The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior

Purpose – The aim of this paper is to examine whether the perception of job insecurity negatively affects procedural fairness, which may relate to the Organizational Citizenship Behavior (OCB). Finally, we have tested whether perceived supervisor support has a moderating effect in the relationship between procedural fairness and OCB. Design/methodology/approach – Data were collected from a sample of 707 employees in a metal-mechanic company in Chile. Regression analysis was used to analyze the mediating and moderating effects. Findings – It was found that procedural fairness fully mediates the relationship between Job Insecurity and OCB, such that Job Insecurity is negatively related to procedural fairness, which in turn is positively related to OCB. Also, perceived supervisor support moderates the relationship between procedural fairness and OCB. Originality/value – The paper provides new theoretical insights on the effects that perceived job insecurity has on OCB, giving an empirical insight of the mediator and moderator effects that aid in the comprehension of this relationship.


Introduction
Recent economic recessions, competitive pressure and a drastic increase in restructuring activities have had a global impact on organizations, which are increasingly using downsizing and mass layoffs in response to economically difficult circumstances (Coile & Levine, 2011;Datta, Guthrie, Basuil, & Pandey, 2010).These environmental and organizational changes have resulted in an increase in job insecurity perceived by workers.Given its importance in current society, it is crucial to carry out empirical studies that provide a deep understanding of the impacts of job insecurity within the organizations.
Perceptions of job insecurity represent a relevant variable for organizational behavior scholars and practitioners, and it has been defined as an employee's feeling or an overall concern that his or her job is at involuntary risk in the near future (Grunberg, Moore, & Greenberg, 2006;Jacobson, 1991).
Wi t h i n t h i s c o n t e x t , a s e r i e s o f investigations have highlighted that the perception of job insecurity has the potential to affect job performance (De Witte, 1999;Gilboa, Shirom, Fried, & Cooper, 2008;LePine, Podsakoff, & LePine, 2005;Probst, 2002).However, scarce research has focused on expanding its relationship with job insecurity.Given the inevitable characteristics of competitiveness in organizations, and the value of team work and cooperation, along with individual initiative (Eatough, Chang, Miloslavic, & Johnson, 2011), it becomes necessary to further investigate how organizational citizenship behaviors are affected when there is a greater perception of job insecurity among workers.
Based on the above, this paper aims to contribute to the literature through the study of the relationship between the perception of job insecurity and the organizational citizenship in the context of social exchange theories (Blau, 1964), a relationship that has received scarce examination in the past (Cheng & Chan, 2008;Staufenbiel & König, 2010;Reisel, Probst, Chia, Maloles, & König, 2010;Stynen, Forrier, Sels, & De Witte, 2015).Such discretionary behaviors help to develop and maintain a favorable social and psychological climate, and the organizational efficiency and effectiveness (Podsakoff, Whiting, Podsakoff, & Blume, 2009).
Another interesting line of discussion that has not yet been academically addressed lies on studying the perception of job insecurity as a phenomenon that can be transformed according to a certain outcome, depending on the evaluation that workers provide about the value of fairness and consistent procedures within the organization.In other words, we propose procedural fairness perceptions as an explaining mechanism why job insecurity is associated with organizational citizenship behaviors.In this sense, only a few empirical investigations (Bernhard-Oettel, De Cuyper, Schreurs, & De Witte, 2011;Ouyang, Sang, Li, & Peng, 2015;Schumacher, Schreurs, Van Emmerik, & De Witte, in press;Zhang, Lepine, Buckman, & Wei, 2014) have related perception of job insecurity with fairness.
In the context of perception of job insecurity, workers closely observe how the organization acts and the treatment that laidoff workers receive.They attribute justice or injustice to these procedures, and these signals are important references on how workers that stay in the organization can be treated in the future (Brockner et al., 1994;Kernan, & Hanges, 2002;Van den Bos, Bruins, Wilke, & Dronkert, 1999).Experiencing injustice when employees have a feeling of job insecurity contributes to negative outcomes, for example in organizational citizenship behavior (Cohen-Charash & Spector, 2001;Moorman, Blakely, & Niehoff, 1998;Niehoff & Moorman, 1993).
On the other hand, a moderating factor that has raised interest for studying The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior organizational research is the workers' perception of supervisor's support.One inquiry is how this perception of support can alter the employees' psychological responses and working behaviors (Bal, Chiaburu, & Jansen, 2010;Dulac, Coyle-Shapiro, Henderson, & Wayne, 2008;Robinson, 1996).On this basis, it has been proposed that two entities or individuals, such as the supervisor and the subordinate, form and maintain an interpersonal relationship that implies reciprocal resources valued by both parts (Dulac et al., 2008;Eisenberger, Stinglhamber, Vandenberghe, Sucharski, & Rhoades, 2002;Huy, 2002;Tepper & Taylor, 2003).In this regard, supervisors have a powerful influence on the employees' interpretation of their work experiences (Piccolo & Colquitt, 2006).
However, investigations have not studied in-depth how the supervisor can alter the perception of procedural fairness.This research proposes to evaluate how the perception of supervisor support can moderate the magnitude and severity of procedural fairness, generated by the perception of insecurity among employees.Thus, the relationship between procedural fairness and OCB will be moderated by perceived supervisor support and this relationship will be stronger for people perceiving high supervisor support.By means of this, we contribute to the existing research on social exchanges by focusing on contextual factors as moderators in these relationships.
Finally, through the results gathered in this research, new knowledge about the effects that job insecurity perception have on workers will be generated, providing a deeper emphasis on how the effects of job insecurity may be explained, and the comprehension of the mediating and moderating mechanisms that underlie this relationship.Also, new evidence will be provided on the comprehension of a relationship that has been poorly studied: perception of job insecurity and procedural fairness.Furthermore, this study contributes to extending our knowledge on the role of supervisor support in an important relationship, such as between the procedural fairness and OCB.This will allow us to create a deeper theoretical background of the psychological dimensions that motivate the organizational performance.
From a practical point of view, this investigation attempts to contribute with information that aids the development of strategies that minimize the negative effects of job insecurity in workers and in the organization.This is achieved by providing evidence that can be used for planning, and effectively executing procedures for the post-dismissal working environment.To prevent these negative consequences, the organization should minimize procedural fairness problems, giving more realistic expectations to workers and managing those sensitive issues of the procedural fairness, such as having fair procedures.Finally, this investigation intends to give new evidence that will help to a better understanding of the value of employees within an organization, so it can achieve goals associated to performance.In this sense, it is important for organizations to focus their efforts on the establishment of a solid psychological contract with their workers.

