Show Summary Details

Page of

Printed from Oxford Research Encyclopedias, Communication. Under the terms of the licence agreement, an individual user may print out a single article for personal use (for details see Privacy Policy and Legal Notice).

date: 19 April 2021

Responsibility in Health and Risk Messagingfree

  • Amber K. WorthingtonAmber K. WorthingtonDepartment of Communication Arts and Sciences, The College of the Liberal Arts, The Pennsylvania State University


Health risk messages may appeal to the responsibility of individuals or members of interdependent dyads for their own or others’ health using many different message strategies. Health messages may also emphasize society’s responsibility for population health outcomes in order to raise support for health policy changes, and these, too, take many different forms. Message designers are inherently interested in whether these appeals to personal, interdependent, and societal responsibility are persuasive. The central question of interest is therefore whether perceptions of responsibility that result from these messages lead to the desired message outcomes. A growing body of empirical research does suggest that there is a direct persuasive effect of perceptions of personal responsibility and interdependent responsibility on health intentions or behaviors, as well as indirect persuasive effects of responsibility on intentions or behaviors via anticipated emotions, specifically regret, guilt, and pride. Research also suggests that perceptions of societal responsibility increase support for public health policy (i.e., the desired message outcome in societal responsibility messages). Important to this area of research is a conceptual definition of responsibility that lends itself toward identifying specific message features that elicit perceptions of responsibility. Specifically, attributions of causation and solution, obligation, and agency are identified as effect-independent message features of responsibility.

Health risk messages communicate information about a potential health threat. These messages oftentimes include appeals to a message recipient’s personal responsibility for their own health or interdependent responsibility for others’ health. Additionally, societal responsibility messages may attempt to persuade readers of society’s responsibility for a particular health outcome to increase public support for health policy changes. In each of these cases, responsibility messages use many different message strategies and variations that fall under the umbrella of “responsibility messages.” This article first provides a conceptual definition of responsibility that is useful to message designers by narrowing responsibility message strategies to specific effect-independent message variables, including attributions of causation and solution, obligation, and agency. Second, this article reviews this vast array of personal, interdependent, and societal responsibility messages and examines the persuasiveness of perceptions of personal, interdependent, and societal responsibility on intended affective and behavioral outcomes.

Responsibility Definition

There are numerous conceptual definitions of responsibility and classification systems that distinguish responsibility from other similar constructs. For example, Hart (1968) provided the following classification: “(1) Role-Responsibility: Whenever a person occupies a distinctive place or office in a social organization, to which specific duties are attached to provide for the welfare of others or to advance in some specific way the aims or purposes of the organization, he is properly said to be responsible for the performance of these duties; (2) Causal-Responsibility: In many contexts … it is possible to substitute for the expression “was responsible for” with the words “caused” or “produced” or some other causal expression; (3) Liability-Responsibility: When legal rules require men to act or abstain from action, one who breaks the law is usually liable, according to other legal rules, to punishment for his misdeeds; and (4) Capacity-Responsibility: The expression “he is responsible for his actions” is used to assert that a person has certain normal capacities … The capacities in question are those of understanding, reasoning, and control of conduct” (pp. 213–230).

Bovens (1998) built upon this previous classification (a classification also seen in Bemelmans-Videc, 2007, p. 22) with the following: “(1) Responsibility as cause. “Responsible for” can in a number of contexts be replaced by “caused” or “to have as a consequence” or some such expression that indicates a causal connection; (2) Responsibility as accountability. Often “being responsible” is used in the sense of political, moral, or legal liability (or in all or some of these senses) for the results … of a given form of behavior or event; (3) Responsibility as capacity. To be responsible in the sense of “accountable,” one must in most cases also have been in a position to exercise a certain amount of responsibility; (4) Responsibility as task. When someone fulfills a given social role, holds an office, or is allotted a task or function in an organization from which competences flow and that entail certain duties in regard to others in regard to the organization, then we mostly speak of his or her competencies and duties as “responsibilities”; and (5) Responsibility as virtue. The concept refers to a virtue [when] it suggests that someone takes his tasks and duties seriously, acts only after due deliberation, and considers himself answerable to others for the consequences of his actions” (pp. 24–26).

Further, Bemelmans-Videc (2007) distinguishes responsibility from other similar concepts (e.g., authority, accountability) by noting, “(1) Authority is the right to act; (delegated) authority presupposes the allocation of commensurate responsibility; (2) Responsibility is the obligation to (properly) perform delegated duties and tasks; and (3) Accountability is the obligation to present an account of and answer for the execution of responsibilities to those who entrusted those responsibilities” (p. 23).

To further define responsibility, it is necessary to differentiate responsibility from other, similar concepts, including blame. To elucidate the relationship between responsibility and blame, it is helpful to further distinguish responsibility as prospective and retrospective. Feelings of responsibility can result either from foreseeing future events that one may cause (or solve) or from looking back over past behavior or events that one caused (or solved). This distinction has also been labeled prospective and retrospective reflectivity (Cummings & Anton, 1990), “before-the-fact consciousness” versus “after-the-fact accountability” (Culbert, 1974) and “assumed responsibility” versus “assigned responsibility” (Graham, 1986). Prospective responsibility is concerned with messages that assign the recipient future responsibility to take care of their own or another’s health (i.e., personal or interdependent responsibility). Research on retrospective responsibility, however, is concerned with the attribution process wherein individuals attribute blame for past actions or outcomes to themselves, others, or some other entity for something that has already occurred (e.g., Fuller, Marler, & Hester, 2006). Indeed, Janoff-Bulman (1979) noted that internal attributions for negative outcomes that have already occurred can include blame directed at one’s character (e.g., ability) and blame directed at one’s behavior (e.g., effort). For example, Beverly and colleagues (2012) found that patients with Type 2 diabetes assumed retrospective responsibility for their unmet treatment goals and expressed a sense of personal failure and blamed themselves for not adhering to recommendations. Accordingly, considerations of blame in responsibility messages focus on this retrospective assignment of responsibility. Responsibility messages should therefore take care to ascribe prospective, not retrospective, responsibility for health behaviors and outcomes in order to avoid increasing perceptions of blame, as perceptions of blame may fail to elicit the desired response. Indeed, self-blame in patients with chronic conditions, for example, is associated with poor psychological adjustment, depression, poor coping, and difficulty managing self-care (Beverly et al., 2012; Voth & Sirois, 2009).

