation faces is important because the crisis response will depend upon this classification. However, assigning a crisis type is challenging because many t ypes exist and none of them are
28 standard (Coombs, 2002; Seeger, Sellnow & Ulmer, 1998). Attempts to create a flexible type that will withstand the test of time continue to this day but given the complex nature of these situations there may never be a perm anent type (Gundel, 2005). A review of the existing crisis types will be discussed and the relevant ones were selected and applied to the CocaCola case focused upon in this study. Crisis Typ es Scholars have approached crisis types in many different ways. Some have focused on how the crisis affects the corporation and those who have invested in it. Others have studied the effect on business practices. Yet others have concentrated on issues of temporality, mainly the amount of time a corporation has to prep are an initial response to the crisis (Gilpin & Murphy, 2008). One such study by Curtis Linke (1989) produced four crisis types: exploding, immediate, building, and continuing. An exploding crisis is an event that occurs suddenly and causes instant damage. An immediate crisis may appear out of nowhere, but its effects do not occur instantly, which gives management a chance to mitigate the damage. Building crises are mishaps that management can expect, and continuing crises are ongoing problems (Linke, 1989) A study by Hwang and Lichtenthal (2000) simplified Linkes (1989) four crisis types into two using a biological theory known as punctua ted equilibria This theory characterizes evolution as a process with long periods of stagnation and short spurts of g rowth brought about by some sort of cataly st The authors drew a parallel between this series of actions and how corporations respond to crises because they typically remain dormant (long period of stagnation) until they encounter a crisis (catalyst) that forces them to undergo change (growt h) The comparison resulted in two crisis typologies
29 abrupt and cumulative. As their names indicate, abrupt crises appear without warning and cumulative crises build steadily until they materialize. The amount of time th at passes before management makes an initial response is an important point of study because this timing is correlated with the crisis duration and intensity (Coombs, 2007 a ). Scholars have also used the origin of a crisis as a basis for creating crisis ty pes (Gilpin & Murphy, 2008). Although countless crisis types exist, many scholars have consistently identified the same ones. Cri sis expert Timothy Coombs (2007b) complied a master list of the most common typologies. They are the natural disasters, ru mor workplace v iolence, malevolence, challenge, technical error accidents, technical error product harm, humanerror accidents, human error product harm, and organizational misdeed. These and their respective definitions as defined by the same author informed this analysis. The first crisis typology, the natural disaster, is premised on Charles Perrows (1984) normal accident. Virtually every scholar that has worked on crisis typologies since has included a natural disaster as one of their categorizations (Coom bs, 2007b, Egelhoff & Sen, 1992, FearnBanks, 1996, Lerbinger, 1997, Marcus & Goodman, 1991). The natural disaster envelops all natural phenomena that can disrupt normal activities like tsunamis, brush fires, and mudslides (Coombs, 2007b). The second type, rumor is defined as false information that is spread and slanders an organization. The third type, workplace violence, involves violence at the workplace by a former or current employee. The fourth type, malevolence, covers events that a person outside o f a corporation intentionally carries out to harm it. Such events include sabotage, blackmail and the sale of trade secrets. Challenge, the fifth type, occurs when a corporations
30 stakeholders state that it is not operating as it should. Examples of challenges are lawsuits, government penalties, and protests. Technical error accident is the sixth type, and it exists when a corporations technology fails and the result is an industrial accident. A crisis that involves a technical error product harm is the s ame as a technical error accident except the result is a product defect. Humanerror accident includes accidents caused by human beings. Humanerror product harm encompasses product defects that are caused by human mistakes. Organizational misdeed encompas ses decisions made by management it knows will harm or place stakeholders at risk for harm without adequate precautions (Coombs, 1999, p. 61). A contemporary example of organizational misdeed may be seen in the Enron and WorldCom scandals that came to light in the early 21st century. As previously noted, the categories mentioned above are not exhaustive but they are the most common types of crises a corporation will face. The main shortcoming of this list is that in some categories it fails to explicitly note the level of control, if any, a corporation has over a crisis. When creating their crisis types, Mitroff, Pauchant, and Shrivastava (1988) avoided this oversight by first identifying the crisis by type and then classifying the crisis as either internal (within corporate control) or external (outside of corporate control). Assessing the level of control a corporation has over a crisis is important because it influences the crisis communication response ( Coombs 2007a). Let us use the rumor typology as an example. If a rumor is generated by a current employee (within a corporation), the public would arguably be m ore receptive to the corporation issuing a denial response because this is the strongest way to distance itself from allegations If the rumor c omes from outside of the corporation, say from a
31 rival the public would arguably be more receptive to the attack the accusers response because the source of the rumor stands to profit from the rumor being accepted by the public. Therefore, evaluating the level of corporate control in each crisis situation is an essential process. However, t his step will not be applied to the Coca Cola crisis because the allegations came from current employees (within the corporation) so it will be assumed Coca Cola had s ome level of control over the crisis. Once a corporation identifies a crisis, the crisis management process moves to the response phase, otherwise known as the containment and resolution phase. This step is the most important step in protecting and saving the corporations reputation because the response it deploys determines the severity and the duration of the crisis effects (Coombs, 2002). Containment and Resolution After a crisis has been detected, the corporation must formulate a response that will c ontain the crisis until it can be resolved. Coombs (1999) divided the response process into four phases: the initial response, reputation management, execution of the crisis management plan, and post crisis communication. The initial response is the first official communication from a corporation in crisis (Coombs, 1999). The timing and the content within this communication must be carefully thought out as both factors will help set the tone for how the stakeholders will respond to the crisis management team. The initial crisis response also sends the public t he message that the corporation is aware of the crisis, is in control of its management and is working to return operations to normal as soon as possible. A poor initial response fails to address these concerns, which could exacerbate the threat to the corporate reputation (Coombs, 1999).
32 Several principles must be kept in mind when creating the initial crisis response. Firstly, timeliness is everything (Coombs, 1999). The initial crisis response must be released as soon as possible because a delay will make stakeholders panic and will encourage speculation. The latter can be far more damaging than the crisis itself because it could frame the crisis in an undesirable way or blow the crisis out of proport ion. Crisis experts note that corporations should make a timely initial crisis response even if it does not have much information. Crisis experts recommend that organizations make up for the lack of information by assuring the public they will provide more information as soon as it is possible (Coombs, 1999). In the initial crisis response, the crisis management team must focus on creating a central message that will be repeated throughout the duration of the crisis (Coombs, 2010b). The message should be s tandardized to eliminate conf usion as to how the corporation is handling the crisis and to ensure that it is viewed favorably. An inconsistent message sends the signal that the corporation is ill equipped to handle the crisis and reduces credibility (Coombs, 2010b ). Throughout the crisis response, but especially during the initial crisis response, the imperiled corporation must put forth an image of honesty and show it wants to keep the public informed (Coombs, 2010b). Being forthcomi ng signals that it has nothing to hide and encourages the public to beli eve the message it is pushing. In addition to honesty, it is crucial for a corporation to show concern for anyone injured by the crisis, as it will inspire feelings of compassion from the rest of society. T hese concerned citizens can become enraged i f they perceive the corporation as callously disregarding any human suffering, wh ich threatens its reputation (Coombs, 2010b; Zaremba, 2010).
33 The final essential feature of a good initial crisis response is what Sturges (1994) called instructing infor mation ( as cited in Coombs, 2010b). This information directs the stakeholders in case they need to take some sort of action. It also serves as a progress report to let the stakeholders understand the circumstances around the initial crisis and how it is being handled (Coombs, 2010b). After the initial crisis response has been made, the focus shifts towards managing the corporate reputation. The main theory used for this purpose is the Situational Crisis Communicatio n Theory (Coombs, 2010). Situational Crisis Communication Theory Coombs ( 2007a) created the Situational Crisis Communication Theory (SCCT) to give communicators evidencebased crisis communication guidance which he thought was lacking in previous researc h (p. 163). He based his theory on Benoits (1995) Image Restoration Theory (IRT), a theory that views communication as a way to rebuild a public reputation after a scandal. IRT divides the rehabilitative process into five categories denial, evading responsibility, reducing offensiveness, corrective action, and mortification (Benoit, 1995). Coombs expanded upon these categories and created a list of nine main crisis response strategies. Before any of these responses may be selected, the crisis management t eam must first analyze the crisis and label it using the crisis types previously discussed (Coombs, 2007b). This is the first step because, according to Attribution Theory which influences SCCT, when a major event occurs people naturally search for someone to ascribe responsibility to As each crisis type carries a different level of responsibility, its selection will determine which crisis response is chosen. Coombs and Holladay (2002) divided the crisis types into three clusters that predict the level of responsibility a
34 corporation is likely to receive. Doing so helps the crisis manager determine which crisis response will protect the corporation most effectively. The three clusters are victim, accidental, and intentional. The victim cluster means that the organization is not going to be blamed for the crisis because it is perceived as the victim. Natural disasters, workplace violence, product tampering, and rumor fall under this category. The accidental cluster means that the corporation will receive some blame, but not a lot because the crisis was not meant to occur. Technical error accident, technical error product harm, and challenge exemplify the accidental cluster. The final cluster, the intentional cluster, attributes the most blame to the corporation because the crisis is considered under its control. Human error accident, humanerror product harm, and organizational misdeed are categorized under this cluster. (Coombs and Holladay, 2002). Once the appropriate cluster has been ascertained, the crisis manager can select the response that will do the most to protect the corporate reputation. Crisis managers must then evaluate a corporations crisis history and prior reputation as these factors will affect the severity of the crisis (Coombs, 2007b). Having a history of crises will make the current one more intense because the public will perceive it as having failed to learn from past mistakes since crises continue to occur. In contrast, a sterling reputation will lessen the reputational damage that could result from the crisis (Coombs, 2007b). Once the crisis type and influential factors have been assessed, the crisis manager can move on to selecting the appropriate crisis response. These include (1) attack the accuser, (2) denial, (3) scapegoat, (4) excus e, (5) justification, (6) reminder,
