1.1.1 Increased Attitude Strength Fosters Greater Advocacy

First, one of the primary drivers of attitudinal advocacy is attitude strength. Attitude strength refers to the extent to which a given attitude is durable and impactful (Krosnick & Petty, Reference Krosnick, Petty, Petty and Krosnick1995). In general, compared to people with weak attitudes, people who hold strong attitudes display more resistance to persuasion, more attitude stability over time, and increased attitude–behavior correspondence. For example, a consumer with a strong rather than weak favorable attitude toward an airline would be more likely to book future flights on the same airline and less likely to switch to a competitor. Most germane to the current review, a wealth of research suggests that the stronger people’s attitudes are, the more likely they are to advocate on behalf of those attitudes (e.g., see Krosnick & Petty, Reference Krosnick, Petty, Petty and Krosnick1995). In recent years, particular attention has been paid to the roles of attitude certainty and moral conviction in shaping advocacy outcomes.

Consider attitude certainty. Attitude certainty refers to the subjective sense of confidence or conviction an individual has about their attitude (Tormala & Rucker, Reference Tormala and Rucker2018; see also Abelson, Reference Abelson1988). Past research has shown that the more certain people feel about their attitudes, the more likely they are to advocate on behalf of those attitudes. Evidence for this relationship is widespread. For instance, Visser, Krosnick, and Simmons (Reference Visser, Krosnick and Simmons2003) found that the more certain people were about their opinions on global warming, the more likely they were to donate money to organizations concerned with global warming and to report attending meetings to discuss the issue of global warming. Similarly, Barden and Petty (Reference Barden and Petty2008) found that attitude certainty predicted people’s willingness to vote and sign petitions. Akhtar, Paunesku, and Tormala (Reference Akhtar, Paunesku and Tormala2013) documented a similar relationship, finding that attitude certainty predicted people’s willingness to share their views with others, as well as their likelihood to try to persuade others. More recently, Philipp-Muller, Wallace, and Wegener (Reference Philipp-Muller, Wallace and Wegener2020) found that the greater people’s feelings of certainty, the more likely they were to engage in advocacy behaviors such as door-to-door canvassing, wearing supportive pins, and sharing attitude-relevant information on social media.

Importantly, though, there is nuance in the certainty–advocacy relationship. For example, Rios, DeMarree, and Miller (Reference Rios, DeMarree and Statzer2014) found that the effect of attitude certainty on advocacy behavior depended on whether certainty was construed in terms of attitude correctness (the sense that one’s attitude is correct) or attitude clarity (the sense that one’s attitude is clear in one’s mind) – a distinction first drawn by Petrocelli, Tormala, and Rucker (Reference Petrocelli, Tormala and Rucker2007). Rios and colleagues found that attitude correctness positively predicted people’s desire to win arguments and persuade others, whereas attitude clarity had no such effect. They reasoned that correctness often implies that one’s attitude is superior and hence might encourage attempts to win arguments and directly influence the opinions of others.

Subsequent research further unpacked the certainty–advocacy relationship, suggesting that the effects of attitude correctness and attitude clarity depend on the type of advocacy in question. As described earlier, we use the term “advocacy” to refer to the general expression of support for or opposition to something. The specific goal of advocating could be to persuade others (i.e., persuasion intentions) or it could be to simply express one’s view or voice one’s opinion (i.e., sharing intentions). Cheatham and Tormala (Reference Cheatham and Tormala2015) investigated how attitude correctness and attitude clarity affected these different advocacy dimensions. They found that correctness predicted both persuasion and sharing intentions, but clarity predicted only sharing intentions. Thus, people are more willing to express their attitudes as their feeling of clarity rises, but the motivation to actually persuade others appears to require a feeling of attitude correctness.

In addition to attitude certainty, moral conviction has been found to predict advocacy. Moral conviction refers to the subjective perception that one’s attitude is rooted in moral values and one’s sense of what is right or wrong (Skitka, Washburn, & Carsel, Reference Skitka, Washburn and Carsel2015). Two people can hold similar attitudes (e.g., a positive attitude toward buying an electric car) but differ in whether their attitudes are based on moral considerations (e.g., buying an electric car is the right thing to do for the environment vs. buying an electric car saves money on gas). In general, the more people believe that their attitudes are rooted in morality, the more likely they are to advocate on behalf of those attitudes. For example, Skitka and Bauman (Reference Skitka and Bauman2008) found that moral conviction predicted voting intentions and behavior, above and beyond other indicators of attitude strength. Similarly, Skitka, Hanson, and Wisneski (Reference Skitka, Hanson and Wisneski2017) found that the greater people’s moral convictions, the more willing they were to engage in advocacy behaviors such as signing a petition, contacting a state representative, and working at a phone bank. Likewise, Philipp-Muller, Wallace, and Wegener (Reference Philipp-Muller, Wallace and Wegener2020) found that moral conviction was predictive of people’s intentions to advocate on behalf of their attitudes in a variety of ways (e.g., sharing attitude-relevant information on social media).

