Rationales and Support for Norms in the Context of Covid-19

Abstract

This study empirically tests whether people invoke moral and prudential rationales when evaluating behavior in a novel context—the early months of the Covid-19 pandemic in the United States—and whether those rationales are associated with their support for a norm. We use data from two online vignette experiments that describe key health behaviors—staying home and masking—and find substantial support. Given the politicization of these behaviors in the U.S. context, we also explore the role of political orientation and find that liberal participants react more strongly to the behaviors.

Keywords

moral norms political orientation prudential rationales

The emergence and spread of Covid-19 in the United States in spring and summer 2020 created a novel circumstance for most Americans, who had little understanding of the disease and little experience with pandemics. Just as policy makers were struggling to understand the virus, its consequences, and appropriate public health measures, so too were members of the public navigating how they should behave and what they should expect from others (Sorokin 1968:157). Guidance was unclear, with frequent changes and contradictions (e.g., López 2020; Tufekci 2020). Thus, the new threat associated with Covid-19 provided a unique opportunity to study norms, in particular, support for emerging norms around health behaviors.

Norms are rules about behavior that are socially enforced (Horne 2001a). When a norm is in place, violations are punished and people expect that others will react negatively to noncompliant behavior. Individuals who support a norm will have attitudes that are consistent with that norm. Here we focus on support for norms—that is, individuals’ own approval or disapproval of a behavior—and the rationales we argue are associated with those reactions. Rationales explain why supporting or adhering to a particular norm makes sense; they provide a reason to support the norm. Two types of rationales—prudential and moral—have been identified (e.g., Mollborn 2017). Because rationales justify norms, they may strengthen those norms and increase adherence and may also contribute to norm emergence, change, and erosion (e.g., Horne and Mollborn 2020). Understanding rationales is therefore potentially important for answering fundamental questions about norms. Despite this, there is little theoretical development or empirical research articulating the connections between rationales and norms.

This research is an initial step to simultaneously assess the roles of prudential and moral rationales in individual support for norms. We examine (1) whether people invoke prudential and/or moral rationales in response to a behavior and (2) whether one or both of those rationales logically underlie individuals’ disapproval of that behavior. We do not make an argument regarding the primacy of prudential or moral rationales. Rather, we examine whether people express these rationales in response to behaviors and whether rationales are associated with approval/disapproval.

We focus on a novel naturally occurring setting—the Covid-19 pandemic in late April 2020. We expect that when an individual engages in a Covid-related behavior, people will assess that behavior in prudential and moral terms (Fiske, Cuddy, and Glick 2006; Mollborn 2017). Those prudential and moral rationales will be associated with and help to explain people’s disapproval of the behavior. We test this argument using two online vignette experiments that describe two health-related behaviors—staying home and masking—and find substantial support.¹ Given the politicization of these behaviors (Horne and Johnson 2021), we also explore potential differences across political orientation of participants. Our findings provide a foundation for future work on the role of rationales in norm emergence, persistence, and erosion.

Theory and Hypotheses

Norms are rules that regulate behavior and are enforced through social sanctions (Coleman 1990; Feinberg, Willer, and Schultz 2014; Horne 2001a). They can encourage or forbid, permit or require, approve or disapprove. When a norm is in place, people sanction violations and expect others to do so as well. Such normative expectations provide an indicator of norms (Horne and Mollborn 2020). Norms are distinct from laws, which are rules enforced by the state. Laws may affect norms, and norms may affect laws (e.g., McAdams 2015). At any given time, laws and norms may be consistent with or contradictory to each other. Individual opinions are distinct from both norms and laws. They may support or contradict a norm (or law), and the strength of their support or disagreement can vary (e.g., Cowan 2014; Cowan and Baldassarri 2018). When individuals personally support a norm, their own reactions toward a behavior will be consistent with their normative expectations about others’ likely reactions to that behavior. Accordingly, explaining individual disapproval can enhance understanding of norms.

There are two main approaches to predicting norms—one that focuses on what is good for the individual and another that focuses on what is good for the group (Horne 2001b; Horne and Mollborn 2020). Research on what is good for the individual includes work showing that people look to others’ behaviors to determine the best course of action (Cialdini 2007; Schelling 1978). People favor behaviors that they think will be beneficial to them. And because they tend to assume that what most others do is likely to have a positive outcome, they view engaging in those behaviors as “wise” or smart (Cialdini 2007:264). For example, it is smart for an individual to drive on the right side of the road because doing so will help them avoid an accident (for related discussion of conventions, see, e.g., Ullmann-Margalit 1977). In turn, those behaviors become normative such that a sense of oughtness attaches to them (Horne and Przepiorka 2021; Opp 2004).

A second body of work focuses on situations in which what is good for the individual conflicts with what is good for the group. In this research, self-interested behavior harms the group (Coleman 1990; Heckathorn 1989). For example, in a social dilemma, self-interested individuals will want to keep their resources for themselves rather than contribute to the group, but if everyone behaves this way, all suffer. Norms are thought to emerge to constrain such self-interested (harmful) behavior and encourage behavior that benefits the collective (Coleman 1990; Ullman-Margalit 1977; see also research on morals and harm; Gray, Schein, and Ward 2014; Schein and Gray 2018).

The behavioral mechanisms emphasized by each approach—individual foolishness/smartness and self-interest/collective interest—are widely recognized as key components of evaluation generally.² For example, work on perception in psychology identifies a competence dimension (that includes traits like intelligence, cleverness, competence, and knowledge) and a moral or intent dimension (that includes traits like helpfulness and generosity; see, e.g., Fiske et al. 2006; Wojciszke 1994; Wojciszke and Klusek 1996). Sociological research on status identifies competence and cooperation (commitment to the group) as relevant for status determinations (e.g., Correll, Benard, and Paik 2007). And trust studies highlight the ability and intent components of trust (e.g., Cook 2005).

In the norms literature, research suggests that people invoke rationales that map onto these two domains (smartness vs. foolishness and self-interest vs. collective interest) to explain their support for norms (Mollborn 2017; Mollborn and Sennott 2015). Rationales that capture the smartness/foolishness dimension are referred to as prudential, and those that capture the self-interest/collective interest dimension are referred to as moral. Rationales are reasons that explain or justify why a norm makes sense. Whether or not they cause an individual to support a norm, they provide an explanation or justification for doing so (for a related discussion of motivating and explanatory reasons, see Alvarez 2017).

Rationales about Staying Home

In spring 2020, many U.S. state governors asked their citizens to stay home and refrain from engaging in nonessential face-to-face activities to reduce the spread of disease, “flatten the curve,” and prevent hospitals from becoming overwhelmed (e.g., Haelle 2020; Stevens 2020). Norms at the time (April 2020) supported staying home and discouraged socializing (Horne and Johnson 2021). Measures limiting social interaction explicitly aimed to protect both individual and public health. Because people could protect public health by staying home, individual decisions about staying home, working, and socializing had implications not just for them but also for their community. People could help to limit the spread of the disease by limiting face-to-face contact. But there were also competing considerations, including the ability of a household to meet its financial needs (Bazelon 2020; Kochhar 2020; Lee 2020; Tankersley 2020). People who could not work remotely needed to work face-to-face to support their families.

Thus, during the April 2020 period, staying home might be both smart and unselfish because it protected the individual and others, whereas attending large social gatherings might be foolish and self-interested because it potentially harmed the individual and others. Working might be smart (or at least smarter than socializing because doing so had financial benefits). But it might also be seen as self-interested because it benefitted the individual/family at the expense of potentially contributing to community spread, thus hurting the collective. Assessments of the worker’s selfishness might be tempered by a sense that not going to work may be excessively costly for an individual (Horne, Dodoo, and Dodoo 2018). Thus, failure to stay home from work may be seen as less self-interested than socializing. Accordingly, we predict that:

Hypothesis 1a: People will view those who attend large parties as more foolish and self-interested than those who stay home and view those who work as somewhere in between.

