Tough times at the top: Occupational status predicts changes in job satisfaction in times of crisis

How do individuals with a higher versus lower occupational status experience major, unexpected changes to their work life? The COVID-19 pandemic has disrupted most areas of work life and, thus, provides a unique opportunity to examine changes in work attitudes in response to a worldwide crisis. We predict that individuals with higher, but not with lower occupational status showed a decline in job satisfaction during the early stages of the COVID-19 pandemic in Germany (1st lockdown; March to May 2020), with subsequent recovery to initial job satisfaction levels. Based on role theory and social-psychological theories of hierarchical differentiation, we argue that, due to the profound work-related changes, individuals with higher (vs. lower) occupational status are more negatively affected in realizing their work goals and, thus, experience decreasing levels of job satisfaction. To test these predictions, we investigated trajectories of job satisfaction between December 2019 and August 2020 (7 measurement waves; N = 1583). Results of piece-wise growth curve models showed that individuals with higher occupational status showed a steeper decline in job satisfaction (followed by recovery) over time, whereas individuals with medium and lower occupational status did not experience a significant change in job satisfaction. In addition, we show that the decline in job satisfaction is moderated by perceived constraints at work associated with the pandemic among individuals with higher occupational status. Overall, these findings contribute to our understanding of the link between occupational status and job satisfaction in times of crisis.


Introduction
After the World Health Organization declared COVID-19 a global pandemic on March 11th 2020, people across the globe were exposed to fundamental social and economic changes. Work has been one of the most exposed and disrupted areas of life. The pandemic removed individuals from their workplaces and isolated them from their colleagues and superiors. Thus, the consequences of the pandemic have not only led to a profound public-health but also to an unprecedented social and economic crisis and many people found themselves confronted with threats and insecurities (e.g., social isolation, loss of income, job loss, changes in demands, bankruptcy; see Kniffin et al., 2021;Rudolph et al., 2021). For example, initial evidence suggests that a national lockdown in response to the pandemic was associated with declines in self-rated job performance . Moreover, studies have shown that people's subjective and occupational well-being decreased during national lockdowns associated with the COVID-19 pandemic (Meyer et al., 2021;Syrek et al., 2021;Zacher & Rudolph, 2021c). After months of lockdowns, organizations have been demanding that workers get back to the office (Johnston, 2022). It thus appears that especially those in higher positions of power have grappled with the suddenly imposed work-from-home restrictions and that many seek to return to pre-pandemic face-to-face work arrangements.
Surprisingly, however, despite the fact that occupational status is one of the most central features at work and in organizations (e.g., Magee & Galinsky, 2008), there is almost no insight into how individuals at different occupational ranks have responded to the dramatic changes associated with the COVID-19 pandemic. Acknowledging that the pandemic has strongly affected those with lowlevel and precarious jobs (Cubrich et al., 2022), we suggest that, nevertheless, the unforeseen occurrence of the pandemic may also have substantially impacted individuals in higher-level positions. Specifically, we argue that many aspects relevant to the work role of higher-status individuals, such as the enactment of power and control, were significantly altered, challenged, and constrained by the pandemic. In the current study, we investigate trajectories of job satisfaction across the early stages of the COVID-19 pandemic in Germany (December 2019 to August 2020), a period that includes the first national lockdown between March and May 2020. We examined whether people with a higher or lower occupational status might differ in their response to the disruption induced by the first national lockdown (i.e., transition and recovery), and specifically in their trajectories of job satisfaction across time (Bliese et al., 2017;Bliese & Lang, 2016). Based on social-psychological theorizing on hierarchical differentiation (Guinote, 2007;Keltner et al., 2003;Sapolsky, 2004;Scheepers & Ellemers, 2018), as well as role theory (Ashforth, 2001;Biddle, 1986Biddle, , 2013, we predicted that people with a higher (vs. lower) occupational status might be at a greater risk to experience a decline in job satisfaction during the lockdown, but also expected that job satisfaction would subsequently recover. We further expected that this effect depends on the extent to which high-status individuals perceive constraints regarding their ability to work because of the pandemic and its consequences.
The current study contributes to and broadens the literature on occupational status as well as on power and status in organizations, which has typically shown that individuals with higher status can better attain their goals, have more positive job experiences, less stress, and better well-being (e.g., Keltner et al., 2003;Robie et al., 1998;Sherman et al., 2012). We demonstrate that higher occupational status can be linked to more profound and, at times negative, changes in job satisfaction across time and through a major crisis. We also point to rank-based differences in the influence of perceived constraints at work due to the pandemic as a potential boundary condition for the hypothesized decline and recovery among those with higher (vs. lower) occupational status.