Job insecurity and organizational citizenship behavior
Changes in the work environment caused by factors such as economic crisis, market regressions, mergers, privatization and technological innovation have generated a series of organizational and management mutations (Cascio, 1993;Coile & Levine, 2011;Datta et al., 2010;Gandolfi, 2010).One of these transformations is a higher number of laid-off workers, which goes hand in hand with the strategy of achieving higher efficiency, productivity and competitiveness through reduction of the work force.These conditions generate a higher sense of job insecurity (Greenhalgh & Rosenblatt, 1984;Lam, Liang, Ashford, & Lee, 2015).
One of the most important definitions of job insecurity is proposed by Greenhalgh and Rosenblatt (1984) who define it as "the perceived Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso powerlessness to maintain desired continuity in a threatened job situation" (p.438).Under this conceptualization, job insecurity is considered a subjective phenomenon based on an individual's appreciation of uncertainty about their current employment (De Cuyper & De Witte, 2007;Hellgren, Sverke, & Isaksson, 1999).
Job insecurity corresponds to a potential chronic threat that has been analyzed by stress theories as a way to comprehend links between experience and employee outcome (Lazarus & Folkman, 1984).This theoretical approach proposes that the possibility of losing one's job can be stressful, depending on the cognitive evaluation made by the individual.During this process, the individual evaluates his/her personal relevance on the situation, and later analyses the concordance between the demands of the situation and the resources to face it, which finally allows the identification of potential threats (Cavanaugh, Boswell, Roehling, & Boudreau, 2000).As such, psychological stress is determined by the individual's evaluation as a threatening situation, highlighting the lack of predictability and control as fundamental elements of the subjective experience of job insecurity (De Witte, De Cuyper, Elst, & Van den Broeck, 2012).
Research shows that job insecurity in this sense yields negative outcomes for the worker and the organization (Sonnentag & Frese, 2003;Sverke et al., 2002).In the subject, evidence shows that job insecurity is correlated to mental or psychological well-being, as well as some physical health markers (Gilboa et al., 2008;Cheng & Chan 2008;Kinnunen, Mauno, Natti, & Happonen, 1999;Sverke et al. 2002).These studies show that job insecurity can have immediate and long-term effects on organizational costs (Sverke et al., 2002).Effects such as job attitudes and satisfaction are shortterm, while job performance and organizational citizenship are affected in the long-term.
In our research, we selected organizational citizenship behavior (OCB), as an example of important valued behaviors, understood as discretional acts that are not directly recognized by the reward system.When employees as a group show OCB, they improve the organizational performance (Podsakoff et al., 2009).
According to the social exchange theory (Blau, 1964;Gouldner, 1960), reciprocity stimulates a feeling of obligation towards others, mainly because the latter have exhibited past behavior that proved beneficial.The same is true for OCB, given that if an organization treats their employees properly, they should in turn respond in the same positive manner towards the institution (Coyle-Shapiro, 2002;Turnley, Bolino, Lester, & Bloodgood, 2003).
By contrast, employees can decrease their level of OCB when the organization violates this exchange relationship (De Cuyper & De Witte, 2006;Robinson & Morrison, 1995).Similarly, it has been shown that when employees feel that organizations have not kept their promises or fulfilled obligations, their OCB is negatively affected (Robinson & Morrison, 1995;Turnley et al., 2003).
Hence, the social exchange theory can be used to see the negative relationship between job insecurity and OCB.Workers who feel their employment is not in danger should respond favorably towards the organization, whereas when their position is threatened, and this causes negative feelings, they will not perceive the same obligation towards the organization, and thus, they exhibit a lower frequency of OCB.Therefore, we can postulate that job insecurity can lead to lower levels of OCB and this leads us to our first hypothesis.