35 (7) ingratiation, (8) compensation, and (9) apology. The excuse category has four subtypes: provocation, defeasibility, accidental, and good intentions which brings the total number of crisis responses to 13 (Coombs, 2007b). The responses are further broken down into postures which are separated by the corporations goal. The main goals are changing how stakeholders view the crisis or altering the perception of the corporation experiencing the crisis (Coombs, 2007c). The four postures are denial, diminishment, rebuilding, and bolstering. The denial posture aims to sever the link between the crisis and the corporation. If this strategy is successfully implemented, then the public will not direct responsibility at the corpor ation, and its reputation will not be jeopardized. The diminishment posture aims to protect the corporate reputation by lessoning the perceived control the corporation had over the crisis or by reducing the negative impact caused by the crisis. The rebuilding posture involves accepting full responsibility for the crisis and offering to make amends. Coombs (2007c) notes that all three of these postures require the corporation to acknowledge some level of responsibility for the crisis and show concern for the victims. The final posture, bolstering, contains complementary strategies that are to be used in conjunction with the strategies in the other postures. SCCT offers crisis managers recommendations on when to use each posture. The relevant suggestions include use rebuilding strategies for any preventable crisis, use denial strategies in rumor crises, use denial strategies in challenges when the challenge is unwarranted, use corrective action in challenges when other stakeholders are likely to support the challenge, and use reinforcing strategies as supplements to
36 the other response strategies (Coombs, 2007c, p. 143). Following these recommendations ensures maximum protection against negative publicity. Once all of these considerations have been made, the next step is for the crisis team to put the crisis management plan into action. The plan should remain in effect until the crisis subsides. There are several signs that a crisis is over: employees are back to their normal routines, business continues to come in, the media do not call and earnings are back on track (Harvard business essentials 2004, p. 91 ; Coombs, 2010a). After the crisis is over, a corporation must be sure that they have given the information they have promised (Coombs, 1999) As p reviously discussed, if a corporation did not have enough information to make an educated comment on something like the cause of the event but promised they would do so at a later date, it should make sure to address this. People w ill remember how it handl ed a cri sis more than the crisis itself Coombs (1999) found post crisis communication also helps to strengthen relationships with stakeholders. Continually giving stakeholders information after the crisis lets the st akeholders know the corporation has com pletely taken care of the crisis and it can move on to repairing itself. In addition to communicating after a crisis, crisis experts recommend debriefing the crisis with all employees and going over the lessons that were learned. Doing so will help prevent similar situations from occurring in the future ( Coombs, 2010a; Fearn Banks, 2007). Journalistic Ethics Because the mass media allow for instant dissemination of information to large populations, they should be involved in the corporate crisis response. (Fearn Banks, 1996). Indeed Coombs ( 2007a) agrees as he asserts that stakeholders mainly receive news about corporations thro ugh the mass media. The media can be a corporations
37 biggest supporter or adversary depending on ones perspective, but no matter w hat attitude is displayed towards a corporation, corporate activity will always be given prominent news coverage. Journalistic professionalism and ethics demand that the news be presented objectively (Jacquette, 2007). This notion comes from the Hutchins Commission of 1947. The purpose of this commission was to evaluate and reform U.S. journalism as it was perceived as being too profit oriented and too biased toward the normative class. Out of this commission emerged the concept of social responsibility. T his philosophy holds, among other things, that the media must keep the public interest in mind when reporting the news and should be truthful, objective, ethical, and professional (McQuail, 2010). These sentiments are echoed by the Society of Professional Journalists. In 1996, the society developed a code of ethics that has been widely adopted by U.S. journalists. Its first tenant advises journalists to Seek Truth and Report It (Society of Professional Journalists, 1996, 2). What this means is that jour nalists must refrain from injecting ideological bias into stories and distorting the truth regardless of any valid justification. While journalists might strive for objectivity, scholars Atschull (1995) and Herman (2000) assert that journalistic objectivit y is merely a faade and that the media has a strong conservative bias. In his book, The Myth of the Liberal Media, Herman (2000) argues that objectivity as practiced in U.S. journalism is not really objective because it is premised on ideological assumpti ons that prejudice the concept. For example, the media presume that responsible capitalism and free trade are both positive but this is not a fact; this is an ideological preference (Herman, 2000, p. 59). By supporting this form of
38 prejudiced objectivi ty, the media inherently provide support to conservative ideology in their reporting. Furthermore, as Altschull (1995) points out, the media are ultimately a profit seeking entity. The emphasis placed on generating profits requires the media to seek comme ntary from official sources as opposed to independent sources because official sources are deemed more credible. Additionally, the media has to be cautious with their content because publishing stories that are critical of advertisers or their affiliates c ould result in a loss of advertising revenue. These concerns prevent the media from offering minorities a public voice and presenting anti corporate information, reinforcing conservative ideals (Atschull, 1995). Not all scholars agree that the media are co nservative. In their study, Groseclose and Milyo (2005) attempt to paint the media as a liberal institution by counting how many times liberal and conservative think tanks are quoted in the print media. The authors found that liberal think tanks were cited more frequently than conservative think tanks, which led them to conclude that the media are predominantly liberal (Groseclose & Milyo, 2005). However, because the authors fail to address the underlying biases present in the journalistic concept of object ivity, their conclusion is debatable. To gauge the medias objectivity in their portray al of CocaColas crisis and its responses, rel evant newspaper articles were studied along with offi cial corporate communications. The following research questions and hypotheses demonstrate other research objectives. Research Questions and Hypotheses The literature on crisis management holds that the type of crisis a corporation experiences will shape the response that the crisis management team will choose
39 (Coombs, 19 99). Because of the crisis typologies influential nature, they must be identified and discussed before a complete analysis of the crisis response typologies may proceed. Therefore, the following question is posed: RQ1: Using the informa tion presented in official Coca Cola communications and newspaper s, h ow can the crisis Coca Cola experienced be classi fied using the crisis types derived from Coombs Situational Crisis Communication Theory? Choosing a crisis response is the most important step in the cri sis management process because the type of response that is deployed will determine the severity and the duration of the crisis effects (Coombs, 1999). This significance prompts the following research questions: RQ2: U sing the information presented in off icial Coca Cola communications and newspapers how can Coca Colas response be desc ribed using the crisis responses identified in Coombs Situational Crisis Communication Theory? How can Coca Colas most frequ ently used crisis response(s) be categorized using the aforem entioned theory ? When a crisis occurs, the media will look to the imperiled corporation for a comment. FearnBanks (1996) notes that the media will continue with its story even if comment is refused so she advises all corporations to prepare some statement. The voices present in the media will help shape the story so the following questions are presented: RQ3: What are the most prominent voices, in terms of corporate voice versus critics voices, included in the news coverage of Coca Colas crisis? Can these voices be described as positive, negative, or balanced? FearnBanks (1996) emphasizes the importance of the media in helping to contain a crisis. She advises viewing the media as an ally and working with it. However, even if a corporati on cooperates with the media, it may not be portrayed in the manner it would like. The following question seeks to evaluate the corporatemedia relationship during the CocaCola crisis by asking:
40 RQ 4: How noticeable was Coca Colas corporate voice in n ews coverage pertaining to the Sinaltrainal controversy and is the tone positive, negative, or balanced? As discussed in the previous rationale, the media play an integral role in managing a crisis, but sometimes it does not present information in the same way a corporation would. To elaborate on this potential disconnect, the following question is set forth: RQ 5: Usin g the crisis responses discussed in Coombs Situational Crisis Communication Theory, how can the crisis response(s) from CocaColas official crisis communications be compared to U.S. newspapers portrayal of Coca Colas crisis response (same or different) ? The allegations made against Coca Cola suggest that there was collusion between the bosses of Coca Cola subsidiaries and right wing pa ramilitaries. The crisis typology that deals with errant decisionmaking on the part of management is called organizational misdeed. Coombs (1999) describes organizational misdeed as decisions made by management it knows will harm or place stakeholders at risk for harm without adequate precautions ( p. 61). Therefore hypothesis one posits that: H 1: Organizational misdeed will be the crisis type most frequently used to describe the CocaCola controversy. Coombs (2002) states that reputation is one of a cor porations most important assets. It follows then that CocaCola will implement crisis responses that are most protective of its reputation. Logically, the most powerful way to distance oneself from an accusation is to refuse any affiliation with it. Hypothesis two argues: H2: Coca Cola will use denial more than any other crisis response. In their respective books, Altschull (1995) and Herman (2000) demonstrate that the U.S. media has a conservative bias as it caters to advertisers and the normative clas s by operating under flawed objectivity. Because of this bias, the following hypotheses are posed:
41 H3: Coca Colas voice will be more prominent than critics voices. H4: The media cover age of the CocaCola crisis will be primarily positive H5: The cris is responses presented in official communications from Coca Cola will be the same as the U.S. newspapers portrayal of Coca Colas crisis responses
42 CHAPTER 3 METHOD My research was conducted to identify the crisis responses CocaCola used to protect its corporate reputation when it was jeopardized by a Colombian labor union. The observations are the official communications from the CocaCola website and relevant U.S. newspaper articles. These communications were collected from two sources and were pr imarily analyzed using a quantitative content analysis informed by Coombs (2010b) Situational Crisis Communication Theory. A secondary method, textual analysis, was used to support or negate the results of the quantitative content analysis. A quantitative content analysis as defined by Riffe, Lacy, and Fico (2005) is the systematic and replicable examination of symbols of communication, which have been assigned numeric values according to valid measurement rules and the analysis of relationships involving those values using statistical methods, to describe the communication, draw inferences about its meaning, or infer from the communication to its context, bot h of production and consumption (p. 25) This studys design followed the steps of content analy sis as developed by Wimmer and Dominick (2006) which included the following steps (a) formulate research questions and hypotheses, (b) define population, (c) select sample from population, (d) select unit of analysis, (e) construct the categories of content, and (g) establ ish the quantifications system. After these steps were completed a textual analysis was conducted following McKees (2003) guidelines. McKee (2003) defined a textual analysis as a way for researchers to make an educated guess at some of t he most likely interpretations that might be made of that
43 text (p. 1). When a textual analysis is conducted, the person relies on his or her intimate knowledge of the culture in which the text was written to guide his or her interpretations. These textual interpretations must be reasonable and make sense within the context they are presented. This means that the interpretation logically flows from the text on which it is premised. It is important to note that a textual analysis is not concerned with making a correct interpretation because, even if the general conclusion is the same, every interpreter will describe their interpretation differently (McKee, 2003). What this means is that there is no such thing as a correct interpretation. A quantitative content analysis was chosen as the primary method for this study because it provides a numeric means of identifying which crisis responses were used most frequently. Isolating the most popular crisis response typologies will influence the textual interpretations because the frequency of the message will provide insight as to what were CocaColas primary crisis responses. Textual analysis was chosen as the secondary method because it allows for the most comprehensive assessment as to why CocaCola could have chosen the strategies it did. The reason the level of comprehensiveness is amplified is because a textual analysis allows external knowledge to be taken into consideration when making educated interpretations. A textual analysis also allows cultural inform ation to influence the interpretations, making the interpretation a more indepth analysis (McKee, 2003). Sampling The observations, U.S. newspapers and official CocaCola communications, were collected through a search on the LexisNexis Academic database and the Coca Cola corporate website.