Occupational status and job satisfaction
Work is usually hierarchically organized with fewer individuals at the top (e.g., executives, managers) and many further down (i.e., employees) in terms of occupational status. Occupational status can be defined as the ranking of formal positions that individuals occupy in a given organizational hierarchy. Hierarchical differentiation within an organization can be beneficial for organizational outcomes as it enables efficient coordination among its members (Berger et al., 1972;Magee & Galinsky, 2008;Ridgeway, 1982). Nonetheless, a body of research suggests that a higher position within a hierarchy is also linked to more resources and many beneficial outcomes at the individual level (Korman et al., 2022). For example, those individuals with a higher occupational status enjoy a wide range of advantages, such as more autonomy and access to career opportunities, higher levels of social status (e.g., more admiration, prestige, and influence in the eyes of other people) and power (asymmetric control over valued resources to influence others), as well as better well-being and health (Fiske, 2010;Singh-Manoux et al., 2005).
There is robust evidence that an advanced position in the hierarchy is linked to higher subjective well-being (see Tan et al., 2020, for a meta-analysis). Higher occupational status has been linked to more favorable work conditions and less stress which, in turn, may contribute to higher levels of job satisfaction (Robie et al., 1998;Sherman et al., 2012). Job satisfaction is one of the most widely investigated work outcomes and represents an indicator of domain-specific subjective well-being in the occupational context. It has been defined as a positive affective state that results from people's appraisal of their job experiences (Brief & Weiss, 2002;Locke, 1976). Job satisfaction is important because it is positively linked to job performance (Judge et al., 2001;Riketta, 2008) and life satisfaction (Iverson & Maguire, 2000;Judge & Watanabe, 1993). There are strong theoretical reasons to assume that higher occupational status is linked to higher job satisfaction. For one, higher-level jobs tend to involve more appealing job characteristics, such as challenging and motivating tasks, better pay, better promotion prospects, and higher degrees of autonomy and responsibility (Robie et al., 1998). In line with this, research has confirmed that holding a leadership position is associated with lower levels of stress (Sherman et al., 2012). Specifically, this study showed that a higher occupational status was associated with a greater sense of power and less anxiety and reduced cortisol reactivity.
Power often results from holding a higher occupational status in an organization (Magee & Galinsky, 2008;Wisse et al., 2019). For example, research suggests that higher ranking members of an organization hold more formal power than lower ranking members, maintaining formal authority to influence subordinates (Yukl & Falbe, 1991). Specifically, holding power entails control over others as well as a sense of assertiveness, autonomy, and being uncontrolled by others. Studies have revealed that the psychological experience of power is associated with agency, action orientation, and goal pursuit (Guinote, 2007;Keltner et al., 2003). For example, power approach-inhibition theory (Keltner et al., 2003) argues that having power activates the behavioral approach system (i.e., focus on rewards and pursuit of goals), whereas a lack of power activates the behavioral inhibition system (i.e., focus on threats and constraints, heightened vigilance). Moreover, situated focus theory of power (Guinote, 2007) proposes that power affords individuals with greater freedom from constraints and greater agency leading to more effective goal pursuit. Taken together, evidence suggests that individuals with a higher occupational status hold higher power that may help them to realize their goals, which may ultimately contribute to higher job satisfaction.
Yet, higher occupational status may not imply that better access to resources and influence are always uncontested. Because hierarchies are subject to changes with people gaining and losing status and power, a large body of research has shown that the rank-wellbeing link may shift depending on the stability of a given hierarchy (Feenstra et al., 2017;Knight & Mehta, 2017;Sapolsky, 2004;Scheepers & Ellemers, 2018;Sivanathan et al., 2008;Weiss & Kunzmann, 2020). Thus, although hierarchies are largely stable and self-reinforcing (Magee & Galinsky, 2008), the composition of hierarchies can be challenged by external influences, for example, through severe events and changing circumstances that affect the privileges that come with higher status (Sapolsky, 2011). Individuals who are at the top of the occupational hierarchy may enjoy more positive outcomes such as higher job satisfaction if they can take their status position and the associated power for granted. However, if these means of influence are challenged, obstructed, or blocked-for example by the changes induced by the COVID-19 pandemic-individuals with higher occupational status may experience a decline in job satisfaction, whereas those with lower occupational status may not.
Arguably, it seems likely that individuals with higher occupational status should have more resources to cope with the constraints imposed by the pandemic. However, according to role theory, the enactment of power and control is a central aspect of holding a higher occupational status role (Biddle, 2013;Yukl & Falbe, 1991). In line with this, research has operationalized power in terms of formal roles such as individuals' position in the organizational hierarchy, which afford them control over resources and opportunities to influence subordinates (Guinote, 2007;Magee & Galinsky, 2008;Wisse et al., 2019;Yukl & Falbe, 1991). For example, face-to-face contact and being able to supervise and monitor the work of others is a major aspect of the work role of higher status individuals (Felstead et al., 2003). The tenets of role theory posit that occupational roles provide individuals with a set of expectations attached to a certain position that establish and generate a scheme of tasks and responsibilities (Ashforth, 2001;Biddle, 1986Biddle, , 2013. When these role-related expectations mismatch with reality due to changing circumstances, individuals feel constrained in enacting their work role. This may lead, in turn, to feelings of ambiguity and conflict impairing their satisfaction with their job. For instance, due to changing circumstances (i.e., the COVID-19 pandemic) individuals with a higher occupational status (e.g., managers) may experience obstacles in their ability to exercise power. Key managerial functions pertaining to exercising power, such as monitoring, evaluating, and correcting work processes and outcomes were significantly altered or obstructed, because many employees were working from home and face-to-face contact with superiors was substantially reduced or completely abolished. If higher ranking individuals feel that their ability to influence others is obstructed, discrepancies between role-based expectations and current circumstances may increase and job satisfaction may suffer.
Therefore, we argue that role conflict is induced by the pandemic as an unforeseen and major stressor that ultimately challenges established and well-known work procedures especially for individuals higher up in the organizational hierarchy (Kahn & Quinn, 1970). As a consequence, realizing that the means that are necessary to fulfill the demands of one's higher status work role are lacking may undermine their subjective well-being and in particular their job satisfaction. In support of this argument, previous research suggests that power holders react negatively to prospective loss of power (Deng et al., 2018) and when they perceive their power position as unstable, they experience elevated levels of stress (Jordan et al., 2011;Sapolsky, 2011). Research also shows that if individuals in positions of power feel that they personally lack power, they are more sensitive to threats, experience more negative emotions, and utilize power in inefficient and less effective ways (Bugental, 2010;Bugental & Lewis, 1999;Fast & Chen, 2009).
The first national lockdown in Germany as a response to the COVID-19 pandemic represented an unprecedented challenge and disruption to people's (work) life. We argue that the profound consequences of the COVID-19 pandemic represented a fundamental and challenging event especially for individuals with a higher occupational status with regard to realizing their goals. Due to the various restrictions that were put in place (e.g., working from home, reduced working hours, social distancing), many aspects pertaining to the enactment of power and control in high status occupational positions were significantly altered, challenged, and constrained, thus significantly changing the opportunity structures and privileges that were normally associated with this position. Therefore, we predict that individuals with a higher (as compared to a lower) occupational status should experience a stronger decrease in job satisfaction during the early stages of the COVID-19 pandemic in Germany (December 2019 to August 2020), including the period of the first national lockdown from March to May 2020. In addition, we assume that, once the pandemic-related restrictions were lifted (i.e., individuals returning to presential work), the job satisfaction of individuals with higher (vs. lower) occupational status would increase again (i.e., recover) over time.
Hypothesis 1. Changes during the early stages of the COVID-19 pandemic in job satisfaction over time depend on occupational status: Employees with a higher occupational status experience a stronger decline in job satisfaction (with subsequent recovery) than employees with lower occupational status.
Second, we suggest that the hypothesized decline in job satisfaction among higher-status individuals depends on perceived constraints at work associated with the pandemic. Due to the pandemic and the subsequent measures that were put in place to control the spreading of COVID-19 (e.g., working from home, telework, shortened work schedules, social distancing), individuals in higher-level jobs were more likely to have been obstructed in their ability to enact their work role and responsibility (e.g., power) than individuals in lower-level jobs. In organizations, taking responsibility and managing people is an essential part of occupying a higher occupational or organizational status that entails an understanding of the structural constraints of the organization (Magee & Galinsky, 2008). During the first lockdown, individuals who occupied higher positions in the occupational hierarchy may have been more or less drastically confronted with hindrances (i.e., obstacles to goal attainment; see Tuckey, 2015) in carrying out their responsibilities associated with their work role as compared to lower status employees. Although some might have experienced more obstacles and greater difficulty in performing essential managerial tasks, others might not have been as affected (e.g., those who are used to teleworking or industries in which in-person work continued despite the lockdowns). Therefore, higher perceived work constraints associated with working during the pandemic should be an important boundary condition to changes in job satisfaction among higher ranking employees.
By contrast, these perceived constraints should be less detrimental to the job satisfaction of individuals further down in the occupational hierarchy, because these constraints should have been less relevant to realizing work-related goals with regard to lower status work roles which are typically less associated with exercising power (e.g., less decision autonomy, less control over valued resources). Therefore, individuals with a higher occupational status who feel that their work role and associated sphere of influence at work was constrained due to the pandemic should report lower levels of job satisfaction.
Hypothesis 2. The decline in job satisfaction among higher-status individuals depends on perceived constraints at work associated with the pandemic: Higher levels of perceived constraints are associated with a steeper decline in job satisfaction.