Rationales about Masking

At the time, there was also advice from the Centers for Disease Control (CDC) regarding mask wearing. Early in the pandemic, people were discouraged from wearing masks and told they were ineffective. Then they were told that they should wear masks and that doing so benefitted others, but not the wearer. On April 3, 2020, the CDC issued a recommendation that people wear masks to protect both themselves and others (Dwyer and Aubrey 2020; Pappas 2020). People were discouraged from purchasing N95 masks to reserve them for health care workers, who needed higher levels of protection while treating patients with Covid (e.g., Achenbach, Sun, and McGinley 2020; Jacobs, Richtel, and Baker 2020). The recommendation to wear face coverings (but not N95s) to protect the wearer and the community was clear. Because disposable masks were not yet widely available, masking largely meant wearing cloth face coverings (homemade cloth masks, scarves, and bandanas).

Wearing a face covering could be seen as smart (because individuals are protecting themselves from illness) and unselfish (because they are reducing spread of the disease). But wearing an N95 might be seen as particularly smart because it is the most effective way to protect oneself (although at this point in the pandemic the superiority of N95 masks, although logically clear from the effort to reserve them for health care workers, may not have been widely recognized). Wearing an N95 mask during a shortage is also likely to be seen as selfish because wearing one deprives health care workers, who are at greater risk and need more protection. Accordingly, we predict that:

Hypothesis 1b: People will view those who do not mask as more foolish and self-interested than those who wear a mask and will see those who wear an N95 mask during a shortage as less foolish and more self-interested than those who wear cloth.

Rationales and Disapproval

These prudential and moral rationales provide people with an explanation for why it makes sense to approve or disapprove of a behavior. That is, they provide a reason to disapprove. This means that stronger rationales are associated with stronger disapproval. Note that rationales do not necessarily causally affect disapproval; they may, but do not necessarily, occur earlier in time than disapproval. An individual might believe that a behavior is selfish and then, in turn, begin to disapprove of that behavior. Or they might disapprove of a behavior and then identify reasons (rationales) for doing so. Rationales and disapproval (and other cognitions and emotions) might occur within a split second of each other or even simultaneously. Understanding potential causal relations between cognitions involves neurobiological dynamics that are beyond our capacity to investigate and are not our primary concern.

Rather, we suggest that rationales are logically essential for disapproval (for discussions of the distinction between motivating and explanatory reasons, see Alvarez 2009, 2017). For people to disapprove, they must have a reason for doing so. Thus, for example, it does not make sense to say, “I disapprove, therefore that behavior is selfish.” Disapproving of a behavior does not provide a legitimate reason for seeing it as selfish. But everyone would understand someone saying, “That behavior is selfish, therefore I disapprove.” The selfishness of the behavior provides a reason that explains or justifies why it makes sense to disapprove of it. The rationale logically (although not necessarily causally) explains the effects of a behavior on others’ reactions to it. Accordingly, we predict that:

Hypothesis 2: Prudential (foolishness) and moral (self-interest) rationales will be associated with disapproval and will at least partially explain the effects of behaviors on disapproval.

Political Orientation

In the case of the Covid-19 pandemic, a key dimension of variation in people’s reactions to Covid-related behaviors has been their political orientation. In the early days of the pandemic, President Trump, prominent Republicans, and conservative media downplayed the seriousness of the virus (e.g., Halon 2020; Mitchell et al. 2020; Steib 2020; Wolfe and Dale 2020). In contrast, prominent liberals and public health experts emphasized the risks (e.g., Chappell 2020). Consistent with this messaging, research finds that conservatives were less concerned about the virus than liberals (Calvillo et al. 2020; for additional research on the effects of political actors on public opinion, see, e.g., Mason 2018; Merkley and Stecula 2021). Our research similarly shows that in spring 2020, norms regulating Covid-related behaviors varied with political orientation (Horne and Johnson 2021).

To the extent that liberals saw the virus as posing greater risks, it is likely that they would have stronger reactions to people who violated public health recommendations and mandates. In addition, although we expect rationales in general to operate in similar ways for liberals and conservatives (in that they provide a reason for disapproval), it is possible that the specific (prudential and moral) rationales that liberals and conservatives attach to behavior may vary. For example, some scholars argue that liberals and conservatives have different moral intuitions (see, e.g., Graham, Haidt, and Nosek 2009). Others disagree, however, claiming that both liberals and conservatives care about harm (e.g., Schein and Grey 2015). Because we have little a priori reason to make specific predictions about differences in prudential and moral rationales across political orientation and because existing literature is inconsistent, we do not propose hypotheses regarding the relative strength of rationales used by liberals and conservatives. Instead, we empirically explore their role. We expect that even as liberals and conservatives may rely on different rationales, those rationales will, in general, support stronger disapproval among liberals than conservatives.

Hypothesis 3: We expect the effects of working and socializing (vs. staying home) and not masking and wearing an N95 during a shortage (vs. wearing cloth) on disapproval to be stronger for liberal than conservative participants.

In sum, we argue that people will frame behavior in prudential and moral terms, that is, in terms of its foolishness and selfishness. For most of the behaviors discussed previously, the same action is likely to invoke both kinds of rationales working in the same direction. But for working and for wearing an N95 mask during a shortage, they may not. Going to work during a pandemic and wearing an N95 mask during a shortage may be seen as smart but selfish. Otherwise, we expect that most of the behaviors we describe will be seen as both foolish and self-interested or smart and caring about the collective good. Our experiments are designed to test people’s reactions to behavior—how prudential or moral they perceive the behavior to be and whether those rationales are associated with disapproval.

Data

We conducted two vignette experiments that allowed us to manipulate the causal factors (health-related behaviors) and measure rationales and disapproval. As discussed earlier, our study does not attempt to disentangle the causal processes linking rationales and disapproval—evaluations of behavior as moral and prudential behavior may foster (dis)approval, (dis)approval may be justified by drawing on notions of what is moral and prudent, and rationales and (dis)approval could even be triggered simultaneously by seeing a behavior. Thus, our analyses are not intended to establish causal order between rationales and disapproval. Instead, they test the argument that rationales provide reasons for disapproval and therefore help to explain disapproval—rationales are logically necessary for disapproval (even as they need not be causally prior). We first look at the rationales that are triggered by the Covid-related behavior described in the experimental conditions. We then examine the association between rationales and disapproval by estimating regression models that show the effect of the experimental conditions on disapproval and assess whether controls for the rationales account for some of those effects. While mindful that this approach inherently assigns rationales and disapproval to being “independent” and “dependent” variables, respectively, we do not intend for this assignment to suggest a causal order. Instead, this approach assesses the extent to which the association between the behaviors and disapproval are tied to their mutual association with rationales. Finally, we include interaction terms to examine the extent to which the political orientation of participants shapes the effects of the experimental manipulation.

We collected our data on April 28, 2020, at a time when people in many states had experienced stay-at-home orders, some of which were beginning to ease, and when distancing and the use of masks or face coverings was being recommended by public health officials. One study focused on stay-home behaviors (a subject of many states’ mandates). The other examined mask wearing (when N95 masks were in short supply for health care providers and for Americans generally).

We conducted both experiments through Prolific, a site much like Amazon’s Mechanical Turk, but designed specifically for academic researchers (Palan and Schitter 2018; Peer et al. 2017). We limited participation to U.S. residents 18 and older. Individuals who chose to participate in our studies clicked on the study link and were taken to the study site. There, they read the vignettes and then answered questions about them. Participants completed the two experiments one after the other. Within each experiment, they were randomly assigned such that the conditions across studies were not correlated. Once they were done, they were automatically directed back to the Prolific site for payment ($2.00 for less than 10 minutes of engagement).

