From Strategic Management to Policy Consensus in a Health-related ...

From Strategic Management to Policy Consensus in a Health-related Crisis:
An Analysis of the National Salmonella Outbreak in the United States

Hyojung Park
hpqk3@mizzou.edu
Hyehyun Hong
hhzrc@mizzou.edu
University of Missouri


Abstract
This study applied the contingency theory to examine how the U.S. government (i.e., FDA and
CDC) managed its stance and strategies during the salmonella outbreak that occurred in the
summer of 2008. A content analysis of 72 news articles revealed that the government primarily
demonstrated advocative stances toward consumers and produce industries while cooperating
with state health departments to pinpoint the source of the outbreak. All publics, with the
exception of the tomato industry, appeared accommodative to the government throughout the
crisis. Regarding contingent factors that influenced the government’s stances and strategies, the
issue under question (e.g., the source of the outbreak) appeared to be the predominant factor. In
accord with the essence of the contingency theory, the results suggest that strategic crisis
management is dynamic and that the stances and strategies of an organization shift over time
along a continuum from advocacy to accommodation according to a given situation. The findings
also suggest that the media may play a supporting role for the government in delivering
immediate, up-to-date information and triggering public attention to problems in the existing
food systems.

Introduction
Health-related crises, such as epidemic diseases and food poisonings, require immediate
attention and rapid dispensing of accurate information, and the government often takes a leading
role in dealing with those critical situations (Jin, Pang, & Cameron, 2006). At the onset of a
health-related crisis, the government’s handling of the situation reflects not only its
responsiveness to public demands, but also its ability to protect its citizens from potential risks
(Lee, 2007). While searching for the causes of and solutions to a problem, the government tries
to communicate information relevant to the issues at hand in a timely manner that reduces
uncertainty or misunderstanding among publics. It also gives guidelines that publics can follow
in order to avoid putting their health at risk. In terms of the contingency theory, such
communication can occur at any point along the continuum—from advocacy to
accommodation—and involve different strategies and tactics for its multiple publics based on its
stance, which moves along the continuum (Cameron, Pang, & Jin, 2007; Cancel, Cameron,
Sallot, & Mitrook, 1997). This dynamism of the contingency theory can be applied to analyze the
U.S. government’s crisis management in the nationwide outbreak of salmonella that occurred in
the summer of 2008.
The first documented illness from salmonella occurred on May 11, 2008, and the
outbreak lasted for more than three months due to the uncertainty about its source (Weise, 2008).
Throughout the crisis, the Centers for Disease Control and Prevention (CDC) and the Food and

---

Drug Administration (FDA) cooperated in their efforts to trace the source of the outbreak and
terminate the spread of illness. The FDA first focused on salmonella-tainted tomatoes and issued
a nationwide warning on June 7 that consumers should not eat raw red tomatoes. Still, the FDA
and CDC searched for the exact source of contamination, and the number of infected people
rapidly increased, soon becoming the largest food-borne illness outbreak of the past decade
(Venkataraman, 2008). On June 27, the FDA and CDC announced that they were investigating
other foods commonly eaten with tomatoes (e.g., jalapeno peppers, cilantro, and salsa), although
they were still concerned about tomatoes. The FDA revoked its warning of avoiding tomatoes on
July 17 and finally found a matching strain of salmonella on a jalapeno pepper at a distribution
center in Mexico on July 21. After causing more than 1300 illnesses, the outbreak of salmonella
was finally terminated on August 28 when the FDA lifted its warning to consumers about
jalapeno peppers. During the course of the crisis life-cycle, many conflicting issues emerged
among different groups of public with respect to perceived responsible parties for the underlying
problems behind this crisis. In the early phase of the crisis, the produce and food industries
cooperated with the FDA by pulling tomatoes from the market. After the FDA’s warning against
eating jalapeno peppers, however, the produce industry, especially tomato growers, accused the
FDA of making a hasty decision to put tomatoes first on the suspect list (Venkataraman, 2008).
On the other hand, the FDA blamed produce industries for not having electronic food tracking
systems, delaying the investigation (Lazo, 2008). As the situation unfolded, the government’s
and other publics’ stances toward each other changed.
Despite growing attention to understanding dynamics in the strategic management of
crises, little research has been conducted on health-related crises using the contingency theory.
This study sought to shed more light on the application of contingency theory in health-related
crises by analyzing the recent crisis of Salmonella Saintpaul. Therefore, the primary purpose of
this study was to explore how the salmonella outbreak was managed and communicated by U.S.
government agencies. Another aim of this study was to identify the significant role that the
media played during the crisis. This study focused on the role of the media as a mediator
between the government and its publics, but it also explored how the formation of public opinion
arose along the stages of the crisis life-cycle and how it affected food-safety policies and
systems. The findings of this study provide important implications for public relations scholars
and practitioners seeking to attain a better understanding of crisis dynamics that possibly lead
public debates and social consensus on critical issues and to develop a strategic communication
approach to mitigating a health-related crisis and resolving related issues. This study also
contributes to advancing a theoretical framework for the application of the contingency theory to
crisis communication.

