Supervisors’ power to deal with employees’ inner resignation: How perceived power of the organization and the supervisor relate to employees’ voluntary and enforced work behavior

Abstract

Power is fundamental in organizations and is exerted on employees by the organization itself as well as by supervisors. In this study, I applied the slippery slope framework (SSF) and interpersonal power interaction (IPI) model to shed light on how power dynamics relate to employees’ inner resignation and contextual performance.

Survey data was obtained from 1102 employees of Austrian and German organizations. In line with expectations, the results of path modeling revealed that perceived coercive power of the organization and supervisors positively relates to employees’ inner resignation. Perceived legitimate power of the organization and supervisors is positively associated with contextual performance and negatively associated with inner resignation. Finally, supervisor reward power further strengthens the beneficial relationship between legitimate organizational power and inner resignation. The results are discussed in light of self-determination theory and the effort-reward imbalance model.

1. Introduction

The importance of power in organizations has been compared to that of oxygen for breathing (Clegg, Courpasson, & Phillips, 2006). Given that power is such an essential part of organizations’ functioning, it is important to investigate how different forms of power exerted by supervisors as well as the organization as a whole (enacted via organizational rules and routines) shape employee work behavior. Since work contexts are becoming more and more uncertain and clear models of optimal behavior do not always exist, organizations must rely on a reflexive and innovative workforce (Batistič; Černe, Kaše, & Zupic, 2016; Schippers, West, & Dawson, 2015). Thus, employee work behavior has often been studied in a way that goes beyond formal job descriptions, which is defined as contextual performance (Demerouti et al., 2014; Frese and Fay, 2001). However, employees’ psychological withdrawal from the organization and ‘work-to-rule’ (i.e., inner resignation) is rather risky for organizations. Although public polls show a high prevalence of inner resignation among employees (Nink, 2014; Williams, 2016), inner resignation remains a relatively unexplored phenomenon in organization behavior literature. It encompasses employees’ experienced frustration and perceived breach of the psychological contract (Wenck, 2013). Since legal obligations might restrict firing such employees, and critically, because such behavior involves much more than a psychological ’defect’ (Fleming & Spicer, 2003), employees’ inner resignation is something employers must deal with. Therefore, it is important to investigate how power in organizations shapes employees’ contextual performance and inner resignation.

While prior research on power within organizations has mostly taken the perspective of power holders (for a review, see Anderson & Brion, 2014), I take the perspective of the power targets (i.e., individual employees) and analyze the relationship between perceived power and employee work behavior. Perceived power in organizations has mostly been studied with regard to interpersonal power exerted by managers as defined in the IPI model (e.g., Hinkin & Schriesheim, 1989; Lunenburg, 2012; Pierro, Raven, Amato, & Bélanger, 2013; Raven, 2008; Schwarzwald, Koslowsky, & Agassi, 2001). Applying the slippery slope framework (SSF) which originated in the tax context, I argue that the bases of power addressed in the IPI model do not necessarily require interpersonal relations and should also enable the examination of power on the organizational level (Hartl, Hofmann, Gangl, Hartner-Tiefenthaler, & Kirchler, 2015; Hofmann, Hartl, Gangl, Hartner-Tiefenthaler, & Kirchler, 2017; Kirchler, Hoelzl, & Wahl, 2009; Wahl, Kastlunger, & Kirchler, 2010). The SSF postulates that coercive power results in enforced compliance whereas legitimate power leads to voluntary cooperation (Kirchler, Hoelzl & Wahl, 2008, 2009). Going beyond previous studies, I consider perceived power exercised by the organization itself in addition to power exercised by the supervisor. Furthermore, I investigate crowding effects on motivation when taking reward power into account (Frey & Jegen, 2001).

Using a sample of 1102 employees of Austrian and German organizations, the purpose of this paper is to investigate how organizational and supervisor coercive and legitimate power relate to contextual performance and inner resignation. In this paper, I particularly highlight the role of inner resignation and investigate how reward supervisor power can crowd-in motivation, strengthening the beneficial effect of legitimate organizational power. Thus, when organizations are perceived as legitimately powerful, the instrumental perspective provided by reward power plays a key role in mitigating employees’ psychological withdrawal in the form of inner resignation. Finally, the results also have practical relevance, as they suggest the possibility to overcome employees’ inner resignation. In particular, supervisor reward power seems to strengthen the beneficial effect of the organization’s perceived legitimate power, making it particularly useful. Thus, supervisors need to adjust their power tactics depending on the organizational power context.

1.1. Perceived organizational power

Studying power in today’s organizations is highly relevant due to the changing nature of power in organizations, involving a shift from bureaucratic to post-bureaucratic principles (Maravelias, 2007), and the expanding spectrum of possibilities to technologically monitor work behavior (e.g., Day, Paquet, Scott, & Hambley, 2012). There is already a vast literature on power in organizations (for a review, see Fleming & Spicer, 2014). Traditionally, power was defined as the potential to influence another party through control of resources (Deutsch & Gerard, 1955; French & Raven, 1959); however, it is now common to take a broader view by studying more implicit forms of power (e.g., Foucault, 1975/2012; Kärreman & Alvesson, 2004).

