Status as Deference: Cultural Meaning as a Source of Occupational Behavior
==========================================================================

* E. K. Maloney
* Kimberly B. Rogers
* Lynn Smith-Lovin

## Abstract

Status is an independent basis of inequality. Cultural meanings create the voluntary esteem and deference that distinguish status inequities from inequalities in power and material resources, as Cecilia Ridgeway and Hazel Markus explain in the introduction to this issue. Here, we use affect control theory (ACT)—a formal theory of culture, identity, and social action—to explore how cultural meanings of occupational identities shape status behavior. ACT assumes that people try to maintain cultural meanings for identities and behaviors on three affective dimensions (evaluation, potency and activity) as they interact with others. We use ACT to define how actors in different status groups—occupations with similar patterns of deference to and from other occupations—act toward one another. We validate our theoretical behavioral predictions with vignette survey data.

*   status
*   affect control theory
*   deference
*   occupational prestige
*   inequality

In classic texts, Max Weber (1978, 2014) argues that status—the sense that someone is worthy of esteem, honor, and respect—is a source of inequality independent of material resources and power, the ability to overcome opposition from others.1 Status directly produces inequality because cultural beliefs about differences between identity groups elicit and justify unequal actions, treatment by others, and resource allocations in social interactions. Status also stabilizes existing inequalities via presumptions that those with greater power and material resources are worthy and deserving of these advantages (Ridgeway 2014; Ridgeway and Markus 2022, this issue).

Despite the clear importance of status for inequality, however, most scholarship focuses on material and structural dimensions of stratification (power and resources) rather than interaction dynamics and cultural forces (status), perhaps because the latter are more challenging to measure. Even the most widely used operationalization of status, occupational prestige, maps more closely onto features of occupations such as income and educational prerequisites than cultural and interactional patterns (Freeland and Hoey 2018). A long-standing criticism that these measures are simply “error-prone estimates” of occupations’ material features (Featherman and Hauser 1976, 405) has encouraged reliance on material measures of occupational importance, and a corresponding reluctance to theorize status as a basis of stratification in its own right.

In this article, we review a new conceptualization and operationalization of status based in affect control theory (ACT)—a formal, mathematical theory of culture, identity, and social action (Heise 2007; Robinson and Smith-Lovin 2018). This approach measures status using ACT predictions about deference between occupational groups in social interactions (Freeland and Hoey 2018) and identifies occupational status groups (OSGs) with similar deference patterns and therefore similar positions within the interactional status structure (Maloney 2020). We use this approach to explore culturally expected behaviors between occupational actors occupying different positions in the status structure by predicting the actions and social treatment of objects from different occupational status groups. In so doing, we illuminate the micro-social dynamics that produce and justify inequalities between groups with differing status. We close by showing that our model predictions conform to people’s actual expectations about unequal interactions among status groups with a vignette survey.

## SOCIAL SCIENCE VIEWS OF INEQUALITY AND SOCIAL CLASS

Weber (1978, 2014) delineates three conceptually independent sources of inequality: status, power, and resources. Social scientists have concentrated most of their efforts on the analysis of resource disparities (usually in terms of income or wealth) and differences in power (usually in terms of organization or mobilization), perhaps because these are easier to measure than cultural features of honor, esteem, and worthiness (status). However, social psychological research shows that beliefs about the worthiness of social identities create and legitimize unequal allocations of power and resources, in that high-status individuals are deemed more deserving of leadership and rewards than those with low status (Correll and Ridgeway 2003; Ridgeway 2014; Ridgeway and Correll 2006).

The most commonly used measure of status at present, occupational prestige, overlaps substantially with measures of the human capital and material rewards associated with specific jobs (see Bukodi, Dex, and Goldthorpe 2011). We contend, however, that measures that rely on the material dimensions of social standing and discount the role of cultural honor, esteem, and respect can lead to incorrect predictions about status dynamics, or a failure to really understand them at all. For example, occupations that offer low material rewards may nevertheless be highly regarded (for example, coal mining and manufacturing versus warehouse jobs, see Koenig 2022; Valentino 2022, this issue), making them desirable in ways that cannot be easily understood by their physical conditions and pay. People often prefer respect to rank position (Anderson et al. 2021).

Recent work has reinvigorated the study of status as an independent source of inequality by developing a new measure of status that better reflects its cultural and interactional character, thereby addressing the shortcomings of prevailing measures of status as occupational prestige. This work uses ACT—a formal, mathematical theory of culture, identity, and social action (Heise 2007; Robinson and Smith-Lovin 2018)—to quantify the expectation that a person in one occupation will defer to someone in another occupation in interaction based on cultural meanings regarding the pleasantness, dominance, and expressiveness of these occupations.2 A deference structure derived from these data identifies groups of structurally equivalent occupations with similar deference patterns and thus similar positions in the status hierarchy (OSGs). Because ACT may not be familiar to a wide interdisciplinary audience, we summarize its core characteristics before discussing the interactional deference structure that motivates our analysis in this article.

### Affect Control Theory

ACT is a mathematical model of social interaction that quantifies the cultural meanings of identities and behaviors and that generates precise predictions about the consequences of these meanings for social action. It uses empirically estimated, culture-specific models of impression formation norms to show how social interactions change our feelings about actors and actions (for reviews of the theory, see Heise 1979, 2007; MacKinnon 1994; Robinson and Smith-Lovin 2018). Widely shared cultural meanings, known in ACT as *fundamental sentiments*, are measured along three dimensions: evaluation (good or bad), potency (powerful or weak), and activity (active or passive), referred to as EPA. These meanings reflect basic cultural knowledge about the social order (Heise 2010; Rogers 2021b), are foundational to socioemotional processing (Scholl 2013), and parsimoniously summarize concept meanings across many cultures (Osgood, May, and Miron 1975; Osgood, Suci, and Tannenbaum 1957). For example, U.S. English-speakers see doctors as good, powerful, and neutral in activity (E = 2.69, P = 2.94, A = 0.37), and patients as neutral in goodness, weak, and passive (E = 0.57, P = –1.49, A = –1.28 on scales that range from 4.3 to –4.3).

The fundamental sentiments that ACT measures map onto cultural features measured in other theoretical models interested in status (see Scholl 2013). This includes, for example, status defined as evaluations of competence within and expectations for contributions to a task group in expectation states theory (Dippong and Kalkhoff 2015; Rogalin, Soboroff, and Lovaglia 2007; Ridgeway and Smith-Lovin 1994) and intergroup judgments of warmth and competence in the stereotype content model (Rogers, Schröder, and Scholl 2013). Here we argue that ACT offers unique advantages for understanding inequalities in status behavior that arise as people endeavor to maintain fundamental sentiments in interactions with others. The theory’s precise measurement of cultural meanings allows for, among other things, a concrete description of how events change these meanings and how people react to that change. In addition, ACT is distinctive in modeling status behavior in a manner that is not constrained to task-oriented interaction settings (Ridgeway and Smith-Lovin 1994) and that f ocuses on interaction dynamics rather than intergroup bias (Rogers, Schröder, and Scholl 2013).