44 Each search on the LexisNexis Academic database was done under the easy search section. Then the news tab was selected. Under the news tab the newspapers and wires option was chosen. Then the box next to U.S. newspapers and wires was checked. This portion of the LexisNexis Academic database searches through 575 independent sources. The sources that returned the most results were Associated Press, Associated Press Online, Associated Press State and Local Wire, Atlanta Journal Constitution Birmingham News, Cox News Service, The Daily News (New York), New York Times, Pittsburgh Post Gazette, PR Newswire, and University Wire. Nine searches were conducted on the LexisNexis Academic database. To isolate the relevant data gathered on the LexisNexis Academic database, all 1,302 documents were put in chronological order. The aim was to eliminate duplicate articles. Articles that had the same date and source were further reviewed for matching headlines. If headlines mat ched, then author and word count were reviewed. If author and the word count were the same, the first paragraph was read to see if the content was duplicative. Any time documents matched, one article was kept and the duplicate(s) was/were discarded. Throug h this process, 818 duplicate documents were eliminated leaving 484 articles left. Reducing the data from 484 to 160 documents was a threestep process. First, the headlines and the first five lines of all 484 documents were read for key terms which were s eparated into five groups. Typically, an entire news article is summarized in the first paragraph, and the first paragraph of both official CocaCola communications and newspaper articles are at least five lines so this is why only the headlines and first five
45 lines were read at first (Arnold & Cook, 2010). The key terms and their groupings were selected because they appeared most frequently when preliminary searches were run. They are as follows (where applicable, all groups include the possessive form of the term): (1) Colombia, Colombian(s); (2) Sinaltrainal; (3) CocaCola, Coke; (4) labor union(s), trade union(s), unionist(s), union(s); and (5) paramilitary, paramilitaries. Finally, the headlines and the first five lines had to contain at least three key terms and each term had to come from a different group for the article to be accepted. At first, articles containing two terms from two separate groups were accepted, but these criteria returned too many irrelevant results. Changing the criteria to thr ee terms dramatically reduced this number. Next, the first 15 lines of all 160 articles were read. Fifteen lines was the chosen cut off point because in many communications and articles, this figure represented roughly half of the line count. Therefore, th e relevance of the article could definitely be ascertained after reading this amount of content. Using the key terms and the groupings discussed in the previous paragraph, the lines were searched for at least four key terms. Four key terms were selected as the minimum because the usage of three terms still brought back irrelevant results. Again, each term had to come from a different group for the article to be accepted. This method reduced the number of documents from 160 to 61. The 61 newspaper articles were then read for official quotes from Coca Cola personnel and statements attributed t o CocaCola. For this study a quote is operationalized as a sentence or phrase enclosed in quotation marks and attributed to Coca Cola or a critic. A statement is o perationalized as a sentence that is attributed
46 to Coca Cola or a critic. (i.e. Coca Cola has said) This process reduced the number of articles from 61 articles to 45 which were kept for analysis. Ten searches were conducted on the CocaCola corporat e website ( http://www.thecoca colacompany.com). The search terms that were used for the LexisNexis Academic search were the same terms that were used for the CocaCola search, but an additional search was run using just the term Colombia. This search wa s added to ensure no relevant communications were missed. Altogether, the searches returned 186 communications. The URL of each publication was saved in a Word document. The text of each URL was singled out and compared to the remaining 185 URLs. If the URLs matched, meaning t he entire URL that was isolated matched any of the 185 other URLs, then the duplicates were eliminated. This comparison reduced the data from 186 to 45. After the first reduction was made, all 45 URLs were opened and searched for key terms from the followi ng five groupings: (1) Colombia, Colombian(s); (2) Sinaltrainal; (3) CocaCola, Coke; (4) labor union(s), trade union(s), unionist(s), union(s); and (5) paramilitary, paramilitaries. Any time one of the key terms appeared, the surrounding text was read f or context to determine if the appearance was relevant to the objective of this study. This process reduced the 45 URLs to 28 publications suitable for a textual analysis. Content Analysis Coding Sheet Construction A coding sheet was developed to facilitate the analysis of the data. It was divided into four sections: source types information, competing voices, crisis type, and crisis responses by Coca Cola based on Coombs (2010b) SCCT. The first section began by recording each sources identification nu mber. Then the date the article was published was written down. The communication type and word count were also noted.
47 The absence or presence of the corporate voice and the critics voice was recorded. Finally, the coder identified which crisis types and crisis responses were present in each observation. Variables This study had three independent variables (1) new spapers (2) official Coca Cola communications, and (3) date. There were thirty one dependent variabl es (1) word count (2) corporate quote, (3) corporate statement, (4) critic quote, (5) critic statement, (6 ) negative, (7) balanced, (8) positive, (9) natural disaster, (10) rumor, (11) workplace violence, (12) malevolence, (13) challenge, (14) technical error accident, (15) technical error product harm, (16) humanerror accident, (17) human error product harm, (18) organizational misdeed, (19) attack the ac cusers, (20) denial, (21) scapegoat, (22) excuse, (23) provocation, (24) defeasibility, (25) accidental, (26) good intentions, (27) justification, (28) reminder, (29) ingratiation, (30) compensation, and (31 ) apology. The following paragraphs will offer conceptualizations of all dependent variables. Conceptualizations The word count is the number of words in the observation, excluding headlines and bylines. A quote is a sentence or phrase enclosed in quotation marks and attributed to Coca Cola or a critic. A statement is a sentence that is attributed to Coca Cola or a critic. (i.e. Coca Cola said). The terms negative, balanced, and positive were used to describe the tone of eac h observation. Tone was determined by reading each paragraph and classifying it as either positive or negative. Then the total number of positive and negative paragraphs was added up and compared. If there were more posit ive than negative paragraphs
48 t hen the observation was labeled posit ive and vice versa. If there were the same number of positive as negative paragraphs then the unit was considered balanced. Negative is defined as the presence of words that indict or crit icize Coca Cola Positive is operationalized as the presence of words that support or defend Coca Cola Balanced is conceptualized as a source that has the same number of paragraphs with a favorable connotation as an unfavorable connotation. The ten cri sis types were conceptualized using Coombs (2007b) definitions. NATURAL DISASTERS. A re acts of nature such as tornadoes or earthquakes (Coombs, 2007b Table 6). MALEVOLENCE. O ccurs when an external agent causes damage to the organization (Coombs, 2 007b Table 6). TECHNICAL ERROR ACCIDENT. Includes equipment or technology failure that cause an industrial accident (Coombs, 2007b Table 6). TECHNICAL ERROR PRODU CT HARM. D escribes equipment or technology failure that cause a product to be defective or potentially harmful (Coombs, 2007b, Table 6). HUMAN-ERROR ACCIDENT. Involve industrial accident caused by human error where as HUMAN-ERROR PRODUCT HARM. D escribes situations where a product is defective or potentially harmful because of human err or (Coombs, 2007b Table 6). CHALLENGE. Is a stakeholder claim that the organization is operating in an inappropriate manner (Coombs, 2007b, Table 6). ORGANIZATIONAL MISDEED. Occurs when there are management actions that put stakeholders at risk and/ or violate the law (Coombs, 2007b, Table 6). WORKPLACE VIOLENCE. Is an attack by former or current employee on current employees onsite (Coombs, 2007b, Table 6). RUMOR. I s false and damaging information being circulated about your organization (Coombs, 2007b Table 6).
49 Coombs (2007b) was also the source for the crisis responses definitions. The first response, attack the accusers occurs when the crisis manager confronts the person or group claiming something is wrong with the organization (Coom bs, 2007b, Table 5). In the denial response a crisis manager asserts that there is no crisis (Coombs, 2007b Table 5). When the scapegoat response is used the crisis manager blames some person or group outside of the organization for the crisis (Coombs 2007b, Table 5). The excuse response is implemented when a crisis manager minimizes organizational responsibility by denying intent to do harm and/or claiming inability to control the events that triggered the crisis (Coombs, 2007b Table 5). The subty pes of excuse are provocation, defeasibility, accidental, and good intentions. Provocation involves responses in which the crisis manager says the crisis was a result of response to some one elses actions (Coombs, 2007b Table 5). The defeasibility response results when the crisis manager asserts that there was a lack of information about events leading to the crisis situation (Coombs, 2007b, Table 5). The accidental response asserts that there was a lack of control over events leading to the crisis s ituation (Coombs, 2007b, Table 5). When the good intentions response is used it will be claimed that the organization meant to do well (Coombs, 2007b, Table 5). Justification is the response type in which the crisis manager minimizes the perceived damage caused by the crisis (Coombs, 2007b, Table 5). The reminder response occurs when the crisis managers tell stakeholders about the past good works of the organization (Coombs, 2007b, Table 5). Ingratiation is implemented when the crisis manager praises stakeholders for their actions (Coombs, 2007b, Table 5). Compensation as a response is used when a crisis manager offers money or other gifts to victims (Coombs, 2007b, Table 5). Finally, the
50 apology response occurs when the crisis manager indicates the organization takes full responsibility for the crisis and asks stakeholders for forgiveness (Coombs, 2007b, Table 5). Measurements Coding was the primary method for answering the research questions and hyp otheses presented in this study Research Q uestio n 1 and Hypothesis 1 were measured by identify ing which crisis type(s) appeared most frequently overall in the units of an alysis. Research Q uestio n 2 and Hypothesis 2 were measured in a similar manner. The number of times each crisis response was present overall determined how the crisis response was described. Research Question 3, Hypothesis 3, and Research Questions 4 were measured by comparing how many times the corporate voice was present overall versus how many times the critics voices were present. To measure the second portion of Research Questions 3 and 4, each paragraph of each article w as marked as positive or negative. The number of positive and negative paragraphs was tallied at the end of each source. If there were more positive than negati ve paragraphs then the observation was considered positive and vice versa. Source types were only descr ibed as balanced if there were an equal number of positive and negative paragraphs. Research Q uestion 5 and Hypothesis 5 were measured by identifying the crisis responses in Coca Colas official crisis communications and counting these responses to determine the top three. The crisis responses cited in U.S. newspapers followed the same process. A comparison of t he top three strateg ies of each source type determined if the same strategies were highlighted in each communication.
51 Inter coder Reliability Intercoder reliability was calculated between the researcher and another graduate student who served as the second coder. A codebook was made to define all relevant terms, and the coder was trained to ensure her understanding of each term. Ten percent of the data collected was isolated for both coders to code. Both coders filled out the coding sheet which included sections on crisis types and crisis responses. For these sections coders marked 0 absence and 1 for presence. Once both coders completed coding the sample, Holstis method, PAo = 2A/(nA + nB), was used to calculate the percentage of agreement (Neuendorf, 2002). A minimum of .80 agreement is required f or coding to be deemed reliable, which the researcher and coder achieved at .96 agreement. Data Analysis Data was analyzed using the Statistical Package for the Soci al Sciences for Windows (SPSS v. 19) Descriptive statistics were run to explain and summarize the characteristics of the 31 dependent variables in the 73 observations. Cross tabulations with c hi square tests were run to see if statistically significant relationships existed between the crisis types and the types of communication as well as the crisis re sponses and the source types Lastly, Analysis of Variance (ANOVA) was used to determine if there was a significant difference between the independent variables (source type) and the length of the observations.