Literature Review
Contingency Theory of Strategic Conflict Management
The notion of public relations as a strategic management function has evolved based on
two major public relations theoretical foundations: the excellence theory and the contingency
theory. Grunig and Hunt’s (1984) excellence theory posits that public relations activities can be
classified into four typologies: (a) press agentry/publicity model, (b) public information model,
(c) two-way asymmetric model, and (d) two-way symmetric model. It emphasizes the two-way
symmetric model as normative theory, which guides practitioners regarding how they should
perform public relations to be most effective and ethical. Noting that public relations is too
complex and versatile to be forced into the four models, however, Cancel et al. (1997) proposed

---

the contingency theory of accommodation as an alternative to the normative excellence theory in
public relations. They argue that a continuum model could explain public relations practices
more accurately and better reflect the dynamics of strategic communication. The contingency
theory assumes that strategic communication occurs at any point along a continuum from pure
advocacy to pure accommodation (Cameron, Wilcox, Reber, & Shin, 2008). Advocacy refers to
the degree to which an organization maintains its own standpoint disagreeing with the public’s
viewpoint, while accommodation means that an organization gives in and takes a position in
favor of its publics (Cameron et al., 2008). In other words, the position or stance that an
organization takes in dealing with conflict can be placed on the continuum from pure advocacy
to pure accommodation. Between these two extremes of the continuum, there are other stances
that involve “different degrees of advocacy and accommodation” (Cancel et al., 1997, p. 37).
Capturing the dynamics of conflict, the continuum identifies the stance of an organization toward
a given public at a given time in a given situation, and the organization’s stance, in turn, affects
its strategies and tactics (Cancel et al., 1997). Based on this theoretical perspective, “true
excellence in public relations may result from picking the appropriate point along the continuum
that best fits the current need of the organization and its publics” (Cancel et al., 1997, p. 35).
There are two basic principles underlying the contingency theory (Cameron et al., 2008).
First, an organization’s stance is determined by various factors when dealing with a conflict or
crisis. The second principle is that the stance of public relations changes as events and factors
evolve. The contingency theory provides 87 factors affecting how an organization responds to
conflict, and those contingent factors are categorized into 11 groups on internal and external
dimensions (Cancel et al., 1997). External variables consist of the following five groups: (a)
external threats, (b) industry-specific environment, (c) general political/social environment, (d)
external public characteristics, and (e) the issue under consideration. Internal variables include
(a) general corporate/organizational characteristics, (b) characteristics of the public relations
department, (c) top management characteristics, (d) internal threats, (e) personality
characteristics of involved organization members, and (f) relationship characteristics. Through
quantification of the contingent variables, Shin, Cameron, and Cropp (2006) showed that all of
the variables are well-combined into a simple matrix of these thematic categories and confirmed
the validity of the contingent variables in order to construct a theory for public relations.
To test the practicality of the contingency theory, Cancel, Mitrook, and Cameron (1999)
interviewed public relations practitioners and further classified these contingent variables into
two dimensions: predisposing and situational factors. Predisposing factors, such as the
characteristics of the dominant coalition and organizational size and culture, affect an
organization’s stance before it enters a situation with a given public. Situational factors, such as
perceived urgency and threat and the feasibility of accommodation, may change the stance of an
organization while it interacts with a particular public (Cancel et al., 1999). By employing in-
depth interviews with public relations practitioners, Cameron, Cropp, and Reber (2001) provided
six proscriptive variables that prevent an organization from accommodating its publics. They
found that the practitioners first described their approaches in a way that was consistent with the
two-way symmetrical model. Deeper glimpses into these approaches, however, revealed less or
nonexistent two-way symmetry in their actual practice; rather, it was the proscriptive factors that
applied and combined to directly affect their decisions: (a) moral convictions, (b) multiple
publics, (c) regulatory constraints, (d) management pressure, (e) jurisdiction issues, and (f) legal
constraints.