The above-mentioned definitions and classification systems are useful for broadly understanding responsibility; however, message designers are interested in specific aspects of responsibility that may translate into message content cues to be used in constructing health risk messages. Indeed, message design scholars should avoid using effect-based definitions of message variables (i.e., defining a responsibility message as one that elicits perceptions of responsibility) and should instead characterize message variations of responsibility in terms of effect-independent message features (see O’Keefe, 2003). For example, Guttman and Ressler (2001) specify three facets of responsibility, including attributions of causation, obligation, and agency. These aspects of responsibility are also prevalent in the above classification schemes: causation (i.e., causal responsibility, Hart, 1968; responsibility as cause, Bovens, 1998), obligation (i.e., role-responsibility, Hart, 1968; responsibility as task, Bovens, 1998; responsibility as obligation, Bemelmans-Videc, 2007), and agency (capacity responsibility, Hart, 1968; responsibility as capacity, Bovens, 1998). Defining responsibility based on these specific components necessitates further examination of each.

Attributions of Causation

Establishing a causal chain of responsibility for a health outcome examines the locus of the cause (as internal or external). Attribution theories are concerned with antecedents to attributions, which refer to the perception or inference of a cause, including factors that lead an individual to attribute a particular event or outcome to one cause rather than another (Kelley & Michela, 1980). As mentioned previously, attribution theory (Heider, 1958) posits that individuals strive to predict their environment, and doing so is achieved by understanding the causes of behavior. The causes of our own or others’ behaviors are attributed to the individual person (internal/dispositional) or to the environment (external/situational) in order to make sense of the world.

These attributions of causation are a well-recognized component of responsibility. For example, Hart (1968) posits that in many contexts, the phrase “was responsible for” can be substituted with the word “caused,” and Bovens (1998) notes “there has to be a certain causal connection between the conduct of the person held to account and the damage done” (p. 29). Likewise, Guttman and Ressler (2001) note that responsibility “inherently assumes causal connections between people’s deeds and health outcomes” (p. 119), and Weiner (2006) notes “responsibility for an outcome implies that one has caused that outcome” (p. 32).

Establishing a causal chain of responsibility for health may also examine the locus of the solution (as internal or external). Indeed, the Theory of Perceived Responsibility and Social Motivation (Weiner, 1993, 2006) distinguishes between causal attributions and solution attributions. Causal attributions are beliefs about who is responsible for causing a particular behavior or outcome (and align with attribution theory), and solution attributions are beliefs about who is responsible for addressing or solving problematic behaviors or ill health outcomes. For example, attributions may be made about a parent’s interdependent responsibility for causing obesity-related behaviors (i.e., the behaviors; e.g., sedentariness or lack of physical activity) or obesity (i.e., the outcome) in their child; separate attributions may be made about a parent’s interdependent responsibility for solving obesity-related behaviors (e.g., promoting physical activity) or obesity outcomes in their child.

Causal and solution attributions are distinct; however, “a close connection exists between claims about patient responsibility for healing a disease and responsibility for its cause. Hence, keeping these responsibilities distinct is difficult” (Kirkwood & Brown, 1995, p. 67). This underscores the importance of examining attributions of causation and solution as separate message features, as attributions of causation may be more important in messages designed to prevent ill health (i.e., for prevention) and attributions of solution may be more important in messages designed to reverse ill health (i.e., for treatment).

Further considerations for message designers include message recipients’ preconceptions about attributions, multiple causal attributions, and the stability of the cause. For example, Kelley and Michela (1980) note that an individual’s perception of new information (i.e., new message content cues) regarding an attribution is “greatly affected by subjects’ preconceptions about cause-effect relations, even being rendered wholly erroneous” (p. 462). This underscores the importance of using, for example, message features that adhere to a two-sided refutational frame. Further, Kelley (1972a, 1972b) makes note of the discounting principle, in which an individual gives less weight to one cause if other plausible causes are also present, and the importance of a causal schema, in which an individual believes two or more causal factors interact to produce a particular event or outcome. Finally, Weiner and colleagues (1971) argued that causes possess an additional property, labeled causal stability. Causal stability refers to whether a cause will continue in the future, in which case a prior effect of the same cause will be anticipated to recur; if the cause changes, then so too might the outcome (see also Weiner, 2010; Weiner, Nierenberg, & Goldstein, 1976). Message design scholars should take care in future work to distinguish message variables that reflect these considerations in appeals to responsibility at the personal, interdependent, and societal levels.


Obligation underscores the responsibility that one has to themselves or to others. Bovens (1998) notes that if someone is to be held responsible for another’s actions, they must have an accepted relationship with the agent: “[without] a particular and close relationship with the agent … the person held to account could not be held responsible” (p. 31). Further, Bemelmans-Videc (2007) notes “responsibility presupposes having the capacity to exercise the tasks (authority), which implies that these duties or tasks have been formally delegated to the person in charge” (p. 23), and Agich (1982) notes that obligation as a component of responsibility is a moral charge to help another person.

Guttman and Ressler (2001) further distinguish three forms of obligation: “(1) The obligation to take care of one’s own health for one’s own sake, for the sake of significant others, and for society as a whole, that is by being a productive member; (2) The obligation to promote the health of others; and (3) The obligation to avoid becoming an unfair burden to others and society as a whole by engaging in risky behaviors that may turn oneself into a dependent” (p. 126). Obligation therefore pertains to “role acceptance” (Cummings & Anton, 1990) and “role responsibility” (Gibson & Schroeder, 2003; Hart, 1968). Indeed, accepting a role as, for example, a parent caring for a child, requires an individual to accept the obligation to perform the “specific, prescribed, often stereotypical, and impersonal expectations” associated with that role (Cummings & Anton, 1990, p. 265).