ACT’s impression-change equations predict the shift in meaning resulting from an event in which an actor does a behavior to an object (Heise 1979, 2010; Smith-Lovin and Heise 1988).3 These new event-contextualized meanings for the actor, behavior, and object-person following a situation are the *transient impressions* created by a social encounter. After a doctor listens to a patient, for example, the doctor seems nicer, less powerful, and less active (E = 3.11, P = 2.03, A = –0.10) than we would expect based on fundamental sentiments alone. The greater the EPA distance between transient impressions and fundamental sentiments, the more culturally misaligned the situation. This misalignment is quantified in ACT as *deflection*, the sum of the squared differences between transient impressions and fundamental sentiments in EPA (Heise 2007; Robinson and Smith-Lovin 2018). Higher deflection events are less likely to occur, less expected, and less culturally normative (Rogers 2021a). Situations that violate cultural expectations prompt restorative actions or reinterpretations of events that realign transient impressions with fundamental cultural meanings for the people and actions involved.4

Crucially, for the cultural nature of the theory (and our view of status presented here), both the cultural sentiments associated with identities and the impression-change processes that occur when these identities are embedded in social events show high consensus across members of a national language culture (Heise 2010; Rogers 2019a, 2019c). Although some variation in meaning is observed in social position, as with occupational prestige and other subjective measures of inequality, these beliefs are widely shared. The exceptions are striking because they are precisely that—exceptions.5

Cecilia Ridgeway and Hazel Markus (2022, this issue) contend that status is distinct from other bases of inequality in being primarily cultural, operating through shared evaluations of a person or group’s worthiness by others; multilevel, affecting the relative worth of persons within groups as well as groups within societies; and relational, in being inherently comparative and emergent through social interactions. We argue that ACT is an especially useful tool for examining status dynamics because it quantifies culturally shared meanings of identities and behavior, reveals how actors’ relative positioning on these dimensions reflects their cultural and interactional standing, and predicts the relational, situated behavior between actors occupying identities that carry status. In addition, ACT is amenable to the study of status dynamics across cultures. Indeed, the data and models already exist to apply the theory in this manner. This is noteworthy because the bases of status differ across cultures. For example, in cultures such as Japan, more traditional occupations might be higher evaluation and potency than they are in the United States because the elderly are seen as sources of wisdom (Schröder et al. 2013). Here we focus on the U.S. context, where status is highly influenced by judgments of competence. However, ACT could also be used to study status processes in cultural contexts where status is more strongly determined by, for example, attributions of morality (Zhao 2022). We expect such cultural features to affect the EPA assessments of different identity groups and the impression-change processes that combine with these assessments to guide social action.

### ACT and the Measurement of Occupational Deference Scores

Robert Freeland and Jesse Hoey (2018) leverage ACT’s approach to modeling cultural expectations to create a new method of operationalizing occupational status. Because ACT is a formal model of how cultural sentiments create expectations for behavior and deference behavior is an expression of an underlying status order, Freeland and Hoey argue that deflections from simulations of deference events provide a quantitative, theoretically motivated indicator of occupational status.

Using ACT, they estimate deflection (cultural dislocation) from simulated deference events of the structure *occupation A defers to occupation B* for all possible combinations of 304 occupational identities as actors and object-persons. For example, the event a surgeon (E = .09, P = 3.14, A = –0.13) defers to (E = –0.15, P = 0.45, A = –0.44) a coal miner (E = 0.78, P = 0.01, A = 0.47) results in a deflection of 7.2.6 A coal miner deferring to a surgeon produces less deflection (5.9), indicating that ACT predicts the latter situation to be more likely, less unexpected, and more culturally aligned than the former (Rogers 2021a). A single deference score was created for each occupation (analogous to occupational prestige scores) by averaging deflections across all events involving interactions with other occupations. For example, the deference score for surgeons was produced by averaging the deflection from all 304 events wherein a surgeon deferred to some other occupation.

Freeland and Hoey’s (2018) deference score summarizes the social, relational, and consensual nature of status that Ridgeway (2019) observes as being culturally and interactionally grounded and relative rather than absolute—the result of an occupation’s position relative to others in the cultural system. It outperforms the standard measure of occupational prestige in predicting Harris poll rankings of occupations’ prestige as well as a variety of affective job outcomes such as workplace attachment, job satisfaction, general happiness, respondents’ feeling that their work is meaningful and that they are respected at work, and their willingness to keep working even if not financially necessary. Thus it has both construct and criterion validity as a measure of occupational status.

#### A New Conceptualization of Social Classes: Occupational Status Groups

Although the new deference score provides a concise summary of status rankings, it ignores a great deal of information contained in the 304 × 304 matrix of simulated interactions by averaging across rows to create a single deference score for each occupational identity. E. K. Maloney (2020) addresses this issue by using network methods to derive occupational status groups with common patterns of deference from the same matrix of deflections.7 Such methods look for structural equivalence, grouping occupational identities that defer to the same occupations and receive deference from the same occupations. Unlike research that uses network methods to search for direct connections between occupations, such as cliques or interconnected positions, this method identifies similarities in occupations’ positions within the cultural status system.

Maloney’s (2020) analysis identifies four OSGs. The first and largest group included 120 *everyday specialists* (mean EPA = 1.35, 0.92, 0.26)—occupations that are respected for their skills and have authority within their specialty, but which are not revered more generally (such as bricklayer or baker). The second status group, *service-to-society* (mean EPA = 1.65, 1.08, 0.85), included fifty-five largely female-typed occupations that are valued but tend not to receive high wages, such as teacher or social worker. The third status group, the *disagreeably powerful* (mean EPA = 0.74, 1.27, 0.26), included 109 largely male-typed occupations with the ability to change others’ actions or influence the course of someone’s day without much collaboration (such as foreman or bailiff). The final status group—the *actively revered* (mean EPA = 1.91, 1.44. 1.44) is the smallest, including only thirteen occupations, but is perhaps the most striking. It includes occupations often seen as heroes and caretakers but that garner lower material rewards than the *everyday specialists* and *disagreeably powerful* (such as nurse or firefighter).

Two recent articles demonstrate that the cultural sentiments associated with occupational identities and the status groups within which occupations fall have important implications for the emotional experiences of their occupants (Maloney 2022; Maloney and Smith-Lovin 2021). This article explores the behavioral implications of these status meanings. More specifically, it identifies the behaviors that are most common, expected, and normative (Rogers 2021a) in interactions within and between the occupations in each status group. These patterns demonstrate how status dynamics contribute to inequality, as higher status groups are culturally entitled to engage in actions and receive social treatment that justifies and reinforces their advantaged position (Rogers 2021b). The article contrasts high- and low-prestige occupations within each status group (using more conventional measures) to develop our understanding of the differences between status as voluntary deference and status as resources or power. Model predictions are validated with a vignette survey.