H1:
Job insecurity has a negative relationship with organizational citizenship behavior.

Perception mediation of job insecurity, procedural fairness and OCB
Procedural fairness theory sustains that workers value fair and consistent procedures within the organization.This is because they hope that fair procedures mean fair results ( AmbroseThe effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior & Arnaud, 2005; Cohen-Charash & Spector, 2001;Lind & Van den Bos, 2002;Moorman et al., 1998).
In the context of perception of job insecurity, workers closely observe organizational actions and the treatment that laid-off workers receive.These signals are important references on how workers that stay in the organization can be treated in the future (Brockner et al., 1994;Kernan & Hanges, 2002;Van den Bos et al., 1999).Particularly, there are six key elements in the theory of procedural fairness that are more valuable in a job insecurity context, i.e., those decisions are consistent, unbiased, accurate, correctable and ethical (Leventhal, 1980;Thibaut & Walker, 1975).
The assessment of procedures fairness is especially important when stressful events or actions, such as a downsizing, occur within an organization.These can be interpreted as unfair, given that employees tend to blame the organization as the main source of stress.This situation results in a severe cognitive assessment of the overall relationship of the worker with the organization, causing these unmet expectations to alter aspects of the employment relationship by creating perceived unfulfilled obligations in it; particularly the expectations that the employer will deliver job security and adequate working conditions, such as fairness of the procedures in a post-dismissal working environment (De Cuyper & De Witte, 2006).
As a result, workers will balance the unfulfilled promises of the psychological contract, diminishing their obligations towards the organization in a behavioral or psychological manner (Bernhard-Oettel et al., 2011;Conway & Briner, 2005;Schumacher et al., in press).One way that workers can do this is by reducing their OCB (Turnley & Feldman, 1999).
By contrast, when organizations carry out practices that damage this relationship of exchange, following the reciprocity norms, workers will not show elevated levels of OCB, feeling less conditioned to behave in a supporting manner towards the institution.Other authors conclude similarly, showing how organizations that have not fulfilled promises or other obligations do not exhibit high levels of OCB (Coyle-Shapiro, 2002;Robinson & Morrison, 1995;Turnley et al., 2003;Zhang et al., 2014).
Hypothesis 2 therefore is: H2: Procedural fairness will mediate the relationship between Job Insecurity and Organizational Citizenship Behavior (OCB), such that Job Insecurity will be negatively related to Procedural fairness, which will in turn be positively related to OCB.

