Too-Much-of-a-Good-Thing Effect of Prosocial Silence and Voice

This study assesses the effects of prosocial silence and voice on organizational citizenship behaviors directed towards individuals under the “Too-Much-of-a-Good-Thing” theory. It is assumed that greater prosocial silence and voice lead to organizational citizenship. However, the theory of too-much-of-a-good-thing suggests that extreme behaviors may perversely have a negative effect raising the possibility that the relationship is curvilinear rather than linear. A similar nonlinear relationship is suggested in this study. Standardized measures of prosocial voice, prosocial silence and organizational citizenship were collected from 381 faculty members from three mid-cycle universities. Regression analyses revealed a significant curvilinear (an inverted U-Shaped) relationship between prosocial voice and organizational citizenship and likewise prosocial silence and organizational citizenship. Too little and, similarly, too much prosocial voice and silence were associated with worse organizational citizenship.


Introduction
From past few decades' research is focused on positive work behaviors that improves the functionality of organizations. Among these variable, the most explored variable is organization citizenship behavior (OCB). Organ (1988) presented one of the pioneering and comprehensive definition of Organizational citizenship behavior as 'Individual behavior that is discretionary, not directly or explicitly recognized by the formal reward system and that in aggregate promotes the effective functioning of the organization' (p. 4).
According to Organ (1988), OCB is an often studied cooperative behavior that positively effects the organizations but cannot be enforced by employers under employment contracts. Although interest in these types of helpful behaviors is increasing, nonetheless there is a great debate among the researchers about the content, causes and effects of OCB (Podsakoff, MacKenzie, Paine, & Bachrach, 2000). An established positive relationship of OCB with employee task and non-task performance has been suggested by researchers. However, most recently there is also a growing debate on whether excess OCB and its offshoot variables improve performance or can we observe a "Too-Much-of-a-Good-Thing" effect. The purpose of this study is to evaluate the relationship between two off shoot variables of OCB i.e. prosocial silence and voice with OCB directed towards individuals through the lens of "Too-Much-of-a-Good-Thing".