In summary, a growing body of research points to attitude strength as a powerful predictor of attitudinal advocacy. Recent work on this topic has focused mostly on attitude certainty and moral conviction, but other dimensions of attitude strength (importance, ambivalence, etc.) have also been found to predict advocacy. As one example, Philipp-Muller, Wallace, and Wegener (Reference Philipp-Muller, Wallace and Wegener2020) reported positive correlations between attitude importance and advocacy behavior. Across multiple dimensions of attitude strength, then, the stronger people’s attitudes are, the more likely they are to advocate.

1.1.2 Compensatory Motives Also Cause People to Advocate

Although the bulk of the literature has shown that the more conviction people have about their attitudes (e.g., attitude certainty, moral conviction), the more likely they are to advocate, some research has shown the opposite – that is, that doubt can be a catalyst for advocacy. Consider the canonical doomsday cult described by Festinger, Riecken, and Schachter (Reference Festinger, Riecken and Schachter1956) who infiltrated a cult whose members believed the world was going to end soon and that they were going to be among the lucky few rescued by aliens. When this prophecy proved untrue, rather than abandon their beliefs, members of the cult increased their efforts to recruit new members and spread the message. In essence, the failed prophecy boosted their advocacy behavior.

One interpretation of Festinger, Riecken, and Schachter’s case study is that when people lose confidence in important and closely held beliefs, they turn to advocacy as a way to affirm those beliefs and restore their confidence in them. Gal and Rucker (Reference Gal and Rucker2010) investigated this interpretation experimentally. In a series of studies, Gal and Rucker found that participants advocated more (e.g., wrote longer persuasive messages) when they were made to feel doubtful (e.g., recalling two situations in which they felt a great deal of uncertainty) as opposed to confident (e.g., recalling two situations in which they felt a great deal of certainty) about a closely held belief. They argued that advocacy played a compensatory role, helping people affirm and restore their confidence in important beliefs after that confidence was shaken.

Rios, Wheeler, and Miller (Reference Rios2012) examined the compensatory role of advocacy in the context of minority opinion expression. They found that when low self-esteem individuals (those most vulnerable to blows to their self-confidence) were made to feel unsure of themselves, they were more likely to express minority opinions. Rios and colleagues argued that when one’s sense of self is shaken, expressing minority opinions can help establish one’s uniqueness and restore self-confidence. These results further suggest that advocacy can serve a compensatory function, helping people deal with feelings of doubt by reaffirming who they are and what they stand for (see Miller & Morrison, Reference Miller and Morrison2009; Rios, Reference Rios2012).

As noted, this evidence for a negative relationship between certainty and advocacy is seemingly at odds with the attitude strength literature, which generally shows a positive relationship. To reconcile these contradictory findings, Cheatham and Tormala (Reference Cheatham and Tormala2017) theorized that the relationship between attitude certainty and attitudinal advocacy might be curvilinear. Across several studies, these authors found that advocacy peaks at high certainty, declines at moderate certainty, and then increases again at very low certainty. The logic was that people with high (compared to moderate) certainty advocate more because they have more conviction and an increased inclination to argue their point. People with low (compared to moderate) certainty advocate more for different reasons. For them, advocacy can act as a means to resolve uncertainty through compensatory mechanisms or by offering the chance to gather useful information from talking to others.

1.1.3 Advocacy Intentions Increase As Perceived Efficacy Rises

Advocacy research also points to perceived efficacy as an important trigger. By perceived efficacy, we mean people’s beliefs that they have the ability to advocate effectively and/or that their advocacy effort will make an impact. Prior research suggests that the more effectively people believe they can advocate or the more impact they think they can have, the more willing they are to advocate on behalf of their views.

For instance, Akhtar, Paunesku, and Tormala (Reference Akhtar, Paunesku and Tormala2013) explored the role of argumentation efficacy in sparking advocacy. Argumentation efficacy refers to people’s perceptions that they can make a convincing case – that is, that they can effectively argue their side of an issue. Akhtar, Paunesku, and Tormala showed that when argumentation efficacy is elevated, people become more likely to advocate on behalf of their opinions. In one study, participants were exposed to proattitudinal arguments, and those arguments were varied to be either weak or strong. Akhtar, Paunesku, and Tormala found that, compared to strong proattitudinal arguments, weak proattitudinal arguments increased feelings of argumentation efficacy, which subsequently increased advocacy intentions. After being exposed to weak (rather than strong) arguments supporting their own political view, for example, participants reported being more likely to try to persuade others to their position and more willing to make phone calls on behalf of a political campaign that they supported. The rationale was that when people see weak (compared to strong) arguments being used to endorse their side, they believe they can make a more compelling case, which boosts their advocacy intentions.