## DATA

Analyses for this article take two forms: theoretical models of status behavior using ACT and a vignette survey validating the predictions. Data for models of status behavior come from four sources. Deference scores are from Freeland and Hoey (2018). Occupational status groups are from Maloney (2020). A crosswalk was used to match occupational identities in the deference dataset to occupations in the General Social Survey (Smith et al. 2017), from which we gathered ISCO-88 classifications to match with Standard International Occupational Prestige Scale scores (SIOPS) (Treiman 1977) and the International Socio-Economic Index of Occupational Status (ISEI) (Ganzeboom and Treiman 1996; Ganzeboom, de Graaf, and Treiman 1992). For some occupations, the SIOPS is an average across several occupations mapped to a single occupational identity (such as different types of teachers). EPA ratings of identities and behaviors were taken from an ACT sentiment dictionary collected between 2011 and 2014 (Smith-Lovin et al. 2016).8

Data for the vignette validation survey were gathered online via Amazon Mechanical Turk in the fall of 2021. After participants who did not pass a manipulation check question and responses with the same IP address were removed, the study included 424 respondents. On average, participation took around ten and a half minutes, and respondents were paid $5 on completion of the study. In our analytic sample, the median age is thirty-six, and the majority of respondents were white (81 percent), male (62 percent), and identified as men (61 percent). The most common education level of our respondents was a bachelor’s degree (58 percent).9

## METHODS

The first analysis elaborates theoretical models of status behavior, using ACT impression-change equations to predict culturally expected behaviors in interactions between all possible pairs of occupational identities in the design. The EPA value identified reflects the action that will maximally confirm the identities of both the actor and object-person (that is, maintain the status of both occupational identities involved in the modeled situation). This EPA value is then matched with the nearest behavior labels in the most recent ACT sentiment dictionary (Smith-Lovin et al. 2016). The analysis was completed using the R package inteRact (Maloney 2021), which implements ACT’s impression-change equations in R. Functions that implement the theory’s U.S. English impression-change equations (Heise 2007) were used to estimate optimal behavior.

To assess the effects of status as a cultural (deference) versus material (occupational prestige) construct, we selected four occupational identities from each status group in Maloney (2020), two low in prestige and two high in prestige. We chose occupational identities with fundamental sentiments as close as possible to the median EPA value of each of the four OSGs to ensure that the identities used in the simulations matched the social positions of their respective OSGs in the status structure. Table 1 displays these sixteen occupational identities, along with their EPA ratings, OSG, prestige (high or low), deference score, distance from the median EPA of the OSG, and SIOPS. The optimal behavior EPA and closest behavior label were estimated for all possible combinations of these identities, 256 in all.

View this table:
[Table 1.](http://www.rsfjournal.org/content/8/7/70/T1)

**Table 1.** 
Occupational Identities Chosen for Simulations

### Vignette Survey

To validate these theoretical projections, we fielded a vignette survey on Amazon Mechanical Turk. Respondents consented to participate in the study, then completed a brief demographic survey. Afterward, they responded to vignette questions that asked them to choose the behavior most likely to occur in an interaction between the two people shown (for example, a surgeon and a nurse). Two response options were provided, the optimal behavior from our theoretical models and a randomly chosen behavior from the most recent U.S. ACT sentiment dictionary (Smith-Lovin et al. 2016). Respondents were randomly assigned to complete subsets of stimuli, answering thirty-two questions each. The study included 512 stimuli in total; approximately twenty-five respondents answered each.10 The interaction (actor-behavior-object) is the unit of analysis.

Two constraints on the random selection of behaviors from the ACT dictionary reduced bias in design. First, the chosen behaviors were required to be a Euclidean distance of five or greater away from the optimal behavior EPA, because distance in EPA space predicts how culturally appropriate an action seems according to ACT; we did not want the randomly chosen behavior to be an affective synonym for the theoretically predicted behavior. Second, we filtered out violent or sexual behaviors and actions clearly constrained to a single profession (such as inject with medicine), which may be viewed as unlikely for cognitive reasons rather than the cultural affective reasons of interest here (Rogers 2021a). This constraint was conservative in the sense that some random behavior choices were still institutionally inappropriate and might not have been chosen for that reason.

To analyze the resulting data, we regressed a binary indicator of choosing the optimal behavior (yes = 1 / no = 0) on the Euclidean distance between the random and optimal behavior in EPA space. We expected that respondents would be more likely to choose the optimal behavior when the randomly chosen behavior was further away from optimal EPA, as predicted by ACT. A multilevel model with questions nested in individuals nested in modules accounted for the underlying grouping of our design. We ran three models: one assessing the base rate of choosing the optimal behavior over the randomly selected behavior (model 1), another including only the distance measure (model 2), and a third including both measures and controls for sex, age, and education (model 3). To help the models converge, both distance and age were standardized such that the mean is 0 and the standard deviation 1. Because the patterns are consistent across these models, only model 3 is discussed in detail here (for additional results, see the online appendix at [https://www.rsfjournal.org/content/8/7/70/tab-supplemental](https://www.rsfjournal.org/content/8/7/70/tab-supplemental)).11

## RESULTS

To illustrate the differences among our core concepts, we show scatterplots (figure 1) of the relationship between deference scores and both occupational prestige (SIOPS) and a standard index of socioeconomic status (ISEI), with status group indicated by color and symbol (Ganzeboom, De Graaf, and Treiman 1992; Treiman 1977). The zero-order Pearson correlation between SIOPS and the Freeland and Hoey deference score is 0.401. The modest association between these measures is predictable from the findings of prior work. Freeland and Hoey (2018) find that conventional prestige scores, which are closely related to education and income for an occupation, are most strongly predicted by their potency (power, dominance), whereas deference scores are most strongly predicted by their evaluation (warmth, esteem). Prestige is also negatively associated with activity (expressivity, engagement), whereas deference scores are positively associated with activity. Put differently, deference and prestige evoke different cultural connotations: high prestige occupations carry connotations of quiet dominance; high deference occupations carry connotations of esteem, efficacy, and agency. Occupational status groups are clearly stratified along the *y* axis of deference (because they are based on deference relations) and also spread widely across the *x* axis of prestige. Notably, *actively revered* occupations are grouped at the top of the deference scale but differ dramatically in prestige. *Disagreeably powerful* occupations span into the upper part of the prestige scale, but never make it past the middle range on deference.