Employees’ perceptions of power in organizations (e.g., Anderson & Berdahl, 2002; Fleming and Spicer, 2014, Foucault, 1975/2012) are shaped by formal and informal control practices (Chenhall, 2003; Ouchi, 1979, 1980) or coercive and legitimate power tactics (Kirchler et al., 2009) that build upon one another Kärreman & Alvesson (2004). Generally, managerial control practices are fundamental in organizations and inherently interweave the material, the social and the symbolic (Orlikowski & Scott, 2008): they coordinate employees’ efforts, enable agreement between managers at different hierarchy levels, serve as a source of motivation by setting up an incentive system, and function as signals to trigger necessary interventions by management (Goold & Quinn, 1990). Distinguishing power from control, I conclude that power can be defined as the potential to control and need not be seen as a ‘conditio sine qua non’ -- in other words, having power is not necessarily tied to the execution of control. Thus, influence takes place regardless of whether power is actively exercised or not (Brass & Burkhardt, 1993).

Although power seems essential in organizations, a balance is paramount, as too much power might create a distrustful atmosphere (Mooijman, van Dijk, van Dijk, & Ellemers, 2019). Furthermore, there is a risk that the organization’s intentions will be misinterpreted or considered as ’hollow promises’ Casey (1999); Gabriel (1999). Therefore, in order to understand the effect of power on work behavior, it is important to incorporate employees’ perceptions of organizational power.

The SSF (Gangl, Hofmann, & Kirchler, 2015; Kirchler et al., 2009, 2008; Kirchler, Kogler, & Muehlbacher, 2014) describes how power perceptions shape individuals’ behaviors. It was originally postulated in the context of tax compliance and describes two different routes for establishing individuals’ compliance with rules: enforced and voluntary behavior.

Enforced behavior is achieved via coercive power, because individuals are forced to act in a certain way by strict control. Coercive power is the most obvious and overt form of power and relies on formal controls based on standard operating procedures, accounting and information systems (Ho, Wu, & Wu, 2014; Langfield-Smith, 1997). Within this perspective, the target’s own desires and beliefs are irrelevant, as compliance is enforced by the power holder either by threatening punishments (i.e., coercive power) for mistakes, failures, or deviations from performance goals.

Voluntary behavior, on the other hand, is triggered by legitimate power. Legitimate power is rooted in the power target’s acceptance of someone’s or something’s influence over himself/herself, which arises from agreed-upon rules (Raven, 1992, 1993), implicit social norms (Berkowitz & Daniels, 1963), reciprocity (Gouldner, 1960), identification with an individual or group, and the internalization of organizational goals (Kelman, 1958).

The distinction between voluntary and enforced behavior is relatively rare because most commonly the distinction between contextual and task performance is made in management and organization behavior literature (Jawahar, Meurs, Ferris, & Hochwarter, 2008; Law, Wong, & Song, 2004; Lindebaum, 2013; Motowidlo & van Scotter, 1994). In this paper, I relate the enforced/voluntary behavior dichotomy to inner resignation and contextual performance in the work context.

Inner resignation describes a situation in which employees feel trapped in their job and do not see adequate options outside of their current place of employment (Wenck, 2013). Employees are cynically detached (Gabriel, 1999) and subjectively distant from the organization (Fleming & Spicer, 2003). Inner resignation is described as the denial to exhibit additional work involvement beyond what is (contractually) mandated in order to re-establish lost equity (Schmitz, Gayler, & Jehle, 2002). Employees experiencing inner resignation have the feeling that the organization has not fulfilled its obligations, and the exchange between the employee and the organization is perceived as unfair. However, no other choice but to comply with organizational rules and policies is seen. Psychological withdrawal serves as a protection against helplessness due to the low level of perceived autonomy Wenck, (2013). Employees adjust their performance accordingly (Shore & Tetrick, 1994). Thus, employees react with inner resignation to the potentially long and complex process of job dissatisfaction (Brinkmann & Stapf, 2005). From the outside, the performance of a person experiencing inner resignation might be similar to that of someone with appropriate task performance, at least in the short term. However, the underlying psychological processes are different, which may lead to different long-term effects. Because employees perceive that their psychological contract (Rousseau, 1995) has been violated by the employer, they exhibit lower engagement and loyalty towards the organization. Often, employees decide to stay with the organization because they see no perspectives outside it or they have to sacrifice too much (Lee, Burch, & Mitchell, 2014). However, employees experiencing inner resignation do not feel committed and do not engage with their work beyond their job description and formal obligations (Wenck, 2013).