Study design, participants, and procedures
We conducted a longitudinal study and measured job satisfaction seven times, first at the beginning (i.e., first week) of December 2019 (Time [T] 1), then again at the beginning of March (T2), April (T3), May (T4), June (T5), July (T6), and August (T7). Additionally, occupational status was measured at baseline (T1), and perceived constraints at work during the first national lockdown were measured at T4 (May; retrospectively for April 2020). Data for this study were collected as part of a larger longitudinal data collection effort, and so far eight other studies based on the same dataset, but with completely different research questions and completely different substantive variables, have been published (Koziel et al., 2021;Rauvola et al., 2022;Rudolph et al., 2021Rudolph et al., , 2022Weiss et al., 2022;Zacher & Rudolph, 2021a, 2021bZacher et al., 2021). A professional panel company was commissioned to recruit participants from a nationally representative online panel in Germany. The company is ISO 26362 certified, which ensures quality of the survey data. To be eligible to participate, participants had to be at least 18 years old and be working full-time at each measurement wave. At T1, 4839 persons in the company's database were contacted. Of these persons, 1583 initiated the survey and provided at least partial responses (e.g., age, gender, job satisfaction; response rate of 32.71 %) to the T1 survey. Of these 1583 persons, 637 provided complete responses at all seven time points (40.24 %).
Participants' ages ranged from 18 to 69 years with a mean age of 43.69 years (SD = 11.18), and 45 % were women. In terms of educational level, a majority held either intermediate secondary school/high school (43.5 %) or college/university or technical college diplomas (56.5 %).
To address systematic patterns of attrition, incomplete responders (n = 1583) were compared to the panel of complete responders (n = 637) on key demographic and substantive variables. Analyses showed small differences between individuals who participated in all seven waves and those who did not participate in all seven waves regarding chronological age and gender (complete responders were slightly younger, d = 0.12 and less likely to be women, 37 % vs. 45 %). No meaningful differences between complete and incomplete responders appeared regarding occupational status (d = 0.06), level of education (d = 0.06), and job satisfaction at T1 (d = 0.05). Given the small differences between complete and incomplete responders, we are confident that attrition was not of principle concern in this study.