![Figure 1.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/7/70/F1.medium.gif)

[Figure 1.](http://www.rsfjournal.org/content/8/7/70/F1)

**Figure 1.** 
Relationship Between Deference and Prestige

*Source*: Authors’ tabulations based on Treiman 1977; Ganzeboom, De Graaf, and Treiman 1992; Freeland and Hoey 2018.

*Note*: Occupational prestige scale is the Standard International Occupational Prestige Scale (Treiman 1977). The socioeconomic index is the International Socio-Economic Index of Occupational Status (Ganzeboom, De Graaf, and Treiman 1992). Deference scores are from Freeland and Hoey (2018).

The plot relating deference and the socioeconomic status index shows almost exactly the same patterns as the occupational prestige plot, which is not surprising; occupational prestige is effectively a subjective measure of objective material resources.

### Theoretical Models of Status Behavior

Table 2 and figure 2 present results from the theoretical models of status behavior. Table 2 displays the mean EPA of optimal behaviors given the status group of the actor (the person enacting the behavior) and object-person (the person receiving the behavior), as well as exemplar behavior labels for the events modeled. The violin plots in figure 2 visualize the probability density of behavior EPA across the three models. They enable us to see the degree of consistency in behavior EPA across the events comprising each cell in the plot. The columns in this plot reflect actor OSG and the rows reflect object OSG.

View this table:
[Table 2.](http://www.rsfjournal.org/content/8/7/70/T2)

**Table 2.** 
Optimal Behaviors

![Figure 2.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/7/70/F2.medium.gif)

[Figure 2.](http://www.rsfjournal.org/content/8/7/70/F2)

**Figure 2.** 
Optimal Behavior Violin Plot by Actor and Object Status Group

*Source*: Authors’ tabulations.

Findings summarized in each cell are affected both by actors’ expected actions based on their status and object-persons’ expected treatment based on their status, because both factors influence cultural expectations about appropriate behavior in the interactions modeled. The top left cell, for example, summarizes the EPA distribution of optimal behaviors for all sixteen modeled events in which *actively revered* actors interact with *actively revered* objects. In that cell, we see that all of the optimal behaviors that *actively revered* actors direct at *actively revered* objects are uniformly positive (the plot for evaluation is concentrated in one value region), whereas the potency and activity of their behaviors are more variable. Our theoretical models consistently predict that when *actively revered* people interact with one another, they will expect and enact behaviors that are extremely positive, pleasant, and congenial.

Our models predict differences in the cultural meanings of optimal behaviors across status group interactions (table 2; for full simulation results, see the online appendix). The most pronounced differences occur based on the actor’s status group. *Actively revered* occupations are expected to engage in actions high in evaluation, potency, and activity (for example, praise, treat, reward). In contrast, the *disagreeably powerful* are expected to engage in moderately powerful actions low in evaluation and activity (such as ask about, show something to, escort). *Service-to-society* occupations are expected to engage in actions that are just below the *actively revered* in goodness and activity and moderate in potency (such as caution, chat with, consult). And *everyday specialists* are expected to engage in actions that are good but low in potency and activity (such as answer, serve, concur with).

Optimal behaviors also differ in some cases based on the status group of the object-person at whom they are directed. This is most noticeable in events with *disagreeably powerful* objects, who are recipients of actions lower in activity and sometimes evaluation than other status groups. We find the greatest consistency in behavior evaluation across the events modeled, with less consistency in potency and activity (figure 2).

Our results point to relationality of status dynamics, as actions are dependent, to some extent, on the status of both the actor and object-person. For example, *actively revered* actors such as firefighters are expected to engage in warmer actions toward those in the more positively evaluated *everyday specialist*, *service-to-society*, and *actively revered* occupations (for example, congratulating a musician, applauding a nutritionist) than toward those in less positively evaluated *disagreeably powerful* occupations (such as coaching a foreman, greeting a bailiff). *Disagreeably powerful* actors such as bailiffs are expected to engage in more powerful actions toward those in the less powerful *everyday specialist* and *service-to-society* occupations (for example, identifying a musician, escorting a nutritionist), than toward those in more powerful *actively revered* and *disagreeably powerful* occupations (such as asking about a surgeon, showing something to a foreman). Whereas firefighters receive honor, deference, and camaraderie from those in most other occupations (for example, defense attorneys toast, pediatricians praise, musicians joke with firefighters), the *disagreeably powerful* have their status affirmed in more relationally dependent ways. *Everyday specialists* (such as dental hygienists, technicians) are expected to serve bailiffs, whereas those in higher status occupations are expected to treat them more warmly (firefighters greet, pediatricians grin at bailiffs).

We also compared optimal behavior EPA for modeled events that involve low- versus high-prestige actors across status groups to determine whether this contributed to the greater variation detected along the potency and activity dimensions (for a summary of the results, see tables A.1 and A.2 and figure A.1). We indeed find that optimal behaviors differ in potency and activity, but not evaluation, for high- versus low-prestige occupations across all status groups. Specifically, behaviors are comparatively high in potency and low in activity for high-prestige actors in most status groups, while the reverse is true for low-prestige actors. The only status group that does not follow this pattern is the *disagreeably powerful*, for whom the opposite is true.12 For example, high-prestige *service-to-society* actors are expected to turn to or seek advice from (2.05, 1.13, –0.33) high-prestige *disagreeably powerful* objects. Low-prestige *service-to-society* actors are expected to call or chat with (2.01, 0.97, 1.22) low-prestige *disagreeably powerful* objects. This is consistent with Freeland and Hoey’s (2018) outlined argument about the relationship of deference scores and occupational prestige with EPA.

Several core predictions emerge from the theoretical models of optimal behavior. First, the models suggest that evaluation (goodness, pleasantness, esteem) has central importance in status behavior. Although the dominance and expressiveness of behavior among occupations varies across status groups, the positivity of the actions differs most strongly. This finding echoes the finding that forgiving someone (a very positive, potent, lively behavior, in our terms) confers more status than the powerful and lively, but nasty act of exacting revenge (Benard et al. 2022). Second, the status of the actor has a larger impact on expected behavior than that of the object-person. The columns of figure 2 show much more consistency than the rows do. This suggests that the esteem in which we hold someone says more about how we think they will treat others than it does about how we expect others to treat them. Third, occupational prestige (as traditionally measured) influences the dominance (potency) and expressiveness (activity) of status behavior but tells us nothing about the largest and most consistent difference in behavioral patterns—how positively (evaluation) people treat one another. Put differently, the effects of prestige on status behavior are always contingent on something prestige cannot directly explain—how good, bad, or evaluatively neutral the behavior seems.