Also relevant to perceived efficacy is the persuadability of the message recipient. Past research suggests that people are more likely to advocate when they believe that message recipients will be receptive and open-minded to their message, as opposed to unwilling to listen or change their minds. In one demonstration, Gal and Rucker (Reference Gal and Rucker2010) found that the effect of inducing uncertainty (the compensatory effect described earlier) depended on whether the target of the advocacy effort was expected to be open- or closed-minded to the message. When the target was open-minded, participants experiencing uncertainty engaged in greater advocacy than those experiencing certainty. However, when the target was closed-minded, there was no difference in advocacy between those who experienced uncertainty and certainty. In general, when advocacy efforts were seen as unlikely to succeed (i.e., when the target was closed-minded), the desire to advocate declined.

As another example, Akhtar and Wheeler (Reference Akhtar and Wheeler2016) investigated people’s implicit theories of attitudes. Implicit theories of attitudes refer to people’s beliefs about the stability or malleability of attitudes (Petrocelli et al., Reference Petrocelli, Clarkson, Tormala and Hendrix2010). Some people hold an entity theory of attitudes and believe that attitudes are inherently fixed and unchanging; others hold an incremental theory of attitudes, believing that attitudes are malleable and dynamic. Akhtar and Wheeler examined the effect of holding an entity versus an incremental theory of attitudes on advocacy. They found a dual effect. First, entity theorists (who believe that attitudes are fixed) tended to assume that other people’s attitudes would be stable and difficult to change, and thus were disinclined to attempt to persuade them. Incremental theorists (who believe that attitudes are malleable) tended to assume that others’ attitudes were more changeable and were more inclined to attempt to persuade them.

Interestingly, though, Akhtar and Wheeler also uncovered the opposite relationship. Specifically, compared to incremental theorists, entity theorists tended to hold their attitudes with greater certainty, which increased their willingness to advocate. Together, these competing forces – believing that other people’s attitudes were difficult to change while also holding one’s own attitude with greater certainty – cancelled each other out, resulting in no overall effect of implicit theories on advocacy. However, perceptions of target persuadability played a critical role in determining people’s advocacy efforts. When participants were induced to focus their attention on their persuasion targets, thereby making salient their belief that others’ attitudes were mutable or immutable, entity theorists were less likely to advocate than incremental theorists. In this case, people’s beliefs about the persuadability of the target shaped their advocacy intentions. When participants were induced to focus on their own attitudes, however, their own certainty played a greater role and entity theorists were more likely to advocate than incremental theorists.

A related factor that has been shown to influence people’s willingness to advocate is the expected impact of their advocacy efforts – that is, how big a difference they believe their advocacy will make. Generally speaking, the greater the expected impact, the more willing people are to advocate. In one study, Bechler, Tormala, and Rucker (Reference Bechler, Tormala and Rucker2020) presented participants with different types of persuasion targets and measured how likely participants were to advocate to each. Specifically, participants were presented with targets whose attitudes toward the focal entity could be changed from very opposed to less opposed, from somewhat opposed to somewhat in favor, or from somewhat in favor to very in favor. Bechler, Tormala, and Rucker found that participants believed the second group – targets whose attitudes would change across rather than within valence – would show the greatest attitude and behavior change, and thus were more likely to advocate to this group than to the others. In other words, attitude change across valence was seen as larger and more impactful than attitude change within valence (see also Bechler, Tormala, & Rucker, Reference Bechler, Tormala and Rucker2019), and the bigger the impact participants believed they could make with their advocacy effort, the more likely they were to advocate (see also Bechler & Tormala, Reference Bechler and Tormala2021).

The results summarized in this section echo findings from research on the relationship between efficacy and collective action (for a review, see van Stekelenburg & Klandermans, Reference van Stekelenburg and Klandermans2013). For example, Klandermans (Reference Klandermans1984) surveyed Dutch union workers contemplating going on strike in order to demand shorter work weeks. Among other factors, Klandermans assessed participants’ beliefs about whether their participation would increase the effectiveness of the strike (e.g., the extent to which they agreed with statements such as, “It really doesn’t make much difference whether or not I take part in action”). Results indicated that the more participants believed their participation would increase the effectiveness of the strike, the more willing they were to take part.