Contextual or extra-role performance is defined as discretionary behaviors by employees that promote the organization’s effectiveness (MacKenzie, Podsakoff, & Fetter, 1991). Thus, contextual actions go beyond formal job descriptions (Demerouti et al., 2014), as employees strive to complete their tasks to the best of their abilities with vigor and zeal (Tyler & Blader, 2000). Although task and contextual performance are seen as two different constructs, the line between them is not very clear (Organ, 1997; Tepper, Lockhart, & Hoobler, 2001), and empirical research shows a considerable overlap (Lindebaum, 2013; Vey & Campbell, 2004). There is widespread agreement that taking an active role in the organization and helping co-workers is a major component of contextual performance.

Applying the SSF to the organizational context (Kirchler et al., 2009), I assume that when organizations rely too heavily on monitoring and sanctioning, employees feel forced to comply with organizational rules and policies. The salience of employees’ instrumental dependence might discourage engagement, reduces psychological attachment to the organization (Gabriel, 1999), and fosters alienation from the organization (Chiaburu, Thundiyil, & Wang, 2014). Thus, drawing on the SSF (Kirchler, Hoelzl, & Wahl, 2008), I hypothesize:

Hypothesis 1

Perceived coercive organizational power positively relates to inner resignation.

However, following the SSF, when legitimate power perceptions are high and the rules and policies are considered legitimate, individuals cooperate voluntarily because they consider it their civic duty and are convinced it is the right thing to do. Thus, legitimate organizational power should stimulate employees’ voluntary cooperation and engagement in extra-role activities, and thus their contextual performance.

Hypothesis 2

Perceived legitimate organizational power positively relates to contextual performance.

Experimental studies testing the SSF by manipulating coercive and legitimate power have repeatedly found that legitimate power reduces an antagonistic climate towards the (tax) authority (Hofmann et al., 2017). In an antagonistic climate, the authority strongly sanctions misbehavior and individuals hide from this authority (Kirchler et al., 2008). When the organization is perceived as legitimate, identifying with the organization is more likely and thus, being subjectively distant is less likely. Therefore, I assume a similarly beneficial effect of legitimate power in the organizational context, and hypothesize:

Hypothesis 3

Perceived legitimate organizational power negatively relates to inner resignation.

1.2. Perceived supervisor power

Supervisor power was described in the interpersonal power interaction model (IPI; Raven, Schwarzwald, & Koslowsky, 1998), based on French and Raven’s (1959) well-received work about the bases of social power (Shafritz, Ott, & Jang, 2011). It describes supervisor power as perceived by subordinates (French & Raven, 1959; Raven 1992, 2008, 1993; Raven et al., 1998), distinguishing among several forms of power that can be subsumed under the broader categories harsh and soft power tactics Elias (2007, 2008), which is defined as coercive and legitimate power in the SSF Gangl et al. (2015); Hofmann et al. (2017).

Harsh power is the most obvious form of power. Within this perspective, the target’s own will and beliefs are irrelevant as compliance is enforced by its holder (e.g., supervisor) via threats of punishments (i.e., coercive power) for mistakes, failures, or deviations from performance goals or via positive incentives (i.e., reward power). Although power is granted to the power holder by the nature of his or her status within an organization (Schwarzwald, Koslowsky, & Ochana-Levin, 2004), some form of performance monitoring is necessary to determine the appropriate level of sanctions and rewards (Raven, 2008). Coercive power can increase productivity, but extensive monitoring and micromanagement impair the supervisor’s relationship with the employee (Grover, Hasel, Manville, & Serrano-Archimi, 2014), impose stress (Bhave, 2013) and reduce employees’ autonomy, which stimulates psychological withdrawal (Wenck, 2013) and thus positively relates to inner resignation Richter (1999). Therefore, I hypothesize:

Hypothesis 4

Perceived coercive supervisor power positively relates to inner resignation.

Legitimate power (Gangl et al., 2015; Hofmann et al., 2017) on the other hand influences employees through supervisor’s persuasion and authority (Turner, 2005). It is exercised not through forced compliance, but rather through voluntary cooperation based on volition and belief. Employees accept the supervisor’s authority because of his/her expertise, charisma or legitimacy and act accordingly. Cooperation occurs even in the absence of monitoring. Legitimate power can be considered risky for supervisors due to the lack of monitoring. Nonetheless, it is also considered more efficient (Elangovan & Xie, 1999; Tyler, 2006a; 2006b), as it encourages employees’ identification with the job and does not require monitoring as cooperation is voluntarily.

Legitimate power might encourage employees’ intrinsic motivation (Pierro, Cicero, & Raven, 2008), as it allows tasks to be carried out more autonomously, which is one of the basic human needs defined in self-determination theory (Ryan & Deci, 2017). When employees are intrinsically motivated, they are highly likely to also exhibit contextual performance (Wang, Lu, & Sun, 2018). Thus, employees who experience a high level of intrinsic motivation will not only fulfill their primary work tasks voluntarily, but will most likely also help out others in the organization. This leads to the following hypothesis:

Hypothesis 5

Perceived legitimate supervisor power positively relates to contextual performance.