The sample was almost 70 percent non-Hispanic white and 46 percent female. Participants ranged in age from 18 to 77 (mean = 34.6), and the average education level, on a scale from 1 = less than high school to 8 = professional or doctorate degree, was 5.1 (equivalent to having some college). We measured political orientation by asking participants, “In general, how would you describe your political views” on a 1 to 7 scale (consistent with the American National Election Survey [ANES]). We reverse-coded the ANES response categories so that higher values reflect greater liberalism (1 = extremely conservative, 7 = extremely liberal). The mean value of liberalism in the sample was 4.7.

Experiment 1: Staying Home

Study 1 tested whether behaviors related to staying home affect prudential and moral rationales and whether those rationales help to explain approval or disapproval. It had three between-subjects conditions—stay home, go to work, or socialize. There were approximately 200 participants per condition for a total of 596.

Vignette conditions

Our vignette described a situation in which the state government had issued a stay-home order and in which going to work and socializing violated that order. For the work condition, we specifically described the individual as opening up their own t-shirt shop—to indicate that the business was not essential and that the individual was not being forced by an employer to go to work. For the socializing condition, we described the individual as going to large parties—to make clear the high level of risk. Each vignette described one of these behaviors; participants read only one version of the vignette:

Assume that the governor of your state has issued a stay-at-home order and banned gatherings of more than 10 people. Certain businesses and services considered “essential” may remain open, but otherwise everyone has been ordered to stay at home unless grocery shopping or going to the doctor or pharmacy.

Someone in your community is [working from home and avoiding gatherings they would normally attend/going to work regularly to open up their t-shirt shop/attending large parties regularly].

Measures

We measured prudential and moral rationales and participants’ disapproval of the individual in the vignette. To assess prudential (foolishness) rationales, we asked participants how intelligent they thought the individual was (0 = not at all intelligent, 10 = very intelligent). We reverse-coded this measure so that higher numbers represent being more unintelligent. To assess moral (self-interest) rationales, we asked how self-centered they thought the individual was (0 = not at all self-centered, 10 = very self-centered). We measured participants’ support for stay-home norms by asking them how much they approved or disapproved of the behavior of the individual in the vignette (0 = strongly approve, 10 = strongly disapprove).³

We included a manipulation check to assess whether participants noticed the manipulation. The manipulation check asked what the person in the vignette was doing—staying home, going to work, or attending parties. About 2 percent of participants failed the manipulation check in each condition. We report the results excluding these cases (N = 584). Results for the full sample (N = 596) were substantively similar.

Results

Table 1 shows mean responses across the experimental conditions: the actor who attended large parties (socialized) was rated as the most unintelligent and self-centered, the actor who stayed home was rated as least so, and the actor who went to work fell in between. Participants also disapproved the least of staying home, disapproved more of going to work, and disapproved the most of attending large parties.

Table 1.

Means by Experimental Condition, Study 1

	Stay home (N = 189)	Work (N = 196)	Socialize (N = 199)
	Mean (SD)	Mean (SD)	Mean (SD)
Unintelligent	1.2 (1.7)	6.9 (2.6)	8.4 (2.1)
Self-centered	2.4 (2.8)	7.7 (2.4)	8.9 (2.0)
Disapproval	.7 (1.6)	7.3 (3.1)	8.2 (3.0)

To test our hypotheses and the significance of these patterns, we estimated a series of ordinary least squares regression models. Hypothesis 1a predicts that people will view actors who fail to stay home as more foolish and self-interested. In Table 2, Models 1 and 2, we present the unstandardized coefficients testing the effects of the experimental conditions on how unintelligent and self-centered the actor was perceived to be. Respondents presented with an actor who went to work and especially those who were presented with an actor who attended large parties (socialized) viewed that actor as more unintelligent and self-centered than those in the stay-home condition (the reference category). The magnitude of these effects is very large (with differences in means representing 1.5 and 1.9 SD for working and socializing, respectively, on the unintelligence scale, and 1.4 and 1.8 SD for working and socializing, respectively, on the self-centered scale).

Table 2.

Ordinary Least Squares Regression Models for Rationales and Support for Norms (Unstandardized Estimates), Study 1

	Rationales		Support for norms
	Unintelligent	Self-centered	Disapproval
	Model 1	Model 2	Model 3	Model 4
Work^a	5.76*** (.22)	5.30*** (.25)	6.63*** (.27)	2.45*** (.35)
Socialize^a	7.24*** (.22)	6.50*** (.25)	7.56*** (.27)	2.32*** (.40)
Unintelligent				.64*** (.05)
Self-centered				.10* (.04)
Constant	1.17 (.16)	2.36 (.18)	.66 (.19)	−.31 (.18)
R ²	.68	.57	.62	.74

Note: N = 582. Standard errors in parentheses.

Reference condition is stay home.

p < .05. ***p < .001 (two-tailed tests).

Hypothesis 1a also predicts that going to work will not be seen as quite as foolish and self-interested as socializing. Post hoc tests confirmed that the coefficients for socialize for each outcome are statistically significantly larger than for work. The magnitude of this difference is smaller than the differences between staying home and going out, however (representing .39 SD in the unintelligence scale and .32 SD in the self-centeredness scale)—still statistically different and meaningful, but much smaller.

Hypothesis 2 predicts that perceptions of foolishness and self-interest will be positively associated with disapproval. It also predicts that rationales will at least partially account for the association between behaviors and disapproval. Correlations between perceptions of the actor as unintelligent and self-centered and disapproval are strong and positive (.84 between unintelligent and disapproval and .71 between self-centered and disapproval). When participants saw the actor as lacking in intelligence or more self-centered, they disapproved more strongly (and vice versa).

We present unstandardized coefficients of two regression models predicting the effects of the experimental conditions on disapproval in Table 2, first without controlling for ratings of the actor’s unintelligence and self-centeredness (Model 3) and then with them (Model 4). Compared to the actor who stayed home (the reference category), participants disapproved of the actor who went to work and even more so the actor who socialized. The magnitude of these differences in disapproval is large (with differences in means representing 1.6 SD for working vs. staying home and 1.8 SD for socializing vs. staying home); the difference between working and socializing was smaller (.21 SD). Importantly, the magnitude of the coefficients was dramatically reduced (although not eliminated) when ratings of unintelligence and self-centeredness were introduced in Model 4. Thus, perceptions of the actor as unintelligent and self-centered help to explain (provide a reason justifying) disapproval of the actor. Post hoc tests also revealed that when rationales were included in Model 4, the difference between the work and socialize coefficients was no longer statistically significant.

Hypothesis 3 predicts that compared to conservatives, liberals will more strongly support staying home norms; they will disapprove more of working and socializing than conservatives. Although we did not have reasons to hypothesize differences in the rationales used by liberals and conservatives, we explored possible differences in our sample. To test Hypothesis 3 and explore political orientation differences in rationales, we examined whether the effects of the experimental conditions on rationales and disapproval differed depending on participants’ political orientation. We present the estimates of the experimental conditions (with staying home left out as a reference group), participant liberalism, and their interaction on ratings of unintelligence (Model 1), ratings of self-centeredness (Model 2), and disapproval in Table 3. For disapproval, we include two models, first without controlling for ratings of unintelligence and self-centeredness (Model 3) and then controlling for them (Model 4).

Table 3.

Ordinary Least Squares Regression Models for Rationales and Support for Norms by Political Orientation (Unstandardized Estimates), Study 1

	Rationales		Support for norms
	Unintelligent	Self-centered	Disapproval
	Model 1	Model 2	Model 3	Model 4
Work^a	3.40*** (.65)	2.86*** (.73)	2.82*** (.79)	.51 (.69)
Socialize^a	5.96*** (.64)	4.08*** (.72)	6.38*** (.77)	2.40** (.71)
Work × liberalism	.49*** (.13)	.52*** (.14)	.79*** (.16)	.44** (.14)
Socialize × liberalism	.28* (.13)	.53*** (.14)	.26 (.16)	.04 (.13)
Liberalism	−.04 (.09)	−.22* (.10)	−.10 (.11)	−.06 (.10)
Unintelligent				.61*** (.05)
Self-centered				.09* (.04)
Constant	1.31 (.46)	3.32 (.52)	1.13 (.56)	.05 (.49)
R ²	.70	.60	.65	.74

Note: N = 575. Standard errors in parentheses.