---

Contingency scholars have shown the application of contingency theory to public
relations practice in diverse fields, such as high-profile conflict resolution (Shin, Cheng, Jin, &
Cameron, 2005), intra- and inter-organizational conflicts (Pang, Cropp, & Cameron, 2006;
Yarbrough, Cameron, Sallot, & McWilliams, 1998), health-related crisis management (Jin et al.,
2006, 2007; Qui & Cameron, 2005), the practitioner-lawyer relationship (Reber, Cropp, &
Cameron, 2001), and the source-reporter relationship (Shin & Cameron, 2003, 2005).
Specifically, Shin et al.’s (2005) study of high profile conflicts supports the dynamics of conflict
management and provides evidence that an organization’s stances and strategies, as well as its
publics’ stances, shift over time along the contingency continuum as situations unfold. In
addition, focusing on dynamics in a health-related crisis, Jin et al. (2006) examined how the
Singapore government strategically managed a crisis of severe acute respiratory syndrome
(SARS) by analyzing its stance and crisis communication strategies toward multiple publics and
influential contingent factors. The crisis management of the Singapore government appeared to
be proactive advocacy; the government took the lead in dealing with the situation and protecting
its citizens from the SARS infection, and thus managed to win public support. Through their
comparison of the Chinese and Singapore governments, Jin et al. (2007) highlight that different
organizations may take different stances and strategies in a given situation according to
influential contingent factors on an organization’s decision. Based on these empirical analyses of
successful crisis management cases, the contingency theory has been elaborated and advanced to
offer useful insights into strategic conflict management and provide theoretical ground for
analyzing crisis-response strategies.
It is important for an organization to develop effective crisis-response strategies that may
reframe the public’s general comprehension of negative issues (Benoit & Pang, 2007) and
generate supportive behaviors as well as collective emotions among various publics (Coombs,
1999). From a contingency theory perspective, an organization’s crisis communication strategies
may be determined by its stance toward a given public (Cameron et al., 2007, 2008) and can also
be described in terms of the contingency continuum (Jin et al., 2006). As an elaboration of his
typology of crisis communication strategies, Coombs (1998) proposed the accommodation-
defensive continuum by integrating various crisis communication strategies. This continuum
includes seven categories: attacking the accuser, denial, excuse, justification, ingratiation,
corrective action, and full apology and mortification. Jin et al. (2006), however, modified
Coombs’ continuum in accord with the contingency framework by adding the strategy of
cooperation and reordering strategies. The modified continuum includes the following strategies
(in order from advocacy to accommodation): (a) attack the accuser (aggressively defending itself
against an accuser), (b) denial (asserting that there was no crisis), (c) excuse (avoiding or
minimizing its responsibility for the crisis by denying any intention to cause the crisis), (d)
justification (explaining why it had to take a certain course of action), (e) corrective action
(fixing the problem and promising to prevent its recurrence by changing its initial positions or
actions), (f) ingratiation (taking action to generate a more favorable public attitude toward the
government), (g) cooperation (making overtures to reach out to the public with the goal of
resolving the problem), and (h) full apology (taking full responsibility, making apologies, and
asking forgiveness for its wrongdoing) (Coombs, 1999).
In accord with the essence of the contingency theory, this study sought to examine the
U.S. government’s management of the salmonella outbreak and communication efforts, focusing
on the complexities and dynamics of strategic management. In this study, the U.S. government
refers to two federal agencies—the CDC and FDA—that were responsible for dealing with this

---

food-poisoning crisis. As the contingency theory notes than an organization may take different
positions toward different publics in a given situation (Yarbrough et al., 1998), the U.S
government strategically dealt with various publics: consumers, the tomato industry, the pepper
industry, the food industry, state health departments, and the Mexican government. Based on the
framework contingency theory, the following research questions were proposed:
RQ1a: What kinds of stances and strategies were used by the U.S. government and its
publics during the various stages of the crisis life-cycle?
RQ1b: What contingent factors appeared to influence the U.S. government’s stances and
strategies toward its multiple publics?