In line with Hypothesis 3 and based on the SSF, I also assume beneficial relationships between legitimate supervisor power and inner resignation. Since inner resignation is defined as resulting from a prior breach of the psychological contract, I assume that employees are less likely to experience inner resignation when supervisors’ legitimate power is perceived as high and employees perceive cooperation as their civic duty.

Hypothesis 6

Perceived legitimate supervisor power negatively relates to inner resignation.

1.3. Supervisor power to strengthen legitimate organizational power

Supervisors are known to buffer the negative effects of job demands (Montani & Dagenais-Desmarais, 2018). Thus, their power might attenuate or strengthen the direct effects of organizational power on employee work behavior. Building on the assumption that perceived legitimate organizational power stimulates contextual performance and reduces inner resignation, I investigate whether reward supervisor power may undermine or strengthen these relationships based on motivation crowding theory (Frey & Jegen, 2001).

Reward power (Goltz, 2003) is rooted in an instrumental view of power Tyler, Schulhofer, & Huq (2010) and encompasses ’control over resources’ (Mintzberg, 1983), with power holders achieving compliance through positive incentives such as pay raise, bonuses or promotions. There is extensive empirical support that the provision of extrinsic rewards crowds out intrinsic motivation (Deci, Ryan, & Koestner, 1999; Frey & Jegen, 2001; Frey & Oberholzer-Gee, 1997; Mikkelsen, Jacobsen, & Andersen, 2017). Based on cognitive evaluation theory (Deci, Koestner, & Ryan, 1999; Ryan, 1982), the underlying rationale is that extrinsic rewards have the potential to thwart satisfaction of the basic human need for autonomy, which consequently reduces employees’ intrinsic motivation (Deci et al., 1999; Ryan, 1982; Ryan & Deci, 2000a). Based on this consideration, I test whether supervisor power to provide tangible rewards suffices for the emergence of crowding-out effects:

Hypothesis 7

Perceived reward supervisor power weakens the positive relationship between legitimate power and contextual performance.

However, when supervisors act in an appreciative and professional manner, employees’ autonomy might not be damaged by reward power (Frey & Jegen, 2002). External interventions such as rewards might even crowd-in motivation if employees perceive them as supportive (Frey & Jegen, 2001). Following self-determination theory, inner resignation can be described as a state of amotivation that is completely non-autonomous and fails to satisfy employees’ basic need for autonomy (Ryan & Deci, 2000b). I argue that when legitimate organizational power is high, the provision of rewards might additionally stimulate motivation. In contrast to organizations as a whole, supervisors are able to spontaneously reward employee behavior in case they have the power to do so. Employees’ awareness of potential rewards might increase their perceived autonomy, as they can now decide whether to engage in additional work tasks in order to receive them. If they do decide to engage in further tasks, it will then also satisfy their need for competence (Ryan & Deci, 2000b). Therefore, I assume a crowding-in of motivation via reward supervisor power, hypothesizing:

Hypothesis 8

Perceived reward supervisor power strengthens the negative relationship between legitimate power and inner resignation.

2. Method

2.1. Participants

A total of 1102 employees of Austrian and German organizations were recruited by a team of five master students to participate in the questionnaire study. Participants had to be employed in an organization. Recruitment was carried out via snowball sampling; in addition, human resources departments in eight organizations distributed the questionnaires among their employees. The majority of participants (55%) were employed full-time. Approximately two-thirds were female (67.5%), and only 17.2% of participants held a leadership position. On average, participants were 39.2 years old (SD = 11.25) and had been working in their organization for 11.5 years (SD = 10.38). Participation in the survey was completely anonymous and voluntary. At the beginning of the questionnaire, participants were introduced to the study and the anonymity and confidentiality of their data were assured. Participants were not required to answer each item in order to continue the questionnaire and could exit the questionnaire at any time without negative consequences.

2.2. Measures

All responses were recorded on Likert scales ranging from completely disagree (1) to completely agree (6), complemented by the option ‘don’t know’.

Pre-testing. To make the questionnaire more attractive for participants, reduce its length, and test the validity, a pre-test with a sample of 123 students was conducted. The pre-test comprised a questionnaire with 169 items. Items on perceived supervisor power from the IPI questionnaire (Raven et al., 1998) were translated into German and then back-translated by a different person in order to assure multilingual validity (Brislin, 1970). Native English speakers checked the semantic equivalency of the two English versions. The organizational power items were adapted from measures based on the SSF and IPI models. Items for the final questionnaire (see Table 1 ) were selected based on high factor loadings in a confirmatory factor analysis while also preserving the sub-dimensions of the scales.

Table 1.

Factor loadings of each item.

