Acquiescent and Defensive Silence in an Indonesian Context

Although employee silence is already well-known to cause harms to both employees and organizations, less is known about the individual and situational factors that can influence it. This study reveals the relationships among acquiescent silence, defensive silence, psychological contract breaches, job-based psychological ownership, voice efficacy, psychological safety and task cohesion. Employing scales with good reliability scores (α between 0.8 to 0.95), we conducted a survey on a sample of of 260 public employees of an Indonesia's government institution. Analysis indicates that (1) individual factors (voice efficacy and psychological contract breach) and situational factors (task cohesion and psychological safety) work hand in hand to affect silence behavior; and (2) job-based psychological ownership has no relationship with acquiescent and defensive silence. This paper discusses (1) the importance incorporating individual and situational factors in understanding silence behavior; and (2) the collectivistic nature of Indonesian people that may contribute to the importance of situational factor (i.e., task cohesion) on silence behavior well and beyond psychological ownership.


Introduction
The attempt of the then chair of Indonesia's national parliament to secure by gift 20% stake (nearly $US4 billion) of the world's most profitable mining companies, which operates in Indonesia, failed because the mining company local Director decided to speak up about it to make sure that public was aware of this move (Mulholland, 2015). The situation would be different had that local director decided to stay silent. The detrimental effect of silence can be seen in the case of the United States' second largest long distance telephone company, WorldCom. Two of its accounting managers were actually aware of WorldCom's financial problem, and that WorldCom's Chief Executive Officer, Chief Financial Officer, and its senior internal auditor had cooked WorldCom's financial report (Akhigbe, Martin, & Whyte, 2005). However, they chose to remain silent, letting WorldCom to lose a total asset of $US 11 billion and declared bankruptcy with 20.000 employees lost their jobs (Akhigbe et al., 2005).
These events are examples to business companies all over the world that apparently silence is not always golden. These cases are only a few of thousands other similar events that tells us how precious the information employees may have and how deadly the silence of employee is, both to company and employees.
Employee silence is employee's behavioral tendency to intentionally withholding relevant ideas, information or opinions that are potentially beneficial for their companies (Van Dyne, Ang, & Botero, 2003). Based on the motivation underlying it, researchers differentiate acquiescent silence and defensive silence (Van Dyne et al., 2003). Acquiescent silence is disengaged behavior based on resignation and low self-efficacy, while defensive silence is a form of self-protective behavior based on fear (Van Dyne et al., 2003). Employees who choose to do acquiescent silence do not have conviction that they are able to change the situations around them. They then tend to extend their tolerance of what they perceive as wrong, and accept that situation as what it is. On the other hand, employees who do defensive silence choose to be silent because they believe that they will receive negative consequences if they speak up.
Previous research has indicated that employee silence may cause negative impacts to companies, like high turnover, slow and ineffective organizational development. It also brings bad consequences among employees, like low motivation, low job satisfaction, withdrawal, low rate of well-being, stress, and strain (Knoll & Van Dick, 2013;Sehitoglu & Zehir, 2010;Vakola & Bouradas, 2005). Discussion related to employee silence becomes more important because eliminating employee silence may bring advantage, such as the potential to identify troubles immediately or to collect constructive ideas for organization success. Those potentials may only happen if employees do not choose to remain silent. In other words, eliminating employee silence could be the key of organization success.
Scholars have investigated how some personal and situational factors affect employee silence. Some of the personal factors are commitment, job satisfaction, wellbeing, strain, trust, work-group identification, perceived justice, and political skills (Knoll & Van Dick, 2013;Tangirala & Ramanujam, 2008). The previously investigated situational factors are supervisor status, organizational climate of silence, punishment, procedural justice, and chance of voice (Knoll & Van Dick, 2013). However, limited studies have integrated personal and situational contexts in order to see their effect on both acquiescent and defensive silence.
Specifically, we observe that some personal factors at individual level have been overlooked or need further exploration. They are psychological contract breach (PCB), job based psychological ownership (JPO) and voice efficacy. Psychological contract breach refers to employees' perception to what extent their organization fails to deliver their obligation (Agarwal & Bhargava, 2013;Ng, Feldman, & Butts, 2014). This construct is based on the concept of psychological contract, which is defined as "… an individual's beliefs regarding the terms and conditions of a reciprocal exchange agreement between [a] focal person and another party.
[This] include[s] the belief that a promise has been made and a consideration offered in exchange for it, binding the parties to some set of reciprocal obligations" (Rousseau, 1989, p.123). Psychological contract is subjective and implicit, and, is usually measured in terms of to what extent employees perceive the occurrence of contract breach (Bal & Vink, 2010).
When employees perceive that the organization has failed to deliver promised mutual obligation, they may find themselves in a state of injustice and lack of trust to their supervisors or organizations. Employees with a state of inequity may be limited in expressing constructive ideas for their organizations. In other words, PCB may potentially induce silence, both, acquiescent and defensive silence. Therefore, we hypothesize that: H1a: Psychological contract breach is positively associated with acquiescent silence. H1b: Psychological contract breach is positively associated with defensive silence.