Interestingly, while the research reviewed so far has highlighted a positive relationship between perceived efficacy and advocacy, there is also some evidence to suggest a negative relationship. Tausch et al. (Reference Tausch, Becker and Spears2011) distinguished between three different types of advocacy behaviors: normative actions (e.g., participating in discussion meetings, signing petitions, demonstrating); non-normative, non-violent actions (e.g., blocking university buildings, blocking the highway); and non-normative, violent actions (e.g., throwing stones or bottles, engaging in arson attacks, attacking the police). In one study, Tausch and colleagues found that while efficacy positively predicted engagement in normative advocacy actions (e.g., signing petitions), it negatively predicted engagement in non-normative, violent actions (e.g., attacking the police). In other words, the more (less) people thought they could do to improve their plight, the less (more) likely they were to consider non-normative, violent actions as a means for advocacy. These results suggest that the relationship between perceived efficacy and advocacy might be more nuanced than previously thought and further research in this area would be worthwhile (see also Ayanian et al., Reference Ayanian, Tausch, Acar, Chayinska, Cheung and Lukyanova2021, who further delineate between types of efficacy and their ability to predict collective action).

To summarize, one of the themes to emerge from recent advocacy research is the importance of perceived efficacy – that is, the likely effectiveness and potential impact of one’s advocacy effort. The greater the chances that people’s advocacy efforts will make an impact – for example, because they believe that they can make a compelling case or because their targets are open to changing their minds – the more willing they are to advocate. Similarly, the more impact people expect to have, the more likely they are to advocate on behalf of their views.

1.1.4 Emotions Affect Advocacy Intentions

Emotions have also been shown to influence people’s willingness to advocate on behalf of their attitudes (for reviews, see Radke et al., Reference Radke, Kutlaca, Siem, Wright and Becker2020; van Stekelenburg & Klandermans, Reference van Stekelenburg and Klandermans2013). For instance, research reveals that anger is linked to advocacy (e.g., Klandermans, van der Toorn, & van Stekelenburg, Reference Klandermans, van der Toorn and van Stekelenburg2008; Leach, Iyer, & Pedersen, Reference Leach, Iyer and Pedersen2006; Lodewijkx, Kersten, & van Zomeren, Reference Lodewijkx, Kersten and van Zomeren2008; Tausch et al., Reference Tausch, Becker and Spears2011; van Zomeren et al., Reference van Zomeren, Spears and Leach2008, Reference van Zomeren, Postmes, Spears and Bettache2011). In one demonstration, van Zomeren and colleagues (Reference van Zomeren, Postmes, Spears and Bettache2011) asked participants to read an article highlighting how a member of a disadvantaged social group has been treated unfairly (e.g., a Muslim woman who was denied a job because of her religious identity). Results indicated that the more anger participants reported feeling in response to the article, the higher their advocacy intentions – for instance, the more willing they were to sign a petition or participate in a demonstration against discrimination. Similarly, Lodewijkx, Kersten, and van Zomeren (Reference Lodewijkx, Kersten and van Zomeren2008) found that moral outrage – anger at the violation of a moral standard – positively predicted advocacy behaviors such as participation in protests. Tausch et al. (Reference Tausch, Becker and Spears2011) also found that anger positively predicted advocacy actions, such as willingness to participate in discussions around a social issue and willingness to partake in demonstrations (see Ayanian et al., Reference Ayanian, Tausch, Acar, Chayinska, Cheung and Lukyanova2021, for a replication of these results in repressive cultural contexts).

In addition to anger, feelings of guilt have been shown to predict advocacy-relevant outcomes. For instance, Mallett et al. (Reference Mallett, Huntsinger, Sinclair and Swim2008) found that among members of advantaged groups (e.g., White Americans), those who were sympathetic to disadvantaged groups (e.g., Black Americans) felt guiltier when they learned about a hate crime committed against those groups (e.g., against a Black American). These feelings of guilt, in turn, predicted a greater desire to advocate on behalf of the disadvantaged groups – for example, increased intentions to write letters that protested hate crimes. Interestingly, in contrast to the Mallett et al. finding, other research has uncovered a more limited role for guilt in shaping advocacy. For example, Leach, Iyer, and Pedersen (Reference Leach, Iyer and Pedersen2006) found that guilt was associated with the abstract goal of compensating disadvantaged groups, but it did not predict intentions to participate in specific advocacy behaviors designed to materially benefit disadvantaged groups. Future research further exploring the relationship between guilt and advocacy would be worthwhile.

The research on emotions raises a number of questions. Most notably, which emotions are likely to increase advocacy and which are likely to decrease it? One answer might be found in past research on emotion specificity and certainty appraisals. This research has revealed that some emotions – for example, anger, disgust, and happiness – are accompanied by a general feeling of certainty, whereas other emotions – such as surprise, fear, and sadness – are associated with a general feeling of uncertainty (Ellsworth & Smith, Reference Ellsworth and Smith1988; Smith & Ellsworth, Reference Smith and Ellsworth1985; Tiedens & Linton, Reference Tiedens and Linton2001). As reviewed earlier, considerable prior research also indicates that certainty boosts advocacy. Therefore, it is plausible that emotions that spark certainty will lead to greater advocacy, while those that undermine certainty will lower advocacy. For example, whereas anger or moral outrage (high certainty emotions) might increase advocacy, fear and sadness (low certainty emotions) might decrease it. Unpacking the relationship between emotions and advocacy would be a worthwhile direction for future work.