	λ
Coercive organizational power
At my company mistakes are penalized.	–
Coercive supervisor power
My supervisor could make things unpleasant for me.	–
Legitimate organizational power
Since I am an employee, I have to comply with my company’s rules. (position)	0.52
I support our corporate philosophy and act on it. (referent)	0.83
My company’s success also depends on me. (dependence)	0.46
The operational procedures in my company are efficient. (expertise)	0.74
My company’s corporate policy is transparent. (information)	0.70
Legitimate supervisor power
My supervisor is deservedly in his position. (position)	0.82
I think highly of my supervisor. (referent)	0.85
My supervisor has accommodated my wishes multiple times in the past, that’s why I feel obliged to him. (dependence)	0.63
My supervisor knows the best way to do my work. (expertise)	0.70
I understand the rationale behind my supervisor’s instructions/remarks. (information)	0.85
Reward supervisor power
A good evaluation from my supervisor could lead to an increase in pay.	0.94
My supervisor decides on bonuses and special benefits.	0.75
Contextual performance
I volunteer to do things that are not required in order to help my organization.	0.77
I make suggestions to help improve the work setting.	0.73
I help my supervisor even when s/he has not asked me.	0.67
Inner resignation
I have worked myself into the ground at my job.	0.69
I used to be much more dedicated to my work.	0.71
I am happy when I can go home after work.	0.51

Coercive organizational power defines compliance resulting from the threat of sanctions (Raven et al., 1998). The coercive power measure was adapted from the tax context (SSF, Hofmann et al., 2017), and asked whether the organization sanctions misbehavior using the following item: “At my company, mistakes are penalized”.

Coercive supervisor power was measured with the item “My supervisor could make things unpleasant for me” from the IPI scale (Raven et al., 1998).

I acknowledge that using only one item for each of the two levels of coercive power is a flaw of the measurement that cannot be changed at this point. This was due to the tax context of the SSF, in which coercive power is operationalized as strict monitoring and control. Supervisors’ control is different, as it can also be seen as something positive, showing that supervisors care about their employees’ work, and it can even enhance trust (Verburg et al., 2018). Empirical studies support this argument; a study reported by Forbes revealed that 65% of employees would appreciate more feedback (Lipman, 2016). Based on this reasoning, and due to the very low correlation between the two items for coercive supervisor power (r = 0.279), I dropped the second item “My supervisor strictly controls my work”. I also dropped the item “At my organization, employees’ working mode is strictly monitored” to maintain consistency across both levels, even though the correlation was much higher here (r = 0.528). All results reported below remain stable regardless of whether I use one or two items for organizational coercive power.

Legitimate organizational power was measured based on Gangl et al. (2015) and Hofmann et al.’s (2017) adaptation of the IPI model for an authority not characterized by interpersonal relations. I adapted the items for the organizational context and measured five aspects of legitimate power (with one item each): (i) legitimate power by position, in which the influencing party has the right to prescribe behavior; (ii) legitimate power by dependence, in which favors need to be returned; (iii) information power, in which the power holder explains how the job should be done; (iv) expert power, in which knowledge is attributed to the influencing party; and (v) referent power, which encompasses identification with the influencing party. Cronbach alpha of the scale was satisfactory, with alpha = .79.

Legitimate supervisor power was measured with equivalent items to those for organizational legitimate power, capturing each of the five dimensions (i.e., position, dependence, information, expert and referent) with one item each drawn from the IPI scale (Raven et al., 1998). Cronbach’s alpha of the scale was excellent, with alpha = .91.

Reward supervisor power refers to compliance resulting from the prospect of receiving incentives in the form of rewards (Raven et al., 1998). It was measured with two items asking whether the respondents’ supervisor is able to help them receive special benefits or a promotion. A sample item is: “A good evaluation from my supervisor could lead to an increase in pay.” Cronbach’s alpha of the two items was .84.

Inner resignation corresponds to enforced behavior in the SSF and was measured with three items from the German Inner Resignation Scale (Schmitz et al., 2002). A sample item is “I used to be more dedicated to my work”; Cronbach alpha was .69.

Contextual performance corresponds to voluntary behavior in the SSF and was measured using three items from Tyler and Blader’s (2000) scale for extra-role behavior. A sample item is “I volunteer to do things that are not required but help my company”; Cronbach’s alpha was .78.

2.3. Analysis

Following Anderson and Gerbing’s (1988) recommendations, I first assessed the validity and reliability of the measures using confirmatory factor analysis (CFA), and then tested the hypothesized relationships using path modeling. The analysis was conducted using Mplus (Version 8; Muthén & Muthén, 1998–2017).