Obligation content cues in responsibility messages should examine specific message features that might elicit obligation (e.g., specific words that connote obligation including the nouns “duty,” “commitment,” and “obligation,” as well as “must,” “bound,” “committed,” “indebted,” and “obligated”). In addition, future work should examine these obligation message features as a function of the intimacy of the relationship, as people report increased social and personal obligations to help in intimate relationships (Roloff, Janiszewski, McGrath, Burns, & Manrai, 1988; Schwartz, 1977), and distinct message variables, including apologies, explanations, inducements, and contingencies, may be important to induce perceptions of obligation in nonintimate relationships (Roloff et al., 1988; see also Roloff & Janiszewski, 1989). Finally, future work should examine how obligation content cues may elicit not only perceptions of responsibility, but also reactance, as obligation may be an uncomfortable state that is made aversive by a restriction on behavioral autonomy and an anticipation of negative social repercussions if one does not comply (Greenberg & Bar-Tal, 1976; Greenberg & Shapiro, 1971).


Agency, which highlights that the message recipient has the knowledge, choice, ability, and control to act on the recommended health behaviors, is also a well-recognized component of responsibility. For example, Guttman and Ressler (2001) state, “a person should be held liable for adverse outcomes of activities … only when these are under their complete volition” (p. 120). Further, Adler and Stewart (2009) note “individuals are responsible for engaging in health-promoting behaviors but should be held accountable only when they have adequate resources to do so” (p. 50). Complete volition and resources include (1) knowledge, (2) choice, (3) ability, and (4) control required to enact the recommended physical activity practices to prevent or reverse obesity in their child or aging parent.


Knowledge highlights a message recipient’s awareness of what they are doing with reference to the advocated health behavior, as well as awareness of the potential ramifications of doing so. For example, Bovens (1998) notes. “the person held to account must at the moment of engaging in the form of behavior under consideration have been in adequate possession of his mental faculties” (p. 30), and Guttman and Ressler (2001) state that individuals must be properly educated about the potential outcomes of their actions or inactions to be considered responsible for their behaviors and outcomes. Additionally, a series of scholars have suggested that foreseeability of consequences is important for responsibility (e.g., Brehm & Jones, 1970; Cooper, 1971; Sogin & Pallak, 1976). Further, the norm activation model (NAM; Schwartz, 1977) posits that awareness of the consequences of particular behaviors influences perceptions of responsibility, and empirical examinations support this proposition (e.g., De Groot & Steg, 2009; Onwezen, Antonides, & Bartels, 2013).


Choice highlights the message recipient’s resources to choose alternative courses of health actions. For example, Weiner (2006) notes that responsibility requires that the causal agent have freedom of choice. Additionally, Bovens (1998) notes that “to be responsible, one must have had a real possibility of acting otherwise than one actually did … [and must have] alternative forms of conduct available” (p. 30), and Bemelmens-Videc (2007) argues that responsibility “presupposes freedom of decision or the ability to make choices” (p. 23). Additionally, Wikler (2002) notes, “ascriptions of responsibility presuppose freedom of action” (p. 50). In one specific example, Zeelenberg, van Dijk, and Manstead (1998) found that participants perceived students to be more responsible when an outcome stemmed from the student’s own choice than when it stemmed from computer reassignment.

Ability and Control

Ability and control are closely related. Indeed, these concepts are often studied together as self-efficacy, which is oftentimes used synonymously with perceived behavioral control. Self-efficacy refers to “beliefs in one’s capabilities to organize and execute the courses of action required to manage prospective situations” (Bandura, 1995, p. 2). Ability refers to the skills required for the message recipient to enact the recommended health behaviors. Indeed, Gustafson and Laney (1968) note that responsibility presupposes the ability and capability to respond. Control refers to a message recipient’s control over a health behavior such that is volitionally alterable. Weiner (1979, 1985) argued that controllability should be included in attribution theory as a separate dimension. As mentioned, internal and external factors refer to the perceived causes of a particular behavior; controllability refers to behaviors that are voluntarily produced (Weiner, 1979, 1985). Internal, individual attributions for outcomes can be viewed as controllable or uncontrollable (Weiner, 1979, 1985). For example, a child’s sedentariness may be believed to be a function of the parent’s laziness or lack of parental control (i.e., controllable) or the child’s sedentariness may be believed to be a function of environmental constraints (i.e., uncontrollable). Likewise, external attributions can be viewed as controllable or uncontrollable (e.g., the physical activity of the child may be considered controllable if the parent has the resources to allow him or her to be physically active, including safe play areas, or uncontrollable if they do not). Message designers should therefore consider specific content cues for knowledge, choice, ability, and control in constructing responsibility messages.

Additional Considerations for Responsibility Message Features

Researchers should also consider message features for responsibility appeals that are used more broadly in message design research. For example, examining message features in guilt appeals appears promising (see O’Keefe, 2002), as guilt is closely related to responsibility (e.g., Berndsen & Manstead, 2007; Kugler & Jones, 1992; Miceli, 1992). For example, Miceli (1992) notes that in order to feel guilt, an individual must feel responsible and know that their action (or inaction) caused harm (for additional information, see Monique Turner’s article on guilt appeals, also in Oxford Research Encyclopedia Communication). Further, feelings (e.g., guilt) are directed by cognitions (e.g., responsibility; see Weiner, 2010), such that responsibility messages may elicit anticipations of emotions such as guilt, which then lead to the desired response (see below). Indeed, research not described as concerned with responsibility messaging may nevertheless be conceptually, and therefore operationally, connected.

Responsibility Messages

Health risk messages take many forms when appealing to the responsibility of individuals or members of interdependent dyads for their own or others’ health. Likewise, messages may emphasize society’s responsibility for population health outcomes in order to raise support for health policy changes, and these, too, appear with a wide array of message strategies and variations. The following section therefore reviews these various personal, interdependent, and societal responsibility messages and provides examples of each.