Occupational status
Occupational status was assessed with a single item developed by Von Hippel et al. (2013) capturing the ranking of an individual's position in the organizational hierarchy: "Please rate your position in your organization using a scale similar to that of a 7-step ladder. It describes the level of your position in the organization. The lowest level (scale value 1) stands for entry-level positions, such as trainees. The top level (scale value 7) represents the highest achievable employment in your company, such as a position within the management or on a supervisory board."

Job satisfaction
Job satisfaction was assessed with a widely-used single item ("All in all, how satisfied were you with your work?"). Participants were instructed to think about the past three months (T1 and T2) and the past month (T3 -T7), respectively, in providing their ratings. The scale was anchored from 1 = very dissatisfied to 7 = very satisfied. Research has demonstrated that single items of job satisfaction represent reliable and valid operationalizations of this rather homogeneous construct (Dolbier et al., 2005;Wanous et al., 1997).

Perceived constraints at work
At T4 (May 2020), participants were asked with two self-developed items whether they felt impaired at work due to the COVID-19 pandemic: "My ability to work is severely impaired by the COVID-19 pandemic" and "Because of the COVID-19 pandemic, I can only carry out my job to a limited extent." The scale was anchored from 1 = strongly disagree to 5 = strongly agree. The two items were highly correlated, r xy = 0.75, p < .001, and we averaged the scores on the items to obtain an overall perceived constraints score.

Demographic characteristics
At T1, chronological age (i.e., years since birth) and gender (i.e., coded as 0 = male, 1 = female) were assessed. We included these covariates in an additional model because research has shown that both occupational rank and job satisfaction are associated with gender and chronological age (Berger et al., 1972;Dobrow Riza et al., 2018;Magee & Galinsky, 2008).