Reference condition is stay home.

p < .05. **p < .01. ***p < .001(two-tailed tests).

Even the least liberal (most conservative) participants rated the actor who went to work and especially the actor that attended large parties as more unintelligent and self-centered than the actor who stayed home (see the work and socialize coefficients in Models 1 and 2, which represent the effects of these behaviors at the lowest values of liberalism). But these effects were stronger among more liberal participants (see the Work × Liberalism and Socialize × Liberalism coefficients, also in Models 1 and 2). That is, compared to conservative participants, liberals saw the noncompliant actors as both more unintelligent and self-centered.⁴

Both conservative and liberal participants disapproved of actors who went out, although not to the same degree for both behaviors. The most conservative participants disapproved more strongly of the actor who went to work, and especially the actor that attended large parties, compared to the actor who stayed home (see the work and socialize coefficients in Model 3, Table 3, which represent the effects of these behaviors at the lowest values of liberalism). Although the strong disapproval of the actor who socialized did not differ significantly across levels of liberalism (see the nonstatistically significant coefficient for Socialize × Liberalism), disapproval of the actor who worked did (see the statistically significant coefficient for Work × liberalism). The more liberal the participant, the more negatively they viewed the actor going to work (see the Work × Liberalism coefficient in Model 3, Table 3). The Work × Liberalism interaction term remained statistically significant when controls for ratings of unintelligence and self-centeredness were introduced in Model 4, although the coefficients for work and socialize were all substantially reduced in magnitude.⁵

To illustrate this interaction and aid in interpretation, Figure 1 shows predicted means in disapproval by political orientation (calculated from Model 3, Table 3). Among the most conservative participants, predicted mean disapproval followed a clear gradient from staying home to working to socializing. In contrast, for the most liberal participants, predicted mean disapproval was higher and about the same for working and socializing compared to saying home. In other words, more liberal participants distinguished less between working and socializing and more between working and staying home.

Figure 1.

Disapproval by Study 1 Condition and Participant Liberalism

These findings are partially consistent with Hypothesis 3. As expected, liberal participants were more disapproving of going to work compared to staying home than were conservatives. But liberals and conservatives alike were strongly disapproving of socializing compared to staying home. Our exploratory investigation of political orientation and rationales also indicates that compared to conservatives, liberal participants viewed people who violated staying home norms (both through going to work and through socializing) as more unintelligent and self-centered. These views substantially, but not wholly, account for the differences in how liberal and conservative participants viewed going to work in a nonessential job.⁶

Experiment 2: Masks

Study 2 tested how people reacted to individuals who wore an N95 mask compared to another face covering (in this case, a scarf) and to no face covering at all in conditions in which there was a shortage or no shortage of N95 masks for the local health care system. The study had a 3 × 2 between-subjects design, crossing masking (no mask vs. cloth vs. N95) by N95 shortages (shortage vs. no shortage). It had approximately 100 participants per experimental condition, with a total of 596.

Vignette conditions

At the time the study was conducted, cloth masks were not yet widely available for purchase and many people improvised. Therefore, our vignette specified that the actor wore either an N95, a scarf, or nothing.

Assume that your local hospital has announced that they have [adequate supplies/a shortage of] personal protective equipment including N95 and surgical masks for nurses, doctors, and other health professionals to wear while treating patients with the coronavirus.

During your recent trip to the grocery store, you see another customer who is [wearing an N95 (medical quality) mask/using a scarf as a mask/not wearing a mask or face covering].

Measures

Our measures of rationales and disapproval were the same as in Study 1.⁷ Again, we included a manipulation check. We asked what the actor was wearing and whether local hospitals were experiencing a shortage of personal protective equipment (PPE). Just over 4 percent of participants failed one or both questions. We report the results excluding these cases (N = 547). Results for the full sample (N = 596) were substantively similar.

Results

Table 4 shows mean responses across the experimental conditions. Participants rated the actor wearing no mask as most unintelligent on average in both the shortage and nonshortage conditions. The actor wearing an N95 mask was rated less intelligent (more unintelligent) than was the actor using a scarf, but only in the shortage condition. Mean ratings of self-centeredness were highest in the condition in which the actor wore no mask, regardless of whether it was the shortage or no shortage condition. The actor wearing an N95 mask was considered more self-centered than one wearing a scarf, particularly when there was a shortage of N95 masks in the local hospitals. Finally, mean disapproval was highest when the actor wore no mask, followed by the actor wearing an N95 mask; disapproval was lowest when the actor wore a scarf. The gap between wearing a scarf and an N95 mask was notably larger in the supply shortage condition than it was when supplies were plentiful.

Table 4.

Means by Experimental Condition, Study 2

	Shortage			No shortage
	No mask (N = 98)	N95 (N = 98)	Scarf (N = 107)	No mask (N = 68)	N95 (N = 71)	Scarf (N = 105)
	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)	Mean (SD)
Unintelligent	6.54 (2.25)	3.97 (2.47)	2.42 (1.88)	7.04 (2.22)	3.17 (2.27)	3.30 (2.28)
Self-centered	6.26 (2.65)	5.73 (2.51)	2.16 (2.22)	6.88 (2.67)	4.41 (2.75)	2.89 (2.48)
Disapproval	6.36 (2.40)	4.71 (2.64)	1.78 (1.79)	7.59 (1.94)	3.37 (2.72)	2.82 (2.49)

To test the significance of these differences and test our hypotheses, we estimated a series of ordinary least squares regression models. In Table 5, we present the unstandardized coefficients from models testing the effects of the experimental conditions on how unintelligent and self-centered the actor was perceived to be. To model the 3 × 2 design, we included face covering conditions (with scarf as a reference category) and supply shortage conditions (with no shortage as the reference) and their interaction (because the expectations around use of an N95 mask in particular has to do with whether there was a shortage in supply for front-line health care workers).

Table 5.

Ordinary Least Squares Regression Models for Rationales and Support for Norms (Unstandardized Estimates), Study 2

	Rationales		Support for norms
	Unintelligent	Self-centered	Disapproval
	Model 1	Model 2	Model 3	Model 4
No mask^a	3.74*** (.35)	4.00*** (.39)	4.77*** (.37)	1.52*** (.27)
No mask × shortage	.38 (.47)	.10 (.53)	−.19 (.49)	−.47 (.32)
N95 mask^a	−.14 (.34)	1.52*** (.39)	.55 (.36)	.39 (.24)
N95 mask × shortage	1.68*** (.46)	2.05*** (.53)	2.39*** (.49)	.88** (.32)
Shortage^b	−.88** (.31)	−.73* (.35)	−1.04** (.32)	−.31 (.21)
Unintelligent				.69*** (.03)
Self-centered				.17*** (.03)
Constant	3.30 (.22)	2.89 (.25)	2.82 (.23)	.05 (.18)
R ²	.37	.33	.42	.75

Note: N = 547. Standard errors in parentheses.

Reference condition is wearing a scarf.

Reference condition is no shortage of supplies.

p < .05. **p < .01. ***p < .001 (two-tailed tests).

Hypothesis 1b predicts that people will view actors who fail to mask as more foolish and self-interested; it also predicts that people will view actors who wear N95 masks during a shortage as smarter and more self-interested than actors who wear cloth face coverings. Consistent with Hypothesis 1b, participants viewed an actor wearing no mask as more unintelligent and self-centered than one wearing cloth (the reference group), and this was true regardless of whether there was a supply shortage in local hospitals (see the significant no mask and nonsignificant No Mask × Shortage coefficients in Models 1 and 2 in Table 5). Post hoc tests indicated participants also viewed the actor wearing no mask as more unintelligent and self-centered than one wearing an N95 mask. These represent large effects. For example, in shortage conditions, the difference in mean ratings of intelligence between not masking and wearing a scarf and not masking and wearing an N95 represent 1.5 and .9 SD on the scale. The difference in mean ratings of self-centeredness for these same comparisons represent 1.3 and .2 SD on the scale.