Job-based psychological ownership (Job-based PO), known as individuals feeling of possession toward their particular jobs (Mayhew, Ashkanasy, Bramble, & Gardner, 2007), is another potential variable affecting silence that needs further investigation. Feeling of ownership is generally experienced toward an object, but according to Pierce, Kostova, and Dirks (2001), it can also be felt toward non physical object, such as to an organization or a job. When individuals feel that their organizations are theirs, they could be willing to pay more attention to their organizations, and to assist and progress their organizations (Beggan, 1992). Psychological ownership triggers employees to commit extra work voluntarily, to protect and to speak up for the benefit of their jobs or their organizations (Pierce, Kostova, & Dirks, 2003;Pierce & Jussila, 2011), even at their own expenses. Van Dyne and Pierce (2004) even demonstrate that psychological ownership is able to predict to what extent employees are willing to speak up over and above job satisfaction or organizational commitment, which have been recognized as two well-established predictors for organizational citizenship behavior (Podsakoff, MacKenzie, Paine, & Bachrach, 2000).
Since, the effect of job-based is stronger than organizationbased psychological ownership in service organization (O'Driscoll, Pierce, & Coghlan, 2006), this research focusses on job-based psychological ownership. Previous studies show that when employees experience psychological ownership toward their job, they tend to feel tied to the job, and thus actively participated in improving the quality of their job (Pierce et al., 2001;2003). Therefore, we hypothesize that: H2a: Job-based psychological ownership is negatively associated with acquiescent silence. H2b: Job-based psychological ownership is negatively associated with defensive silence.
The third variable is voice efficacy, which refers to the extent to which employees believe that they are capable of speaking up (Tangirala, Kamdar, Venkataramani, & Parke, 2013). Efficacy potentially plays a key factor in triggering employees to speak up, considering perceived competence of oneself is particularly important in challenging the status-quo (Bandura, 1994). As demonstrated by McAllister, Kamdar, Morrison and Turban (2007), voice efficacy is related to taking charge, and one way to do it is by not remaining silent. When employees decide, not to remain silent but, to express their ideas or suggestions, they would never be sure if their ideas would be accepted or instead, would cause their peers to feel offended and retaliate. In line with that, Ashford, Rothbard, Piderit, and Dutton (1998), in their massive sample study, find that efficacy significantly affects challenging behavior, such as issue selling, while Tangirala et al. (2013) find that voice efficacy enhances positive relationship between duty-oriented and speaking up. Therefore, we hypothesize that: H3a: Voice efficacy is negatively associated with acquiescent silence. H3b: Voice efficacy is negatively associated with defensive silence.
Besides individual level variables, we argue that employees' experience with their working group will also have substantial impact on silence. Two potential variables that may affect silence are: (1) how much employees in a group are committed to achieve the group goals (i.e., task cohesion) and (2) how safe they feel to execute actions leading to the achievement of group goals (i.e., psychological safety). For the first aspect, cohesion is a dynamic process that reflects to what degree members of a group remain in their group to pursue its goal and/or to fulfill member's affective need (Carron & Brawley, 2012). It consists of task cohesion and social cohesion. While social cohesion focuses on member's affective needs, task cohesion emphasizes task-orientation behaviors among the group members. Because of the task oriented behavior, task cohesion may substantially affect employees' behavior to improve the quality of their tasks.
A longitudinal study demonstrates that task cohesion is a strong predictor of group performance and stimulates the sense of task-oriented among group members (Chang, Duck, & Bordia, 2006). Similarly, task cohesion triggers creativity of a group of employees, which come along with constructive behaviours (Joo, Song, Lim, & Yoon, 2012) that lead employees to actively contribute to the group, including by expressing their thoughts to achieve the group's goals. In contrast, employees tend to keep silent if there is no task cohesion. It is then argued that employee silence depends on the perceived task cohesion within the group with whom employees are working. Therefore, we argue that: H4a: Task cohesion is negatively associated with acquiescent silence. H4b: Task cohesion is negatively associated with defensive silence.
The concept of psychological safety refers to individual perception that s/he is able to express themselves without fear of negative consequences to his/her self-image, status, or career (Nembhard & Edmondson, 2006). More than individual consequences, Edmondson and Lei (2014) observe that psychological safety refers to the perceptions of the consequences of taking interpersonal risks particularly in a workplace. Previous research supports the idea that employees with high level of psychological safety feel safe to express their ideas and opinions (May, Gilson, & Harter, 2004). Siemsen, Roth, Balasubramanian and Anand (2009), in a research conducted in three different industries, also found that employees with higher psychological safety communicate more frequently than employees with lower psychological safety. In contrast, employees tend to remain silent when they perceive a possibility to be ignored or confronted by other group members (Nembhard & Edmondson, 2006). Employees who experience a low level of psychological safety, may tend to restrict risktaking behavior to avoid any negative consequence.
Lacking of the psychological safety may induce the fear of being viewed or labeled negatively, as well as the possibility of having bad career or receiving punishments including social punishment such as being ridiculed. Therefore, we hypothesize that: H5a: Psychological safety is negatively associated with acquiescent silence. H5b: Psychological safety is negatively associated with defensive silence.