Personal Responsibility Messages

Appeals to personal responsibility take many forms. Kirkwood and Brown (1995) examined these personal responsibility appeals and proposed a framework of audiences (including the undiagnosed, the diagnosed, and respondents) and message strategies for responsibility communication. Responsibility messages targeted toward those who are at risk for a disease, or the undiagnosed public, attempt to motivate and foster health-promoting behavior that reduces the likelihood or risk of disease by focusing attention on controllable factors (Kirkwood & Brown, 1995). Strategies to do so include these: (1) emphasize people’s power to influence their health, and (2) define health as a question of choice. For example, messages to motivate obesity-prevention behaviors often stress the individual’s power to influence health: “You can counteract most risk factors [for obesity] through diet, physical activity and exercise, and behavior changes” (Mayo Clinic Staff, 2015a). Additionally, the language of choice often appears in obesity-prevention responsibility messages: “Overeating and a lack of exercise both contribute to obesity. But you can change these lifestyle choices” (Johns Hopkins Health Library, 2015, emphasis added).

Responsibility messages targeted toward those who already have a disease, or the diagnosed public, attempt to (1) relieve the sick of feelings of guilt about having a disease, and (2) empower the sick by fostering a feeling of personal control (Kirkwood & Brown, 1995). In order to ease guilt, these messages may invoke the use of a message strategy that discusses responsibility for the cause of the ill health by attributing responsibility of the causes of the disease beyond individual control. To alleviate guilt, responsibility messages to obese individuals may thus stress genetic contributors to disease. For example, Health Central (2010) focuses on the genetic contributors to obesity, which may alleviate feelings of guilt in overweight or obese individuals: “Genetic factors may play some part in 70–80% of obesity cases … [and] may directly contribute to severe obesity in people with family histories.” Additionally, messages oftentimes expand the list of possible causes of obesity, focusing on biological (e.g., age, race) and socioeconomic risk factors instead of lifestyle choices to ease guilt. For example, the CDC (2014) notes, “Obesity affects some groups more than others. Obesity is higher among middle age adults, 40–59 years old (39.5%) than among younger adults, age 20–39 (30.3%) or adults over 60 or above (35.4%) … higher income women are less likely to have obesity than low-income women.”

To empower the sick (i.e., the diagnosed), responsibility messages may foster a feeling of personal control instead of powerlessness. To do so, responsibility messages may (1) directly assign responsibility to the sick for their disease to convince patients that they can influence their health; or (2) assign responsibility for the solution of the disease, not the cause (Kirkwood & Brown, 1995). This latter strategy emphasizes that the sick are not responsible for causing their disease, but they are responsible for how they respond. Kirkwood and Brown (1995) provide the following example: “If you fight for recovery, you may enhance the possibility of that recovery, whether or not you were to blame for the development of the disease.” These responsibility messages attempt to promote an individual’s involvement in their own care and treatment, without implying that they caused their disease, thus empowering the individual.

Responsibility messages targeted toward caretakers of individuals with disease, or respondents (e.g., family members, healthcare workers, close friends), attempt to motivate favorable treatment of these sick individuals by, for example, assigning responsibility to causes beyond the control of the sick (Kirkwood & Brown, 1995). Message strategies to do so include these: (1) identify the sick as innocent victims, or (2) define the behavior as a disease. For example, messages may attempt to motivate favorable treatment of obese individuals by family members or close friends by noting that (some) obese people are innocent victims. Indeed, the Food Research and Action Center (2010) notes, “Due to the additional risk factors associated with poverty, food insecure and low-income people are especially vulnerable to obesity.” Messages may also define obesity-related behaviors as a disease to promote positive treatment of obese individuals by healthcare providers. For example, the Mayo Clinic Staff (2015b) states that consuming excess calories, which is associated with overweight and obesity, has been labeled binge-eating disorder.

Interdependent Responsibility Messages

The above-mentioned framework of responsibility messages offered by Kirkwood and Brown (1995) comprehensively addresses appeals to personal responsibility. However, it does not effectively consider messages that focus on interdependent responsibility. Social ties can “instill a sense of responsibility and concern for others that then lead individuals to engage in behaviors that protect the health of others” (Umberson & Montez, 2010, p. S56), and responsibility appeals can therefore address responsibility of the undiagnosed, diagnosed, and respondents with respect to dyads.

Prevention messages to the undiagnosed attempt to foster not only individual health behaviors, but also attempt to promote interdependent health-promoting behaviors by calling on the responsibility of, for example, parents, spouses, adult children, and employers to encourage health behaviors in children, spouses, aging parents, and employees, respectively. Likewise, messages to the diagnosed public attempt to ease guilt and empower not only sick individuals, but also parents or caretakers who are oftentimes blamed or seen as responsible for another’s health (e.g., their child or spouse, respectively). Further, messages to respondents attempt to motivate favorable treatment not only toward individuals, but also toward interdependent dyads. The following section provides an example of such interdependent responsibility messages by focusing on the parent-child dyad.

Example: Parent-Child Responsibility Messages

Parents are important forces in changing children’s behaviors (e.g., food and activity behaviors; Lindsay, Sussner, Kim, & Gortmaker, 2006), and appeals to the interdependent responsibility of parents to undiagnosed their children oftentimes (1) stress the power of parents to influence their child’s health, and (2) define their child’s health as the parent’s choice in order to foster health-promoting behavior. For example, the CDC (2015) notes, “You can help children learn to be aware of what they eat by developing healthy eating habits [and] looking for ways to make favorite dishes healthier.” The same message also uses language of choice: “You (and your child) can … reach or keep a healthy weight through physical activity and healthy food choices!”

Parents of obese children (i.e., the diagnosed) oftentimes feel guilty and worry about being blamed for their child’s weight (e.g., Turner, Salisbury, & Shield, 2011), especially mothers (Jackson, Mannix, Faga, & McDonald, 2005). These perceptions of guilt likely increase amid messages accusing parents of obese children of child abuse (e.g., Saul, 2014) and public perception polls indicating 87% of adults believe parents are responsible for childhood obesity (Associated Press-NORC, 2013). In order to ease guilt, responsibility messages may thus attempt to assign responsibility to causes beyond the parent’s control, but may also try to empower them to take responsibility for the solution. For example, Carlson (2014, emphasis in original) notes: “If your child is overweight or obese, it is NOT your fault. It is your responsibility to seek the proper help and guidance for your family though.”