Too Much of a Good Thing (TMGT) Effect
"Too much can be worse than too little" Chinese aphorism.
"Everything in moderation; nothing in excess," Western aphorism.
The above aphorisms state that excess of everything is bad and is acceptable across cultures and religions. The modern philosophers have studied this activism for moderation rather than extremism under the doctrine of golden mean. For them achieving the golden mean of moderation is their moral as well as practically imperative. The effects of this doctrine have been studied under the TMGT effects in all forms of research ranging from pure sciences to social sciences. This effect occurs when ordinarily beneficial predictor variables reach infliction level and the linear and positive relationship with criterion variable ceases. Beyond these infliction levels the relationship becomes either non-significant (no additional benefit) or undesirable outcomes emerge (e.g. Lack of OCB, commitment, performance).
In the management literature little focus is given to the concept of finding balance between deficiency and excess. Majority literature focuses on the former i.e. negative consequences due to lack of a relationship rather than the latter i.e. negative consequences of an excess of relationship. This approach has led to theory development and application that focuses on the assumption that "More is better." This in turn generates linear relationships between the predictors and criterion variables with the motivation to improve the desired variable. The confirmation and approval of these hypotheses reinforces our assumption that "More is better" and linear methodology best explains organizational relationships which may not be the reality.
According to Pierce and Aguinis (2013) 'TMGT effect occurs when ordinarily beneficial antecedents reach inflection points after which their relations with desired outcomes cease to be linear and positive, instead yielding an overall curvilinear pattern.' They further argue that TMGT is much more complex than the linear paradigms; enabling it to make valuable contributions towards theory and practice by explaining numerous inconsistent and paradoxical results in organizational theory e.g. organizational identification, level of morale, trust, autonomy, team dynamics and group size are some of the constructs that in excess have negative impact on citizenship behaviors. Excessive OCB has also been identified by researchers to have negatively related with performance related constructs (Bergeron, 2007;Klotz & Bolino, 2013). Furthermore, studies have also identified nonlinear relationship between performance based constructs and OCB (Kernodle, 2007;Ng & Feldman, 2011). In summary, the major objective of this study is to suggest a feasible yet unexplained nonlinear relationship between Prosocial voice and silence with OCB directed towards individuals.
Hypothesis Development. Prosocial Silence (PS) and OCB. Prosocial silence (PS) is defined as 'withholding work-related ideas, information, or opinions with the goal of benefiting other people or the organization based on altruism or cooperative motives' ( Van Dyne, Graham & Dienesch 1994. p. 1368. Korsgaard, Meglino, and Lester (1997) present PS as an others oriented, discretionary, proactive and intentional behavior. Podsakoff et al. (2000) have identified seven dimensions of OCB among which sportsmanship is most closely related to PS. Organ (1997) defines sportsmanship as 'the Prosocial absence of complaints; tolerating the inevitable inconveniences and impositions of work without whining and grievances' (p. 88).
Researchers have identified absence of complaints and withholding of grievances as PS. Furthermore, similar to other prosocial behaviors PS is based on showing patience and courtesy to others. PS previously has been recognized as an off shoot of OCB yet most recently literature presents a non-significant (Kılınç & Ulusoy, 2014) or negative (Çınar, Karcıoğlu, & Alioğulları, 2013;Fatima, Salah-Ud-Din, Khan, Hassan, & Hoti, 2015) relationship between these two variables. As literature proposes contrasting relationships between two ordinarily beneficial behaviors that is PS and OCB it is proposed that these could be explained through the TMGT Effect. Hence the nonlinear hypothesis developed for this relation is as follows: Hypothesis 1: The relationship between prosocial silence and organizational citizenship behavior exhibits an inverted U-shape relationship such that prosocial silence positively impacts organizational citizenship behavior to a specific point; beyond this point, the relationship between prosocial silence and organizational citizenship behavior becomes negative.
Prosocial Voice (PV) and OCB. Voice has been positioned as a positively intended behavior in majority of its literature. Hence to differentiate with other positively positioned voice behaviors, the other-oriented voice behavior is termed as Prosocial Voice (PV). Similar to PS it is also other-oriented, proactive and intentional. Dyne, Ang, and Botero (2003) define PV as 'expressing work-related ideas, information, or opinions December 2017 | Vol. 21 | No. 2 based on cooperative motives' (p. 1371). Similarly LePine and Van Dyne (2001) presents PV as a set of non-required expressions to bring about change in the organizations with the purpose of improving the situations. Organ (1988) has presented PV as the noblest form of OCB as it challenges the status-quo of the organizations and can result in personal risk and repercussions. However researchers have critiqued the notion that voice is predominantly prosocial construct (Barry & Wilkinson, 2015). Lin and Johnson (2015) shows a negative relationship between voice based on prosocial motives and OCB. Harlos (2001) states that research has neglected the negative contexts of PV. These negative contexts result in employees feeling disadvantaged and harmed from voice systems which are often assumed beneficial for both individuals and the organizations. Similar to PS literature, PV and OCB literature also propose contrasting relationships between these two ordinarily beneficial behaviors. Hence it is proposed that these could also be explained through the TMGT Effect. The following hypothesis is developed based on these propositions:

Hypothesis 2:
The relationship between prosocial voice and organizational citizenship behavior exhibits an inverted U-shape relationship such that prosocial voice positively impacts organizational citizenship behavior to a specific point; beyond this point, the relationship between prosocial silence and organizational citizenship behavior becomes negative.

Methods
Procedure and Sample. The model of the study was tested in three public sector universities in Pakistan. These universities were selected based on their respective lifecycle. A panel of experts evaluated all the universities of the region based on the Lester, Parnell, and Carraher (2003) 5 stages of organizational lifecycle. Three universities were selected, one from each life cycle category as all of the universities were categorized in middle three stage of lifecycle and none characterized in the top and bottom stages by the experts. The population of the study was 1039 individuals while 381 individuals completed questionnaires were used for data analysis (Response rate = 36.5%). The sample included 313 males (82.3%) and 68 females (17.8%), full time permanent employees were 266 (69.8%) while 115 (30.2%) were part time contract employee, lastly 326 (85.6%) were non-supervisory staff while 55 (14.4%) were supervisory staff as they reported 1 or more employees directly report to them. The average age of the sample was 35.45 years (SD = 7.76 years), mean education in years was 18.80 years (SD = 1.94 years) while average experience in years 8.35 years (SD = 7.76).