Taken together, our theoretical predictions provide a different view of the behavioral implications of status hierarchy than those derived from the traditional focus on wealth and power. Our findings highlight the important role doing good (or not) can play in the actualization of one’s status through behavior viewed as appropriate to the situation—especially to the cultural meaning of one’s own identity. Attempts at dominance should be more effective, according to ACT, when culturally aligned with the situation at hand and the identities of those involved. Thus *actively revered* actors should have their status affirmed when their dominance is paired with warmth and expressiveness, and *disagreeably powerful* actors should have theirs affirmed through dominant actions that are less nice and more reserved. In other words, situated action can contribute to the (re)production of social hierarchies through multiple behavioral pathways. We now move to seeing whether our predictions are supported by the expectations of real people (participants in a Mechanical Turk vignette survey).

### Vignette Survey

On average, participants in the vignette survey had a 71 percent probability of choosing the theoretically optimal behavior over one randomly selected from a large corpus of potential social actions. The distance between the optimal and randomly chosen behavior also had a significant effect, meaning that respondents were more likely to choose the optimal behavior when random behaviors were more divergent in EPA from (culturally dissimilar to) the optimal behavior (see figure 3). Both the base rate of choosing the optimal behavior and the effect of distance between the two behavior choices indicate that the behaviors predicted by the theoretical models follow respondents’ two cultural expectations about status behavior.

![Figure 3.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/7/70/F3.medium.gif)

[Figure 3.](http://www.rsfjournal.org/content/8/7/70/F3)

**Figure 3.** 
Deflection by Distance of Random Behavior from Optimal Behavior

*Source*: Authors’ tabulations.

### Implications for Inequality

Ridgeway and her colleagues (Ridgeway 2014, 2019; Ridgeway and Markus 2022, this issue) argue that status contributes to inequality through two main pathways. First, status creates inequality as cultural beliefs about essential differences between identity groups generate interaction dynamics that produce and reproduce social hierarchies. Second, status stabilizes existing inequalities by legitimizing unequal allocations of power and resources based on attributions of greater worthiness to powerful and resource-rich actors. Here we briefly discuss the implications of our findings for status-based inequality generated through each of these pathways and consider the utility of our approach for understanding status dynamics in other domains.

In regard to the first pathway, evidence is consistent that occupational status as measured culturally with ACT stratifies how we expect individuals in different status groups to enact their status advantage or disadvantage in social encounters with others. By capturing occupational status in the three-dimensional cultural structure of goodness, potency, and activity, ACT enables us to measure the relative status of occupations within the broader cultural system (Freeland and Hoey 2018; Maloney 2020). The theory demonstrates how status differences produce behavior dynamics that uphold the social order (the findings here about the expected behaviors among status groups). We specifically show that patterns of relational deference emerge from the cultural meanings of occupational identities, producing distinctive behavioral norms for different status groups. Although the powerful actions of the *actively revered* are warm and expressive, the *disagreeably powerful* enact dominance in ways less nice and more reserved. Others honor, praise, toast, and applaud the *actively revered* but serve, answer, and turn to the *disagreeably powerful*. In the aggregate, these behavior patterns not only reify the occupational status structure but also have downstream consequences. For example, they can shape decisions to enter into particular professions (such as by influencing which identities are seen as desirable to occupy, or align with gendered self-sentiments) as well as workplace experiences and outcomes (for example, treatment by others, emotions, retention). Our work centers the relationality of occupational status as enacted in reference to and through interactions with other occupational identities.

In regard to the second pathway, deference dynamics based in ACT offer a novel approach to studying status in that it both contributes to and arises from inequalities in power and resources. Access to material resources and cultural scripts can afford actors greater ability to influence the definition of an interaction (Cast 2003). If individuals with high material resources want to gain status as shown here, they may try to create situational contexts that conform with the behavioral patterns expected (and even prescribed) for our highest status group, the *actively revered*. Basically, one can “buy” status when material resources or power allow one to control interactions in ways that create certain affective meanings. The rich can afford to be pleasant and supportive, if they choose to be so; they have the ability to acquire status by enacting behaviors that confer those meanings and pushing others into social roles that support them. Similarly, we see that high-status people may tend to move upward in prestige (material resources and power) over time as others respond to them more positively and afford them social opportunities to gain wealth and power. Our empirically validated theoretical models suggest at least two pathways by which high-prestige actors (who have greater material resources) might effectively actualize their advantage through status behavior in social interactions—via powerful actions that are warm and expressive or dispassionate and reserved. The pathway likely to be most effective will depend on the cultural meanings of the actor’s identity and, to a lesser extent, the object-person with whom they are interacting (Robinson, Smith-Lovin, and Zhou 2020).

This article focuses on behavior dynamics emergent from the occupational status structure, but the same methods can be applied to analyze status structures and dynamics for social identities (such as race, class, gender), role identities related to the major institutions that structure social life (parent-child, boss-employee), and even traits (assertive, agreeable) (Rogers 2019b). Maloney’s (2020) method for detecting OSGs can be used to uncover groups of identities with distinctive interaction patterns in a network of simulations involving identities of any type. Our approach can be used to understand the behavior and emotion dynamics emergent from structurally equivalent identity groupings (Maloney 2022; Maloney and Smith-Lovin 2021; Rogers 2021b). This is because all identities carry cultural meanings with implications for social norms regarding the appropriate behavior, treatment, and emotions of identity groups.

Kimberly Rogers (2019b, 2021b), for example, finds that privileged social identities (such as white, rich, cisgender, heterosexual) carry distinctive cultural meanings that are higher than all other social identity groups in potency but lower than all other groups in evaluation and activity. These cultural meanings affect behavior dynamics and emotional experiences in interactions between the privileged and members of other identity groups. Privileged actors are expected to engage in significantly nicer behavior toward powerful than weak object-persons and to direct the least dominant actions toward in-group members. They experience the most positive and activated emotions in interactions with highly evaluated object-persons closest to them in power and the least positive emotions in interactions with weak object-persons. These stratified behavior and emotion dynamics reproduce and justify inequalities between social identity groups much as the occupational status dynamics observed here do.

## DISCUSSION

This article shows how deference scores, as a cultural measure of status, differentiate occupations in a distinct way from prior operationalizations of prestige that more closely map onto material features of occupations, such as income or education. Deference scores based in ACT leverage culture-specific identity and behavior meanings and impression-change models to predict relational status behavior and can be used to identify status inequalities at both the cultural and interactional level. They therefore capture the cultural, relational, and multilevel nature of status (Ridgeway and Markus 2022, this issue). The lower correlation of deference versus prestige with material resources also suggests that deference is better suited to disentangling inequalities based on status and resources.

Using theoretical models of status behavior, we predict which behaviors are most culturally expected in interactions between actors and objects belonging to differing status groups and examine how predictions differ for occupations with low versus high prestige. We find appreciable differences in predicted behavior EPA based on the status group of the actor and, less often, object-person involved in an event. These behaviors reflect the cultural meanings associated with occupational identities. For example, *actively revered* actors are predicted to engage in warm, expressive, and powerful actions, and *disagreeably powerful* actors are expected to be less nice and more reserved as they enact their power. More powerful, less lively behaviors were expected for high- than low-prestige occupations across most status groups; the goodness of predicted behavior did not differ appreciably by prestige. Our predictions were validated through a vignette survey. Respondents were more likely to report that behaviors which were close to culturally optimal were more likely to occur between a given pair of occupations.