52 Table 31 LexisNexis search term r esults Term(s) Documents Returned Sinaltrainal 75 Sinaltrainal, Coca Cola 67 Paramilitaries, Colombia, Coca Cola 223 Labor unions Colombia, Coca Cola 305 Sinaltrainal, Coke 57 Pa ramilitaries, Colombia, Coke 192 Labor unions, Colombia, Coke 275 Trade unions, Colombia, Coke 47 Trade union, Colombia, Coca Cola 51 Total 1,302
53 CHAPTER 4 FINDINGS Research Question 1 Using the information presented in official Coca Cola com munications and newspaper s, how can the crisis CocaCola experienced be classified using the crisis types derived from Coombs Situational Crisis Communication Theory? Ten crisis types from Coombs (2010b) Situational Crisis Communication Theory were identified. Frequencies and cross tabulations with chi square were run on each crisis type to test for significance. Figure 41 and Table 41 show the results of these analyses. For a crisis type to be considered significant it had to have an alpha score less tha O nly four of the 10 crisis types appeared in the source type. They are challenge (59%, N = 43), organizational harm (58%, N = 42), rumor (40%, N = 29), and human error product harm (3%, N = 2). See Figure 41. Table 41 demonstrates how many times crisis as an overall concept was present in the source type. The results indicate that crisis appeared in 100% (N = 45) of newspaper articles and 36% (N = 10) Coca Cola communications. The chi square in Table 41 shows that there is a significant relationship between crisis and the source apparent in both newspapers and Coca Cola communications. E ach crisis type was individually tested against both newspapers an d CocaCola communications. The dependent variables, natural disasters, workplace violence, malevolence, technical error accident, technical error product harm and human error
54 accident were explored, but no statistical relationships were found. These crisis types were absent because they did not apply to the CocaCola crisis. Human error product harm and rumor both appeared in the source types but did not generate statistically significant relationships. This means that although these crisis types were present, the presence was not notable. Human er ror product harm appeared in 4% (N = 2) of newspaper s. As seen in Table 41, the chi square score for human error product harm was not conducted because it not have enough observations to run the analysis. Rumor was present in 26% (N = 19) of newspaper art icles and 36% (N = 10) of CocaCola communications The chi square score for rumor was [ = .31; p = .581] Two crisis types showed a statistically significant relationship with the source types. They are challenge and organizational misdeed. Both of these crisis types appeared only in newspapers. Challenge appeared in 43 newspaper articles (59%), and as depicted in Table 41, its chi indicates that challenge had a very noticeable presence in newspapers. Organizational misdeed appeared in 58% (N = 42) of newspaper articles. The chi square for presence was also very prominent in newspapers. Research Question 2 Using the information presented in official Coca Cola communications and newspaper s, how can Coca Colas response be described using the crisis responses identified in Coombs Situational Crisis Communication Theory? How can Coca Colas most frequently used crisis response(s) be categorized using the aforementioned theory?
55 Thirteen cr isis responses were derived from Coombs Situational Crisis Communication Theory. Frequencies and cross tabulations with chi square were run on all crisis responses to test for significance. For a crisis response to be considered significant, it had to hav e an alpha score less than or equal to .05. Of the 13 crisis responses, 10 were present in the observations as Figure 42 and Table 42 show They are denial (66%, N = 48), reminder (37%, N = 27), attack the accusers (23%, N = 17), excuse (23%, N = 17), s capegoat (21%, N = 15), accidental (16%, N = 12), ingratiation (8%, N = 6), compensation (3%, N = 2), defeasibility (1%, N = 1), and justification (1%, N = 1). The association between provocation, good intentions, and apology was explored but no statistically significant relationship was found. There were five crisis responses that had a presence i n the source types but did not have a significant statistical relationship with the independent variables. The first response, attack the accusers, was present in 29% (N = 13) of newspapers and 14% (4) of Coca Cola communicati ons. The chi 73) = 2.06; p = .151], which means this crisis type did not have a significant presence in the type of communication. Defeasibility in 2% (N = 1) of newspapers only. Its chi 1, N = 73) = .63; p = .427], which shows that defeasibility did not have a significant presence in newspapers or Coca Cola communications (Table 42). Justification was present in 2% (N = 1) of newspapers. Its chi = .63; p = .42 7], which indicates that justification did not have a noticeable presence in the communications studied. Ingratiation appeared 7% (N = 3) of newspapers and 11% (N = 3) of Coca Cola communications. Its chi
56 .540], demonstrating that ingratiation did not have a strong presence in the types of communication. Lastly, compensation appeared in 2% (N = 1) of newspapers and 4% (N = 1) of Coca Cola communications. Compensation had a chi 73) = .12; p = .731] which indicates that compensation did not have a notable presence in Coca Cola communications. There were five crisis responses that had significant statistical relationships with the source types. The first one, denial, was present in 89% (N = 40) of n ewspapers and 29% (N = 8) of Coca Cola communications. Its chi p = .000], which demonstrates that denial had a very significant presence in both newspapers and Coca Cola communications. This result may be seen in Table 42. Reminder appeared in 11% (N = 5) of newspapers and 79% (N = 22) of Coca Cola communications. Reminder had a chi showing that it also had a very significant presence in both sources. Scapegoat was present in 29% (N = 13) of newspapers and 7% ( N = 2) of Coca Cola communications. Its chi that scapegoat had a small but noticeable presence in both newspapers and Coca Cola communications. Accidental was present in 24% (N = 11) of newspapers and 4% (N = 1) of Coca Cola communications. The chi 73) = 5.48; p = .019], which indicates that accidental and the source type had a significant relationship. Finally, excuse had a presence in 31% (N = 14) of newspapers and 11% (N = 3) of Coca Cola communications. The chi = 73) = 4.02; p = .045]. This result was very close to the rejection threshold of .05, which means excuses presence is questionable but still considered signific ant.
57 Research Question 3 What are the most prominent voices, in terms of corporate voice versus critics voices, included in the news coverage of Coca Colas crisis? Can these voices be described as positive, negative, or balanced? The researcher initiall y identified two main categories of voices: critic(s) voice(s) and the corporate voice (or Coca Colas voice). Further reading of the source types created two subtypes for each voice: quotes and statements. Frequencies and cross tabulations were run to test the significance of each voice in the observations. For a voice to have a significant presence in each communication type, it had to have an alpha score less than or equal to .05. As Figure 43 illustrates, the corporate voice as a whole was present in 72 (99%) of the observations. Corporate quote was noted in 36 (49%) communications where as corporate statement appeared in 57 (78%) of the source types. The critic voice as a whole was present in 49 (67%) of the communications. The critic quote was present in 36 (49%) communications, and the critic statement was published in 43 (59%) of the observations. The corporate voice appeared in 98% (N = 44) of newspaper articles as Table 43 shows. The corporate quote appeared in 64% (N = 29) of newspaper articles, and the corporate statement appeared in 87% (N = 39) of the same communication. The critic voice was also present in 98% (N = 44) of newspaper articles. When broken down, the critic quote was published in 80% (N = 36) of the articles while critic statement appeared in 98% (N = 40) of the same communication. In official Coca Cola communications, the corporate voice as a whole appeared 100% (N = 28) of the time. The corporate quote was present in 25% (N = 7) of the
58 corporate communications and the corporate statement was published in 64% (N = 18) documents. The critic voice as a whole appeared in 18% (N = 5) of the communications. Critic quotes were present in 0% (N = 0) of official Coca Cola communications, and critic statements were published in 11% (N = 3) publications (Table 43). As Table 43 illustrates, t he chi square for the = .631; p = .427], which is not significant. This means that the presence of the corporate voice was not related to the observations. The chi square for the corporate quote was corporate quote varied by the type of communication. The chi square for corporate ich shows that there is a somewhat significant relationship between the presence of the corporate statement and the type of communication. The chi .000], proving the presence of the critic voice was very strongly related to the type of communication. The chi square for t he critic quote was .000]. The alpha score also shows a very significant relationship between the critic quote and both newspapers and Coca Cola co mmunications. Lastly, the chi square the presence of the critic statement was very strongly related to the type of communication. A one way ANOVA was run to test if there was a significant relationship between source type and tone (Table 44, Figure 44). A significant relationship was found between the tone and the source type [F(1, 70) = 328.850, p = .000]. The pvalue
59 between negative articles and source type was .000. The pvalue between positive and source type was .000. These results mean that the source type impacted the tone. The chi square for tone as it relates to the source type is depicted in Table 45. had a very strong relationship with the source type. Research Question 4 How noticeable was Coca Colas corporate v oice in news coverage pertaining to the Sinaltrainal controversy and is the tone positive, negative, or balanced? The corporate voice appeared in 99% (N = 72) of the communications (Figure 43). As Table 43 shows, 98% (N = 44) of these instances were in newspapers, and 100% (N = 28) were in CocaCola communications. These results demonstrate that Coca Colas voice was very prominent in both newspapers and CocaCola communications. Table 45 illustrates that the negative tone was present in 51% of t he publications (N = 37). The balanced tone appeared in 8% (N = 6) of the sources, and the positive tone was present in 41% (N = 30) of the communications. Because the negative tone appeared most frequently, the tone in the news coverage toward Coca Cola can be described as negative. Research Question 5 Using the crisis responses discussed in Coombs Situational Crisis Communication Theory, how can the crisis response(s) from Coca Colas official crisis communications be compared to U.S. newspapers portr ayal of Coca Colas crisis response (same or different)?
60 The presence of Coombs 13 crisis responses in each observation was measured using a frequency test. The results in Table 42 show that newspapers published 10 of the 13 crisis responses and Coca C ola communications contained eight of the crisis responses. The responses present in the newspapers were denial (89%, N = 40), excuse (31%, N = 14), attack the accusers (29%, N = 13), scapegoat (29%, N = 13), accidental (24%, N = 11), reminder (11%, N = 5), ingratiation (7%, N = 3), defeasibility (2%, N = 1), justification (2%, N = 1), and compensation (2%, N = 1). The Coca Cola communications implemented reminder (79%, N = 22), denial (29%, N = 8), attack the accusers (14%, N = 4), excuse (11%, N = 3), i ngratiation (11%, N = 3), scapegoat (7%, N = 2), accidental (4%, N = 1), and compensation (4%, N = 1). The top three crisis responses present in newspaper articles were denial, excuse, and attack the accusers. Coca Cola communications implemented reminder, denial, and attack the accusers. The frequency distribution in Table 42 shows that denial was used far more frequently in newspapers than in Coca Cola communications. The main strategy CocaCola used was reminder. Although both communications had at tack the accusers in common, this response had more appearances in newspapers. These results indicate that the crisis responses were not the same in the source types. Additional Findings A one way ANOVA was run to compare the effect of tone on word count in negative, balanced, and positive conditions. As Table 46 and Figure 4 4 show tone had a significant effect on word count at the p < .05 level for the three conditions [F (2,70) = 3.77; p = .028]. This means that word count varied due to tone. More spe cifically, as word count increased, the article became more negative.