Methods
Participants in this study were 181 male and 79 female employees (Mage = 29.92, SD = 4.73) from an Indonesian government institution that had applied the 'whistleblowing' system. This system allows every employee to report any illegal, immoral or illegitimate actions observed in the workplace anonymously, and is known to decrease employee silence (Vakola & Bouradas, 2005). The average organizational tenure of participants was 8.15 years (ranged from 1 to 21 years).
Most participants had attained a bachelor's degree (52.90 %), 19.90% held a master's degree, and 18.10% had an associate degree.
Employee Silence. Acquiescent silence and defensive silence were assessed using scales developed by Van Dyne et al., (2003). To get more valid result, the scales were adapted from the original supervisor-report form into self-report form for employee silence is best measured with self-report method (Knoll & Van Dick, 2013). The scales consists of five items to measure acquiescent silence (α= 0.89) and six items to measure defensive silence (α= 0.88). A sample item measuring acquiescent silence is "There is no point of expressing any ideas, because the organization will not change anyway. A sample item for defensive silence is "I don't want to express my opinion in a meeting because I'm afraid of ruining my interpersonal relationship with coworkers". Psychological Safety. Psychological safety was assessed using nine items adapted from by Nembhard and Edmondson (2006). A sample item is "Employees in this division do not respect my effort".
Voice Efficacy. We measured voice efficacy using items adapted from Spreitzer (1995). We added four 4 items to this scale to increase the measurement quality of the scale. A total number of eight items used in this scale (α= 0.85) with a sample item "I am confident about my ability to speak up on work-related issues in my organization".
Task Cohesion. Task cohesion among the participants was assessed using six items adapted from Group Environment Questionnaire (Carron, Widmeyer, & Brawley, 1998). The scale has reliability coefficient of 0.80. A item sample is "Our team is united to reach its goals".

PCB.
PCB was assessed using scale adapted from Robinson and Morrison (2000). The scale consists of five items (α= 0.95). A sample item of this scale is "My employer has broken many of its promises to me even though I have kept my side on the bargain".

Job-based Psychological Ownership.
Job-based psychological ownership was measured using six-items scale adapted from Mayhew et al. (2007) with α= 0.88. We modified this scale by adding new instruction for participants to briefly write down some activities that they do as part of their job. This modification was intended to provide a context for employees to answer each item. A sample item of this scale is "This is my job".
Harman's single-factor test was conducted to examine the potential issues related to common method variance (CMV) bias due to one-time data collection (Podsakoff, McKenzie, & Podsakoff, 2012). Additionally, the unrotated factor solution involving all exploratory factor variables was also analysed. Results suggested that no single factor accounted for the majority of the covariance in the independent and criterion variables, suggesting no common method bias.

Control Variables.
Previous research has shown that demographic variables (i.e., gender, age, education, and organizational tenure) have influences to the tendency of employees to speak up or to remain silent (Stansbury & Victor, 2009;Whiteside & Barclay, 2013). We also measured to what extent employees have ideas (having ideas variable). It is important to know that employees are silent because they intentionally want to do it, not because they don't have opinions or ideas to bring up. The 'intentionally remain silent' element is necessary in the concept of employee silence (Knoll & Van Dick, 2013;Van Dyne et al., 2003;). Having ideas was measured using two items adapted from Burris, Detert, and Chiaburu (2008). The items are "I have ideas about how to make this company better" and "I have ideas about how my job could be done better" (α= 0.65).
The availability of opportunities to speak up is negatively correlated with employee silence (Vakola & Bouradas, 2005;Knoll & van Dick, 2013), thus this factor also needs to be controlled. The more existing medias employees can use to communicate, the less their tendency to remain silent. Frequency of employee communication via email and informal meeting, were each measured using a single item. Table 1 displays the bivariate correlations for all variables under investigation. It shows that there is no correlation between demographic variables and both types of silence. Therefore, demographic variables will not be included in the hierarchical regression analysis. Table 1 also shows that there is negative correlation between frequency of employee communication via email and both acquiescent silence (r= -0.12, p< 0.05), and defensive silence (r= -0.11, p< 0.5), and frequency of communication in formal discussion with acquiescent silence (r=-0.13, p< 0.5), and with defensive silence (r=-0.12, p< 0.5). We also found moderate negative correlations between acquiescent silence and having idea towards both organization (r= -0.16, p< 0.01) and job (r= -0.20, p< 0.01).