In all, we endeavored to show how status differences between occupational identities shape interactions differently than material resource (prestige) levels. Status expectations guide interactions between occupational identities, shaping what sorts of behaviors are enacted. These behavioral expectations, then, uphold status asymmetries.

We see several possible avenues for future research on this subject. Here, we use dictionaries and impression-change models developed for U.S. English language culture, but future research could examine whether status dynamics operate differently in other cultural contexts where, for example, judgments of morality are a more important basis of status than competence. Future work could also examine the extent of consensus in deference scores by social position in a similar manner to Lauren Valentino’s work on occupational prestige (2022, this issue). Although the fundamental sentiments in EPA used here are highly consensual, variation (especially in potency and activity ratings) by social position may still be enough for different behavioral patterns by race or class. In addition, scholarship building on our approach to examine the efficacy of different behavior pathways for legitimizing power and resource advantages through status could be particularly fruitful. Given that ACT has historically been used to generate both expectations about and normative prescriptions for behavior (Heise 2007), this line of theoretical development seems particularly promising.

Finally, this article treats occupational interactions as a function of general occupational status identities. Most occupational interactions, of course, occur within narrower institutional frameworks, and involve interactions with nonoccupational actors. For example, doctors interact in medical settings with patients as well as other medical personnel, such as nurses, physical therapists, lab workers, secretaries, hospital administrators, insurance claim processors, and so on. The interactions among status groups analyzed here are more like those among occupational occupants outside work, such as when a doctor and a bill collector serve on a jury together or meet at a party or neighborhood association meeting. Their occupational status is relevant there in the same ways that other status identities such as race-ethnicity, gender, sexuality, and physical attractiveness might be. However, future research should explore the more contextualized, institutionally anchored interactions among occupational roles within work settings.

## Appendices

View this table:
[Table A.1.](http://www.rsfjournal.org/content/8/7/70/T3)

**Table A.1.** 
High-Prestige Optimal Behaviors

View this table:
[Table A.2.](http://www.rsfjournal.org/content/8/7/70/T4)

**Table A.2.** 
Low-Prestige Optimal Behaviors

![Figure A.1.](http://www.rsfjournal.org/https://www.rsfjournal.org/content/rsfjss/8/7/70/F4.medium.gif)

[Figure A.1.](http://www.rsfjournal.org/content/8/7/70/F4)

**Figure A.1.** 
Optimal Behavior Violin Plot

*Source*: Authors’ tabulations.

## FOOTNOTES

*   1. Power can be derived from control over economic and social resources, including status (Weber 1978, 2014).

*   2. Robert Freeland and Jesse Hoey (2018) use a Bayesian version of ACT that allows for variation and uncertainty in the cultural meanings associated with occupational identities and behaviors (for a description of BayesACT and its properties, see Schröder, Hoey, and Rogers 2016). Given that occupational identities usually have well-institutionalized and widely shared cultural meanings (Heise 2010; Ridgeway 2019), we use the non-Bayesian ACT that assumes a point estimate in three-dimensional space for our analyses in this article.

*   3. The estimation of these impression-change equations from simple actor-behavior-object event vignettes is described in detail elsewhere (see Smith-Lovin and Heise 1988; Heise 2010; Morgan, Rogers, and Hu 2016).

*   4. ACT also has an explicit model of attribution and emotion, but we do not describe it here because we do not use this aspect of the theory in this article.

*   5. Prior work has generally found high consensus in EPA ratings within cultures, but some differences in these ratings are associated with race and class (see Ambrasat et al. 2014; Rogers 2019c). This mirrors the findings of Valentino (2022, this issue) about different “status lenses” on occupational prestige. We identify this as an area for future research in the discussion section.

*   6. Ridgeway (2019) has in her recent status work more explicitly brought in motivations to further group interests, but this has not historically been a major feature of expectation states theory. The principle that status is a reward for high-competence people devoting themselves to positive, shared group outcomes is, however, core to that theory.

*   7. All EPA profiles come from a compilation of ratings by undergraduates at two universities in a collaborative project conducted from 2012 through 2014 (Smith-Lovin et al. 2016).

*   8. For a list of all 303 occupational identities including their EPA ratings, GSS occupational markers, deference scores, SIOPS, and occupational status group, see online appendix A ([https://www.rsfjournal.org/content/8/7/70/tab-supplemental](https://www.rsfjournal.org/content/8/7/70/tab-supplemental)). Maloney (2020) eliminates one occupational identity, dishwasher, because respondents interpreted this identity as a machine rather than a person. For a full replication repository, see [https://github.com/ekmaloney/rsf\_occ_status](https://github.com/ekmaloney/rsf_occ_status) (accessed June 19, 2022).

*   9. A table with descriptive statistics for the validation sample and a more detailed description of the data collection process is in section B of the online appendix.

*   10. A full list of the stimuli, including events, the optimal behavior term, the randomly selected behavior term, and the distance between the two are in section C of the online appendix.

*   11. For a more detailed summary of modeling choices and results from all three models as well as robustness checks, see section D of the online appendix.

*   12. This pattern has some exceptions, the most noteworthy being the slightly lower evaluation levels found among low-prestige actors in the *everyday specialists* and *service-to-society* status groups.

*   © 2022 Russell Sage Foundation. Maloney, E. K., Kimberly B. Rogers, and Lynn Smith-Lovin. 2022. “Status as Deference: Cultural Meaning as a Source of Occupational Behavior.” *RSF: The Russell Sage Foundation Journal of the Social Sciences* 8(7): 70–88. DOI: 10.7758/RSF.2022.8.7.04. The authors thank Robert Freeland and Jesse Hoey for providing the deference scores used in this research. We also thank Kieran Healy, Mark Chaves, the Social Psychology and Worldview Workshops at Duke University, and the Writers’ Workshop at Dartmouth College for their helpful feedback on this article. This research was supported by NSF GRANT #DGE-1644868. Direct correspondence to: E. K. Maloney, at em.k.maloney{at}duke.edu, Department of Sociology, Box 90088, 417 Chapel Drive, Duke University, Durham, NC 27708-0088, United States; Kimberly Rogers, at krogers{at}dartmouth.edu, Department of Sociology, Dartmouth College, 20 North Main Street, Hanover, NH 03755, United States; or Lynn Smith-Lovin, at smithlov{at}duke.edu, Department of Sociology, Box 90088, 417 Chapel Drive, Duke University, Durham, NC 27708-0088, United States.

Open Access Policy: *RSF: The Russell Sage Foundation Journal of the Social Sciences* is an open access journal. This article is published under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.