Statistical analyses
We used latent growth modeling (LGM; McArdle, 2009) to estimate intercept and slopes of job satisfaction. In order to disentangle the non-linear growth trajectories and to capture the different phases of the pandemic, we applied piece-wise latent growth models defining two separate trajectories of job satisfaction across time (decline and recovery; Bliese et al., 2017;Bliese & Lang, 2016;Bollen & Curran, 2006). Specifically, we used a three-step modeling approach: In a first LGM, we tested for change in job satisfaction across the seven measurement occasions by analyzing intercept as well as linear slopes (i.e., decrease and increase) modeled as latent variables. In a second set of LGMs, we tested our first hypotheses predicting intercept and decrease and recovery slopes of job satisfaction by occupational status (i.e., treated as a continuous variable). In doing so, we applied two-slopes piecewise growth-curve models modeling change in job satisfaction including an intercept, a pre-lockdown slope (from T1-T3), and a post-lockdown slope (T5-T7). To supplement this analysis and to better illustrate the effects we specified a multi-group LGM. We used a tercile split to divide the sample into three occupational status groups of equal size (low: 1-3; mid: 4; high: 5-7; n ls = 514, n ms = 536, and n hs = 533) to demonstrate that linear slopes with regard to decrease and recovery are only significant in the high, but not in the medium or low occupational status groups. Finally, in a third LGM, we tested Hypothesis 2, which posits that perceived constraints at work moderate the effect of occupational status on decline in job satisfaction across time by analyzing the interactive effect of occupational status and perceived constraints at work.
Analyses were carried out using Mplus 7.3 (Muthén & Muthén, 2015). Model evaluation was based on values of the chi-square statistic, the comparative fit index (CFI), the root-mean-square error of approximation (RMSEA), and the standardized root-meansquare residual (SRMR). Missing data were handled using the FIML (full information maximum likelihood) estimation procedure, which includes all available data (Bollen and Curran, 2006). Thus, the sample size used was n = 1583 in the main analysis and n = 1049 in the moderation analyses, representing the panel of respondents who provided at least partial responses across time.

Results
Means, standard deviations, and bivariate correlations of the variables are reported in Table 1. Occupational status and perceived constraints at work were not significantly associated. However, both were significantly associated with measures of job satisfaction.

Latent change in job satisfaction
We computed a quadratic LGM as well as piece-wise, two-slope LGMs to capture the predicted decrease and increase of changes in job satisfaction across the seven measurement occasions spanning the early stages of the COVID-19 pandemic. Compared with 10 alternative models (see Table S1 in the supplementary material section), we identified the best fitting model with T4 as breakpoint (minimum) separating the two slopes with different rates of change in a pre-crisis (pre-lockdown) and recovery (post-lockdown) slope (see Bliese & Lang, 2016). Specifically, the model defined a negative linear slope (T1-T3: − 3, − 2, − 1, 0, 0, 0, 0) with a breakpoint at T4 and a significantly positive increase thereafter (T5-T7 Variances of the intercepts and slopes were also significant indicating substantial variation in initial mean level as well as the linear slopes (p < .001). Finally, intercept and slopes were significantly correlated, indicating that participants with a higher intercept in job satisfaction experienced a steeper decline (with linear slope 1: r = 0.49, p < .001), but also a stronger recovery (with linear slope 2: r = − 0.37, p < .001) over time. Finally, the two slopes were negatively correlated (r = − 0.63, p < .001).

Occupational status predicts decline and recovery in job satisfaction
Next, we tested the hypothesized effect of occupational status (i.e., treated as a continuous variable) on the intercept as well as decline (slope 1) and recovery (slope 2) of job satisfaction across time. The LGM fit the data well [N = 1583; χ2(31) = 36.86, p < .001, CFI = 0.998, RMSEA = 0.011, 90 % CI 0.001, 0.023, SRMR = 0.024; AIC = 23,426; BIC = 23,560]. Occupational status had a significantly positive effect on the intercept of job satisfaction (B = 0.07, SE = 0.03, p < .001). In support for Hypothesis 1, we found a significant effect of occupational status on decline in job satisfaction (slope 1; B = − 0.15, SE = 0.04, p < .001), indicating that 1 Note that there was a longer time lag (3 months) between the first (December 2019) and second (March 2020) measurement occasion as compared to the one-month lags between the other measurement occasions. The reason for this was that the study was initially planned as a longitudinal study with 3-month time lags, but was adapted in March/April 2020 to monthly measurement occasions. Given that the first lockdown started not before the end of March 2019, the model reveals only an attenuated decrease in job satisfaction between the first and second measurement occasion which is reflected in the defined rate of change. The alternative models adjusting for the longer time lag resulted in worse model fit (see models 6 and 11 in supplementary material).  individuals with a higher (but not those with a lower) occupational status experienced a significant decline in job satisfaction until T4. In addition, we found a significant effect of occupational status on the recovery in job satisfaction after T4 (slope 2; B = 0.10, SE = 0.05, p < .05) suggesting that individuals with a higher occupational status tended to recover in job satisfaction after the first lockdown.