2.3.1. Confirmatory factor analysis

The factor structures of all scales were tested with a CFA (see Table 1 for factor loadings). A good model fit was achieved (χ² (123) = 469.09, p < .001, RMSEA = . 05, SRMR = 0.05, CFI = 0.94). Table 2 shows means, standard deviation, inter-construct correlations, average variance extracted (AVE), and composite reliability (CR). The CRs ranged from 0.67 to 0.88, surpassing the minimum acceptable level of 0.60 (Fornell & Larcker, 1981). However, the AVEs for inner resignation (AVE = 0.41) and legitimate organizational power (AVE = 0.44) were below the 0.50 threshold. According to Fornell and Larcker (1981), AVE may be a more conservative estimate, and CR can be used alone to judge internal consistency (Lam, 2012; Huang, 2013). As shown in Table 2, the square root of the average variance extracted for each construct exceeded the corresponding inter-construct correlations (Fornell & Larcker, 1981). Therefore, discriminant validity was achieved.

Table 2.

Means, standard deviation, correlations, AVE and CR.

	Items	M	SD	1	2	3	4	5	6	7	8
1. Male	1	33%	–
2. Leadership position	1	13%	–	.26∗∗
3. Coercive organizational power	1	2.90	1.52	-.13∗∗	−0.02
4. Coercive supervisor power	1	4.21	1.59	−0.06	−0.03	.22∗∗
5. Legitimate organizational power	5	4.44	0.96	.12∗∗	.09∗∗	-.12∗∗	-.18∗∗	.67. (.79)
6. Legitimate Supervisor power	5	4.22	1.22	0.03	.08∗	-.14∗∗	-.23∗∗	.44∗∗	.78. (.88)
7. Reward supervisor power	2	3.44	1.87	.24∗∗	.32∗∗	−0.01	0.02	.21∗∗	.10∗∗	.85. (.84)
8. Contextual performance	3	4.43	1.11	.26∗∗	.25∗∗	-.14∗∗	−0.06	.40∗∗	.31∗∗	.26∗∗	.72. (.77)
9. Inner resignation	3	3.44	1.31	-.12∗∗	−0.03	.27∗∗	.31∗∗	-.35∗∗	-.32∗∗	-.09∗∗	-.26∗∗	.64. (.67)

Note. M = Mean. SD = Standard deviation; Sub-diagonal entries are latent construct inter-correlations. The first entry on the diagonal is the square root of the average variance extracted (AVE), while the second entry (in brackets) is the composite reliability.

∗∗p < .01. ∗p < .05.

Since all measures were rated at the same time by the respondents, the likelihood of common method bias is high. Therefore, we tested for common method bias (Podsakoff & Organ, 1986) using the latent method factor approach. All items were allowed to load on the latent method factor. The AVE by the common method factor was only 0.13 and therefore, we conclude that common method bias was unlikely to be problematic. Furthermore, results building on interactions should be unaffected of common method bias (Siemsen, Roth, & Oliveira, 2010).

2.3.2. Path modeling

I tested the hypotheses using MLR (maximum likelihood estimation robust to non-normality) and modeled four exogenous power variables (coercive and legitimate power; each on the supervisor and organizational level). Inner resignation was regressed on coercive and legitimate power at both levels and contextual performance was regressed on legitimate power at both levels. The proposed model fit was good (χ² (12) = 44.57, p < .001, RMSEA = . 05, SRMR = 0.04, CFI = 0.94). With regard to control variables, the analysis revealed that both gender (β = 0.18, p < .001) and leadership (β = 0.17, p < .001) affect employees’ contextual performance, but not their inner resignation. To keep the model as parsimonious as possible, I excluded the non-significant control variables. Male participants and employees holding a leadership position were more likely to report contextual performance than female participants and employees without a leadership position. Overall, in this model, the predictor variables explained almost 23% of the variance in contextual performance (R ² = 0.231) and 24% of the variance in inner resignation (R ² = 0.242).

The results for organizational power were in line with expectations. Coercive organizational power related positively with inner resignation (H1: β = 0.19, p < .001). Legitimate organizational power was positively associated with contextual performance (H2: β = 0.29, p < .001) and negatively associated with inner resignation (H3: β = −0.23, p < .001).

The results for supervisor powers also supported the hypotheses. Coercive supervisor power related positively to inner resignation (H4: β = 0.20, p < .001). Furthermore, legitimate supervisor power was positively related to contextual performance (H5: β = 0.17, p < .001) and negatively related to inner resignation (H6: β = −0.16, p < .001).

When testing the interaction effect of reward supervisor power, I found support for H8, but not H7. Reward supervisor power did not weaken the relationship between legitimate organizational power and contextual performance (H7: β = −0.23, p = .117). However, simple slope analysis revealed that the relationship seemed to become weaker in tendency when reward supervisor power was perceived as high. In this model, the predictor variables explained almost 37% of the variance in contextual performance and 23% of the variance in inner resignation (see Fig. 1 )

Fig. 1.

Interaction between organizational legitimate power and supervisor reward power on contextual performance.

Regarding the motivation crowding-in hypothesis, I found that reward supervisor power strengthened the negative relationship between legitimate organizational power and inner resignation (H8: β = −0.44, p < .001). When reward power was also taken into account, legitimate power alone was no longer significant (β = −0.06, p = .362), but reward power had a strong direct influence on inner resignation (β = 0.35, p < .001). This model explained 24% of the variance for contextual performance and 41% for inner resignation (see Fig. 2 ).