## REFERENCES

1.  Ambrasat, Jens, Christian von Scheve, Gesche Schauenburg, Markus Conrad, and Tobias Schröder. 2014. “Consensus and Stratification in the Affective Meaning of Human Sociality.” Proceedings of the National Academy of Sciences 111(22): 8001–06.
    
    [Abstract/FREE Full Text](http://www.rsfjournal.org/lookup/ijlink/YTozOntzOjQ6InBhdGgiO3M6MTQ6Ii9sb29rdXAvaWpsaW5rIjtzOjU6InF1ZXJ5IjthOjQ6e3M6ODoibGlua1R5cGUiO3M6NDoiQUJTVCI7czoxMToiam91cm5hbENvZGUiO3M6NDoicG5hcyI7czo1OiJyZXNpZCI7czoxMToiMTExLzIyLzgwMDEiO3M6NDoiYXRvbSI7czoxOToiL3JzZmpzcy84LzcvNzAuYXRvbSI7fXM6ODoiZnJhZ21lbnQiO3M6MDoiIjt9) 

2.  Anderson, Cameron, Robb Willer, Gavin J. Kilduff, and Courtney E. Brown. 2021. “The Origins of Deference: When Do People Prefer Lower Status?” Journal of Personality and Social Psychology 102(5): 1077–88.
    
    

3.  Benard, Stephen, Long Doan, D. Adam Nicholson, Emily Meanwell, Eric L. Wright, and Peter Lista. 2022. “To Forgive Is Divine? Morality and the Status Value of Intergroup Revenge and Forgiveness.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(6): 122–39. DOI: [https://doi.org/10.7758/RSF.2022.8.6.07](https://doi.org/10.7758/RSF.2022.8.6.07).
    
    

4.  Bukodi, Erzsebet, Shirley Dex, and John H. Goldthorpe. 2011. “The Conceptualisation and Measurement of Occupational Hierarchies: A Review, A Proposal and Some Illustrative Analyses.” Quality & Quantity 45(3): 623–39.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1007/s11135-010-9369-x&link_type=DOI) 

5.  Cast, Alicia D. 2003. “Power and the Ability to Define the Situation.” Social Psychology Quarterly 66(3): 185–201.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.2307/1519821&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000222046500001&link_type=ISI) 

6.  1.   John Delamater
    
    Correll, Shelley J., and Cecilia L. Ridgeway. 2003. “Expectation States Theory.” In *Handbook of Social Psychology*, edited by  John Delamater. New York: Klewer Academic/Plenum Publishers.
    
    

7.  Dippong, Joseph, and Will Kalkhoff. 2015. “Predicting Performance Expectations from Affective Impressions: Linking Affect Control Theory and Status Characteristics Theory.” Social Science Research 50(1): 1–14.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1016/j.ssresearch.2014.11.001&link_type=DOI) 
    
    [PubMed](http://www.rsfjournal.org/lookup/external-ref?access_num=25592917&link_type=MED&atom=%2Frsfjss%2F8%2F7%2F70.atom) 

8.  Featherman, David L., and Robert M. Hauser. 1976. “Prestige or Socioeconomic Scales in the Study of Occupational Achievement?” Sociological Methods and Research 4(4): 403–22.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/004912417600400401&link_type=DOI) 

9.  Freeland, Robert E., and Jesse Hoey. 2018. “The Structure of Deference: Modeling Occupational Status Using Affect Control Theory.” American Sociological Review 83(2): 243–77.
    
    

10. Ganzeboom, Harry B.G., Paul M. De Graaf, and Donald J. Treiman. 1992. “A Standard International Socio-Economic Index of Occupational Status.” Social Science Research 21(1): 1–56.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1016/0049-089X(92)90017-B&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=A1992HJ53500001&link_type=ISI) 

11. Ganzeboom, Harry B.G., and Donald J. Treiman. 1996. “Internationally Comparable Measures of Occupational Status for the 1988 International Standard Classification of Occupations.” Social Science Research 25(3): 201–39.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1006/ssre.1996.0010&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=A1996VD26900001&link_type=ISI) 

12. Heise, David R. 1979. *Understanding Events: Affect and the Construction of Social Action*. New York: Cambridge University Press.
    
    

13. Heise, David R. 2007. *Expressive Order: Confirming Sentiments in Social Actions*. New York: Springer.
    
    

14. Heise, David R. 2010. *Surveying Cultures: Discovering Shared Conceptions and Sentiments*. New York: John Wiley & Sons.
    
    

15. Koenig, Biko. 2022. “Politicizing Status Loss Among Trump Supporters in 2020.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(6): 69–86. DOI: [https://doi.org/10.7758/RSF.2022.8.6.04](https://doi.org/10.7758/RSF.2022.8.6.04).
    
    

16. MacKinnon, Neil J. 1994. *Symbolic Interactionism as Affect Control*. Albany: State University of New York Press.
    
    

17. Maloney, Em K. 2020. “The Social Psychology of Occupational Status Groups: Relationality in the Structure of Deference.” Social Psychology Quarterly 83(4): 463–75.
    
    

18. Maloney, Em K. 2021. “inteRact: Implements Affect Control Theory Equations in R.” Accessed June 18, 2022. [https://ekmaloney.github.io/inteRact/index.html](https://ekmaloney.github.io/inteRact/index.html).
    
    

19. Maloney, Em K. 2022. “The Influence of Occupational Identity on Emotional Experience.” *American Behavioral Scientist*. First published online: January 2022. DOI: [https://doi.org/10.1177/00027642211066053](https://doi.org/10.1177/00027642211066053).
    
    

20. Maloney, Em K., and Lynn Smith-Lovin. 2021. “The Emotional Implications of Occupational Deference Structures.” Advances in Group Processes 38(1): 1–21.
    
    

21. Morgan, Jonathan H., Kimberly B. Rogers, and Mao Hu. 2016. “Distinguishing Normative Processes from Noise: A Comparison of Four Approaches to Modeling Impressions of Social Events.” Social Psychology Quarterly 79(4): 311–32.
    
    

22. Osgood, Charles E., William H. May, and Murray S. Miron. 1975. *Cross-Cultural Universals of Affective Meaning*. Urbana: University of Illinois Press.
    
    

23. Osgood, Charles E., George J. Suci, and Percy Tannenbaum. 1957. *The Measurement of Meaning*. Urbana: University of Illinois Press.
    