The moderating role of perceived constraints
Hypothesis 2 states that the effect of occupational status on changes in job satisfaction is moderated by perceived constraints at work, such that a decline in job satisfaction among employees with higher (vs. lower) occupational status is moderated by their perception of constraints at work associated with the pandemic. We specified another two-slope LGM including the predicted interaction effect between occupational status and perceived constraints on decline in job satisfaction across time. This model fitted the data well [N = 1049; χ2(39) = 47.18, CFI = 0.997, RMSEA = 0.014, 90 % CI 0.001, 0.027, SRMR = 0.025; AIC = 18,975; BIC = 19,128]. We found a significant interaction effect of occupational status and perceived constraints on the decline of job satisfaction (slope 1; B = − 0.44, SE = 0.19, p = .02). Specifically, supporting Hypothesis 2, this suggests that the effect of occupational status on decline in job satisfaction during the first period of the pandemic was moderated by perceived constraints at work. Please note that all models and effects reported above were stable with and without including age and gender as covariates.

Discussion
Research has shown that a higher rank in an organization often entails higher status and power and comes with many advantages that generally contribute to higher levels of job satisfaction and occupational well-being at the between-person level (Robie et al., 1998;Sherman et al., 2012;Keltner et al., 2003). In line with this body of research, we found that individuals with higher occupational status generally report higher job satisfaction than individuals with lower occupational status. However, the current study further reveals that, in times of crisis and at the within-person level, this relationship can be disrupted and even reversed. Specifically, our results suggest that the profound consequences of the COVID-19 pandemic represented a threat particularly to individuals with high (vs. low) occupational status undermining their job satisfaction during the early stages of the COVID-19 pandemic. In other words, we found that although those at the top of occupational hierarchies are generally more satisfied with their jobs, their level of job satisfaction over time (during the early stages of the COVID-19 pandemic) was more strongly disrupted by the pandemic. In addition, and as predicted these deleterious consequences for the job satisfaction of high-ranking individuals during the first lockdown were modulated by perceived constraints at work due to the pandemic. These results challenge current understandings of the psychological consequences of having high power and status in organizations.
The current study adds to our understanding about how job satisfaction changes over time, and how ranks in the occupational hierarchy are linked to job satisfaction by taking external, unforeseen events and circumstances into account (see Morgeson et al., 2015). In a stable hierarchy with no external disruptions, managers and executives seem to be more satisfied with their work lives as compared to those lower in the occupational hierarchy. However, when confronted with unexpected work-related changes that came with the COVID-19 pandemic, those at the top of the occupational hierarchy became less satisfied with their job over time, whereas employees lower in the hierarchy seemed to be relatively untroubled. In addition, when pandemic-related restrictions were lifted and Fig. 1. Estimated trajectories (two-slope model: decrease & recovery) of job satisfaction across the first period of national lockdown during the COVID-19 pandemic in Germany for individuals with a low, medium, and high occupational status. discrepancies between higher status role-based expectations and current circumstances decreased, the job satisfaction of individuals with higher occupational status recovered. Consistent with role theory and theories of hierarchy dynamics (Ashforth, 2001;Biddle, 1986Biddle, , 2013Keltner et al., 2003;Knight & Mehta, 2017;Sapolsky, 2011), this was true for individuals with a higher occupational status who experienced constraints due to the pandemic but not for those with a relative lower occupational status.