61 An independent samples t test was conducted to also evaluate if a significant relationship existed between word count and the source type. The results in Table 47 show that there was a statistically significant relationship between the word count and newspaper (M = 644.93, SD = 418.84) as well as word count and Coca Cola communications (M = 405.54, SD = 360.01); t (71) = 2.492, p = .015). This means that word count varied with the sourc e type. The effect size was also calculated using the formula for Cohens standard, d = x1x2 /s. The score was .590 whic h means the effect size is medium Finally, a bar graph was created to assess when the most coverage occurred. Figure 45 indicates t hat there were five dates which saw an increase in news coverage. They are July 20, 2001; April 2, 2003; January 19, 2006; January 25, 2006; October 4, 2006. Hypothesis 1 Organizational misdeeds will be the crisis type most frequently used to describe the Coca Cola controversy. With a chi square of [ significance level of .000, organizational misdeed had a very significant presence in the communications. However, challenge had a chi square of [ and a significance level of .000, making it also a frequently used crisis type. Therefore, H ypothesis 1 is partially supported. Hypothesis 2 Coca Cola will use denial more than any other crisis response. With a chi square of [ significance of .000. denial was the crisis response mo st frequently used by CocaCola. Consequently, H ypothesis 2 is supported.
62 Hypothesis 3 Coca Colas voice will be more prominent than critics voices. The corporate voice had a chi corporat e quote had a chi The corporate statement had a chi level of .025. The critic voice had a chi signi ficance level of .000. The critic quote had a chi and a significance level of .000. The critic statement had a chi = 43.58] and a significance level of .000. These scores indicate that the criti c voice, critic quote and critic statement had a stronger presence in the source types than their corporate counterparts. Hence, H ypothesis 3 is rejected. Hypothesis 4 The media coverage of the Coca Cola crisis will be primarily positive. The ANOVA betwee n tone and source type shows that the negative tone and newspapers had a si gnificant relationship. Hence, H ypothesis 4 is rejected. Hypothesis 5 The crisis responses presented in official communications from Coca Cola will be the same as the U.S. newspap ers portrayal of Coca Colas crisis responses. newspapers. Refer to Table 42. In Coca Cola c 2.06] were the most common crisis responses. The fact that there was at least one
63 difference in response (scapegoat and reminder) and differences in the frequencies of the common responses, denial and attack the accusers, H ypothesis 5 is rejected.
64 Figure 41 Bar graph illustrating the crisis types that were present in the source t ype s
65 Figure 42 Bar graph illustrating the pr esence of crisis responses in the source types
66 Figure 43 Bar graphs illustrating the presence of the corporate voice and critic v oice in the source types
67 Figure 44 Illustration of the relationship between tone and source types
68 Figure 45 Illustration of the r ela tionship between t one and word count
69 Figure 46 Bar graph i llustrating the source types publication date and how m any communications were published on those days
70 Table 41 Frequency and significance of the c r isis types in the source t ypes Crisis Type Newspaper (N = 45) % Coca Cola Comm. (N =28) % > .05) Crisis 45 100 0 36 38.40 .000 Human Error Product Harm 2 4 0 0 1.28a .258 Rumors 19 26 10 36 .31 .581 Challenge 43 59 0 0 65.11 .000 Organizational Misdeed 42 58 0 0 61.54 .000 Table 42 Frequency and significance of the c risis responses in the source t ypes Crisis Response Newspaper (N = 45) % Coca Cola Comm. (N = 28) % > .05) Attack the Accusers 13 29 4 14 2.06 .151 Defeasibility 1 2 0 0 .63 .427 Justification 1 2 0 0 .63 .427 Ingratiat ion 3 7 3 11 .38 .540 Compensation 1 2 1 4 .12 .731 Denial 40 89 8 29 27.89 .000 Scapegoat 13 29 2 7 5.00 .025 Accidental 11 24 1 4 5.48 .019 Reminder 5 11 22 79 33.70 .000 Excuse 14 31 3 11 4.02 .045 Table 43 Frequency and significance of corporate and critic voices in the source t ypes Voice Newspapers (N = 45) % Coca Cola Comm. (N = 28) % > .05) Corporate Voice 44 98 28 100 .631 .427 Corporate Quote 29 64 7 25 10.74 .001 Corporate Statement 39 87 18 64 5.05 .025 Critic Voice 44 98 5 18 49.96 .000 Critic Quote 36 80 0 0 44. 20 .000 Critic Statement 40 98 3 11 43.58 .000
71 Table 44 Relationship between tone and source t ypes Sum of Squares df Mean Square F Sig. Between Groups 54.551 1 54.551 328.850 .000 Within Groups 11.778 71 .166 Total 66.329 72 Table 45 Frequency and significance of tone in the s our ce t ypes Tone Newspaper (N = 45) Coca Cola Comm. (N = 28) Total > .05) Negative 37 0 37 65.11 .000 Balanced 6 0 6 65.11 .000 Positive 2 28 30 65.11 .000 Table 46 Relationship between word count and t one Sum of Squares df Mean Square F Sig. Between Groups 1187209.684 2 593604.842 3.773 .028 Within Groups 1.101E7 70 157313.579 Total 1.220E& 72 Table 47 Group statistics for word count as it relates to the source t ypes Mean Standard Deviation Degrees of Freedom Significance Newspaper 644.93 418.84 71 .015 Coca Cola Comm. 405.54 360 .01
72 CHAPTER 5 DISCUSSION This study was conducted to evaluate how a multinational company responds to its home public when a crisis originating in a foreign country becomes major news. Summary of the Coca Cola Crisis In December of 1996, a CocaCola worker named Isdro Segundo Gil was shot at the gates of a Colombian bottling plant. His killers were right wing paramilitaries who allegedly had an agreement with the plants management: terrorize unionists and prevent them from organizing and receive c ompensation. Unionists who witnessed Gils death as well as those who were members of his union, Sinaltrainal, sought solidarity with union workers in the United States who eagerly rallied behind them. Through the assistance of U.S. unionists, Sinaltrainal was able to bring a lawsuit against CocaCola and its Colombian bottlers. Meanwhile, labor activist Roy Rogers started the Campaign to Stop Killer Coke which raised awareness about Sinaltrainals plight on college campuses. Students responded to the campaign by demanding that their universities end their contracts with CocaCola. After highprofile universities such as the University of Michigan and New York University terminated these contracts, CocaCola launched a public relations strategy meant to c lear its name in the court of public opinion and protect its reputation. Summary of Crisis Types Used to Portray the Coca Cola Case The F indings reveal that the top three crisis types in order of frequency were challenge, organizational misdeed, and rumor ( Figure 41). These were used to describe the CocaCola case because they were the only applicable crisis types.
73 Challenge was the most popular crisis type, and it was only used in newspapers. Organizational misdeed was the second most popular crisis type, and it also only appeared in newspapers. These crisis types were probably mentioned in newspapers and not CocaCola communications because they both attribute some level of blame to the corporation. Newspapers have at least two major interests in reporting news where a well known corporation is accused of wrongdoing. The first interest involves social responsibility, or the newspapers duty to publish news that is in the publics best interest. As the allegations were never proven to be false, it was incumbent upon the media to present the allegations in all its derivative forms. This includes the possibility that CocaCola was guilty of conspiring with paramilitaries. The second interest is fiscal as reporting salacious news about a prominent corpor ation is likely to generate a lot of public interest. The heightened interest would increase readership and consequently, profits. Because newspapers ultimately exist to make a profit, portraying the crisis as a challenge or an organizational misdeed would benefit the newspaper. Text that supports the characterization of the crisis as a challenge proved abundant. For example, the Boston Herald reported that Ray Rogers, the founder of the Campaign to Stop Killer Coke has spent years trying to make the CocaCola Co. take responsibility for what he says are a series of gruesome labor related murders in Colombia, where the international giant has bottling plants ( McConville, 2010, 2). This quote names a specific individual with an interest in CocaCola and discusses the situation which he believes demonstrates that the corporation is not operating as it should.
74 This characterization balances the interests of the competing voices as it refrains from indicting Coca Cola by attributing the claims to an independent source and including words that express uncertainty about the claims. Thus, allegations that could damage CocaColas reputation are made public (critic interest), but they are never confirmed or advanced (corporate interest). The accommodation of both voices demonstrates the journalists desire to follow the journalistic value of objectivity. As neither side is fully supported by the newspapers, it would appear that the status quo remains unchanged. CocaCola remains a multinational corporation facing accusations of wrongdoing rather than an unethical multinational or a victim of activists agendas. Similarly, those bringing the allegations remain activists rather than instigators or champions of social justice. Because change either way is not encouraged and CocaCola maintains its status, Hermans (2000) assertion that objectivity supports corporate interests and in effect conservative ideology is supported. The second most popular crisis type described in the source types was organizational m isdeed. This crisis type is defined as management actions that put stakeholders at risk and/or violate the law (Coombs, 2007b, Table 6) Lawsuits against a corporation for its alleged wrongdoing are an example of organizational misdeed, which makes this crisis type sound similar to challenge. Although subtle, there is a distinction between the two types. An organizational misdeed identifies management actions as essential to the definition ( Coombs, 2007b, Table 6). What this implies is that the actions were deliberate so the corporation can justifiably be blamed. In contrast, challenge describes stakeholder claims which creates uncertainty as to whether or not
75 these claims are true, making it difficult to ascribe blame to the corporation (Coombs and Holladay, 2002). Journalists who wrote articles using organizational misdeed are condemning Coca Cola instead of dismissing or minimizing the allegations. This demonstrates that corporations are not always supported by news agencies. Figure 41 and Table 4 1 show that natural disasters, technical error accident, technical error product harm, humanerror accident, workplace violence, and malevolence had no presence in the observations and hence no significance. These crisis types were absent because they did not apply to the CocaCola crisis. Coca Cola communications refrained from portraying the crisis as either a challenge or an organizational misdeed because the corporation had no interest in blaming itself for a tragic situation. If it would have done so, it could have encouraged lawsuits and negative publicity both of which would have hurt the corporate reputation and profits. Rumor was the third most popular crisis type, and it was mainly used in CocaCola communications. It was probably employed by Coca Cola because the rumor crisis type is the only one of the three applicable types that does not ascribe blame to the corporation. If blame for a crisis cannot be attributed to an organization, then the corporate reputation will not be threatened. Rumor probably appeared in a few newspaper articles because those newspapers either had some sort of relationship with Coca Cola, which could have been jeopardized if the newspaper had implicated the corporation, or they did not believe the corporation was guilty.