1.1.5 The Way People Frame Their Attitudes Shapes Their Likelihood of Advocating

Advocacy research has also examined attitude framing effects. Specifically, Catapano and Tormala (Reference Catapano and Tormala2021) examined the effect of support–oppose framing on advocacy-relevant outcomes. Catapano and Tormala argued that when people hold attitudes, they can think of those attitudes in terms of what they support or what they oppose (see also Bizer et al., Reference Bizer, Larsen and Petty2011). For example, a person with a favorable attitude toward guns could construe and express their attitude as either: “I support allowing guns” or “I oppose banning guns.” Likewise, a person with an unfavorable attitude toward guns could construe and express their attitude as either: “I support banning guns” or “I oppose allowing guns.” Catapano and Tormala investigated the effect of this support–oppose framing on advocacy – in particular, on attitude sharing intentions. Across a number of attitude issues, they found that people were more likely to share (or express) their attitudes when those attitudes were framed in support rather than oppose terms.

Catapano and Tormala argued that this effect occurs for two reasons. First, people tend to define who they are in terms of what they support rather than what they oppose. As a result, support-framed attitudes feel more value-expressive, which increases the likelihood that people will share them. Second, people want to be liked by others, and believe that agreement facilitates liking. Because support-framing is more agreement-focused than oppose-framing, people are more likely to express views framed in terms of support rather than opposition. Consistent with this theorizing, Catapano and Tormala found that value expression and impression management motives played a dual role in driving the effect of support–oppose framing on sharing intentions.

1.1.6 Affective and Cognitive Bases Influence Advocacy Intentions

Just like attitudes can vary in how they are framed, attitudes can vary in their underlying basis – that is, in the extent to which they are rooted in affect or cognition (e.g., Crites et al., Reference Crites, Fabrigar and Petty1994). For example, two consumers who both hold positive attitudes toward plant-based meats could differ in whether their attitudes are based primarily on affect and feelings (e.g., “I like plant-based meats because it makes me happy to eat meat without hurting animals.”) or cognition and beliefs (e.g., “I like plant-based meats because I believe they are a useful tool to combat climate change.”).

Teeny and Petty (Reference Teeny and Petty2018) investigated the effect of attitude bases on attitudinal advocacy. Specifically, they examined how people’s perceptions of the extent to which their attitudes are rooted in affect versus cognition influenced two different types of advocacy behaviors: requested advocacy and spontaneous advocacy. Requested advocacy refers to advocacy efforts that are undertaken in response to an explicit request – for example, endorsing a particular restaurant after a friend asks for a recommendation. By contrast, spontaneous advocacy refers to advocacy efforts undertaken without any prompting from others – for example, spontaneously texting a friend to recommend a particular restaurant. Teeny and Petty found that people who perceived their attitudes to be rooted in affect were more likely to engage in spontaneous advocacy, whereas people who perceived their attitudes to be rooted in cognition were more likely to engage in requested advocacy. They proposed that people associate emotions with spontaneous activity, thereby advocating more spontaneously when they believe their attitudes are grounded in emotions (e.g., feeling excited after trying out a restaurant and subsequently recommending it to friends). However, when advocacy is requested, it prompts expectations of thoughtful and well-reasoned inputs. Thus, people advocate more in response to direct requests when they view their attitudes as grounded in cognition.

1.1.7 Individual Differences in the Need to Evaluate Predict Advocacy

Recent research also suggests that individual differences in the need to evaluate may impact people’s tendency to advocate. Specifically, Xu et al. (Reference Xu, Petty, Wright and Briñol2021) revisited the classic construct of Need to Evaluate, a trait measure that captures people’s likelihood of engaging in evaluation and possessing attitudes (Jarvis & Petty, Reference Jarvis and Petty1996). Xu and colleagues argued that the original Need to Evaluate scale was focused on intrapersonal aspects of evaluation and proposed two new interpersonal dimensions to complement the original scale. The first reflects the motive to express or share one’s evaluations with others and the second reflects the motive to learn about the opinions of others. Xu et al. found that people who scored highly on the motive to express their evaluations appeared to engage in more advocacy-relevant behavior. For example, people high in the motive to express their evaluations were more likely to seek roles that allowed them to share their opinions than were those low in this motive. Thus, there are individual differences in the tendency to express attitudes, and these may map onto diverse advocacy behaviors.