Views of the actor wearing an N95 mask compared to cloth varied by shortage conditions. Contrary to our prediction, participants viewed the actor wearing an N95 mask as less intelligent (more unintelligent) than the actor wearing cloth, but only in the shortage condition (see the nonsignificant effect of N95 and the significant effect of N95 × Shortage in Model 1). Consistent with our expectation, participants viewed the actor wearing an N95 mask in the shortage condition as more self-centered than the actor wearing cloth (see the significant N95 × Shortage coefficient). They also, unexpectedly, saw the actor even in the nonshortage condition as self-centered, although this effect was smaller (see the N95 coefficient in Model 2). The difference in mean ratings of self-centeredness between cloth and N95 in shortage conditions is again large, representing 1.2 SD on the scale.

Overall, these findings offer mixed support for Hypothesis 1b. For ratings of intelligence (Model 1), masked actors were seen as more intelligent than nonmasked actors, consistent with the hypothesis. However, contrary to our prediction, actors wearing N95 masks during a shortage were seen as more unintelligent than actors wearing cloth. For ratings of self-centeredness (Model 2), the findings are largely supportive of the hypothesis. The minor exception is that participants rated the actor wearing an N95 mask as more self-centered than one wearing a scarf even when there was no local shortage of such masks (although the effect was stronger in the shortage condition).

Hypothesis 2 predicts that perceptions of the actor as foolish and self-interested will be associated with disapproval and that these rationales will at least partially account for the association between the behaviors and disapproval. Correlations between perceptions of the actor as unintelligent and self-centered and disapproval are strong and positive (.83 between unintelligent and disapproval and .66 between self-centered and disapproval). When participants saw the actor as lacking in intelligence or more self-centered, they disapproved more strongly (and vice versa). We present unstandardized coefficients of two regression models predicting the effects of the experimental conditions on disapproval in Table 5, first without controlling for ratings of the actor’s unintelligence and self-centeredness (Model 3) and then with them (Model 4).

Participants disapproved much more strongly when an actor wore no mask than when the actor wore a scarf (the reference condition); this did not vary by shortage condition (see the no mask coefficient in Model 3). Post hoc tests show that they also disapproved more of wearing no mask than wearing an N95. In addition, participants disapproved more strongly when an actor wore an N95 mask compared to a scarf only when there was a shortage of such masks for local hospitals (see the nonsignificant effect of N95 and the statistically significant effect of N95 × Shortage in Model 3). These effects are generally quite large, with differences in mean disapproval representing 1.5 SD in the scale for no mask versus scarf, .54 SD for no mask versus N95, and .96 SD for scarf versus N95. The magnitude of these differences was substantially reduced (although not eliminated) after controlling for ratings of unintelligence and self-centeredness in Model 4. Thus, disapproval of the actor’s behavior and assessments of the actor as unintelligent and self-centered were closely tied.

Hypothesis 3 predicts that compared to conservatives, liberals will be more disapproving of nonmasking and wearing an N95 mask when supplies are low. We tested this hypothesis and also explored differences by the political orientation of our study participants in ratings of self-centeredness and intelligence across the experimental conditions.

In Table 6, Models 1 and 2, we present the estimates of the experimental conditions (with wearing a scarf and no shortage of supplies as the reference categories), participant liberalism, and their interaction on ratings of unintelligence and self-centeredness. (We also examined whether liberalism interacted with shortage conditions and whether there were three-way interactions between face covering, shortage, and liberalism. Because they were not statistically significant, we did not include them in the presented model.)

Table 6.

Ordinary Least Squares Regression Models for Rationales and Support for Norms by Political Orientation (Unstandardized Estimates), Study 2

	Rationales		Support for norms
	Unintelligent	Self-centered	Disapproval
	Model 1	Model 2	Model 3	Model 4
No mask^a	1.95* (.76)	.90 (.86)	2.16** (.80)	.67 (.53)
No mask × shortage	.39 (.47)	.18 (.53)	−.15 (.49)	−.44 (.32)
N95 mask^a	−1.82* (.75)	.06 (.85)	−1.84* (.79)	−.59 (.52)
N95 mask × shortage	1.71*** (.46)	2.09*** (.52)	2.36*** (.48)	.85** (.32)
Shortage^b	−.89** (.31)	−.76* (.35)	−1.05** (.32)	−.31 (.21)
No mask × liberalism	.37** (.14)	.64*** (.16)	.54*** (.15)	.18 (.10)
N95 mask × liberalism	.35 (.14)*	.30 (.16)	.49** (.14)	.21* (.10)
Liberalism	−.16 (.09)	−.23* (.11)	−.21* (.10)	−.06 (.06)
Unintelligent				.69*** (.03)
Self-centered				.16*** (.03)
Constant	4.08 (.52)	4.05 (.58)	3.85 (.54)	.40 (.38)
R ²	.38	.35	.44	.76

Note: N = 541. Standard errors in parentheses.

Reference condition is wearing a scarf.

Reference condition is no shortage of supplies.

p < .05. **p < .01. ***p < .001 (two-tailed tests).

The effects of face covering on perceptions of the actor’s lack of intelligence varied with participants’ political orientation (Model 1). Wearing no mask was seen as more unintelligent than wearing a scarf; this effect was stronger for more liberal participants (see the statistically significant no mask and the No Mask × Liberalism coefficients). Also, as participants’ liberalism increases, they saw the actor wearing an N95 mask as more unintelligent compared to the actor wearing a scarf (see the N95 Mask × Liberalism coefficient). For the most conservative participants in the nonshortage condition, the actor wearing a scarf was seen as more unintelligent than one wearing an N95 mask (see the negative N95 Mask coefficient, which is the effect of N95 mask compared to scarf when liberalism is at its lowest value and when there was no shortage of supplies). A figure illustrating the conditional effects of masking behaviors on ratings of unintelligence by study participants’ political orientation appears in the online supplement.

The effects of face covering on perceptions of the actor’s self-centeredness also varied with participants’ political orientation (Model 2). The results indicate that the most conservative participants did not see wearing no mask as more self-centered than wearing a scarf (see the statistically insignificant no mask coefficient); the more liberal the participant, the more they viewed not masking as self-centered (see the statistically significant No Mask × Liberalism coefficient). Both liberal and conservative participants saw wearing an N95 mask during a shortage as self-centered (see the statistically significant N95 Mask × Shortage coefficient and the statistically insignificant N95 mask and N95 Mask × Liberalism coefficients).⁸

Table 6 also presents the results of analyses examining political orientation and disapproval. We include two models, one without controlling for ratings of unintelligence and self-centeredness (Model 3) and the other including them (Model 4). Participants disapproved more strongly of wearing no mask (vs. a scarf). This effect was larger for more liberal participants (see the no mask and No Mask × Liberalism coefficients, Model 3). Liberalism is also associated with greater disapproval of wearing an N95 mask—for the most conservative participants, in the nonshortage conditions, wearing an N95 mask garnered less disapproval than wearing a scarf (see the negative and significant N95 mask coefficient). But disapproval of wearing an N95 strengthened with participant liberalism (see the N95 × Liberalism coefficient), surpassing that of wearing a scarf among liberals.

The predicted means presented in Figure 2 illustrate these conditional effects (calculated from Model 3, Table 6). Note that because there were no statistically significant interactions between shortage condition and participant liberalism or three-way interactions between face covering, shortage conditions, and participant liberalism (and thus these effects were not included in the model), each pair of lines for the different masking behaviors (shortage conditions in solid lines and nonshortage conditions in dash lines) are parallel. The differences in disapproval between the three masking behaviors were larger for the more liberal participants. In other words, the effects of the masking behaviors were stronger among more liberal than conservative participants. This pattern of findings supports Hypothesis 3.

Figure 2.