Fig. 2.

Interaction between organizational legitimate power and supervisor reward power on inner resignation.

3. Discussion

The goal of this study was to draw conclusions about how organizations and supervisors can foster employees’ contextual performance and limit employees’ inner resignation through their power tactics and related interactions. This paper provides suggestions for dealing with the often-neglected phenomenon of employees’ inner resignation: strengthening legitimate organizational power by explaining the rationale of policies and cautiously using rewards. It contributes to the literature in several ways.

First, I tested whether the SSF (Kirchler et al., 2008, 2009) from the tax context also applies to the organizational context. The SSF enables me to study power beyond the interpersonal level, as perceived coercive and legitimate power in organizations are usually studied with regard to supervisors only Raven et al. (1998). In line with expectations, I found that perceived coercive power is positively related to inner resignation, while perceived legitimate power is positively related to contextual performance and negatively related to inner resignation. These results confirm the importance of current efforts to make organizations flatter (e.g., Darrell, Sutherland, & Takeuchi, 2016; Laloux, 2014; Robertson, 2015).

Second, supervisor power has been researched with regard to their effect on employees’ desired work behaviors (Pierro, Kruglanski, & Raven, 2012; Pierro et al., 2013). This study sheds light on power dynamics with regard to counterproductive work behavior: employees’ inner resignation. Although contextual performance and inner resignation negatively correlate with each other, they should not be seen as opposites. Motivation theory typically draws a distinction between extrinsic and intrinsic motivations (Ryan & Deci, 2000b). Inner resignation is neither described by intrinsic nor extrinsic motivation, as it refers to a state of denial and withdrawal and might be best compared to amotivation, which is defined as the absence of intentional regulation Gagné & Deci (2005). Alternatively, drawing on Fleming and Spicer (2003), one could argue that inner resignation constitutes a more active force of resistance than amotivation; however, no motivational equivalent to this active resistance exists within self-determination theory (Gagné & Deci, 2005).

Third, although it is commonly acknowledged that a mix of power tactics are at play in organizations Raven et al. (1998), this study represents the first empirical attempt to jointly analyze perceived supervisor and organizational power. Investigating both levels makes it possible to compare their effects on employee work behavior. Inspecting the size of beta weights.I conclude that the effect of legitimate power seems to be larger when exerted by the organization than by the supervisor. This finding emphasizes the importance of considering the organizational context, which has been mostly neglected when studying work behavior despite its relevance as a shaper of meaning (Johns, 2006, 2018).

Fourth, the use of (tangible) rewards has been criticized in a variety of contexts for undermining intrinsic motivation due to the exertion of control (Deci et al., 1999). Since intrinsic motivation seems to be becoming more and more important in today’s business environments, many organizations aim to increase employees’ autonomy (e.g., by creating flatter hierarchies or flexibility with regard to the time and place of work). In this study, I investigated the effect of reward supervisor power on the relationship between legitimate organizational power and inner resignation, assuming that intrinsic motivation is very low when employees experience inner resignation. Thus, when analyzing inner resignation, the question is not only whether employees’ motivation to work is crowded-out by reward power, but – more importantly – whether their motivation is crowded-in by adding rewards to a context of legitimate power. Studies from economics have demonstrated crowding-in effects in the context of welfare systems Künemund & Rein (1999) and unconditional grants (Caldeira & Rota-Graziosi, 2014), but this assumption has not yet been tested in the work context. My results support the crowding-in effect in the work context as well, as reward supervisor power was found to strengthen the beneficial effect of legitimate organizational power on employees’ inner resignation.

I conclude that rewards are not so bad after all – particularly when supervisors apply rewards as a legitimate form of appreciation in ways that do not impair employees’ self-determination. I further support my argument for reward power using the effort-reward imbalance (ERI) model (Siegrist, 1996; Sigrist, Siegrist, & Weber, 1986). The basic assumption of this model (which has been tested in various organizational contexts) is that “a mismatch between high extrinsic or intrinsic efforts and low reward may lead to adverse health” (Van Vegchel, de Jonge, Bosma, & Schaufeli, 2005, p. 1119). Since inner resignation results from the perception that the psychological contract has been breached, the resulting denial or withdrawal from work and/or the organization might serve as a form of self-protection with regard to employees’ physical and/or mental health. Future studies should therefore apply a process perspective and investigate how employees experiencing inner resignation can be ’brought back’ to being healthy and engaged employees. I revealed initial evidence that the provision of rewards and therefore the restoration of balance between effort and reward might help with this endeavor, but longitudinal studies are needed to fully understand and test the underlying process.