1.2.1 What Do People Say When They Advocate?

First, what do people do or say when they advocate on behalf of their own position? Research has begun to answer this question by investigating the factors that affect how people advocate – that is, by exploring some of the variables that shape the messages people generate when they advocate for their views. Here, we highlight some initial findings in this domain.

As a starting point, research has moved beyond advocacy intentions to assess the messages people generate when they feel certain or uncertain of their attitudes. For example, Cheatham and Tormala (Reference Cheatham and Tormala2017) examined the relationship between attitude certainty and message length. They found a curvilinear relationship, such that both high and low certainty individuals wrote longer messages on behalf of their attitudes than did moderate certainty individuals. Interestingly, though, they found that the content of the messages written by high versus low certainty individuals was markedly different. Messages crafted by people with high certainty contained more arguments, more emotional language, more moral content, and had a more judgmental tone. In contrast, messages crafted by people with low certainty included more qualifiers, hedges, and questions, and expressed more interest in understanding the opposing side of the issue at hand. In short, high certainty individuals’ messages were more forceful, whereas low certainty individuals’ messages tended to be more tentative and inquisitive and reflected a desire to seek information from others (for related findings, see Cutright et al., Reference Cutright, Wu, Banfield, Kay and Fitzsimons2011).

In addition to certainty, past research suggests that feelings of power influence not only people’s willingness to advocate (e.g., share their opinions; Anderson & Berdahl, Reference Anderson and Berdahl2002), but also the content of their advocacy messages. Dubois, Rucker, and Galinsky (Reference Dubois, Rucker and Galinsky2016) assigned participants to experience high or low power and asked them to generate persuasive messages. They found that high power participants tended to create messages that conveyed competence, while low power participants tended to craft messages that conveyed warmth. These authors argued that feelings of high (low) power tilt people’s thinking towards competence (warmth) and these competence- and warmth-related thoughts, in turn, influence the content of their messages.

Also relevant, Nguyen et al. (Reference Nguyen, Wang, Li and Cotte2021) investigated the effect of expertise on message content. They documented a tendency for experts to refrain from extremes. That is, compared to novices, experts were less likely to take extreme positions (e.g., assign an extremely positive or extremely negative star rating) and their advocacy messages were less likely to contain extreme sentiment (e.g., extremely positive or negative words). Nguyen et al. argued that experts’ restraint from extremes was driven by the number of attributes they considered. Basically, experts considered more attributes than did novices, making it unlikely that the product or service reviewed would score consistently highly or consistently poorly. As a result, experts ended up giving less extreme evaluations and recommendations.

Relatedly, Rocklage, Rucker, and Nordgren (Reference Rocklage, Rucker and Nordgren2021) found that experts, compared to novices, tend to be more “emotionally numb.” For example, Rocklage, Rucker, and Nordgren compared film reviews written by film critics (experts) to those written by everyday consumers (novices). The authors found that reviews written by experts were less emotional than those written by novices, even after controlling for other factors such as the valence and extremity of the language used. Rocklage, Rucker, and Nordgren argued that experts develop cognitive structures that are more analytical in nature and that relying on those structures shifts their focus away from emotions, thus rending the content of their messages less emotional. Together, these recent findings suggest that expertise can influence the extremity and emotionality of people’s advocacy messages.

Recent research has also begun to investigate the influence of modality on message content. For example, Berger, Rocklage, and Packard (Reference Berger, Rocklage and Packard2022) found that people generally express less emotional attitudes when writing rather than speaking their messages. In one experiment, participants were asked to share their attitudes toward their favorite restaurant with a close friend either in writing or by recording a voice message. Results indicated that participants included less emotional content in their messages when they wrote rather than spoke them. Berger and colleagues argued that writing reduces emotionality because it is more premeditated and allows for greater deliberation.

In a related line of research, Melumad, Inman, and Pham (Reference Melumad, Inman and Pham2019) found that the device people use when writing a message (i.e., writing a message using a laptop computer or a smartphone) can influence the content of the message. Using a dataset of more than 60,000 customer-generated restaurant reviews from TripAdvisor.com, the authors found that messages written from a smartphone were more emotional than messages written from a laptop. Melumad, Inman, and Pham replicated these results using online and lab experiments and argued that because smartphones are more physically constrained than laptops, smartphones encourage people to focus on the gist of their experiences, which leads to greater emotionality of their messages.

Recent work has also examined the impact of incentives. Perhaps unsurprisingly, incentivizing advocacy generally increases people’s willingness to advocate, but does it alter the content of their advocacy messages? Woolley and Sharif (Reference Woolley and Sharif2021) examined this question and found that incentivizing customer reviews (both financially and nonfinancially) increased the proportion of positive relative to negative emotion expressed in those reviews. The authors argued that this enhanced positivity was the result of affect transfer, whereby positive affect from receiving the incentive was transferred to the experience of writing the review, making it more enjoyable and resulting in a more positive sentiment in the actual review content. In a similar vein, Rocklage, Rucker, and Nordgren (Reference Rocklage, Rucker and Nordgren2018) found that participants used more emotional language when incentivized to write a review compared to when asked to write a review without incentive.