    

24. Ridgeway, Cecilia L. 2014. “Why Status Matters for Inequality.” American Sociological Review 79(1): 1–16.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/0003122413515997&link_type=DOI) 

25. Ridgeway, Cecilia L. 2019. *Status: Why Is It Everywhere? Why Does It Matter?* New York: Russell Sage Foundation.
    
    

26. Ridgeway, Cecilia L., and Shelley J. Correll. 2006. “Consensus and the Creation of Status Beliefs.” Social Forces 85(1): 431–53.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1353/sof.2006.0139&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000243493900020&link_type=ISI) 

27. Ridgeway, Cecilia L., and Hazel Rose Markus. 2022. “The Significance of Status: What It Is and How It Shapes Inequality.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(7): 1–25. DOI: [https://doi.org/10.7758/RSF.2022.8.7.01](https://doi.org/10.7758/RSF.2022.8.7.01).
    
    

28. Ridgeway, Cecilia L., and Lynn Smith-Lovin. 1994. “Structure, Culture and Interaction: A Comparison of Affect Control Theory and Expectations States Theory.” Advances in Group Processes 11(1): 213–39.
    
    

29. 1.   Peter J. Burke
    
    Robinson, Dawn T., and Lynn Smith-Lovin. 2018. “Affect Control Theories of Social Interaction and Self.” In *Contemporary Social Psychological Theories*, 2nd ed., edited by  Peter J. Burke. Stanford, Calif.: Stanford University Press.
    
    

30. 1.   Richard T. Serpe, 
    2.   Robin Stryker, and 
    3.   Brian Powell
    
    Robinson, Dawn T., Lynn Smith-Lovin, and Jun Zhao. 2020. “The Role of the Other: How Interaction Partners Influence Identity Maintenance in Four Cultures.” In *Identity and Symbolic Interaction: Deepening Foundations, Building Bridges*, edited by  Richard T. Serpe,  Robin Stryker, and  Brian Powell. New York: Oxford University Press.
    
    

31. Rogalin, Christabel L., Shane D. Soboroff, and Michael J. Lovaglia. 2007. “Power, Status, and Affect Control.” Sociological Focus 40(2): 202–20.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1080/00380237.2007.10571306&link_type=DOI) 

32. Rogers, Kimberly B. 2019a. “Do You See What I See? Testing for Individual Differences in Impressions of Events.” Social Psychology Quarterly 81(2): 149–72.
    
    

33. 1.   Jan E. Stets and 
    2.   Richard T. Serpe
    
    Rogers, Kimberly B. 2019b. “Identity Meanings and Categorical Inequality.” In *Identities in Everyday Life*, edited by  Jan E. Stets and  Richard T. Serpe. New York: Oxford University Press.
    
    

34. Rogers, Kimberly B. 2019c. “Sources of Consensus and Variegation in Cultural Affective Meanings.” Social Currents 6(3): 219–38.
    
    

35. Rogers, Kimberly B. 2021a. “Event Likelihood Judgments Revisited.” Social Psychology Quarterly 84(2): 177–88.
    
    

36. Rogers, Kimberly B. 2021b. “Separate and Unequal: Predicting Intergroup Behavior and Emotions from Social Identity Meanings.” Advances in Group Processes 38(1): 23–52.
    
    

37. Rogers, Kimberly B., Tobias Schröder, and Wolfgang Scholl. 2013. “The Affective Structure of Stereotype Content: Behavior and Emotion in Intergroup Context.” Social Psychology Quarterly 76(2): 125–50.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/0190272513480191&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000319526600002&link_type=ISI) 

38. Scholl, Wolfgang. 2013. “The Socio-Emotional Basis of Human Interaction and Communication: How We Construct Our Social World.” Social Science Information 52(1): 3–33.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/0539018412466607&link_type=DOI) 
    
    [Web of Science](http://www.rsfjournal.org/lookup/external-ref?access_num=000315169900001&link_type=ISI) 

39. Schröder, Tobias, Jesse Hoey, and Kimberly B. Rogers. 2016. “Modeling Dynamic Identities and Uncertainty in Social Interactions: Bayesian Affect Control Theory.” American Sociological Review 81(4): 828–55.
    
    [CrossRef](http://www.rsfjournal.org/lookup/external-ref?access_num=10.1177/0003122416650963&link_type=DOI) 

40. Schröder, Tobias, Kimberly B. Rogers, Shuichirou Ike, Julija N. Mell, and Wolfgang Scholl. 2013. “Affective Meanings of Stereotyped Social Groups in Cross-Cultural Comparison.” Group Processes and Intergroup Relations 16(6): 717–33.
    
    

41. 1.  Principal Investigator, 
    2.   Tom W. Smith; Co-Principal Investigators, 
    3.   Peter V. Marsden, and 
    4.   Michael Hout, NORC ed
    
    Smith, Tom W., Michael Davern, Jeremy Freese, and Michael Hout. 2017. *General Social Surveys, 1972–2016*. [machine-readable data file]. Principal Investigator,  Tom W. Smith; Co-Principal Investigators,  Peter V. Marsden, and  Michael Hout, NORC ed. Chicago: NORC.
    
    

42. Smith-Lovin, Lynn, and David R. Heise. 1988. *Analyzing Social Interaction: Advances in Affect Control Theory*. New York: Taylor & Francis.
    
    

43. Smith-Lovin, Lynn, Dawn T. Robinson, Bryan C. Cannon, Jesse K. Clark, Brent H. Curdy, Robert Freeland, Jonathan H. Morgan, and Kimberly B. Rogers. 2016. “Mean Affective Ratings of 929 Identities, 814 Behaviors, and 660 Modifiers by University of Georgia and Duke University Undergraduates in 2012–2014.” University of Georgia, Affect Control Theory. Accessed June 19, 2022. [https://affectcontroltheory.org/resources-for-researchers/data-sets-for-simulation/](https://affectcontroltheory.org/resources-for-researchers/data-sets-for-simulation/).
    
    

44. Treiman, Donald J. 1977. *Occupational Prestige in Comparative Perspective*. New York: Academic Press.
    
    

45. Valentino, Lauren. 2022. “Status Lenses: Mapping Hierarchy and Consensus in Status Beliefs.” RSF: The Russell Sage Foundation Journal of the Social Sciences 8(7): 89–110. DOI: [https://doi.org/10.7758/RSF.2022.8.7.05](https://doi.org/10.7758/RSF.2022.8.7.05).
    
    

46. 1.   Guenther Roth and 
    2.   Claus Wittich
    
    Weber, Max. 1978. *Economy and Society: An Outline of Interpretive Sociology*, vol. 1, edited by  Guenther Roth and  Claus Wittich. First published 1918. Berkeley: University of California Press.
    
    

47. 1.   Hans H. Gerth and 
    2.   C. Wright Mills
    
    Weber, Max. 2014. *From Max Weber: Essays in Sociology*, edited by  Hans H. Gerth and  C. Wright Mills. New York: Routledge.
    
    

48. Zhao, Jun. 2022. “Modeling Impression Formation Processes Among Chinese and Americans.” American Behavioral Scientist. First published online: January 22, 2022. DOI: [https://doi.org/10.1177/00027642211066025](https://doi.org/10.1177/00027642211066025).