Disapproval by Study 2 Condition and Participant Liberalism

The coefficients for the experimental conditions were substantially reduced (and some no longer significant) in Model 4 when ratings of unintelligence and self-centeredness were included. Wearing an N95 mask during a shortage continued to garner more disapproval than wearing cloth, and the effect of wearing an N95 was stronger among more liberal participants.^9,10

Discussion

We have argued that rationales provide logical reasons for supporting a norm (Horne and Mollborn 2020; Mollborn 2017). We examine this argument in a novel context, the early months of the Covid-19 pandemic, testing whether people assess newly relevant health-related behaviors in prudential and moral terms and whether those assessments explain how much they disapprove of the behavior. Our findings support this rationales argument.

Two findings were unexpected, however. First, we expected that people would view an individual wearing an N95 mask as self-centered when there was a shortage. The results are consistent with this prediction but also show that people saw wearing an N95 mask as self-centered (although less so) even in the absence of a shortage. The effect in the no shortage condition may reflect the context. Our vignette described a local shortage of PPE. It may be that in the context of spring 2020, people were aware that there were shortages nationwide and so evaluated the actor’s behavior in light of their own understanding of that larger national context. Norms research is not clear on how people decide which others matter, for example, whether those others live in the same community or in the same country. Early norms research pointed to the potential role of reference groups (e.g., Merton 1968; Saxena 1971). More recent work highlights the importance of interdependence among actors who are affected by a behavior (e.g., Horne 2001a), but little current work explicitly assesses just who matters.

Second, we expected that wearing an N95 mask might be viewed as safer for the individual than wearing a scarf and thus considered smarter to use. We found that participants viewed both wearing a scarf and wearing an N95 mask as more intelligent than not masking (as predicted). But contrary to our expectation, people viewed those who wore an N95 mask as less intelligent compared to using a cloth face covering when PPE supplies were short in local hospitals. One interpretation of this finding is that participants’ evaluations were driven by CDC recommendations at the time—that they saw noncompliance as both foolish and selfish. However, if evaluations were driven by such a desire to comply with government recommendations or mandates, we would expect people to have reacted equally negatively to working a nonessential job and socializing (both of which violated the government mandate described in the vignette). They did not—they reacted more negatively to noncompliant socializing than to noncompliant working. We interpret our results as suggesting that participants thought the actor wearing an N95 failed to understand the implications of their behavior for the collective (and not only for themselves). If so, then it appears that people evaluated others’ smartness in collective, not only individual terms. Prudential rationales may support behavior that benefits the individual and/or behavior that benefits the collective. Contrary to popular wisdom, helping others may not be seen as foolish (Phillips and Taylor 2010).

We also argued that liberal participants would be more disapproving of the Covid-related behaviors in our vignettes. Our findings in both the staying home and masking studies generally support this argument. Representing an interesting exception, disapproval of attending large parties did not vary by political orientation. Rather, the conditional effects by liberalism in the staying home study were specifically about going to work at a nonessential job. Liberalism was associated with stronger disapproval of working compared to staying home; the most liberal participants disapproved of working as much as attending large parties. That the magnitude of this conditional effect was dramatically reduced after controlling for the rationales measures suggests that it is tied to liberal participants’ understanding of how smart and self-interested work was compared to socializing in the context of the vignette.

Limitations and Future Research

Our study aimed to assess whether rationales support disapproval. Our results provide initial evidence that they do. Future research should further assess the role of rationales. It could examine, for example, when different rationales align (or not) to support the same behaviors, the relative primacy of rationales across contexts, whether people invoke different rationales depending on the domain, and when different groups of people invoke different rationales. In addition, we focus here on rationales as providing a logical reason for support for norms (disapproval of a behavior) and have not addressed their potential causal role. Future research should assess not only whether rationales provide a reason to support a norm but also whether they causally affect that support and/or vice versa. More broadly, it is possible that communicating rationales may support or prevent erosion of norms. Research shows that gossip and statements about what people ought to do affect compliance (e.g., Shank et al. 2019). Effective communication of rationales may provide a justification both for complying with the norm and for sanctioning violations as well as reason to believe that others support the norm.

Future research could also consider more deeply the nature of moral and prudential concerns. We found that rationales were associated with support for norms (disapproval). Our measures of rationales were narrow—using only single items and asking about only one of multiple possible dimensions of moral and prudential considerations. Prudential and moral rationales are likely more complex than we captured in this study. Qualitative research that explores rationales would help to more fully understand the relevant dimensions of prudential, moral, and perhaps other not yet identified rationales. It could also help to understand how rationales are developed and negotiated and how they apply in given contexts (for early work on these processes, see Mollborn 2017; Mollborn and Sennett 2015).

In addition, we focused on rationales related to the behavior (whether the behavior was foolish and selfish). We did not examine approval or rationales related to government mandates. But the illegality of a behavior along with perceptions that a law or mandate is a legitimate exercise of government authority may provide another source of rationales. For example, a belief that government stay-home mandates are an illegitimate infringement on individual freedom provides a rationale for approving of those who violate the mandate. Our study does not address the legitimacy of government mandates. Future research could simultaneously address and disentangle rationales about behaviors and about government mandates regarding those behaviors.

We also focus on support for norms (disapproval) rather than on normative expectations or sanctioning. Normative evaluations (individual disapproval or expectations about others’ disapproval) are likely to rest on different causal factors and mechanisms than sanctioning. However, we expect that rationales provide people not only with reason to disapprove of a behavior and support norms constraining that behavior but also with reason to punish violations. Future research could test this possibility.

Given the politicization of public health directives over the course of the Covid-19 pandemic (e.g., Halon 2020), we explored whether the political orientation of participants shaped their interpretation of behavior via the rationales they evoke. Even in the early months of the pandemic, Republican leaders downplayed the risks of the virus. To the extent that conservatives did not see the virus as a threat, there was probably less concern about shortages of PPE (e.g., N95 masks) for health care workers and less concern about working and socializing outside the home. In such a context, it makes sense that compared to conservatives, liberals would develop stronger prudential and moral rationales. It is not clear from our results whether liberals, in general, are more judgmental than conservatives or if these attributions of unintelligence and self-centeredness are unique to the study context. Future research should more explicitly assess the role of rationales for liberals and conservatives. It could assess prudential and moral rationales around issues typically seen as liberal (e.g., climate change, antiracism) and conservative (e.g., gun ownership, antiabortion). Furthermore, our results are based on an online convenience sample of participants who may not be representative of liberals and conservatives generally. And political orientation is only one characteristic that might alter people’s reactions to the experimental conditions. Work status (whether in an essential occupation or not and whether able to work from home or not), religiosity, and related views on the importance of church attendance may also have implications for how people use rationales. Future research should assess the use of rationales with nationally representative samples and with diverse socioeconomic groups.

Finally, we expect that rationales and disapproval reflect social structural conditions (e.g., political context, status hierarchies, and network structures; Désilets, Brisson, and Hétu 2020; Horne and Mollborn 2020; Paluck, Shepherd, and Aronow 2016). In the Covid-19 context, compliance with public health mandates has over time become a marker of (political) group coalitions. Future research should analyze just how rationales and norms become aligned with social structural boundaries.

Conclusion

In sum, we argue that prudential and moral rationales provide people with reasons for disapproving of a behavior. Consistent with this theory, we find that exposure to Covid-related behaviors triggers prudential and moral rationales and that those rationales are associated with disapproval. Theoretically, our study highlights the potential role of rationales in norm emergence, maintenance, and erosion. Substantively, it has implications for understanding variation in public health behaviors and support for and compliance with public health mandates.