As an aside, in addition to investigating factors that influence the content of people’s messages, past research has also examined how the content of a message influences sharing intentions. In particular, research on word-of-mouth – informal communications about the ownership, usage, or characteristics of particular goods and services (Westbrook, Reference Westbrook1987) – has shed light on how the content of a message affects the likelihood that people will share it. This work has revealed that people are more likely to share content that is entertaining, useful, self-concept relevant, emotional, and unique (for a review, see Berger, Reference Berger2014). As one example, the greater the emotional intensity of movies and news articles, the more likely they were found to be discussed and shared (Berger & Milkman, Reference Berger and Milkman2012; Luminet et al., Reference Luminet, Bouts, Delie, Manstead and Rimé2000). As a more recent example, Rathje, Van Bavel, and van der Linden (Reference Rathje, Van Bavel and van der Linden2021) found that the ingroup–outgroup focus of a message influences its sharing likelihood. These authors examined over 2.5 million posts on Facebook and Twitter and found that posts about the outgroup were shared or retweeted twice as often as posts about the ingroup. Indeed, language mentioning the outgroup was found to be the strongest predictor of virality on social media, above and beyond other established predictors such as negative affect and moral–emotional language.

In summary, an active and growing area within advocacy research focuses on charting the factors that shape the content of people’s advocacy messages. Given the rapid development and deployment of natural language processing tools (e.g., Eichstaedt et al., Reference Eichstaedt, Kern and Yaden2021), we see this area as ripe for further research. In particular, while some of the antecedents outlined at the start of this chapter have already been shown to impact how people advocate (e.g., certainty; Cheatham & Tormala, Reference Cheatham and Tormala2017), other antecedents have yet to be examined. For instance, how does the perceived affective or cognitive basis of people’s attitudes influence the way they advocate, or what they actually say in their messages? What about perceived efficacy: Do people not only advocate more, but also advocate differently as their feelings of efficacy rise? We hope future research will revisit the known antecedents of advocacy to determine how those factors might also affect the specific manner in which people advocate.

1.2.2 What Other Actions Do People Undertake to Advance Their Views?

Related to the question of how people advocate, recent work has examined other actions people engage in to advance their views or positions that have yet to be linked to the literature on attitudinal advocacy. First, some people advocate by boycotting brands – that is, by refusing to use or purchase brands that express views or opinions contrary to their own. Consumers might refuse to eat at a particular restaurant chain, for instance, because it endorses a religious or political ideology at odds with theirs. Second, some people advocate by censoring counterattitudinal perspectives and information. With the proliferation of social media and online discussion forums with user moderators, individuals increasingly find themselves in a position to censor – that is, block or remove – opinions and information that support the opposing side of a particular issue. For instance, the moderator of an anti-vaccination discussion forum online might remove comments citing the efficacy of vaccines as a way to advance their own position or agenda. Likewise, everyday social media users might delete comments in response to their own posts if they disagree with those comments or find them objectionable.

We submit that boycotting and censorship are relevant to our understanding of attitudinal advocacy in that they constitute attitude-relevant actions that people undertake to oppose one position and, thus, support another. Ultimately, when people boycott or censor entities or content that they find objectionable, they are expressing their views and arguably seeking to influence (or prevent influence on) others’ attitudes or behaviors. The target might vary – for instance, one might boycott a brand to compel a company to change its ways or censor a comment to prevent that comment from influencing readers’ opinions – but at their core both boycotting and censorship can be used as advocacy mechanisms. Nevertheless, to our knowledge, neither has been studied through the lens of attitudinal advocacy. In this section, we review some of the relevant work on boycotting and censorship. In each case, we highlight the potential connection to attitudinal advocacy, discuss relevant antecedents, and note potential future directions. Our aim is not to be exhaustive in our coverage of these areas, but rather to highlight links to attitudinal advocacy and call for more research on these topics.

First, consider boycotting. What triggers it? Like other forms of advocacy, research on boycotting suggests that the perceived efficacy of a boycott is a major factor driving people’s willingness to participate in it. For instance, Sen, Gürhan-Canli, and Morwitz (Reference Sen, Gürhan-Canli and Morwitz2001) found that the more consumers believed a boycott had a chance of succeeding (i.e., of changing a brand’s contested behavior), the more likely they were to engage in the boycott. Klein, Smith, and John (Reference Klein, Smith and John2004) observed a similar result and additionally documented that the more egregious consumers found the action committed by the brand to be, the more likely they were to participate in the boycott. Other research has investigated how individual-level differences, such as ideology, impact the tendency to engage in boycotts. Jost, Langer, and Singh (Reference Jost, Langer and Singh2017) found that liberal consumers were more likely than conservative consumers to have boycotted products, controlling for other demographics like gender, age, income, education, and party identification.