The Role of Media in Crisis
When a crisis breaks out on a large scale, the government should communicate important,
up-to-date information with its publics in a timely manner due to the rapid evolution of the
situation and the uncertainty about its impact (Jin et al., 2006). Thus, the media may play a
significant role in informing the public of what has happened and guiding them to avoid potential
risks in times of crisis. Particularly in outbreaks of serious diseases, such as SARS, the
government can strategically deal with its multiple publics in cooperation with the media and
resolve critical situations without serious damage to its image (Jin et al., 2006, 2007; Qui &
Cameron, 2005). In doing so, the media can help draw public attention to the reform of related
health policies and facilitate a social consensus on setting new regulations and enforcing existing
laws (Dorfman, 2007; Gamson & Modigliani, 1989; Jernigan & Wright, 1996).
As a result of recent health-related crises (e.g., spinach contaminated with E. coli and
outbreaks of salmonella poisoning), the public’s awareness of food-safety issues and food
tracking systems has been considerably increased. Since the media have reported problems with
the fresh food distribution system that might have worsened the crises, many public groups (e.g.,
consumer unions and restaurants) have called for the development of new food safety plans
along with reform of current regulations (Weise, 2008). It is crucial for the government to be
aware of public opinion and attentive to its voice, especially when latent public opinion and
sentiments trigger public events such as protests and vigils (Heath, 1997; Sturges, 1994). In
responding to public opinion and action, the government may search for solutions and encourage
sponsor organizations to take the initiative to enhance problematic situations (Heath, 1997).
According to Sturges (1994), the process of group opinion formation can be described by the
following series: (a) latent issues emerge
(b) an event occurs
(c) pro and con factions form
(d) debates occur
(e) time lapses
(f) public opinion forms
(g) social actions take
place
(h) social norms form. As a facilitator, the media contributes to this process of
generating the collective opinion among publics (Page & Shapiro, 1987). Since the public may
tend to pay more attention to the media after a crisis breaks out, the media can serve as a useful
tool to trigger public policy initiatives. Jernigan and Wright (1996) note that the media can be an
effective instrument for educating the public and policy makers and garnering public support for
policies to promote a healthier society.
Although issues presented in the media cause the generation of public opinion on those
issues, media discourse and public opinion can interact with each other as parallel systems
(Gamson & Modigliani, 1989). In other words, both media discourse and public discourse may
be parts of the process of constructing meaning in a large context. Gamson and Modigliani
(1989) specify that “media discourse is part of the process by which individuals construct

---

meaning, and public opinion is part of the process by which journalists and other cultural
entrepreneurs develop and crystallize meaning in public discourse” (p. 2). While trying to
understand emerging social issues that touch their own lives, people may also rely on media
messages to help them make sense of those issues and construct underlying meanings
(Viswanath & Demers, 1999). Regarding some social issues, such as policy reforms and disease
prevention, however, the media may take a larger part than the public counterpart could in the
social construction of meaning and the creation of initiatives to resolve social problems. That is,
the potential of the media in influencing public opinion may vary from issue to issue (Ball-
Rokeach & DeFleur, 1982). When it comes to health-related crises, publics may be more
dependent on media accounts because critical issues pertaining to the situation (e.g., the
treatment, prevention or possible causes of illnesses; problems in the current health system) are
too volatile or complicated for them to easily understand. Therefore, media discourse during a
crisis and in its aftermath may be the key to understanding public opinion, and it at least
contributes to public discussion that can lead to social change (Viswanath & Demers, 1999).
As the major information channels in the public sphere, mass media are considered
essential for understanding the formation of public opinion and the emergence of social
consensus on important issues (Habermas, 1991). In the salmonella outbreak, while the media
were promoting the flow of relevant information about food contamination occurrences, they
were also drawing public attention to the reform of related health policies. The media may also
have facilitated the formation of public opinion and social consensus on the need for new
regulations and enforcement of existing laws to enhance food traceability and public safety.
Further, social consensus ultimately leads to social change with moderate adjustments in the
current social system (Viswanath & Demers, 1999). In terms of Viswanath and Demers’ (1999)
typology of relationship between social control and social change, many health-related
movements fall into the category of moderated change because a gradual change is made to some
aspect of the system while the dominant values remain the same. A content analysis of news
articles about a health-related crisis might reveal key issues in media messages as well as
emerging discussion on policy reform or other changes among lawmakers in the opinion
formation process (Malone, Boyd, & Bero, 2000). In this sense, this study focuses on the role of
the media in encouraging the formation of public opinion as well as in dispensing important,
accurate information speedily in the outbreak of salmonella. Therefore, the second research
question asks the following:
RQ2: What role did the media play in the salmonella outbreak with respect to dispensing
information and covering issues that possibly influence public opinion?