Supplemental Material

sj-docx-1-spq-10.1177_01902725221107594 – Supplemental material for Rationales and Support for Norms in the Context of Covid-19

Supplemental material, sj-docx-1-spq-10.1177_01902725221107594 for Rationales and Support for Norms in the Context of Covid-19 by Christine Horne and Monica Kirkpatrick Johnson in Social Psychology Quarterly

Supplemental Material

sj-pptx-2-spq-10.1177_01902725221107594 – Supplemental material for Rationales and Support for Norms in the Context of Covid-19

Supplemental material, sj-pptx-2-spq-10.1177_01902725221107594 for Rationales and Support for Norms in the Context of Covid-19 by Christine Horne and Monica Kirkpatrick Johnson in Social Psychology Quarterly

Footnotes

Acknowledgements

We thank Stefanie Mollborn for comments on an earlier draft of this article.

ORCID iDs

Christine Horne

Monica Kirkpatrick Johnson

Supplemental Material

Supplemental material for this article is available online.

1

Study materials (instructions, vignettes, and questions) and code for both are available at Harvard Dataverse.

2

A version of these two dimensions is also seen in behavioral mechanisms that link internal states to action. For example, instrumental rationality (rational self-interest) is a mechanism that is widely used in social research (see, e.g., discussion of instrumental rationality).

3

As described earlier, when an individual supports a norm, their own disapproval or approval of a behavior is consistent with their normative expectations—whether they expect others to disapprove or approve. Participants’ disapproval was highly correlated with another measure not used in this analysis, normative expectations of others’ disapproval (r = .71). In the present article, we focus on the rationales held by individuals and the associations between those rationales and individuals’ own approval/disapproval.

4

Figures illustrating these patterns are available in an .

5

We also explored whether the association between rationales and disapproval vary by political orientation. To do so, we estimated an additional model for disapproval (not shown), including interaction terms between liberalism and self-centered and liberalism and unintelligent. Neither of these interaction terms were statistically significant.

6

Results in Study 1 were not sensitive to additional controls for the geographic location of residence of participants (measured by census region) or for Residence × Experimental Condition interactions.

7

In this study, as in Study 1, participants’ disapproval was highly correlated with expectations of others’ disapproval (r = .77).

8

A figure illustrating the conditional effects of masking behaviors on ratings of self-centeredness by study participants’ political orientation appears in the .

9

We also explore whether the association between rationales and disapproval varied by political orientation. To do so, we estimated an additional model for disapproval (not shown), including interaction terms between liberalism and self-centered and liberalism and unintelligent. Neither of these interaction terms were statistically significant.

10

Results in Study 2 were not sensitive to additional controls for the geographic location of residence of participants (measured by census region) or by any Residence × Experimental Condition interactions.

Bios

Christine Horne is a professor of sociology at Washington State University. She studies the emergence, enforcement, and application of social norms. Recent research focuses on norms related to household energy use and gender relations in Ghana.

Monica Kirkpatrick Johnson is a professor of sociology at Washington State University. Her work focuses on understanding the role of values, expectations, and aspirations on behavior and life-course outcomes, particularly during the transition to adulthood.

References

Achenbach

Joel

Sun

Lena H.

McGinley

Laurie

. 2020. “CDC Considering Recommending General Public Wear Face Coverings in Public.”Washington Post, March 20. https://www.washingtonpost.com/health/cdc-considering-recommending-general-public-wear-face-coverings-in-public/2020/03/30/6a3e495c-7280-11ea-87da-77a8136c1a6d_story.html.

Alvarez

Maria

. 2009. “How Many Kinds of Reasons?”Philosophical Exploration 12(2):181–93.

Alvarez

Maria

. 2017. “Reasons for Action: Justification, Motivation, Explanation.”In Stanford Encyclopedia of Philosophy (Winter 2017 edition), edited by Zalta

E. N.

https://plato.stanford.edu/cgi-bin/encyclopedia/archinfo.cgi?entry=reasons-just-vs-expl.

Bazelon

Emily

. 2020. “Restarting American Means People Will Die. So When Do We Do It?”New York Times, April 10. https://www.nytimes.com/2020/04/10/magazine/coronavirus-economy-debate.html.

Calvillo

Dustin P.

Ross

Bryan J.

Garcia

Ryan J. B.

Smelter

Thomas J.

Rutchick

Abraham M.

2020. “Political Ideology Predicts Perceptions of the Threats of COVID-19 (and Susceptibility to Fake News about It).”Social Psychological and Personality Science 11(8):1119–28.

Chappell

Bill

. 2020. “Coronavirus: Covid-19 Is Now Officially a Pandemic, WHO Says.”National Public Radio, March 11. Retrieved July 13, 2021. https://www.npr.org/sections/goatsandsoda/2020/03/11/814474930/coronavirus-covid-19-is-now-officially-a-pandemic-who-says.

Cialdini

Robert B.

2007. “Descriptive Norms as Underappreciated Sources of Social Control.”Psychometrika 72(2):263–68.

Coleman

James S.

1990. Foundations of Social Theory. Cambridge, MA: Belknap Press.

Cook

Karen Schweers

. 2005. “Networks, Norms, and Trust: The Social Psychology of Social Capital.”Social Psychology Quarterly 68(1):4–14.

10.

Correll

Shelley J.

Benard

Stephen

Paik

. 2007. “Getting a Job: Is There a Motherhood Penalty?”American Journal of Sociology 112(5):1297–338.

11.

Cowan

Sarah K.

2014. “Secrets and Misperceptions: The Creation of Self-Fulfilling Illusions.”Sociological Science 1:466–92.

12.

Cowan

Sarah K.

Baldassarri

Delia

. 2018. “‘It Could Turn Ugly’: Selective Disclosure of Attitudes in Political Discussion Networks.”Social Networks 52:1–17.

13.

Désilets

Élise

Brisson

Benoit

Hétu

Sébastien

. 2020. “Sensitivity to Social Norm Violation Is Related to Political Orientation.”PLOS ONE 15(12):e0242996. doi:10.1371/journal.pone.024996.

14.

Dwyer

Colin

Aubrey

Allison

. 2020. “CDC Now Recommends Americans Consider Wearing Cloth Face Coverings in Public.”National Public Radio, April 3. https://www.npr.org/sections/coronavirus-live-updates/2020/04/03/826219824/president-trump-says-cdc-now-recommends-americans-wear-cloth-masks-in-public.

15.

Feinberg

Matthew

Willer

Robb

Schultz

Michael

. 2014. “Gossip and Ostracism Promote Cooperation in Groups.”Psychological Science 25(3):656–64.

16.

Fiske

Susan T.

Cuddy

Amy J. C.

Glick

Peter

. 2006. “Universal Dimensions of Social Cognition: Warmth and Competence.”Trends in Cognitive Science 11(2):77–83.

17.

Graham

Jesse

Haidt

Jonathan

Nosek

Brian A.

2009. “Liberals and Conservatives Rely on Different Sets of Moral Foundations.’’Journal of Personality and Social Psychology 96(5):1029–46.

18.

Gray

Kurt

Schein

Chelsea

Ward

Adrian F.

2014. “The Myth of Harmless Wrongs in Moral Cognition: Automatic Dyadic Completion from Sin to Suffering.”Journal of Experimental Psychology 143(4):1600–15.

19.

Haelle

Tara

. 2020. “Why Everything Is Closing for Coronavirus: It’s Called ‘Flattening the Curve.’”Forbes, March 13. https://www.forbes.com/sites/tarahaelle/2020/03/13/why-everything-is-closing-for-coronavirus-its-called-flattening-the-curve/?sh=3638ee716e2b.

20.

Halon

Yael

. 2020. “Stephanie Grisham: Democrats Using Coronavirus ‘as a Tool to Politicize Things and Scare People.’”Fox News, March 2021. https://www.foxnews.com/media/stephanie-grisham-calls-on-dems-to-stop-using-coronavirus-to-scare-public-for-political-points.

21.

Heckathorn

Douglas D.

1989. “Collective Action and the Second-Order Free-Rider Problem.”Rationality and Society 1(1):78–100.

22.

Horne

Christine

. 2001a. “The Enforcement of Norms: Group Cohesion and Metanorms.”Social Psychology Quarterly 64(3):253–66.

23.