Messages about childhood obesity also attempt to motivate favorable treatment of these parent-child dyads by respondents (e.g., by healthcare providers). For example, the authors of a recent article published in The Lancet obligated “clinicians to go beyond mere recommendations to eat less and move more” (Ochner, Tsai, Kushner, & Wadden, 2015). These messages are important, as parents oftentimes feel blamed when seeking help from healthcare providers for their child’s weight (Edmunds, 2005), and healthcare providers oftentimes do not tell parents of overweight children that their child is overweight (Perrin, Skinner, & Steiner, 2012). Responsibility messages may also appeal to other interdependent dyads, including, for example, the caregiver-spouse, adult child-aging parent, and employer-employee dyads.

Societal Responsibility Messages

Recent national survey data suggests that the public believes that individuals are largely responsible for their overall health (Robert & Booske, 2011), obesity (Associated Press-NORC, 2013), and childhood obesity (Wolfson, Gollust, Niederdeppe, & Barry, 2015). Health disparities may therefore be perpetuated by lack of public awareness and support for obesity-related public policies (e.g., CDC, 2011). Thus, in addition to messages appealing to personal or interdependent responsibility for health, recent work in communication examines societal responsibility message strategies that can be used to persuade the public of the role of society in health conditions like obesity to bolster support for public health policies to reduce health disparities.

Attribution theory (Heider, 1958) provides one useful perspective to identify societal responsibility message strategies to raise awareness of health disparities (Niederdeppe, Bu, Borah, Kindig, & Robert, 2008). Attribution theory posits that individuals strive to predict their environment by attributing the cause of others’ observed behaviors to internal of external factors (i.e., by assigning responsibility to personal or societal factors). Weiner’s (1993, 2006) extension, the Theory of Perceived Responsibility and Social Motivation, posits that attributions of causation influence attributions of solution, which in turn shape support for public policy. Drawing on this framework, Niederdeppe and colleagues (2008) identified three message strategies to influence personal or societal responsibility beliefs about causes of and solutions to health conditions like obesity, including message framing, narratives, and visual images.

Message framing is “to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation” (Entman, 1993, p. 52). Message framing can function to define problems, diagnose causes, make moral judgments, suggest solutions (Entman, 1993; see also Iyenger, 1991), and influence social movements (Kolker, 2004). A message “can [therefore] frame a social issue as being caused by internal factors or external factors, which in turn influences how people think about who is responsible for causing [and addressing] a societal problem” (Niederdeppe et al., 2008, p. 489). As such, responsibility messages about obesity may emphasize individual or societal responsibility for health. An example of an individual responsibility frame is as follows:

Research shows that these differences in health are caused by people’s behaviors, since people can prevent illness by taking personal responsibility to make healthy choices. Researchers believe these behavioral choices are the most important influences on whether or not people get sick.

(Gollust & Capella, 2014, p. 510)

An example of a societal responsibility frame is as follows:

Research shows that these differences in health are caused by social factors, such as people’s income, their levels of education, and the environment where they live and work. Researchers believe these social factors are the most important influences on whether or not people get sick.

(Gollust & Capella, 2014, p. 510)

These one-sided causal frames (i.e., individual or societal responsibility), however, may not be effective because the public perceives a role for both individual and societal factors in conditions like obesity. Instead, to increase perceptions of societal responsibility, message designers should consider a message that “(1) acknowledges a role for individual decisions but (2) refutes the idea that individual behavior and medical care alone cause poor health and (3) emphasizes that unemployment, racial discrimination, and poverty shape individual behaviors” (Niederdeppe et al., 2008, p. 492). This is known as a two-sided refutational frame. Gollust and Capella (2014) provide examples of both one- and two-sided messages. The one-sided message includes information only about societal responsibility (as seen above in the societal responsibility frame example). The two-sided refutational message, however, acknowledges the role of personal responsibility, but refutes this explanation for disease in favor of societal responsibility. An example of a two-sided refutational frame is as follows:

Research shows that these differences in health are caused by social factors, such as people’s income, their levels of education, and the environment where they live and work. People can also prevent illness by taking personal responsibility to make healthy choices, but social factors still limit the choices that people can make. Even though the choices people make are important to their health, researchers believe that social factors are the most important influences on whether or not people get sick.

(Gollust & Capella, 2014, p. 510)

Competing frames, or multiple messages presenting divergent perspectives, may also influence public perceptions of societal responsibility (Druckman & Leeper, 2012). Chong and Druckman (2007) have argued that researchers should move away from experimental designs that compare one frame against another and instead test combinations of frames that more closely mimic the messaging environment in the real world by considering how competing frames in the public discourse influence opinion. For example, “anti-policy” messages disseminated by industries that promote products linked to health issues undermine regulatory support by emphasizing strongly held American values (e.g., personal responsibility, threat of big government) that resonate with the public (Barry, Niederdeppe, & Gollust, 2013). These “anti-policy” framed messages then compete with messages distributed by health organizations that attempt to raise support for health policy changes. For example, Niederdeppe, Gollust, and Barry (2014) include two anti-tax messages about sugar-sweetened beverages; the first describes a proposed sugar-sweetened beverage tax as an arbitrary target as obesity is complex and caused by many factors, and the second describes the issue of a tax in terms of government overreach. Niederdeppe and colleagues (2014) also include two pro-tax messages: the first describes sugar-sweetened beverages as the single largest contributor to obesity, and the second describes the tax of sugar-sweetened beverages as protecting children by raising funds for obesity prevention.