Methods
Study Design
In order to examine the proposed research questions, this study employed a quantitative
content analysis of major U.S. newspapers’ crisis coverage regarding salmonella-tainted
tomatoes. Media may reflect dynamics of crisis situations and thus serve as a useful tool in
examining an organization’s stances and strategies, as well as publics’ reactions in times of crisis
(Martinelli & Briggs, 1998). Three newspapers (i.e., The New York Times, USA Today, and The
Washington Post) were selected based on the circulation size and availability.

Data Collection

---

News articles were downloaded from Lexis-Nexis News Database. A key word search
using the word “salmonella” in news publication during the six months from April 1 to
September 30, 2008, generated 62 articles in The Washington Post, 36 articles in The New York
Times, and 27 articles in USA Today. Because the first salmonella occurrence was reported to the
state health department on May 11 and the CDC announced the end of the outbreak on August
28, the timeframe of six months (from one month before the report until one month after the end
of the outbreak) was reasonable for comprehensively examining the evolution of the crisis. Some
of the articles retrieved, however, mentioned salmonella in a peripheral way (e.g., discussion on
food irradiation) or addressed other salmonella cases; these articles were excluded from the
analysis. Both editorials and feature news, including news briefings, were also included. The
deletion process resulted in a total of 72 articles.

Coding Procedure and Inter-coder Reliability

Two graduate students were trained to code the news articles. After training sessions, two
coders separately coded 15 percent of the sample for an inter-coder reliability test. Scott’s pi
scores for each variable ranged from .81 to 1.0, indicating that the agreement between the coders
was acceptable (Wimmer & Dominick, 2006). Two coders then coded the rest of the news
articles independently.

Coding Categories
The unit of analysis was an individual news story, and the codebook was developed
based on Jin et al.’s (2006) research. The coding categories were composed of four sections: (1)
general publication information (e.g., newspaper source, date, section, length, phase, etc.); (2)
stance changes of the government and involved publics, as well as and contingent factors; (3)
crisis response strategies of the government; and (4) sources cited and information addressed by
the government and other sources.
Publication information. The date of publication, the page where an article appeared, and
the length of the article (in number of words) were coded from the information provided by the
Lexis-Nexis Database.
Phase. To examine the evolution of the crisis, four phases were identified: (a) tomato-
warning phase, (b) jalapeno-warning phase, (c) matching-strain phase, and (d) post-crisis phase.
Although the time frame of this study includes the date the first salmonella case was reported to
the Mexico State Health Department (May 11), newspapers first reported the salmonella
outbreak on June 8, when FDA announced a national consumer warning not to eat certain types
of red raw tomatoes (June 7). Therefore, the actual time span for the analysis began with the first
news report on the outbreak (June 8), with no news articles found prior to the crisis.
The phase was identified by examining the critical events during the life-cycle of the
crisis. The tomato-warning phase (coded as 1) included publications from June 8 (the first news
report regarding the salmonella outbreak) to June 26. News articles in this phase presented the
FDA’s national consumer warning and its initial investigative efforts. The jalapeno-warning
phase (coded as 2) included publications from June 27 to July 20, when the FDA expanded its
investigation into other sources of contamination, such as peppers, and then initiated its public
warning on jalapeno and Serrano peppers. As the FDA found the matching bacterial strain in
Texas on July 21, the crisis moved to the matching-strain phase (from July 21 to August 27;
coded as 3). Finally, the post-crisis phase was defined as the aftermath of the CDC’s
announcement of the end of the outbreak on August 28 (coded as 4). The four phases divided the