Horne

Christine

. 2001b. Sociological Perspectives on the Emergence of Social Norms. Pp. 3–34 in Social Norms, edited by Hechter

Opp

K.-D.

New York, NY: Russell Sage.

24.

Horne

Christine

Dodoo

Naa Dodua

Dodoo

F. Nii-Amoo

. 2018. “The Conditionality of Norms: The Case of Bridewealth.”Social Psychology Quarterly 81(4):319–39.

25.

Horne

Christine

Johnson

Monica Kirkpatrick

. 2021. “Testing an Integrated Theory: Distancing Norms in the Early Months of Covid-19.”Sociological Perspectives 64(5):970–87.

26.

Horne

Christine

Mollborn

Stefanie

. 2020. “Norms: An Integrated Framework.”Annual Review of Sociology 46:467–87.

27.

Horne

Christine

Przepiorka

Wojtek

. 2021. “Technology Use and Norm Change in Online Privacy: Experimental Evidence from Vignette Studies.”Information, Communication, and Society 24(9):1212–28. doi:10.1080/1369118X.2019.1684542.

28.

Jacobs

Andrew

Richtel

Matt

Baker

Mike

. 2020. “‘At War with No Ammo’: Doctors Say Shortage of Protective Gear Is Dire.”New York Times, March 19. https://www.nytimes.com/2020/03/19/health/coronavirus-masks-shortage.html.

29.

Kochhar

Rakesh

. 2020. “Unemployment Rose Higher in Three Months of Covid-19 Than It Did in Two Years of the Great Recession.”Pew Research Center. https://www.pewresearch.org/fact-tank/2020/06/11/unemployment-rose-higher-in-three-months-of-covid-19-than-it-did-in-two-years-of-the-great-recession/.

30.

Lee

Don

. 2020. “Reopening the Economy vs Keeping It Shut Longer. What’s More Costly.”Los Angeles Times, April 23. https://www.latimes.com/politics/story/2020-04-23/whats-worse-reopening-the-economy-or-keep-it-shut-down-longer.

31.

López

Canela

. 2020. “Celebrities Like Gwyneth Paltrow and Kate Hudson Were Criticized for Wearing Face Masks Early in the Pandemic. Here’s What Changed.”Business Insider, July 7. https://www.businessinsider.com/why-celebrities-were-criticized-for-wearing-masks-and-what-changed-2020-7.

32.

Mason

Lilliana

. 2018. Uncivil Agreement: How Politics Became Our Identity. Chicago, IL: University of Chicago Press.

33.

McAdams

Richard

. 2015. The Expressive Power of Law. Boston, MA: Harvard University Press.

34.

Merkley

Eric

Stecula

Dominic A.

2021. “Party Cues in the News: Democratic Elites, Republican Backlash, and the Dynamics of Climate Skepticism.”British Journal of Political Science 51(4):1439–56. doi:10.1017/S0007123420000113.

35.

Merton

Robert K.

1968. “Continuities to the Theory of Reference Groups and Social Structure.” Pp. 335–440 in Social Theory and Social Structure, edited by Merton

R. K.

New York, NY: Free Press.

36.

Mollborn

Stefanie

. 2017. Mixed Messages: Norms and Social Control around Teen Sex and Pregnancy. New York, NY: Oxford University Press.

37.

Mollborn

Stefanie

Sennott

Christie

. 2015. “Bundles of Norms about Teen Sex and Pregnancy.”Qualitative Health Research 25(9):1283–99.

38.

Mitchell

Amy

Jurkowitz

Mark

Oliphant

J. Baxter

Shearer

Elisa

. 2020. “Americans Who Rely Most on White House for Covid-19 News More Likely to Downplay the Pandemic.”https://www.journalism.org/2020/05/20/americans-who-rely-most-on-white-house-for-covid-19-news-more-likely-to-downplay-the-pandemic/.

39.

Opp

Karl-Dieter

. 2004. “‘What Is IS Always Becoming What Ought to Be.’ How Political Action Generates a Participation Norm.”European Sociological Review 20:13–29.

40.

Palan

Schitter

2018. “Prolific.ac – A Subject Pool for Online Experiments.”Journal of Behavioral and Experimental Finance 17:22–27.

41.

Paluck

Elizabeth Levy

Shepherd

Hana

Aronow

Peter

. 2016. “Changing Climates of Conflict: A Social Network Experiment in 56 Schools.”PNAS 113(3):566–71.

42.

Pappas

Alex

. 2020. “Trump Says CDC Wants Americans to Cover Faces with Cloth Amid Coronavirus.”Fox News, April 3. https://www.foxnews.com/politics/trump-cdc-recommending-face-covering-voluntary.

43.

Peer

Eyal

Brandimarte

Laura

Samat

Sonam

Acquisti

Alessandro

. 2017. “Beyond the Turk: Alternative Platforms for Crowdsourcing Behavioral Research.”Journal of Experimental Social Psychology 70:153–63.

44.

Phillips

Adam

Taylor

Barbara

. 2010. On Kindness. London: Picador.

45.

Saxena

Dhirendra Prakash

. 1971. “The ‘Reference Group’ Concept.”Social Science 46(3):155–64.

46.

Schelling

Thomas

. 1978. Micromotives and Macrobehavior. New York, NY: WW Norton.

47.

Schein

Chelsea

Gray

Kurt

. 2015. “The Unifying Moral Dyad: Liberals and Conservatives Share the Same Harm-Based Moral Template.”Personality and Social Psychology Bulletin 41(8):1147–63.

48.

Schein

Chelsea

Gray

Kurt

. 2018. “The Theory of Dyadic Morality: Reinventing Moral Judgment by Redefining Harm.”Personality and Social Psychology Review 22(1):32–70.

49.

Shank

Daniel B.

Kashima

Yoshihisa

Peters

Kim

Yang

Robins

Garry

Kirley

Michael

. 2019. “Norm Talk and Human Cooperation: Can We Talk Ourselves into Cooperation?”Journal of Personality and Social Psychology 117(1):99–123.

50.

Sorokin

Pitirim A.

1968. Man and Society in Calamity. New York, NY: E.P. Dutton.

51.

Steib

Matt

. 2020. “Fox Business Network: Covid-19 Is a ‘Coronavirus Impeachment Scam.’”NY Magazine, March 10. https://nymag.com/intelligencer/2020/03/fox-business-covid-19-is-a-coronavirus-impeachment-scam.html.

52.

Stevens

Harry

. 2020. “Why Outbreaks Like Coronavirus Spread Exponentially, and How to ‘Flatten the Curve.’”Washington Post, March 14. https://www.washingtonpost.com/graphics/2020/world/corona-simulator/.

53.

Tankersley

Jim

. 2020. “US Is Nowhere Close to Reopening the Economy, Experts Say.”New York Times, April 6. https://www.nytimes.com/2020/04/06/business/economy/coronavirus-economy.html.

54.

Tufekci

Zeynep

. 2020. “Why Telling People They Don’t Need Masks Backfired.”New York Times, March 17. https://www.nytimes.com/2020/03/17/opinion/coronavirus-face-masks.html.

55.

Ullmann-Margalit

Edna

. 1977. The Emergence of Norms. Oxford, UK: Oxford University Press.

56.

Weber

Max

. 1978. Economy and Society, edited by Roth

Wittich

Berkeley: University of California Press.

57.

Wolfe

Daniel

Dale

Daniel

. 2020. “‘It’s Going to Disappear’: A Timeline of Trump’s Claims That Covid-19 Will Vanish.”CNN, October 31. https://www.cnn.com/interactive/2020/10/politics/covid-disappearing-trump-comment-tracker/.

58.

Wojciske

Klusek

1996. “Moral and Competence Related Traits in Political Perception.”Polish Psychological Bulletin 27:319–24.

59.

Wojciszke

1994. “Multiple Meanings of Behavior: Construing Actions in Terms of Competence or Morality.”Journal of Personality and Social Psychology 67:222–32.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.03 MB

0.37 MB