In addition to message framing, narratives may also be used in societal responsibility messages. Narrative is defined as “a representation of connected events and characters that has an identifiable structure, is bounded in space and time, and contains implicit or explicit messages about the topic being addressed” (Kreuter et al., 2007, p. 222). Narratives can facilitate message recall and comprehension, reduce counterarguments, and facilitate attitude and behavior change through transportation of readers into the narrative (Hinyard & Kreuter, 2007; Kreuter et al., 2007), with recent meta-analytic data suggesting that narratives are more effective than didactic messages at increasing behavioral intentions (Zebregs, van den Putte, Neijens, & de Graaf, 2015). Additionally, narratives integrate information to communicate causal sequences (Green, 2006), and “stories may be particularly well equipped to channel attention toward the situational determinants of individual action” (Strange, 2002, p. 276). Policy narratives in particular focus on cause and effect and illustrate the causes of social problems and policy solutions to address them (Jones & McBeth, 2010; Shanahan, Jones, McBeth, & Lane, 2013), and narratives appear to be well suited to convey the complexity of health, societal responsibility, and health interventions (Lundell, Niederdeppe, & Clarke, 2013a, 2013b). For example, Niederdeppe, Heley, and Barry (2015) used a narrative message that told a story about Cynthia, a mother of two, who has faced struggles with her daughter’s weight problem due to her inability to control her daughter’s food and beverage choices. An excerpt from this example narrative is as follows:

Despite her best efforts to encourage these healthier options at home, Cynthia found that her attempts to improve her daughter’s drink choices were mostly unsuccessful. She noticed that sodas were constantly available to her daughter outside of her home—in schools and at several stores she passed every day on her way back and forth from school. And her daughter saw what seemed like a never ending stream of soda advertising and promotion on television, the Internet, and in local stores.

(Niederdeppe et al., 2015, online supporting materials)

This narrative moves past merely emphasizing individual or societal responsibility (as in the aforementioned examples) and instead provides a personal story of someone who faces societal constraints when attempting to provide healthy beverage options for her overweight child.

Visual images can also be used as persuasive tools in societal responsibility messages by inviting generalizations and causal interpretations (Messaris, 1997). For example, images of an exemplified individual may lead to sustained changes in beliefs about the prevalence of social risks (Zillmann, 2006). Images can also be used to influence estimates of the rates of disease for particular subgroups (e.g., ethnic groups; Gibson & Zillmann, 2000). Importantly, images can influence attributions relevant to societal responsibility and health disparities by inviting causal interpretations (Niederdeppe et al., 2008). Statistical images may be particularly useful for inviting generalizations by using inferential statistics and might also invite causal interpretations by showing associations between societal responsibility and health; however, participants might see these images as oversimplified versions of a complex reality (Lundell et al., 2013b) thus reducing their credibility (see Parrott, Silk, Dorgan, Condit, & Harris, 2005). For example, a specific type of image, GIS mapping, holds promise in health communication and public health policy making (see Parrott, Hopfer, Ghetian, & Lengerich, 2007).

Perceptions of Responsibility and Persuasion

Message designers are, of course, interested in whether these responsibility appeals are persuasive. These personal, interdependent, and societal responsibility messages attempt to elicit perceptions or beliefs about personal, interdependent, or societal responsibility in the message recipient. The central question of interest to message designers thus considers whether these perceptions of responsibility result in the desired message outcomes. To address this question, the next section reviews the available empirical literature by considering the direct persuasive effects of perceptions of personal responsibility and interdependent responsibility on health intentions or behaviors, as well as the indirect persuasive effects of perceptions of responsibility via, for example, anticipated emotions. This next section also reviews the empirical work on the persuasiveness of perceptions of societal responsibility on support for public health policy, as the desired message outcome in these societal responsibility messages is policy support.

Perceptions of Personal Responsibility

There is some empirical evidence that an individual’s feelings of personal responsibility motivate self-oriented health intentions and behaviors (e.g., Rothman, Salvoey, Turvey, & Fishkin, 1993; Niederdeppe, Roh, Shapiro, & Kim, 2013; Williams-Piehota et al., 2004; for an exception, see Ziff, Conrad, & Lachman, 1995). For example, Rothman and colleagues (1993) found that women who viewed a video emphasizing internal attributions of responsibility for health were more likely to believe they were personally responsible for cancer prevention and then obtain a mammogram than those who viewed an external attributions or information-only presentation. Likewise, Niederdeppe, Roh, Shapiro, and Kim (2013) found that high personal responsibility increases intentions to engage in health behaviors, including eating fruits and vegetables, engaging in regular exercise, and dieting to lose weight. Further, several meta-analyses have found that the inclusion of perceived moral obligation, which Ajzen (1991) defines as “personal feelings of … responsibility to perform, or refusal to perform, a certain behavior” (p. 199), in the “Theory of Planned Behavior” (Ajzen, 1988, 1991) explains an additional 3% to 4% of the variation in intentions (Conner & Armitage, 1998; Rivis, Sheeran, & Armitage, 2009).

Perceptions of Interdependent Responsibility

There is also evidence that an individual’s perceptions of interdependent responsibility motivate them to engage in other-oriented behaviors (e.g., Bryan & Hershfield, 2012; Morrison & Phelps, 1999; Tennen, Affleck, & Gershman, 1986; for an exception, see Yun, Silk, Bowman, Neuberger, & Atkin, 2009). For example, Tennen and colleagues (1986) found that perceptions of responsibility are associated with positive health behavior changes in mothers of children with prenatal complications. Further, Bryan and Hershfield (2012) conducted an interesting experiment that examined people’s sense of responsibility to their future self (i.e., participants viewed their future self as an “other” that was distinct from their current selves). The authors found that participants exposed to an other-oriented responsibility message that emphasized their responsibility to a future self (i.e., to an “other”) were more likely than those exposed to a self-oriented message to save money for retirement. Importantly, Rivis and colleagues (2009) found that perceived moral obligation, which, as mentioned is defined as feelings of responsibility, is more strongly correlated with intentions for behaviors that have consequences for the well-being of others, suggesting an important role for perceptions of interdependent responsibility in other-oriented health behaviors.

Indirect Effects of Perceptions of Personal and Interdependent Responsibility

Perceptions of personal and interdependent responsibility may also influence intentions and behaviors indirectly. For example, responsibility may indirectly impact intentions and behaviors via anticipated emotions (for a meta-analysis, see Rivis et al., 2009), specifically regret, guilt, and pride. Perceptions of responsibility are associated with anticipations of regret, as responsibility is considered a necessary precondition for regret (see Zeelenberg, van Dijk, & Manstead, 2000), and empirical work demonstrates a clear relationship between regret and responsibility. For example, Frijda, Kuipers, and ter Schure (1989, study 2) found that a perception of responsibility was a typical appraisal item for regret. Further, Gilovich and Medvec (1995) had participants list regrets, and, as very few participants listed regretting outcomes beyond their control, the authors concluded that “a sense of personal responsibility appears to be central to the experience of regret” (p. 383). Additionally, in an interesting series of four papers (consisting of numerous empirical studies), Connolly and colleagues (Connolly, Ordóñez, & Coughlan, 1997; Ordóñez & Connolly, 2000) and Zeelenberg and colleagues (Zeelenberg, van Dijk, & Manstead, 1998, 2000) eventually concluded that responsibility for a decision is indeed associated with regret.