Measures.
Organizational Citizenship Behavior was measured by 7 item OCBI (Organizational citizenship behavior Individual) scale (e.g. I provide cover up for absent coworkers) developed by Williams and Anderson (1991) with reported reliability of α= 0.91.
Prosocial silence was measured by a 5 item scale (e.g. I protect confidential information relevant to my coworkers) developed by Dyne et al. (2003). Kılınç and Ulusoy (2014) report α= 0.899 for this scale.
For Prosocial voice, a 5 item scale (e.g. I communicate my opinions about work issues even if others disagree) developed by Dyne et al. (2003) was employed with reliability coefficient α= 0.87 reported by Lee, Diefendorff, Kim, and Bian (2014).
Control Variables: The following demographic variables known to impact OCB are age , gender, education in years, job contract, experience in years and supervisory status were used as control were collected. Furthermore, organizational identification (OI) was also used as a control variable for effective testing of our hypotheses by decreasing the potential effects of identification based display of OCB. OI has an established positive relationship with OCB e.g. (Qureshi, Shahjehan, Zeb, & Saifullah, 2011;Shahjehan & Yasir, 2015) and other facets of prosocial behavior including PS and PV (Hsieh, 2014;Knoll & van Dick, 2013;Shahjehan & Yasir, 2015;SHI & WANG, 2014). Age, education and experience were measure in years, while dichotomous scale was used for gender (1 Male, 2 Female), job contract (1 Part time Contract, 2 Full time permanent) and supervisory status (1 supervisory, 2 non-supervisory). Finally to assess OI A six items scale developed by Edwards and Peccei (2007) was used with reliability α coefficients ranging from 0.87 to 0.93 across different samples.
Analytic Strategy. Hypotheses of the study were test with the following regression equation: Two models were created with separately regressing OCBI (Y) with linear (X) and quadratic (X 2 ) terms of PV and PS to estimate the linear and quadratic effects.
To reduce the multicollinearity effects, all the variables were mean centered as recommended by Aiken, West, and Reno (1991). For both models, significant b 2 values would suggest support for our hypotheses. A three step hierarchical multiple regression analysis was computed based on the above mentioned equation. In the first step, control variables such as age in years, gender, education in years, job contract, experience in years, supervisory status and organizational identification are entered. In the second step the linear term and lastly the quadratic terms for PS and PV are entered.

Results
Before testing the hypotheses, a confirmatory factor analysis was conducted to evaluate the distinctiveness of the constructs used in this study. Table 2 shows the Fornell and Larcker's (1981) index of the average variance extracted, Jörekog's index of internal consistency reliability, correlation and Cronbach alpha coefficients for the study variables. The rhô statistic of internal consistency for all the 4 constructs was more than 0.7 benchmark suggested by Fornell and Larcker (1981). The Fornell and Larcker's (1981) index of the average variance extracted ρvc(η) ranging from 0.55 to 0.76 is also above the threshold of 0.5. The Cronbach alpha value for the constructs ranges from 0.73 to 0.92 which is above the acceptable value of 0.7.
As the results of table 3 shows that Quadratic term main effects of both the PS (β = −0.12, p < .01) and PV are (β = −0.12, p < .01) are significant thus confirming a curvilinear relationship between PS, PV and OCBI. These nonlinear relationships are displayed in figure 1 displaying inverted U-shaped relationships of both PS and PV with OCBI thus supporting both hypotheses of the study.