76 Summary of Coca Colas Crisis Responses Denial as the Primary Response Based on the information contained in newspapers and official Coca Cola communications, the crisis response that CocaCola used the most w as denial. Coombs (2007a) suggests t hat denial is the strongest way to distance a corporation from negative allegations because it severs the link between the crisis and the corporation. Once the connection is gone, then the public will not assign blame to the corporation. Without blame, the corporate reputation does not suffer. Since CocaCola wanted to protect its reputation as much as possible, denial is the best strategy for it to use, especially when the allegations could not be definitely proven. Denial was used in various ways by the beverage giant. At the beginning of the crisis, the corporation denied the allegations by calling them false and asserting that Coca Cola had no connection to the tragedy. As time went on, CocaCola adjusted its message by refusing to claim ownership of the bottlers while emphasizing the contractual nature of their relationship. By doing so, CocaCola could explain its connection to the bottlers while showing that it did not legally control the bottlers actions. This adjustment probably occurred becaus e it was undeniable that Gil had been murdered on a CocaCola bottlers premises, so the public may not have believed that there was no link between the bottler and the paramilitaries. CocaCola would have had to find a strategy that is credible in order to protect its reputation. It is interesting to note that denial appeared mostly in newspaper articles rather than CocaCola communications. This is probably because newspapers have access to large groups of people that CocaCola could not reach on its own. If Coca Cola
77 wanted to protect its reputation and prevent people from believing the allegations then it would make sense to use denial in the media. Evidence that supports denials presence is as follows. A CocaCola communication asserted the cl aims in the suit filed against The CocaCola Company and two bottlers in Colombia are inaccurate and based on distorted versions of events (Miami court dismisses Colombia lawsuit, 2006, 3). A similar communication proclaimed, The allegations made against us in Colombia are absolutely false and they are repugnant to all of us at The CocaCola Company (Notice of annual meeting of shareowners, 2006, p. 65, 4). These quotes contain strong words like repugnant and distorted and phrases such as it goes without saying and do not reflect the facts which attempt to create as much distance between CocaCola and the allegations as possible. The choice of words also suggests the corporations confidence in its innocence. Consumers may be assured by t his confidence and as a result support CocaCola. Coca Cola probably refrained from using denial as much in its own communications because if a consumer is making the effort to access the corporations website, it is probably because they want to hear more about Coca Colas perspective than what is given in the media. This is the perfect opportunity for the corporation to draw attention away from the negative accusations and direct it towards positive messages. Then the consumer can repeat these messag es to the people he or she interacts with, which could counter the effects of negative publicity. Additional Responses The denial response was supplemented with excuse, accidental, scapegoat, and attack the accusers. These four crisis responses are related as they all involve
78 removing the blame from the corporation and directing it to a third party or extraneous circumstances. When Coca Cola blamed Colombias political environment for Gils death or asserted that it was not responsible because it did not own the subsidiaries, it was using excuse, accidental, and scapegoat. Excuse and accidental both involve a corporation claiming that it could not control the circumstances that caused the crisis. If CocaCola could get the public to believe that the violent political environment was responsible for the murder, then it would be absolved from any culpability because no logical person would assert that CocaCola is responsible for Colombias civil unrest. If CocaCola could convince people that it merely contracted with the subsidiaries but did not directly control them, then again, the corporation would not be blamed. Blaming Colombias political situation and emphasizing the nature of a contractual relationship were also examples of scapegoat as this resp onse entails placing responsibility for the crisis on an independent third party. If the outside party receives the blame, then the corporation does not have to worry about its reputation being threatened. Attack the accusers was also used to supplement denial when Coca Cola argued that the unionists and their allies were using the corporations name in the lawsuit to attract attention and advance their own political agenda. By employing this strategy, the responsibility for the crisis was deflected away f rom Coca Cola and placed on the critics shoulders. If the public accepts this strategy then CocaColas reputation remains intact.
7 9 Provocation was not present in any of the observations because it involves claiming responsibility for the crisis while as serting that the crisis was the result of someone elses actions. This response would not appeal to CocaCola because it admits involvement in the crisis, and CocaCola did not want any link to it as evidenced by its frequent use of denial. Similarly, good intentions and apology also admit responsibility for the crisis, which explains why these responses did not appear. The following crisis responses did not have a significant presence. They are defeasibility, justification, ingratiation, compensation, and attack the accusers. Defeasibility probably was not used very much because it emphasizes that the corporation did not have enough information to prevent the crisis. Because the violent political situation in Colombia has been occurring for about 50 year s, it would be implausible for CocaCola to claim that it could not have had enough information to prevent such a crisis. At the very least, employing defeasibility would make the corporation look like it did not do its due diligence and was irresponsible. The public is unlikely to respond well to an admission of negligence especially when a life was lost. Justification involves making the crisis seem less significant. This response did not have a significant presence because since the unionists were symp athetic regardless of Coca Colas involvement, it would look callous if it tried to downplay the situation. The compensation response requires the corporation to pay the victims of the crisis. Coca Cola created several social programs in Colombia to hel p victims of violence, so it indirectly compensated some Colombian citizens. However, overtly stressing that CocaCola was going to directly pay the victims would make it appear
80 guilty. As the corporations goal was to be completely separated from the all egations, employing this response would be counterintuitive. Ingratiation was used a couple of times in the source types. Mainly this response was used to thank the courts for dismissing the lawsuits against CocaCola. This response probably was not used much because there were not many public figures supporting the corporation, so there were few opportunities to express appreciation. Attack the accusers probably was not used frequently by CocaCola because, as mentioned in the justification paragraph the persons bringing the allegations were very sympathetic. Attacking them would make the corporation look like a bully and decrease the publics good will. Reminder as Coca Colas Primary Response Reminder was the third most popular strategy overall and the primary strategy within Coca Cola communications. This strategys goal is to call attention to a corporations good deeds Reminder probably was not used frequently in the newspapers because one of this publications goals is to be objective. Highlighting a corporations positive deeds would make the newspaper appear biased towards CocaCola. As biases are not tolerated in U.S. media, the prominent use of reminder would not support the newspapers interest. In contras t, reminder could have appealed to CocaCola because it would allow it to be viewed favorably by the public. If Coca Cola is seen in a positive light then the public may get distracted from the allegations and continue to support the corporation. In this scenario, CocaColas reputati on does not suffer, and its profits are unaffected. Coca Cola s used reminder when it discussed the ways it has helped its Colombian workers, mainly by offering them secure transportation and cell phones to use to call for
81 help. The corporation also discussed a social program it started which funded education for victims of violence. These examples were likely meant to show that CocaCola could not have been involved in Gils death because the corporations actions demonstrate that it supports the workers and has no interest in hurting them. Application of Situational Crisis Communication Theory Coombs (2007c) offers several suggestions about applying Situational Crisis Communication Theory (SCCT). He states that the denial response should be used when a c orporation experiences an unwarranted challenge or rumor. Because challenge was the most frequent crisis type, it can be argued that CocaCola adhered to SCCT as it used denial more than any other strategy. However, one important caveat for employing the denial strategy is that it should not be used when a corporation is experiencing a challenge that stakeholders are likely to believe (Coombs, 2007c). Although no analysis was undertaken to determine if stakeholders believed the challenge, it can be assume d that some stakeholders did given the protracted nature of the crisis. The recommended response for this type of challenge is to engage in corrective action, or fix the original cause of the crisis to prevent it from reoccurring. Evidence that suggests Coca Cola took corrective action includes its provision of cell phones and transportation to its Colombian workers, and its creation of a global human rights policy. These measures were meant to prevent employees from experiencing the danger that caused the original crisis. As Coca Cola responded to both forms of challenge with the suggestions advocated by Coombs, it is apparent that the corporation was executing SCCT. Organizational misdeed was the second most popular crisis type and it falls under the preventable cluster. Coombs (2007c) advises using a rebuilding strategy for crises
82 of this nature. Responses that fall under the rebuilding strategy are compensation and apology. CocaCola never used apology and when it issued compensation, it was limited to charity. As money was never directed at the victims, classifying the money that was put into foundations as compensation may be improper. In which case, CocaCola did not follow the recommended strategies when facing the organizational misdeed frame. Rum or was the third overall crisis type, and the most frequent c risis type present in Coca Cola communications. As with challenge, Coombs recommends using denial to combat rumor. While denial was the most common response overall, reminder was used more than any other type of response in the corporate communications. This suggests that reminder was used to combat rumor. Tests were not run to examine which response corresponded to each crisis type so this assumption is not definite. If it is accurate, then Coca Cola did not use the suggested response to deal with a rumor crisis. However, Coombs (2007c) indicates that rebuilding strategies, such as reminder, should be used to supplement other responses. If reminder was not intended as a main response, then CocaCola executed SCCT appropriately when addressing rumor. Coca Cola also used several other crisis responses to a lesser degree. They are excuse, accidental, scapegoat, and attack the accusers. In his suggestions, Coombs notes that denial strategies should not be used in conjunction with any of the strategies listed in the other postures. As excuse and scapegoat are part of the diminishment posture, CocaCola did not heed this recommendation. However, this oversight does not appear to have affected the overall campaign.
83 Evaluation of Coca Colas Crisis Responses and Recommendations The analyses indicate that CocaCola mostly followed SCCT when responding to the labor rights crisis. Its campaign appears to have been successful because its profits have not suffered and the two most prominent schools to boycott CocaCola, the University of Michigan and New York University, have both resumed their relationships with the corporation. However, the crisis has not completely vanished as a documentary critical of the corporations Colombian business practices emerged in 2010, prompting a new wave of protests. As all multinationals face scrutiny from critics, it is not necessarily Coca Colas fault that the crisis has not completely dissipated, but CocaCola could alter its previous strategy to increase its reputational protection. Throughout its campaign, CocaColas primary focus was on avoiding legal liability for the crisis. As the crisis itself and the political context in which it occurred are both sensitive, Coca Cola would have been wise to express more compassion towards the victims, in addition to its denial response, as this would have helped it appear more human and evoked favorable attitudes toward it from the public. Additionally, the provision of cell phones and transportation and the creation of a global human rights policy were both commendable efforts, but they are not powerful enough to convince the more cynical stakeholders of the corporations sincerity to promote change. An independent audit o f the bottlers (to see if they were guilty of conspiring with paramilitaries) could have persuaded such stakeholders, but commissioning such an investigation could have invited other lawsuits. In sum, CocaColas responses were effective, but there is room for some improvement.