Additional studies have, in turn, shown that anticipations of regret lead to desired personal and interdependent responsibility message outcomes, including intentions and behavior change. Indeed, a recent meta-analysis found that anticipated regret added significantly and independently to the prediction of intentions over and above variables in the “Theory of Planned Behavior” (Sandberg & Conner, 2008). Further, anticipated regret had a direct and significant impact on prospective behavior (Sandberg & Conner, 2008). Additionally, several studies have found that anticipated regret strengthens intentions by associating inaction with aversive affect, and, additionally, anticipated regret increases the consistency between intentions and behavior (Abraham & Sheeran, 2003, 2004; Sheeran & Orbell, 1999).

Responsibility is also associated with anticipations of guilt, as Miceli (1992) notes that in order to feel guilt, an individual must feel responsible and know that their action (or inaction) caused harm. Further, several scholars have noted that guilt arises when one feels personally responsible for the violation of personal, social, or moral norms (Berndsen & Manstead, 2007; Kugler & Jones, 1992; Tangney, Miller, Flicker, & Barlow, 1996) and when failure is attributed to internal, controllable causes (i.e., agency, a component of responsibility; Weiner, Russell, & Lerman, 1978, 1979).

Like anticipated regret, studies have also shown that anticipated guilt is associated with the desired message outcomes of intentions and behavior change. This may be particularly relevant in interdependent dyads, including, for example, the parent-child and adult child-aging parent relationships, as guilt is a moral emotion that is based in social relationships and can occur when individuals imagine doing something wrong or perceive that their action (or inaction) might adversely impact another (Haidt, 2003; Lazarus, 1991). O’Keefe (2002) notes that people will avoid actions that they anticipate will make them feel guilty, thus serving as an influence mechanism. Some initial examinations of anticipated guilt indicate that anticipated guilt functions to motivate people to alter their behavior after reading a message about a threat to family members or unknown others (e.g., bone marrow donation; Lindsey, 2005; Lindsey, Yun, & Hill, 2007). Likewise, Xiao (2011) found that anticipated guilt increased the variance in intentions to register as an organ donor and to discuss organ donation with family members, after controlling for variables in the “Theory of Planned Behavior.”

Finally, responsibility also appears to be associated with anticipations of pride. Like guilt, pride is also a self-conscious emotion that is evoked by evaluating one’s self after following (or failing to follow) personal or social norms (Tracy & Robins, 2004a). An individual feels pride when they attribute their achievements to the self (i.e., they had agency) and evaluate themselves positively with respect to both personal standards and social standards (Exline & Lobel, 2001; Tracy & Robins, 2004b; Williams & DeSteno, 2008). Importantly, pride is integral to perceptions of responsibility, as Mascolo and Fischer (1995) define pride as an emotion that is “generated by appraisals that one is responsible for being a socially valued person” (p. 66), and Harth, Leach, and Kessler (2013) found that in-group perceptions of responsibility for prosocial behaviors elicited pride.

Anticipated pride is also associated with intention and behavior changes, as “pride can be considered a desired end state that individuals strive to reach and maintain” (Onwezen, Bartels, & Antonides, 2014, p. 55). Tangney, Stuewig, and Mashek (2007) note that anticipated pride plays an important role in promoting behaviors that conform to social standards. Further, pride affects behavior over time (Patrick, Chun, & MacInnis, 2009). Importantly, pride encourages prosocial behaviors, including caregiving activities (Tracy & Robins, 2004b). Further, Goei and Boster (2005) found that gratitude, a positive, desirable state associated with pride, contentment, and happiness, increased compliance with a request, while obligation did not. Additionally, there is some evidence that perceptions of responsibility indirectly influence anticipated pride, which, in turn, influences intentions (e.g., Onwezen et al., 2013).

Together, this evidence suggests that personal and interdependent responsibility may indirectly influence intentions and behaviors via anticipated emotions. Specifically, perceptions of personal or interdependent responsibility appear likely to influence anticipations of regret, guilt, and pride, which, in turn, appear likely to influence positive health intentions and behavior changes. Indeed, some initial empirical evidence suggests this is the case. For example, Onwezen and colleagues (2013) found that perceptions of responsibility indirectly influence anticipations of guilt and pride, which, in turn, influence health behavior intentions.

Perceptions of Societal Responsibility

There is some empirical evidence that perceptions of societal responsibility that stem from societal responsibility appeals increase support for public health policy (i.e., the intended outcome of these societal responsibility messages). For example, Niederdeppe, Shapiro, and Porticella (2011) found that the relationship between societal cause attributions and support for public policies was partially mediated by societal solution attributions. Additionally, Wolfson, Gollust, Niederdeppe, and Barry (2015) found that attributions of external responsibility for solving childhood obesity (i.e., to schools, the food and beverage industry, and the government) was strongly and positively related to support for policies designed to prevent childhood obesity (e.g., prohibit advertising of unhealthy foods during children’s television programming). There is additional evidence that the societal responsibility message strategies reviewed previously, including framing, narratives, and visual images, evoke the intended responses.