Theoretical and practical implications
The findings of our study have several implications for theory development and organizational practice. First, by showing that the COVID-19 pandemic is associated with changes in job satisfaction among high-status individuals in the workplace, our findings advance organizational theories on events, roles, and transitions (Biddle, 2013;Bliese et al., 2017;Morgeson et al., 2015). These theories suggest that novel, critical, and disruptive events lead to changes in employee experiences and behaviors as well as their work roles, particularly if events are located at a global level and persist over time, such as the COVID-19 pandemic. Based on theorizing on role theory and hierarchy dynamics (Ashforth, 2001;Biddle, 1986Biddle, , 2013Keltner, et al., 2003;Knight & Mehta, 2017;Sapolsky, 2011), we demonstrate that differences in occupational status interact with the effects of a major event, entailing a decrease and subsequent recovery in job satisfaction among high-status individuals. Importantly, our comparison to alternative linear and quadratic LGMs showed that breaking up the growth curve of job satisfaction in separate trajectories (decline vs. recovery with the break point at T4) best describes the data (Bliese & Lang, 2016).
Second, most studies on the effects of the COVID-19 pandemic-arguably a rather distal and macro-level event-are relatively descriptive in nature, in that they report changes in employee experiences (e.g., emotional exhaustion) and behavior (e.g., coping) across the various stages of the crisis (e.g., Meyer et al., 2021;Syrek et al., 2021). We advance theory development in this area by testing a pandemic-related moderator, perceived constraints at work due to the pandemic, as a more proximal boundary condition of changes in job satisfaction. Thus, our findings suggest that it is not the global event per se that leads to changes in work-related outcomes, but depends on the extent to which employees perceive that their ability to carry out their work tasks is constrained (Kahn & Quinn, 1970). In the current study, we find that particularly higher-ranking employees report lower job satisfaction during the first national lockdown in Germany as they struggled with such constraints. Future theory development should consider additional mechanisms (e.g., appraisals, changes in work characteristic) that influence the effects of other external events on work outcomes. Some of the work-related changes imposed by COVID-19 Pandemic are likely to be implemented on a regular basis in the foreseen future (e.g., teleworking, work from home). Thus, our research contributes poignant knowledge in the context of new ways of working and increasing flexibility, autonomy, and digitalization. Third, recent research suggests that uncertain and volatile circumstances induce instability of managerial positions, which is associated with reduced power sharing (i.e., delegating responsibility and authority and by not incorporating subordinates' ideas into decision making; see Feenstra et al., 2020).
In addition, we contribute to a growing body of research that investigates the impact of hierarchical status on attitudes and behavior at work. Status hierarchies are key to organizational functioning and an individuals' position within the hierarchy is a significant predictor of work attitudes and behavior (for an overview see Magee & Galinsky, 2008). For example, studies have shown that if individuals lose status and/or power, they may become highly dissatisfied, as they feel incompetent and even respond aggressively to others (Fast et al., 2012;Pettit et al., 2010). Our research contributes and expands this line of research, which is often carried out with student samples, by showing that a highly disruptive external event such as the COVID-19 pandemic leads to profound decreases in job satisfaction among high-, but not among lower-ranking individuals. Although, we do not investigated changes occurring in one's status position, we foreground temporal dynamics in a key outcome (i.e., job satisfaction) affected by status. Notably, however, higher ranking individuals in our study started out with the highest levels of job satisfaction and only decreased to some extent, with the lowest point during the national lockdown still remaining at a somewhat higher level compared to lower-and medium-status individuals. Although job satisfaction among individuals with a low and medium occupational status was significantly lower at the beginning of the study, job satisfaction centered around a mean of 4.9 to 5.04 on a scale between 1 and 7 across time. Trajectories of job satisfaction that were identified in this study might be also attributable to different baseline levels of job satisfaction that might be influenced by different factors in which people of different occupational status differ, such as personality (e.g., neuroticism), as well as material and interpersonal resources. For example, one possibility is that because individuals with a higher occupational status have more material, psychological, and social resources, they also have more to lose when confronted with a major, unexpected crisis (Wanberg et al., 2020). Likewise, individuals with a lower occupational status might have difficulties to accumulate resources in the first place that are important for the maintenance of job satisfaction. As a consequence, individuals with a lower occupational status might experience a smaller decrease in job satisfaction after a crisis, because they have lower levels of job satisfaction to begin with. In addition, it is likely that government wage subsidies which were paid to lower-income individuals in Germany during the pandemic might have buffered their job satisfaction during the lockdown and beyond.
Future theorizing could focus on how the instability of work environments of high-ranking hierarchical positions and work roles in times of crisis may impact not only indicators of work-related attitudes and occupational well-being, such as job satisfaction, but also behavior toward subordinates. Individuals with a relatively lower occupational status have fewer resources and might also experience significant role conflicts and interruption due to the pandemic, for example, experiencing work-family role conflicts when working from home. However, given that lower ranking individuals are generally confronted with more constraints, experience a lack of autonomy and control, and show more inhibition, they appear to be more susceptible to experiencing lower job satisfaction to begin with (Guinote, 2007;Keltner et al., 2003). Therefore, experiencing work-family role conflicts during the lockdown may have not affected their job satisfaction as much, because obstacles with regard to their work goals are common among individuals with a lower occupational status.
Regarding practical implications, our findings suggest that potential threats to job satisfaction must be considered in the context of organizational hierarchy. Increasing and maintaining job satisfaction is important given its moderate associations with, and time lagged effects on, job performance (Judge et al., 2001;Riketta, 2008) and broader well-being outcomes, such as life satisfaction (Iverson & Maguire, 2000;Judge & Watanabe, 1993). Thus, although job satisfaction should be protected and improved for individuals across all hierarchical levels of an organization, individuals with higher occupational status might be at greater risk for relative decreases in job satisfaction over time than those with lower occupational status. Based on our findings, organizational practitioners should focus particular attention to removing constraints to individuals' ability to carry out their work in times of crisis. This is particularly important when considering immediate effects for employees with higher occupational status, as these employees likely have greater responsibility for other people and organizational success.