84 Summary of General Tone Table 45 shows that the negative tone appeared most frequently in newspaper articles whereas the positive tone was mostly present in the CocaCola communications. Because the number of articles marked negat ive was more than those marked as positive, the observations as a whole were considered negative. The newspapers probably used the negative tone the most as they are responsible for reporting news that is critical of a major corporation if doing so is in the publics best interest. In this case, ensuring that CocaCola acted ethically abroad is in the publics best interest so newspapers were obligated to report the situation. As the circumstances surrounding Gils death are not positive, even from an obj ective standpoint, by default words used in the report would be negative so this could explain why so many newspapers articles had this tone. Another possible explanation is that newspapers do not necessarily have an interest in portraying CocaCola posit ively unless a business relationship exists with the corporation. Using negative tones would create a more interesting story than using positive tones and would attract more readers. An increase in readership would increase the newspapers profits which obviously benefits them. The negative tone was arguably so frequent because the balanced tone was conceptualized poorly. To recall, balanced was defined as an observation with the same number of positive and negative paragraphs. The researcher conceptualized balanced in this way because upon a preliminary reading of the observations, it was discovered that many of them gave irrelevant information. Classifying irrelevant information as balanced could potentially skew the results, which was unfavorable. The res earcher thought about asking the coder only to mark paragraphs that were relevant to the
85 Colombian crisis, but as CocaCola was simultaneously battling allegations coming from India, many of the observations intertwined the two situations into one paragraph. It became impossible to define what was relevant without going through each source with the other coder, which also would have impacted the results. Therefore, balanced was conceptualized in the aforementioned way. T he negative tone was absent in CocaCola communications, and the positive tone was overwhelmingly present because the corporation has no interest in threatening its own corporate reputation. CocaCola exists to sell products and make money so to publish negative messages about itself woul d go against its fundamental purpose. The use of positive tones makes sense because this would allow CocaCola to build itself up and encourage the public to support it and buy its product. Summary of the Word Count Tone Relationship Table 47 shows that word count and tone had a significant relationship. Figure 44 indicates that the mean of articles with a negative tone was the greatest. The balanced tone had the second highest number of words, and the positive toned articles had the fewest amount of words. The negative toned articles probably had the most words because there was plenty of interest in the crisis and many people were interested in giving their opinion about the situation. Given the sympathetic nature of the unionists, most people unaffil iated with CocaCola probably had a negative view of the corporation, so negative commentary was abundant. Also, the newspapers not affiliated with CocaCola may have had a financial interest in publishing negative stories about a well known corporation because it would increase readership and hence, profits. As newspaper articles do not have unlimited space to publish stories, it would be more
86 prudent for them to dedicate the space they do have to content that will help their profits. On the other hand, as discussed in the previous section, newspapers must report news that is in the publics best interest. As this situation is anything but positive, it would be very difficult to describe it using positive words. Articles with a balanced tone had the second highest word count probably because the newspapers that published these articles wanted to adhere to the U.S. journalists ideal of objectivity. To be objective, a journalist must present all sides to a story so that he or she does not appear biased. Whe n one undertakes a thorough analysis, the resulting text is likely to be long so this is probably why balanced articles had so many words. T he positive articles probably had the least words because the majority of them originated from the CocaCola websit e. For consumers to actively seek out the corporate website in order to find more information is good for CocaCola but the corporation needs to create publications that are interesting to the consumer so they will remember the message. Writing long paragraphs praising CocaCola will probably result in the reader losing interest in CocaColas message, and then the corporation loses its opportunity to use the consumer as an unofficial spokesperson. By delivering a short, positive and memorable message, the consumer is likely to be engaged and repeat the message. Those that visit the corporate website are probably proCoca Cola or at least willing to hear its side, and CocaCola needs to capitalize on this chance. Summary of Competing Voices Figure 43 and Table 4 3 show that the corporate voice and critic voice both had a strong presence in the observations. Upon closer inspection, it was revealed that newspapers gave almost equal space to both voices so the newspaper articles should
87 have been classifie d as balanced overall. By offering equal space to the corporate and critic voices, the newspapers were encouraging the acceptance of the status quo. According to Herman (2000) maintaining the status quo panders to corporate interests and conservative ideol ogy. His argument is further supported by the fact that newspapers only quoted named sources that were officially sanctioned instead of including independent sources. By doing so, the newspapers only legitimized official voices and dismissed independent voices as not credible. The CocaCola communications overwhelmingly quoted CocaCola employees and supporters and barely allowed the critics to voice their thoughts. This is unsurprisingly considering the purpose of the CocaCola communications was to pr otect and build up its reputation. Allowing critics to speak in the corporations official communications could have strengthened CocaColas communications because it would have appeared less biased, but it is not really expected that corporations would do so. Summary of Publication Date and Frequency The observations confirm that publications increased on July 20, 2001, April 2, 2003, and October 4, 2006 because of developments in the lawsuit. The first date was important because this was the date of the initial filing by Sinaltrainal against CocaCola. On April 2, 2003, a judge ruled that the lawsuit could continue against the Colombian bottlers (one of which was owned by a U.S. citizen) even though CocaCola had been dismissed. On October 4, 2006, the lawsuit against the bottlers was dismissed. The coverage increased in January of 2006 because the University of Michigan had publically stated that it would not carry CocaCola products until the corporation agreed to an independent thirdparty investigation. Also, Coca Cola
88 spokespersons went on news shows to discuss the allegations concerning their bottlers which explains this surge in coverage. In sum, coverage increased when a major development in the public relations battle surfaced. Implications fo r Public Relations and Communication Management This study used Coombs (2010b) Situational Crisis Communication Theory to evaluate CocaColas crisis responses to a Colombian labor rights situation that generated substantial negative publicity for the beverage corporation. The findings of this study validate SCCT as they show that using denial to combat rumor and challenge was effective for CocaCola. This suggests that identifying the crisis type and its inherent level of attributable responsibility bef ore choosing a crisis response from the relevant posture does secure the maximum possible reputational protection for the corporation. Similar studies should be conducted to verify this studys findings. Support for Molleda, Connolly Ahern, and Quinns (2 005) Cross National Conflict Shifting Theory was also found as the crisis studied demonstrates that incidents that occur abroad can have serious repercussions in the home country. This means that public relations professionals need to be knowledgeable about their corporations performance in all areas where it is doing business, as doing so will allow them to contain a potential crisis before it reaches a threatening level. Finally, the findings in this study underscore the need for corporations to pay clos er attention to how collegeaged students perceive them as these students tend to have high levels of activism. This crisis shows that even the strongest corporation can be threatened by small groups of zealous activists.
89 Limitations and Suggestions for Future Study As with all research, this study had several limitations. Firstly, the researcher should have categorized the balanced tone more precisely to allow for more articles to receive this label. The definition used in this study was the unit of analy sis has the same number of paragraphs with a favorable connotation as an unfavorable connotation. Because this definition did not allow any paragraphs to be classified as balanced, the definition may have biased the results. The research should have creat ed a definition that allowed balanced paragraphs to be labeled as such. Secondly, the researcher must acknowledge that newspapers could have portrayed the crisis as a rumor because they were trying to protect themselves from liability, and not because they believe the allegations were false. Thirdly, several of the crisis types and crisis responses overlapped. Efforts should be made to streamline both in order to allow for clearer analyses in the future. Four thly, the results in this study may have been di fferent if the researcher had coded more articles. Articles from U.S. publications and CocaCola were limited so expanding the sample to include international publications would have allowed for stronger results. When writing the discussion section, it w as apparent that the researcher should have identified which voice used each response. For example, it would have been helpful to the analysis if the attack the accusers response had been attributed to a Coca Cola spokesperson or the corporations attorney. Identifying a connection between the voice and the response would have provided more insight into why the particular response was selected. The researcher also should have noted the name of the publication that used each crisis type. By doing so, the researcher could have drawn stronger conclusions about each newspapers motive for framing the crisis in a
90 particular way. Finally, the researcher should have analyzed what crisis response was used for each crisis type. This would have allowed for a stronger assessment of SCCTs effectiveness.
91 APPENDIX A CODING SHEET I. Unit of analysis information 1. Unit of analysis identification number: ____________ 2. Date ______________ 3 What ty pe of communication was the unit of analysis ? 1 Newspaper 2 A commu nication from the Coca Cola website 4 How many words were in the unit of analysis ? ___________________ II. Competing voices Unless otherwise indicated, mark 0 for No and 1 for Yes. 5 Was the corporate voice present in the unit of analysis ? 6. Was the c orporate voice quoted in the unit of analysis? 7. Was any statement attributed to the corporate voice in the unit of analysis? 8. Was/were any critic(s) named in the unit of analysis? 9. Was/were any critic(s) quoted in the unit of analysis? 10. Were any statements attributed to critics in the unit of analysis? 11. How can the voices in the article as a whole be described? 0 Negative 1 Balanced 2 Positive III. Crisis type Mark 0 for No and 1 for Yes 12. Did the article describe the crisis? If yes, answer questions 1322. If no, move on to question 25. 13. W as the crisis portrayed as a natural disaster? 14. W as the crisis portrayed as rumors? 15. W as the crisis portrayed a s workplace violence? 16. W as the crisis portrayed as a malevolence? 17. Was the cri sis portrayed as a challenge? 18. W as the crisis portrayed as technical error accidents ? 19. W as the crisis portrayed as technical error product harm? 20. W as the crisis portrayed as humanerror accidents? 21. W as the crisis portrayed as humanerror produc t harm?
92 22. W as the crisis portrayed as an organizational misdeed? IV Crisis responses by CocaCola based on Coombs SCCT (2007) Mark 0 for No and 1 for Yes. 23. Was attack the accusers used as a crisis response? 24. Was denial used as a crisis respo nse? 25. Was scapegoat used as a crisis response? 26. Was excuse used as a crisis response? This category includes the provocation, defeasibility, accidental and good intentions subcategories. 27. Was provocation used as a crisis response ? 28. Was defeasi bility used as a crisis response? 29. Was accidental used as a crisis response? 30. Was good intentions used as a crisis response? 31. Was justification used as a crisis response? 32. Was reminder used as a crisis response? 33. Was ingratiation used as a crisis response? 34. Was compensation used as a crisis response? 35. Was apology used as a crisis response?
93 APPENDIX B CODE BOOK Unit of analysis ID number: is the number assigned to each unit of analysis. Date: is the day, month, and year the unit of analysis was published. Unit of Analysis: is either a newspaper or official Coca Cola communication. Word Count: is the unit of analysis word count, excluding headlines and bylines. Corporate voice: quotes and statements made by Coca Cola or on beh alf of Coca Cola Statement: is a sentence that is attributed to Coca Cola or a critic (i.e. Coca Cola said). Quote: is a sentence or phrase enclosed in quotation marks and attributed to Coca Cola or a critic. Critics voice: quotes or statements given by anyone critical of Coca Cola Positive: the p resence of words that support or defend CocaCola. Negative: the presence of words that indict or criticize CocaCola. Balanced: this means that the unit of analysis has the same number of paragraphs with a favorable connotation as an unfavorable connotation. Stakeholders: people with an interest in the organization. The interest can either be financial, emotional, social, or legal. Crisis: circumstances that created negative publicity for Coc a Cola or threatened its image. Crisis Types Natural disasters: acts of nature s uch as tornadoes or earthquakes (Coombs, 2007d, Table 6). Malevolence: external agent causes damage to the organization (Coombs, 2007d, Table 6). Examples: product tamp ering, espionage, and terrorism. Technical error accident: equipment or technology failure that cause an industrial accident (Coombs, 2007d, Table 6).