Discussion
The purpose of this study was to empirically evaluate TMGT effect in the relationships between PS and PV with OCBI. This effect has been confirmed by the inverted U-shaped relationships of both PS and PV with OCBI. Interestingly the linear relationship of OCBI were found to be significantly negative with both PS (r=-0.32, β= -0.16 & -0.36) and PV (r=-0.33, β= -0.19 & -0.37) providing further evidence of TMGT effects as PS, PV and OCB are ordinarily beneficial variables yet there exists a significant negative relationship between them. On a secondary note the curvilinear analysis also explained the unexpected significant negative linear relationship between PS and PV with OCBI.
Literature shows that prosocial behaviors would lead to enhanced display of citizenship behaviors (Detert & Burris, 2007;Dyne et al., 2003;LePine & Van Dyne, 2001). However, the results of our study show that the effects of prosocial motive based silence and voice are more complicated as nonlinear relation is found i.e. too much or too little of PS and PV may lead to decrease in the display of citizenship behaviors. This challenges the narrative of "the more the better" in the context of prosocial behaviors. Previous research have pointed out that nonsignificant or even negative relationships can exist between prosocial motive based silence, voice and OCB (Barry & Wilkinson, 2015;Çınar et al., 2013;Fatima et al., 2015;Harlos, 2001;Kılınç & Ulusoy, 2014;Lin & Johnson, 2015).
The application of TMGT is widely applied in the field of organizational behavior and employee relations. To simplify the concept of TMGT, it is a phenomenon in which good things lead to unexpected negative relations or consequences. Pierce and Aguinis (2013) have suggested three main implications that need to be considered when applying and analyzing this concept in management (1) Identify the location of construct specific infliction point, (2) Analyzing the theory boundary conditions through control or moderating variables and (3) Specifying the shape of nonlinear relationship whether it is asymptotic (U-shape) or negative (inverted U-shape).
In the study at hand these implications have been taken into consideration. The location of the infliction points for this study are near the origins, for both PS and PV as the linear relationships are negative and the curvilinear curves are inverted U-shaped it can be inferred that the initial positive relationships are being neutralized by the later strong negative relationship resulting in the movement of the infliction point towards the left side of the graph. The boundary conditions of theory are satisfied by the induction of control variables i.e. demographic variables and OI, which act as important situational factors for the relationships of the study.
Lastly before analysis, based on literature an inverted U-shape relationship was proposed between prosocial silence, voice and citizenship behavior which was proven by significant results in hierarchical multiple regression analysis and plots shown in figure 1. This result of the study conceptually and empirically contributes towards the TMGT effects in the area of organizational citizenship behaviors.
This study has major practical implications for organizational leaders and managers. The study challenges the notion that prosocial organizational behaviors would lead to citizenship behaviors. Especially excessive voice and silence behaviors based on prosocial motives have a negative relationship with OCB to the extent that it would neutralize OCB's initial positive effects. As Barry and Wilkinson (2016) point out that the prosocial behaviors are viewed through a unitarist lens that is 'what is good for the firm must be good for the worker. ' Detert and Burris (2007)  This study enhances our understanding about the relationships of prosocial silence, voice and organizational citizenship behaviors, yet it also has some limitations. First this is a cross-sectional study which does not fully encompass the casual relationship between the variables of the model. It is therefore recommended that for future studies longitudinal or experimental methodologies may be employed to address the issue of causality. Second and most important the data was collected in Pakistani organizations; therefore, the effect of culture could not be ruled out. Future studies should use data sets from other cultures to verify the generalizability of our finding and also report the effects of cultures on the results of our study.
Despite the limitations this study enhances our understanding of prosocial behaviors in general and prosocial voice, prosocial silence and organizational citizenship behavior in specific. We have provided empirical evidence in support of nonlinear relationships between these three prosocial motive based behaviors. Although citizenship behaviors and prosocial behaviors have received a lot of interest, recent literature has been calling for empirical studies to test curvilinear relationship between them. This study has recognized this gap and demonstrates nonlinear relationship between these three prosocial behaviors. Finally, this study points out towards the usefulness of TMGT framework in the study of prosocial behaviors. Additional research in the field is crucial to enhance our understanding of prosocial behaviors.

Conclusions
In this study, we have investigated a nonlinear relationship of PS and PV with OCB to broaden the theoretical and empirical foundations of extra-role behaviors. This study is an attempt to draw attention to a more general possibility that a nonlinear relationship may exist between OCB and its antecedents. By doing so we partly explain why previous researches that overlooked the nonlinear relationship between OCB and its antecedents presents poorly validated and inconsistent results.
If this study had also tested these relationships for linear effects we wouldn't have detected the Too Much of a Good Thing effect between PS, PV and OCB. Furthermore, complex behaviors such as OCB could be explained effectively if research explored theoretically justified nonlinear relationships. Lastly extra-role behaviors literature, constructs that improve interpersonal affiliation and cooperation at work are thoroughly studied, however our study point out to extra-role behaviors like PS and PV that challenges organizational status quo thus expanding its relationship with change oriented behaviors.