Limitations and future research
Despite its strengths (e.g., 7-wave longitudinal design with baseline assessment before the COVID-19 pandemic; exploration of boundary conditions of job satisfaction change), our study has several limitations that could be addressed in future research. First, all measures were self-reported by employees, which may raise concerns about artificially inflated associations due to common method bias. However, such concerns may be alleviated by our within-person design that involved repeatedly assessing job satisfaction over several months. Nevertheless, future research could collect additional outcome measures in times of crisis, including physiological markers of employee health and wellbeing or supervisor or peer ratings of employee performance.
Second, despite its longitudinal nature, our research design does not allow for conclusions regarding the causes of the changes in job satisfaction over time. To draw stronger inferences (at least with respect to temporal precedence), future research also needs to repeatedly assess occupational status and perceived constraints due to the pandemic over time. Third, our moderator, perceived constraint at work, was pragmatically measured using only two self-developed items, which may raise concerns about construct validity. For instance, we were not able to disentangle whether perceived constraints are linked to loss of power or perceptions of not being able to take responsibility for other or the organization. Thus, further theoretical and empirical work is necessary to explore the nomological net of this pandemic-related variable. Third, although a single-item measure represents a simple, reliable, and valid assessment of occupational status, the self-reported nature of the measure and the single dimension may restrict the assessment of the complex nature of the construct (Matthews et al., 2022). Thus, future research needs to include multi-source measures (e.g., organizational records, supervisor or peer reports) that capture the multifaceted nature of occupational rank.
Finally, while we considered the COVID-19 pandemic as a broad contextual factor, we did not take the more proximal organizational and national context into account when investigating within-person changes in job satisfaction. Regarding the organizational context, it may be interesting to examine further which factors (e.g., social support, team size) may buffer or intensify higher-status individuals' drop in job satisfaction. Regarding the broader national context, it is important to consider that our study was conducted in Germany and, therefore, all participants, independent of their occupational status, remained in full-time employment during the study period. For instance, workers in Germany are protected by a unique social security system and many jobs were secured during the pandemic with "Kurzarbeit" (i.e., short term work schedules and government wage subsidies). Future research should, therefore, investigate how occupational status relates to job satisfaction in times of crises in other national contexts, such as the United Kingdom or the United States, where the social safety net is traditionally weaker.

Conclusion
Research has shown that occupying a higher occupational rank is associated with high levels of job satisfaction. We found that these interindividual differences also hold in times of crisis presented by the profound workplace-related changes during early stages of the COVID-19 pandemic. We further examined how individuals with a higher versus lower occupational status experienced this worldwide crisis including major, unexpected changes to their work life. Across this time period, we found that individuals with a higher occupational status experienced a significant disruption with regard to their within-person development in job satisfaction, with initial declines during the first national lockdown in Germany and a subsequent recovery. By contrast, individuals with a lower occupational status experienced no such changes in job satisfaction across time. The results further suggest that the decline in job satisfaction among individuals with a higher occupational status was modulated by perceived constraints at work that were associated with the pandemic. Together, these findings uncover the context-dependent nature of the relationship between occupational status and occupational wellbeing, emphasizing the importance of considering volatile and changing circumstances as well as major events.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
The data presented in this study and analyses code are openly available [OSF; https://osf.io/23zy6/?view_ only=cbf4c084536e4c98a2883a1ac4e2e795].