Moderation of the perception of supervisor support in the relationship between procedural fairness and OCB
The perception of supervisor support is the degree to which employees perceive that supervisors value their contribution and manifest concern for their wellbeing.This includes the manner in which they communicate and treat employees with respect and recognition (Eisenberger et al., 2002).
Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso From this point of view, supervisors are considered relevant to the development and maintenance of labor relations (Coyle-Shapiro & Conway, 2005;Griffin, Patterson, & West, 2001), as well as being responsible for giving information about objectives and values of the organization, implementing policies, managing workers, establishing standards and performance evaluations (Chen & Chiu, 2008;Vecchio & Bullis, 2001).
Via the supervisor, the employee receives different types of support.This can include emotional support through the demonstration of empathy, acceptance and care.It can also include informational support, related to receiving feedback, orientation or advice about work-related knowledge and skills (Griffin et al., 2001).Furthermore, material support is understood as assistance in preparing proposals and related tangible collaboration, measured by job performance (Chen & Chiu, 2008;Einsenberger et al., 2002;Tepper & Taylor, 2003;Zagenczyk, Gibney, Kiewitz, & Restubog, 2009).
However, one of the aspects that have received less attention in current research is the study of how the supervisor can be an agent that can modify workers' perceptions and their contribution to the organization.In this research, we suggest to study how the supervisor can alter the relationship procedural justice-injustice and OCB (Cohen-Charash & Spector, 2001;Colquitt, Conlon, Weson, Porter, & Yee, 2001).
With respect to the above, previous research has pointed out that the supervisor may have a role of social support that minimizes or dampens the negative experiences of workers (Cohen & Wills, 1985).However, more recent research has incorporated a new perspective called betrayal framework (Bordia, Restubog, & Tang, 2008;Elangovan & Shapiro, 1998;Gillespie & Dietz, 2009;Zagenczyk et al., 2009), which argues that a high perception of support from the supervisor could aggravate the negative effects that workers have on their attitudes and work behaviors (Kiewitz, Restubog, Zagenczyk, & Hochwarter, 2009;Suazo & Stone-Romero, 2011).Hence, it has been reported that the most significant act of betrayal is the one that involves the most important and close relationships, among which is the one with the supervisor.This relationship can generate higher expectations associated to their role and greater pressures and obligations in a context of increasing job insecurity (Morrison & Robinson, 1997).
Moreover, according to the betrayal perspective frame (Bordia et al. 2008;Elangovan & Shapiro, 1998;Gillespie & Dietz, 2009;Zagenczyk et al., 2009), employees that perceive a high supervisor support show a sharp fall in performance as job expectations about the organization decreases, for example, in terms of procedural fairness.Indeed, the supervisor's failure to prevent instances of decreased justice can be considered as an act of betrayal, product of the violations of rules and expectations inherent to a good relationship between the workers and their supervisors (Elangovan & Shapiro, 1998;Jones & Burdette, 1994;Zagenczyk et al., 2009).In this sense, workers have the expectation that their supervisor will provide emotional support and resources, which will lead to attitudes of respect, trust and mutual obligations (Lynch, Eisenberg, & Armeli, 1999).However, the failure of the supervisor to prevent instances of decreased justice can be considered an act of betrayal, in which workers will respond by decreasing their work behaviors, for example, OCB (Bordia et al., 2008;Coyle-Shapiro, 2002;Harris & Kacmar, 2006;Kobasa & Puccetti, 1983).
On the other hand, the hypothesis of expectation-violation says that people react more strongly to the actions of the other party, when they violate their previous expectations about how the other should behave (Brockner, Tyler, & Cooper-Schneider, 1992;Restubog & Bordia, 2006).Thus, workers who reported high levels of perceived supervisor support had the expectation that they would be helped to face a significant negative work experience, characterized by a higher level of injustice in the process.However, The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior the supervisor from whom they expected support and protection failed to fulfill this role, and the supportive relationship becomes part of a context which aggravates the problematic experience of injustice in the procedures (Anderson & Schalk, 1998;Eisenberger & Justin, 2003;Morrison & Robinson, 1997).Thus, according to this framework, employees would not work as effectively and will decrease their extra-role behaviors.
Finally, an additional theoretical explanation for this hypothesized relationship comes from the theory of cognitive dissonance (Festinger, 1957), which states that it is highly probable that dissonant or incongruent cognitions produce states of tension and stress.When workers perceive the source of support and stress as the supervisor, they may experience conflictive cognitions.In this sense, the failure to comply with the fairness of the procedures, due to the discrepancy between what was promised and what was actually given to them, in the context of a supportive relationship, may be more evident, and therefore have a greater impact on corporate citizenship behavior, because it is unexpected (Baumeister, Bratslavsky, Finkenauer, & Vohs, 2001).This is exacerbated by the supervisor's trivialization of the worker's stressful experience.
On the other hand, in situations of low supervisor support, the theory mainly suggests that the relationship between procedural fairness and work behaviors is not significant (Bal et al., 2010;Bal, De Lange, Jansen, & Van der Velde, 2008), because the decrease in procedural fairness is another sign that states that the supervisor does not care about the worker as a member of the organization, does not value him/her and does not give him/her sufficient resources to cope with adverse conditions at work (Aselage & Einsenberger, 2003).Therefore, the slope representing workers with low social exchange with their supervisor is essentially flat through low and high levels of procedural fairness; thus the OCB would remain the same and would not be affected.
According to the arguments presented in this section, it is proposed that under conditions of high perceived supervisor support, as the procedural fairness decreases, workers reduce their OCB.Moreover, under conditions of low perceived supervisor support, as the procedural fairness decreases, workers do not alter their OCB.
Hypothesis 3 therefore is: H3: The relationship between Procedural fairness and OCB will be moderated by Perceived Supervisor Support, such that this relationship will be stronger for people with high levels of Perceived Supervisor Support.