There is some empirical support for the use of message framing to increase perceptions of societal responsibility and subsequent policy support. For example, framing the causes of obesity by emphasizing society (i.e., thematic) compared the individual (i.e., episodic), has been shown to increase societal attributions of responsibility for obesity (Major, 2009) and support for policies (Barry, Brescoll, & Gollust, 2013; Coleman, Thorson, & Wilkins, 2011). However, these results are inconsistent (e.g., Coleman et al., 2011 found no effect on societal attributions), and Gollust, Lantz, and Ubel (2009) found that although societal cause frames are effective for Democrats, they foster reactance among Republications. Importantly (and consistent with the above review), meta-analytic data shows that a two-sided refutational frame, or a message that articulates a position and then refutes opposing arguments, is more persuasive than a one-sided or nonrefutational two-sided message (O’Keefe, 1999). Indeed, Niederdeppe, Roh, and Shapiro (2015) found that a message that fails to acknowledge individual responsibility (i.e., a one-sided frame) invites counterarguing and reduces feelings of empathy that, in turn, undermine support for obesity policies. Gollust and Capella (2014) likewise found that messages that emphasized societal responsibility but acknowledged personal responsibility (two-sided) reduced counterarguing among Republicans. However, the effect of two-sided and one-sided messages may be moderated by the focus of the message on the individual or the community via effects on simple elaboration (i.e., thoughts on societal responsibility for obesity; Niederdeppe, Kim, Lundell, Fazili, & Frazier, 2012).

With respect to competing frames, Niederdeppe, Gollust, and Barry (2014) examined the role of competing anti-tax and pro-tax frames about sugar-sweetened beverages using inoculation messages as a proactive strategy to anticipate counterframes; results suggest that exposure to a weakened form of an oppositional message influences short-term policy support, but does not result in sustained resistance to later anti-tax messages. Niederdeppe, Heley, and Barry (2015) also examined competing frames by employing pro-policy inoculation messages and industry anti-policy messages; results suggest that inoculation messages conferred resistance to persuasion with later exposure to an anti-policy message and increases in policy support.

Empirical work also suggests that using narratives may be an effective strategy for societal responsibility appeals. For example, narratives about societal responsibility for obesity are likely to reduce counterarguing and increase societal solution attributions among liberals (Niederdeppe, Shapiro, & Porticella, 2011) and increase policy support among conservatives, depending on the level of personal responsibility also acknowledged within the story (Niederdeppe, Shapiro, Kim, Bartolo, & Porticella, 2013; see also, Niederdeppe, Roh, & Shapiro, 2015). Niederdeppe, Heley, and Barry (2015) also found that narratives about societal responsibility for obesity reduce counterarguing and increase policy support.

Initial work with visual images also indicates the potential persuasiveness of this societal responsibility message strategy in increasing support for health policy changes. For example, Niederdeppe, Roh, and Dreisbach (2015) found that the addition of a GIS map of food deserts to an individual focused narrative reduced counterarguing and increased simple elaboration of the persuasive topic. These two cognitive processes related to increased support of obesity-prevention policy (Niederdeppe et al., 2015).

Conclusions and Future Directions

This article provided a conceptual definition of responsibility that is conducive to identifying effect-independent message variables for responsibility message design, including attributions of causation and solution, obligation, and agency. This article then reviewed the vast array of health and risk responsibility messages at the personal, interdependent, and societal level and examined the persuasiveness of perceptions of personal, interdependent, and societal responsibility on intended affective and behavioral outcomes. Future work that uses these types of responsibility appeals should adhere to the conceptual definition provided that characterize message variations of responsibility in terms of effect-independent message features (see O’Keefe, 2003) instead of using effect-based definitions of message variables (i.e., defining a responsibility message as one that elicits perceptions of responsibility). Doing so will assist in specifying what exactly a responsibility message is, as well as the effects each message variable has on persuasive outcomes. Together, this review highlights the potential for efforts to design responsibility messages that arouse perceptions of responsibility to be used to effectively design health and risk messages; however, it also underscores the need for future work to further validate these empirical claims.

First, future work should strive to specify specific message features for attributions of causation and solution, obligation, and agency in appeals to personal, interdependent, and societal responsibility. These message features may impact responsibility message outcomes both directly and indirectly via specific affective and cognitive outcomes. Future work should also empirically establish the persuasive role of perceptions of responsibility for the intended message outcomes.

Future work should also examine whether perceptions of responsibility differ across individuals of different ethnic and cultural backgrounds and whether responsibility functions differently across these groups. For example, there is some work that indicates that the Norm Activation Model (Schwartz, 1977) is applicable across cultural groups (e.g., Milfont, Sibley, & Duckitt, 2010); however, future work should aim to determine if this also holds for responsibility messages and their outcomes. Additionally, some research indicates that there are differences in the self-regulatory role of anticipated emotions between collectivistic and individualistic cultures. Self-conscious emotions such as guilt and pride are based upon the self (Tracy & Robins, 2004a). Individuals from individualistic versus collectivistic cultures may differ in their construal of the self (Markus & Kitayama, 1991), so the function of these emotions as mediators between perceptions of responsibility and related outcomes may be sensitive to cultural differences (Tracy & Robins, 2004a).

Finally, and importantly, future work should also examine potential iatrogenic effects (i.e., negative, unintended consequences). Responsibility appeals may evoke perceptions of, for example, self-blame. Perceptions of blame may not only fail to elicit the desired changes in intentions or behaviors, but may also negatively impact the message recipient. Likewise, responsibility messages that stress the power of the individual in personal or interdependent responsibility may elicit stigma perceptions. Indeed, understanding responsibility is central to the formation of stigma beliefs, as people stigmatize those who choose immoral beliefs, attitudes, or actions more than those who do not have control (Smith, 2007). Message design scholars should therefore take care to uphold a basic ethical tenet of medicine, “primum non nocere,” or “first, do no harm.”

Further Reading

  • du Pré, A. (2005). Communicating about health: Current issues and perspectives (2d ed.). New York: McGraw-Hill.
  • Guttman, N. (2000). Public health communication interventions: Values and ethical dilemmas. Thousand Oaks, CA: SAGE.
  • Maibach, E., & Parrott, R. P. (Eds.). (1995). Designing health messages: Approaches from communication theory and public health practice. Thousand Oaks, CA: SAGE.
  • O’Keefe, D. J. (2016). Persuasion: Theory and research. Thousand Oaks, CA: SAGE.
  • Thompson, T. L., Parrott, R., & Nussbaum, J. F. (Eds.). (2011). The Routledge handbook of health communication. New York: Routledge.