94 Technical error product harm: equipment or technology failure that cause a product to be d efective or potentially harmful (Coombs, 2007d, Table 6). Human error accidents: industrial accident caused by human error (Coombs, 2007d, Table 6). Human error product harm: product is defective or potentially harmful because of human error (Coombs, 2007d, Table 6). Challenges: stakeholder claim that the organization i s operating in an inappropriate manner (Coombs, 2007d, Table 6). Organizational misdeed: management actions that put stakeholders at risk and/or violate the law (Coombs, 2007d, Table 6). Workplace violence: attack by former or current employee on current employees onsite (Coombs, 2007d, Table 6). Rumors: false and damaging information being ci rculated about your organization (Coombs, 2007d, Table 6). Crisis Responses Attack accu sers: crisis manager confronts the person or group claiming something is wrong with the organization (Coombs, 2007d, Table 5). Example: CocaCola attacks critics or vice versa. Denial: crisis manager asserts that there is no crisis (Coombs, 2007d, Ta ble 5). Example: claiming the facts causing the crisis are wrong or untruthful. Scapegoat: crisis manager blames some person or group outside of the organization for the crisis (Coombs, 2007d, Table 5). Example: blaming anyone not from CocaCola for the crisis. Excuse: crisis manager minimizes organizational responsibility by denying intent to do harm and/or claiming inability to control the events that triggered the crisis (Coombs, 2007d, Table 5). Provocation: crisis was a result of response to some one else s actions (Coombs, 2007d, Table 5). Defeasibility: lack of information about events leading to the crisis situation (Coombs, 2007d, Table 5). Example: claiming ignorance about alleged paramilitary/bottler relationship.
95 Accidental: lack of control over events leading to the crisis situation (Coombs, 2007d, Table 5). Good intentions: organization meant to do well (Coombs, 2007d, Table 5). Justification: crisis manager minimizes the percei ved damage caused by the crisis (Coombs, 20 07d, Table 5). Reminder: crisis managers tell stakeholders about the past good works of the organization (Coombs, 2007d, Table 5). Ingratiation: crisis manager praises stakeholders for their actions (Coombs, 2007d, Table 5). Compensation: crisis manager offers money or other gifts to victims (Coombs, 2007d, Table 5). Apology: crisis manager indicates the organization takes full responsibility for the crisis and as ks stakeholders for forgiveness (Coombs, 2007d, Table 5).
96 LIST OF REFERENCES Altschull, J. H. (1995). Agents of power: The media and public policy White Plains, NY: Longman Publishers, USA. Arnold, C., & Cook, T. (2010, April 25). The inverted pyramid structure. Retrieved January 11, 2011 from http://owl.english.purdue.edu/owl/resource/735/04/ BBC. (2010, February 27). Colombia country profile Retrieved from http://news.bbc.co.uk/2/hi/americas/country_profiles/1212798.stm Benoit, W. L. (1995). Accounts, excus es, and apologies: A theory of image restoration strategies. Albany, NY: State University of New York Press. Bucheli, M. (2005). Bananas and business: The United Fruit Company in Colombia, 18992000. New York, NY: New York University Press. Chomsky, A. (2008). Linked labor histories: New England, Colombia, and the making of a global working class Durham, NC: Duke University Press. Coca Cola. (2005, July 9). 2004 citizenship report. http://www.thecoca colacompany.com/ourcompany/pdf/2004_citizenship_report.pdf Coca Cola. (20062010). The Chronicle of Coca Cola. [Company History]. Retrieved from http://www.thecoca colacompany.com/heritage/chronicle_birth_refreshing_idea.html Coc a Cola. (2006, October 3). Miami court dismisses Colombia lawsuit. [Press release]. Retrieved from http://www.thecoca colacompany.com/presscenter/viewpoints_colombia_oct2006.html Coca Cola. (2006, March 10). Notice of annual meeting of shareowners. Retr ieved from http://www.the coca cola company.com/investors/2006_coca cola _proxy.pdf Coombs, T W. ( 1999) Ongoing crisis communication: Planning, managing, and responding. Thousand Oaks, CA : SAGE. Coombs, T.W. (2007a). Protecting organization reputations d uring a crisis: The development and application of Situational Crisis Communication Theory. Corporate Reputation Review. 10 (3), 163 176. Coombs, T.W. (2007b). Crisis management and communications Retrieved January 24, 2011 from http://www.institutefor pr.org/essential_knowledge/detail/crisis_ management_and_communications/#Crisis_Response Coombs, T.W. (2007c). Ongoing crisis communication: Planning, managing, and responding (2nd ed.). Thousand Oaks, CA: SAGE.
97 Coombs, T.W. & Holladay, S. J. (2002). Helping crisis managers protect reputational assets. Management Communication Quarterly 16(2), 165186. Coombs, T W. & Holladay, S.J. (2010a). PR strategy and application: Managing influence. West Sussex, United Kingdom: Wiley Blackwell. Coombs, T.W. & Holladay, S.J. (2010b). The handbook of crisis communication. West Sussex, United Kingdom: Wiley Blackwell. Egelhoff, W. G., & Sen, F. (1992). An information processing mode of crisis management. Management Communication Quarterly. 5 443 484. Fea rn Banks, K. (1996). Crisis communications: A casebook approach Mahwah, NJ: Lawrence Erlbaum Associates. FearnBanks, K. (2007 ). Crisis communications: A casebook approach. ( 2nd ed.). Mahwah, NJ: Lawrence Erlbaum Associates. Garamone, J. (2003, August 14). Colombia: Another front in war on terror. Retrieved July 25, 2010 from http://www.defense.gov/news/newsarticle.aspx?id=28599 Gilpin, D.R., & Murphy, P.J. (2008). Crisis management in a complex world. New York, NY: Oxford. Globalization. (2010). In The Stanford Encyclopedia of Philosophy Retrieved January 11, 2011, from http://plato.stanford.edu/archives/sum2010/entries/globalization/ Groseclose, T., & Milyo, J. (2005). A measure of media bias. The Quarterly Journal of Economics. 120( 4), 1191 1237. Gundel, S. (2005). Towards a new typology of crises. Journal of Contingencies and Crisis Management, 13 (3), 106 115. Harvard Business Essentials. (2004). Crisis ma nagement: Master the skills to prevent disaster. Boston, MA: Harvard Business School Press. Herman, E. (2000). The myth of the liberal media New York, NY: Peter Lang. Hwang, P., & Lichtenthal, J. D. (2000). Anatomy of organizational crisis. Journal of Contingencies and Crisis Management. 8 (3), 129140. International Labor Rights Forum. (2009). Freedom at work: Colombia. Retrieved January 14, 2011 from http://www.laborrights.org/end violenceagainst tradeunions/Colombia
98 International Trade Union Confederation. (2010, June 9). ITUC annual survey: 101 trade unionists murdered in 2009; Pressure on workers rights grows as crisis hits jobs Retrieved January 14, 2011 from http://www.ituc csi.org/ituc annual survey.html Jacquette, D. (2007). Journalistic ethics Upper Saddle River, NJ: PearsonPrentice Hall. Jones, G. (2005). Multinationals and global capitalism: F rom the nineteenth to the twenty first century. Oxford, United Kingdom: Oxford University Press. Lapan, T., & Harris, R. (n.d.). Colombia: The Coca Cola Controversy Video retrieved from http://www.pbs.org/frontlineworld/fellows/colombia0106/index.html Lerbinger, O. (1997). The crisis manager: Facing risk and responsibility Mahwah, NJ: Lawrence Erlbaum Associates. Lehrer, J. (n.d.). Colombias civil war. Retrieved from http://www.pbs.org/newshour/bb/latin_america/colombia/timeline.html Linke, C. G., (1989). Crisis Dealing with the unexpected. In B. Cantor & C. Burger (Eds.), Experts in a ction: Inside public relations, (pp.166178) New York, NY: Longman. Marcus, A.A., & Goodman, R.S. (1991). The dilemma of presenting corporate policy during a crisis. The Academy of Management Journal. 34( 2), 281305. McConville, C. (2010, February 5). Film claims Coke is killer; Beverly native: Co. works with paramilitary groups. The Boston Herald Retrieved November 12, 2010, from LEXIS NEXIS Academic database. McKee, A. (2003). Textual analysis: A beginners guide London, United Kingdom: Sage Publications. McQuail, D. (2010). McQuails mass communication theory. (6th ed.) Thousand Oaks, CA: SAGE. Mitroff, I. I., & Anagnos, G. (2001). Managing crises before they happen: What every executive and manager needs to know about crisis manageme nt. New York, NY: Amacom Books. Mitroff, I. I., Pauchant, T.C., & Shrivastava, P. (1988). The structure of manmade organizational crises: Conceptual and empirical issues in the development of a general theory of crisis management. Technological For ecasting and Social Change. 33 83 107.
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), 87 102. Neuendorff, K. (2002). The content analysis guidebook. Thousand Oaks, CA: SAGE. Osterhammel, J., & Petersson, N.P. (2003). Globalization. Princeton, NJ: Princeton University Press. Perrow, C. (1984). Normal accidents New York, NY: Basic Book s, Inc. Perry, R. (2007). What is a disaster? In H. Rodriguez, E.L. Quarantelli, & R.R. Dynes (Eds.), Handbook of Disaster Research (pp. 115). New York, NY: Springer. Pew Research Center. (2006, July 30). Online papers modestly boost newspaper readership: Maturing Internet news audience broader than deep. Retrieved January 23, 2011 from http://peoplepress.org/report/282/onlinepapers modestly boost newspaper readership Ray ner, J. (2004). The United States and the war on trade unions in Colombia: A call for solidarity. Retrieved February 24, 2010 from http://henningcenter.berkeley.edu/gateway/colombia.html Regester, M., & Larkin, J. (2008). Risk issues and crisis management in public relations: A casebook of best practice London, United Kingdom: Kogan Page. Riffe, D., Lacy, S., & Fico, F.G. (2005). Analyzing media messages: Using quantitative conte nt analysis in research. Mahwah, NJ: Lawrence Erlbaum Associates. Rogers, R. (20042011). The Campaign to Stop Killer Coke Retrieved January 9, 2011 from http://killercoke.org Safford, F., & Palacios, M. (2002). Colombia fragmented land, divided society New York, NY: Oxford University Press. Society of Professional Journalists. SPJ code of ethics. Retrieved January 14, 2011 from http://www.spj.org/ethicscode.asp Seeger, M.W., Sellnow, T.L., & Ulmer, R.R. (1998). Communication, organization, and crisis Communication Yearbook 21, 231275. Simons, G. (2004). Colombia a brutal history. London, United Kingdome: SAQI. Wimmer, R.D., & Dominick, J.R. (2006). Mass media research: An introduction Boston, MA: Wadsworth.
100 Zaremba, A.J. (2010). Cr isis communication: Theory and practice Armonk, NY: M.E. Sharpe, Inc.
101 BIOGRAPHICAL SKETCH Mary Elizabeth Ruiz is from Fort Lauderdale, Florida. She first came to the University of Florida in 2003 to pursue a Bachelor of the Arts in criminology. Mary ret urned to Gainesville in the Fall of 2009 to obtain a Master of the Arts in Mass Communication. She specialized in International/Intercultural Communication with an emphasis on Latin America. Marys research interests include corporate social responsibility social media and corporate reputation, and cross national conflict shifting. She is pursuing a career in corporate communications.