Design
Given the nature of the research's objectives, a cross sectional design was carried out.

Participants
The participants were obtained through a convenience sample.The sample consisted of 707 Chilean workers that remained employed after organizational restructuring of a private enterprise pertaining to the metal mechanic industry.This sector was chosen because of its importance in the national economy and because they belong to a highly competitive industry that has undergone many changes in the short term.
During the past five years, the metal mechanic industry has undergone several restructuring processes, due to the organizational adjustments that have been made in order to face the 25.4% decrease in production demand.This lower economic activity has caused companies in the industry, mainly family businesses, to partially or totally close their operations, given their diminished ability to compete with foreign companies.This is attributed to the increase in administration, salary, electricity and production costs.
At the same time, companies in the metal mechanic industry have registered a change Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso in their business model, most of which have become import companies, motivated by the loss of competitiveness of the local industry and a 400% increase in the import of Chinese low cost metalworking products.
These circumstances have led the metal mechanic industry to adjust their workforce, leading to a loss of 5,300 jobs.According to official statements, similar processes are still expected to occur in the near future.
Of the 1010 employees, 707 returned completed questionnaires (response rate 70%).The sample was representative for employees in the metal mechanic sector.
Of the total sample, 75.4% were male and 24.6% were female, with ages between 18 and 75 years-old with an average age of 35.57years (DS=10.89).Out of all the participants, 29.0% correspond to operator level, 22.8% had administrative positions, 19.8% were executives, 20.0% were supervisors, and 2.6% were managers or assistant managers.Regarding the educational level of the participants, 3.2% of them had a maximum of elementary education, 30.5% had completed high school, 35.1% mentioned having completed some type of professional or technical training, 26.9% were university graduates, and 4.5% had postgraduate studies.The mean organizational tenure was 7.97 years (SD=8.20)whose range varied from a minimum of one month to a maximum of 45 years in the company.Average work experience was 14.31 years (SD=10.35),with a minimum value of 6 months and a maximum of 48 years.

Procedures
The research team contacted the organization's human resources area to request voluntary participation of workers in the study.Workers, who agreed to participate in the investigation, signed a letter of informed consent.
The instrument was applied in the meeting room of each organization at the beginning or end of working hours to minimize any interference with their daily workload.
The application of the questionnaires was conducted by a team of research assistants, properly trained for the procedure.Research assistants introduced the project to every group, explained the purpose and value of the study, and highlighted participants' rights to anonymity, confidentiality, and voluntary participation.In all surveys, a cover letter accompanying the questionnaire indicated that the survey was being conducted solely for scientific purposes.
Also, participants were recruited by sending an email with an invitation to participate in the study.This email provided a link to access the online questionnaire.This email also described the main goal of the study, the confidentiality and voluntary conditions of it.
Adopting the proposals of Brislin (1970), all the measures used in the study were translated and back-translated between English and Spanish by two of the authors, working independently.In cases of translation discrepancies, the two translators discussed the implications of these differences and defined together a final version for the ambiguous translations to reduce the risk of losing meaning.A pilot test of the Chilean version of the questionnaires using 30 employees was conducted to assess their usability.