---

time frame in a relatively balanced way: tomato-warning phase (3 weeks), jalapeno-warning
phase (3 weeks), matching-strain phase (5 weeks), and post-crisis phase (4 weeks).
Conference call follow-up. To examine the relationship between the government and
media in the crisis, it was coded whether a news article was written based on the conference call
or news releases. The dates of news releases and conference calls were identified from the FDA
website. If a news article was published within one day of the government’s conference call or
news release, the article was coded as 1; otherwise, it was coded as 0.
Types of publics. Seven major publics involved in the crisis were identified: (a)
government (i.e., FDA and CDC), (b) state health departments, (c) consumers/consumer
organizations, (d) tomato industry, (e) pepper industry, (f) food industry (i.e., food retailers,
restaurants), (g) Mexican government, and (h) others. Each article was coded if one or more
types of publics were presented in addressing the crisis (1=presented; 2=non-presented).
Stance. The stance variable measured how each group of publics took a stance toward the
other groups of publics on the advocacy-accommodation continuum. Since this study focused on
the government’s role in the crisis, this study mainly measured (1) what stance the government
took toward the other seven types of publics and (2) what stance the other groups of publics took
toward the government. The stance was measured on a 5-point scale, ranging from 1 (pure
advocacy) to 5 (pure accommodation), based on Cameron et al.’s (2008) operational stances on
the contingency continuum. For example, a stance reflecting arguing or avoiding was coded as 2,
a stance reflecting comprising or negotiation as 3, and a stance reflecting collaboration or
cooperation as 4.
Contingent factors. Since news coverage of crisis situations may not reflect internal
contingent factors, such as organizational characteristics and culture, only external factors that
may have affected the government’s stances or strategies were coded (1=presented; 2=not
presented) under six sub-categories. These factors included (a) threats, such as a rapid
dissemination of illness or possible reputational damage to the government; (b) industry
environment, such as recent changes in the produce and food industries or new safety measures
recently implemented in restaurants; (c) general political/social environment/cultural
environment, such as consumers’ support or opposition to the government’s current policy or
new proposals in Congress; (d) external public, such as consumer organizations’ requests; (e)
issue under question, such as arguments about the possible sources of the outbreak; and (f)
others.
Crisis-response strategy. The government’s crisis response strategies were coded for the
seven groups of publics, according to Coombs’ (1998) crisis-response strategy continuum.
However, adopting Jin et al.’s (2006) modification of Coombs’ continuum, the strategies were
measured in terms of the following eight categories: (1) attack the accuser (e.g., FDA and CDC
accusing the tomato industry of not having the food tracking system while defending itself
against tomato producers’ criticism), (2) denial (e.g., asserting that their delayed investigation
was not due to their negligence), (3) excuse (e.g., avoiding or minimizing its responsibility for
the delayed investigation to track the source of the outbreak), (4) justification (e.g., explaining
why it took a long time to pinpoint the cause of the crisis), (5) corrective action (e.g., correcting
the source of the problem and promising to prevent its recurrence), (6) ingratiation (e.g., stating
that its state-of-the-art technology was used to find the source of the outbreak or announcing
possible financial support to produce industries for profit loss), (7) cooperation (e.g., working
with state health departments to resolve the situation), and (8) full apology (e.g., taking full

---

responsibility for the rapid spread of illness or making apologies to consumers or tomato
growers).
Source. The sources cited in the news articles were coded (1=cited; 2=not cited) under
seven sub-categories: (a) federal agencies (e.g., representatives of the FDA and CDC), (b) state
health departments (e.g., Indian Health Service), (c) consumers/consumer organizations (e.g.,
individual consumers, Center for Science in the Public Interest, Consumer Union), (d) produce
industry (e.g., individual farmers, Produce Marketing Association, California Tomato Farmers),
(e) food industry (e.g., National Restaurant Association, restaurant owners/spokespersons), (f)
food safety experts (e.g., professors or researchers), and (g) others. A direct quotation with
quotation marks and a citing phrase, such as “according to,” were examined to point to the news
sources. Additionally, verbs indicating one’s verbal statement were used to identify sources in
the news articles: said, noted, reported, criticized, announced, told, recommended, advised,
reiterated, asserted, warned, acknowledged, and declared. Two or more sources cited in the same
category were also coded as 1.
Problem and solution. These variables measured whether or not each group of sources
(i.e., source variable) addressed (1) the problems regarding the existing food safety system,
policy, or regulation and (2) the solutions to resolve the problems embedded in the existing food
safety system, policy, or regulation (1=addressed; 2=not addressed).
Information provided by the government. This variable measured what kind of
information was presented in the news article being provided by the government sources. A news
article was coded as to whether information in each sub-category was presented (coded as 1) or
not (coded as 2). Sub-categories included (a) general information about salmonella (e.g., type of
disease, symptoms of disease), (b) potential causes of the outbreak (e.g., source of
contamination, region of the outbreak’s origin), (c) figures and statistics (e.g., the number of
illnesses; spread of illness), (d) investigation updates (e.g., CDC’s efforts, discovery of the
outbreak strain), (e) FDA alerts and recommendations, (f) impact on the industry (e.g., financial
loss), (g) previous outbreaks (e.g., spinach with E. coli in 2006), (h) problems (e.g., problems
with food tracking system), (i) solutions (e.g., new traceability standards, electronic records,
advanced produce distribution system), and (j) others.