3.1. Limitations

As with all empirical studies, this study also has some limitations. First and foremost, the number of items measuring the latent constructs was too low; most notably, one-item measures were used for coercive power. The goal was to design an attractive questionnaire that could be filled in easily by a high number of employees. This resulted in sacrificing the general ’rule’ that each construct should be measured with at least three items. Future research should therefore use expanded measurements in order to capture the underlying constructs with adequate reliability.

Second, since answering each questionnaire item was completely voluntary and made salient by the option ’I don’t know’, the data contained a considerable number of missing values. Power perceptions are a sensitive topic (particularly when employees feel inner resignation). Since human resources departments also helped to distribute the questionnaires, some employees might have been reluctant to provide their honest opinion and decided to skip the respective items – despite the assurance of anonymity and the very selective use of socio-demographic questions. The original aim was to reduce social desirability bias, which is considered a way to deal with the threat of common method bias Conway & Lance (2010). This gave respondents the chance to opt out of some questions without having to refrain from answering the survey completely, but resulted in a high number of missing values.

Third, all measures obtained are self-reports relying on participants’ honesty and ability to comprehend and answer the items correctly. Furthermore, there is a risk of response bias and social desirability bias. Since I was interested in analyzing employees’ power perceptions and not actual enactments of power per se, I selected a questionnaire design as the most suitable method. Nonetheless, comparing the questionnaire responses to actual power enactments in organizations would have added further value and might represent an interesting avenue for future research.

Finally, because of the correlative nature of the data, causal interpretations of the results must be treated with caution. Strictly speaking, it is not valid to conclude that reward supervisor power can be used to ’bring employees back to the path of greater engagement’ when they are already experiencing inner resignation. The other causal direction also seems plausible: Organizations and supervisors can adjust their power enactments based on their evaluation of employees’ work behavior. I conclude that reciprocal effects seem most likely and stress the need for longitudinal studies to answer the question of causality.

3.2. Practical implications

Since today’s organizations are undergoing constant changes (Schwarzmüller, Brosi, Duman, & Welpe, 2018), organizations have come to strongly rely on employees’ contextual performance (Aronson, 2018; Johnson, 2001). In line with expectations, this study shows that coercive and legitimate power by supervisors and even more importantly, by organizations are of critical significance for stimulating employees’ contextual performance. Organizations need to provide an environment that enables employees to take responsibility for their actions and understand the rationale behind organizational rules and policies, while refraining from exercising force and coercion via punishments. Although this may require significant effort, organizations and supervisors should always strive to legitimize their actions and explain the rationale behind their decisions. This seems to not only foster employees’ contextual performance, but also inhibit inner resignation.

Human resource managers are more and more aware of the potentially detrimental effects of rewards Kohn (2018), given the vast literature showing that the provision of rewards crowds out employees’ intrinsic motivation (Gagné & Deci, 2005). This paper reveals that reward supervisor power could also crowd-in motivation, as it strengthens the inhibiting relationship between organizational legitimate power and employees’ inner resignation. But how should organizations deal with these ostensibly contradictory guidelines for engaged and disengaged employees? I argue that the provision of rewards is not bad in principle, but rather that the specific rewards employed need to be chosen wisely in order to support organizational legitimacy. This is always easier on an interpersonal level, where the intention and underlying rationale for the reward can be negotiated between supervisor and employee.

Self-determination theory (Deci & Ryan, 1980; Deci and Ryan, 1996; Ryan & Deci, 2000b) states that rewards are not detrimental for employees’ well-being if they are used to support the satisfaction of employees’ basic needs for autonomy and competence. When employees are amotivated, the provision of rewards might even foster satisfaction of the need for autonomy, as it enables employees to decide whether they want to direct their behavior towards pre-defined indicators. Furthermore, the exertion of reward supervisor power might also restore the balance between employees’ work efforts and rewards (Siegrist, 1996). If employees have psychologically withdrawn from their job, the provision of rewards might even allow supervisors and employees to improve their relationship quality, since giving out rewards creates good will and facilitates communication about successful events.

Finally, I highlight the importance of supervisors for shaping employees’ perceptions of organizational power. Supervisors should take an active role in explaining the rationale of organizational policies, as they are considered representatives of the organization (Eisenberger et al., 2010). Since organizations are becoming flatter and employees are working more autonomously, perhaps even working more from home (e.g., due to the COVID-19 crisis), supervisors are of paramount importance for helping employees understand the organization’s principles. This fosters employees’ perceptions of legitimate supervisor and organizational power.

3.3. Conclusion

This study analyzes how perceived power relates to employees’ contextual performance and inner resignation. When coercive power is perceived as high, employees are more likely to develop inner resignation. On the other hand, higher legitimate power is associated with higher contextual performance and lower inner resignation. I highlight the paramount role of supervisors in supporting employees’ perceptions of legitimate organizational power, which can further be strengthened by their instrumental use of rewards. This can particularly motivate employees who have already resigned internally through a crowding-in mechanism. By shedding light on inner resignation, this paper contributes to our understanding of the effects of power dynamics on counterproductive work behavior.