Measures
The research provided a self-report questionnaire to collect data comprising the following self-reporting scales: Job insecurity was measured by using the scale developed by De Witte (2000) to measure insecurities about the work continuity.This scale is composed of a six-item Likert scale with five answers each, (1 = Completely disagree, 5 = Completely agree) where higher values mean a higher level of insecurity.Survey participants indicated their degree of agreement with items such as "I worry about keeping my job."Reliability of this scale (Cronbach Alpha) was 0.79.
Procedural Fairness was measured by using a scale by Niehoff and Morrman (1993).The measurement was composed of a six-item Likert The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior scale with five alternatives for each answer (1= Completely disagree, 5 = Completely agree).Higher values meant higher levels of procedural fairness.Participants indicated their degree of agreement with items such as "Job decisions are made by the general manager in an unbiased manner".Cronbach's Alpha was 0.86.
Pe rc e i ve d Su p e r v i s o r Su p p o r t w a s measured using a scale developed by Eisenberger, Huntington, Hutchison and Sowa (1986).The instrument was composed of an eight-item Likert scale with five alternatives for each question.Higher values meant a higher level of perceived support.Participants gave their degrees of agreement with statements such as "My supervisor would forgive an honest mistake on my part".Cronbach's Alpha was 0.89.
Organizational Citizenship Behavior (OCB) was evaluated by using the scale developed by Williams and Anderson (1991).The instrument was composed of a seven-item Likert scale with five choices for each question (1 = Completely disagree, 5 = Completely agree).Higher numbers meant higher levels of OCB by workers.Participants indicated their degrees of agreement with statements such as "Conserves and protects organizational property."Cronbach's Alpha was 0.73.
Controls.Additionally, two positive and negative effect control variables were incorporated (Watson, Clark, & Tellegen, 1988) due to the possible effect of the affective states of the workers on their answers.The instrument was composed of a seven-item Likert scale with five answers for each one (1 = Completely disagree, 5 = Completely agree).The scale was presented with the following statement "Indicate to what extent you have felt these emotions" and the participants indicated to what degree they had felt items such as "enthusiastic" or "distressed."Cronbach's Alpha was 0.79 for the positive affect and 0.65 for the negative affect.Along with the formerly mentioned variables, the effects of variables such as age, work experience were also examined.However, these variables were not included in the analysis, given that they did not correlate to the independent variables of the study.

Data analysis
Data analysis was approached in three phases.In the first phase, the factorial validity of the proposed model was evaluated through confirmatory factorial analysis and reliability analysis (Brown 2006;Byrne, 2012) using the program MPLUS 6 (Muthén & Muthén, 2010).These analyses were carried out to estimate the robustness of the proposed scales, supported by the procedure described by Bolger, Davis and Rafaeli (2003).In the second phase, mediating effects were tested (Baron & Kenny, 1986).Subsequently, the SOBEL statistical test was performed to confirm the mediation effect (Preacher & Hayes, 2004).In this test, the reduction between the variables involved in the mediation was evaluated to test its statistical significance.Finally, in the third phase, the moderation hypothesis was evaluated, through moderated multiple regression (MMR), as described by Aiken and West (1991).Study variables were centered before calculating the interaction with the purpose of reducing multicollinearity (Aiken & West, 1991).Finally, the sign and significance were analyzed from the interaction terms to identify direction patterns of the variables, and the slopes were represented from the relationship between the independent and dependent variables of each one of the moderators that were significant, shown by values ± 1SD from the respective mean.These analyses were carried out by using the macro MODPROBE developed by Hayes and Matthes (2009) for the estimation of interactions in MMR.