Censorship offers another means of advocating on behalf of one’s attitudes. Indeed, removing information and opinions that support the opposing side of an issue can act as an expression of one’s attitude and may constitute an attempt to affect others’ views. Research on censorship has identified some of its key antecedents, including attitude position, attitude strength, political ideology, and various personality dimensions. First, attitude position matters. Perhaps unsurprisingly, people are more likely to censor counterattitudinal content than proattitudinal content. For example, Boch (Reference Boch2020) examined the extent to which liberals and conservatives supported banning an extreme speaker from giving a talk at a university campus. Boch found that liberals (conservatives) were more likely to ban the speaker when the speaker advocated for ideas from the far right (left) compared to ideas from the far left (right). In other words, people were more likely to censor counter-rather than proattitudinal speakers.

Ashokkumar et al. (Reference Ashokkumar, Talaifar and Fraser2020) found similar results in the context of online censorship. These authors developed a paradigm in which participants were asked to imagine moderating an online discussion forum (akin to a Facebook Group). Participants were then shown comments that expressed support for or opposition to a social issue (e.g., gun control). Consistent with the Boch (Reference Boch2020) results, Ashokkumar and colleagues found that participants were more likely to censor counter-rather than proattitudinal posts. In addition to attitude position, Ashokkumar et al. (Reference Ashokkumar, Talaifar and Fraser2020) investigated the role of attitude strength in predicting censorship decisions. They found that the stronger people’s attitudes were – for example, the more certain participants reported feeling or the more important they found their attitudes to be – the more likely they were to censor counter-compared to proattitudinal content. Thus, both attitude position and attitude strength appear to play a role in censorship decisions.

In addition to examining the factors that predict specific censorship decisions, researchers have assessed people’s support for censorship as a general enterprise. In this work, political ideology has proven to be important, though findings regarding the exact nature of the relationship are somewhat mixed. Some research has shown that conservatives are more likely than liberals to support censorship (Fisher et al., Reference Fisher, Lilie and Evans1999; Hense & Wright, Reference Hense and Wright1992), whereas other research has found that conservatives and liberals are equally likely to support censorship (Suedfeld, Steel, & Schmidt, Reference Suedfeld, Steel and Schmidt1994). More recently, Boch (Reference Boch2020) found that, contrary to prior work, conservatives are somewhat more tolerant of extreme left-wing speakers than liberals are of extreme right-wing speakers. It seems that the relationship between ideology and support for censorship is complex and warrants further investigation.

Finally, past research also investigated how some of the big-five personality dimensions relate to censorship. Specifically, Lambe (Reference Lambe2008) investigated how neuroticism, openness to experience, and extraversion were associated with support for government censorship. For example, Lambe measured the extent to which participants supported or opposed city officials’ (hypothetical) decision to prevent a weekly Neo-Nazi call-in program from appearing on TV. Lambe found that openness to experience negatively predicted support for censorship, whereas neuroticism positively predicted support for censorship. There was no relationship between extraversion and support for censorship.

In addition to boycotting and censorship, other forms of advocacy appear to be on the rise as well. For example, when Texas passed an anti-abortion law and set up a website for anonymous tips on violations in 2021, pro-choice activists flooded the website with fake tips, making it difficult to discern which tips were genuine and which were fabricated (Perlroth, Reference Perlroth2021). Although the activists’ actions do not reflect attitude expression or persuasion efforts as we might traditionally define them, those actions were driven by advocacy-like goals and certainly reflected deeply held attitudes and values. Thus, these actions seem advocacy-relevant and offer yet another example of attitude-relevant action that could be studied through the lens of attitudinal advocacy.

We see these and other new forms of advocacy as exciting directions for future research in this area. In particular, linking the antecedents discussed in this review to new advocacy behaviors such as boycotting and censorship would be a useful next step. Which dimensions of attitude strength predict boycotting behavior? Would people with affective or cognitive attitudes be more likely to censor? How does support–oppose attitude framing affect these outcomes? As noted, people are more likely to express their attitudes when they frame those attitudes in terms of support rather than opposition (Catapano & Tormala, Reference Bechler and Tormala2021). Would the opposite be true for boycotting and censorship given their inherently oppositional focus? In addition to these questions, contrasting boycotting and censorship with other advocacy actions, such as merely expressing one’s views or writing a persuasive message, might yield important insights into the nature of these behaviors. Specifically, identifying factors that uniquely predict some but not other forms of advocacy could be an important direction for future work in this domain. This is a rich area of research with many open questions.