Findings


Table 1. Number of articles analyzed by newspaper and phase
Phase
Tomato
Jalapeno
Matching
Post crisis
Total
Newspapers
warning
warning
strain
New York Times
10 (52.6)
4 (21.1)
4 (21.1)
1 (5.3)
19 (100)
USA Today
3 (18.8)
7 (43.8)
5 (31.2)
1 (6.2)
16 (100)
Washington Post
15 (40.5)
11 (29.7)
9 (24.3)
2 (5.4)
37 (100)
Total
28 (38.9)
22 (30.6)
18 (25.0)
4 (5.6)
72 (100)

Of a total of 72 news articles analyzed, 19 articles (26.4%) were from The New York
Times, 16 articles (22.2%) were from USA Today, and 37 articles (51.4%) were from The
Washington Post. Regarding the four phases, news articles were generated most during the

---

tomato-warning phase (38.9% of the entire publications), followed by the jalapeno-warning
phase (30.6%), the matching-strain phase (25.0%), and the post-crisis phase (5.6%) (Table 1).

The first research question regarded the stances and strategies employed by the U.S.
government and its publics over the four stages of the salmonella outbreak life-cycle. Over the
six-month period of the outbreak, federal agencies appeared to be involved in all of the articles
analyzed (n=72, 100%). Consumers (n=57, 79.2%) and the tomato industry (n=38, 52.8%) were
frequently addressed as being involved in the crisis, followed by the pepper industry (n=27,
37.5%), the food industry (n=27, 37.5%), state health department (n=14, 19.4%), and the
Mexican government (n=7, 9.7%). Chi-square analysis revealed differences of involvement of
certain types of publics (i.e., state health department and food or produce industry) over the
identified four stages, while federal government agencies and consumers were addressed as
being involved in the crisis regardless of the phase (Table 2). Specifically, state health
departments and the food industry were mostly discussed in the tomato-warning phase, and
mention of them seemed to disappear over the next phases. On the other hand, the engagement of
the tomato and pepper industries tended to increase over the phases.

Table 2. Type of publics involved
Phase
Tomato
Jalapeno
Matching
Post
χ2
Publics
warning
warning
strain
crisis
Total
p-value
Government
28 (100)
22 (100)
18 (100)
4 (100)
72 (100)
NA
State health dept.
10 (35.7)
2 (9.1)
2 (11.1)
0 (0)
14 (19.4)
.046*
Consumers
23 (82.1)
18 (81.8)
13 (72.2) 3 (75.0)
57 (79.2)
.846
Tomato industry
8 (28.6)
14 (63.6)
14 (77.8) 2 (50.0)
38 (52.8)
.007*
Pepper industry
0 (0)
9 (40.9)
14 (77.8)
4 (100)
27 (37.5)
.000*
Food industry
16 (57.1)
6 (27.3)
5 (27.8)
0 (0)
27 (37.5)
.033*
Mexican gov’t
1 (3.6)
2 (9.1)
4 (22.2)
0 (0)
7 (9.7)
.183
Others
8 (28.6)
6 (27.3)
6 (33.3)
1 (25.0)
21 (29.2)
.973
N
28 22
18
4 72
Note. * significant at p < .05

On a five-point scale, ranging from 1 (pure advocacy) to 5 (pure accommodation), the
government mostly demonstrated advocacy stances toward consumers (M=2.23, SD=.890), the
tomato industry (M=2.81, SD=1.009), and the pepper industry (M=2.33, SD=.734), while it
appeared to be accommodative to state health departments (M=3.85, SD=.555). The ANOVA
test resulted in statistically significant changes of government stances toward consumers, the
tomato industry, and the pepper industry, in an accommodative direction as the crisis approached
the endpoint (Table 3).

---