Results and Discussion
As expected, we found that all variables under investigation were significant predictors of acquiescent silence and defensive silence (see Table 1 for details). These findings allowed us to test the hypotheses using hierarchical regression analysis. For each dependent variable, we conducted one hierarchical regression analysis. Based on correlation testing, we controlled for frequency of employee communication and having idea.
Hierarchical regression analyses were executed twice to explain each type of employee silence. There were only four of five predictors (PCB, voice efficacy, task cohesion, and psychological safety) that explained 56% variance of acquiescent silence (Table 1 & 2). Task cohesion was the strongest contributor among four predictors (β= -0.54), while job-based PO did not significantly contribute to acquiescent silence. As for defensive silence, 53% of variance was explained by job-based PO, voice efficacy, task cohesion and psychological safety, but not PCB. Voice efficacy was found to be the strongest predictor comparing to the other three (β= -0.53).

Table 2. Hierarchical Regression Analysis for Acquiescent Silence
Step 1 Step 2 Step 3 Step 4  Step 1 Step 2 Step 3 Step 4 Table 3). The model explains 53% variance of defensive silence. The only predictor that does not contribute to defensive silence is psychological contract breaches (PCB) (H1b rejected). As seen in Table 3, both situational factors are significantly associated with defensive silence (H4b dan H5b accepted). Table 3 also shows that voice efficacy is the strongest predictor of defensive silence This current research has advanced knowledge on silence behaviour, in at least three areas. First, the current study shows that individual and situational variables, specifically voice efficacy, task cohesion, and psychological safety, are working hand in hand in influencing both acquiescent as defensive silence. This current research demonstrates that these individual and situational variables contribute to more than 50% of variance for acquiescent as well as defensive silence. In other words, this current research suggests that, in addressing silence, scholars and practitioners should focus on both levels -individual and group levels. Only by focusing on variables at both levels, scholars may get a better understanding of silence, and practitioners may well reduce the tendency of acquiescent as well as defensive silence.
Second, the result of this current research questions the dynamic of how psychological contract breach is associated with silence behaviour. Results show that psychological contract breach is positively associated with acquiescent silence but not defensive silence. The more employees experience psychological contract breach the more they perceive that they have to accept the situation as it is and choose to be silent. This occurs, most likely, because perceived contract breach induces the perception of inequity and may strengthen employees' perception that they will not be able to change the situation and, thus, not expressing their concern. Interestingly, results show that psychological contract breach is not associated with defensive silence. This occurs, most likely, because the level of perceived contract breach in this research setting is not high (i.e., 2.93 in the range between 1-5). The results could be different in a context of high perceived contract breach. Therefore, more investigations need to be done to understand silence behaviour.
Third, the current research contributes to organizational behaviour theory by indicating that job-based psychological ownership is not associated with, either acquiescent nor with defensive silence, suggesting that the individual perception of ownership is not a powerful predictor for silence behaviour. This result is surprising given previous research indicate the strong effect of psychological ownership on voice behaviour (Van Dyne & Pierce, 2004). One explanation perhaps related to the argument that voice and silence maybe two different constructs that have different antecedents. In other words, although psychological ownership is strongly associated with voice behavior, more research needs to be done to delineate the antecedents of silence behavior and to see how it differs with the antecedents of voice behavior.
Another possible explanation for this unique result could be related to the the collectivistic nature of Indonesian people. In line with Munawaroh, Riantoputra and Marpaung (2013) who argued for the importance of interconnectedness and interdependence among group members in Indonesia, the results of this current research also suggests that a sense of duty toward one's group is more important than factors at individual level. This suggestion is evidenced in the association between task cohesion and psychological safety, which are group level factors, and silence behavior. Apparently, the perception that the group is glued to certain tasks may induce employees to be willing to express ideas for the benefit of the group. This group level factor has stronger association with silence than individual level perception of ownership. Thus, this current research calls for more studies to investigate the relationship between job-based psychological ownership and silence, especially in collectivistic culture.

Conclusion
In conclusion, this research shows that voice efficacy, psychological safety, task cohesion and perceived contract breach influence the occurrence of silence behaviour. These variables occur at both individual and group levels variables, suggesting the importance of these two aspects in understanding employee silence. Specifically, this current study indicates that task cohesion, a group level variable, is a more powerful predictor than job-based psychological ownership, an individual level variable. This result suggests that group level variable may be more important in explaining risk taking behaviour, such as silence. It also may indicate that group level variable plays a more critical role in a collectivistic culture, such as Indonesia. Both alternatives may need further investigation.