Results
Confirmatory factor analysis (CFA) was carried out with the objective of evaluating the validity of the construct of the study variables, given that the scales used were Spanish translations of validated constructs in English.Specifically, a model was tested with five factors with all of the scale items used: job insecurity, Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso organizational fairness, perceived supervisor support and organizational citizenship behavior.These analyses reached a poor goodness of fit from the model (χ2 = 1299.752,df = 447, p = 0.0000; RMSEA = 0.052; SRMR = 0.058; CFI = 0.89; TLI = 0.88).
The modification indexes showed high levels of residual covariance errors (M.I. of more than 10; Byrne, 2012) from items theoretically expected as pertinent to the same scales.These residual covariance errors could be caused by certain items that presented redundant information.This could also indicate bias in the answers, where the answer to one item was biased by the answer to another.Research has claimed that these cases of deletion of redundant items are justified in order to achieve a well-fitting model (Byrne, 1991).In this manner, based on the information given by the modification indexes, a new model was tested.In this model, items with low factor loading were removed in the latent factors, or in those which presented cross-loading.The eliminated items were principally negative items.Specifically, two items on the job insecurity scale were deleted.Additionally, correlations of residual errors were controlled from the scale items that presented inconveniences.These modifications resulted in an improvement of the model fit (χ2 = 730.176,df = 359, p = 0.0000; RMSEA = 0.038; SRMR = 0.046; CFI = 0.95; TLI = 0.94).Results showed empirical support for the validation of the construct of the six-factor model (job insecurity, procedural fairness, perceived supervisor support, organizational citizenship behavior, positive affect and negative affect).
Table 1 shows the means, standard deviations, inter-correlations and Cronbach's Alpha for all study variables.Zero-order correlations were in the expected directions.Correlation matrix shows that the relationship between job insecurity is significant and negative for the variables procedural fairness, perceived supervisor support, and organizational citizenship behavior.The mediating variable procedural fairness is positively and significantly related to organizational citizenship behavior.Also, the moderating variable showed significant relationships with organizational citizenship behavior, job insecurity and organizational fairness.
The results for the H1 are shown in Table 1.Job insecurity is negatively related to organizational citizenship behavior (r=-0.10,p ≤ 0.01).Therefore, H1 is supported.
H2 stated that procedural fairness would have a mediating effect between job security and organizational citizenship behavior.Consistent with this hypothesis, the significant and negative relationship between job insecurity and organizational citizenship behavior disappears when the mediator is introduced into the regression equation (Step 3, B = -0.031,p = 0.238), meaning a complete mediation.With the objective of evaluating the significance of the mediation, a Sobel test was carried out to evaluate indirect effects.Results show that the reduction in the relationship between the involved variables in the mediation was statistically significant (Z = -2.61,p < 0.01).Thus, H2 was completely supported (see Table II).
H3 suggested that perceived supervisor support would moderate the relationship between procedural fairness and organizational citizenship behavior.Results show that the interaction term was significantly related to organizational citizenship behavior (Step 3, B = -0.08,p ≤ 0.01).From these results, it can be observed that the relationship between procedural fairness and organizational citizenship behavior is strong and significant at high levels (+1SD) of perceived supervisor support.This relationship was weak and non-significant at low levels (-1SD) of perceived supervisor support (see Figure 1).These results support H3 (see Table 2).
The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior

Discussion
Our study had three main objectives: First, we evaluated the effect that the perception of job insecurity had on organizational citizenship behavior.As shown in previous research works, and supporting our first hypothesis, we found that there is a negative relationship between job insecurity and OCB.These results are consistent with the theory of social exchange (Blau, 1964;Gouldner, 1960), which delivers an explanation of why job insecurity reduces OCB.Specifically, when an organization fulfills its obligations to workers by providing job security, they may Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso respond favorably increasing their contributions to the organization, such as OCB.However, with increasing job insecurity, workers perceive that their organizations are providing less benefit and, therefore, no longer feel so compelled to reciprocate positively, resulting in lower levels of OCB (Turnley & Feldman, 1999).
The second objective was to determine the mediator effect that procedural fairness had on the relationship between job insecurity and OCB.Results from hypothesis 2 were also confirmed.This means that we found support for the proposed underlying mechanisms of procedural fairness that enables a deeper understanding of the conditions that translate perception of job insecurity into lower organizational citizenship behavior.From a social exchange theory point of view, work conditions that employees consider as unfair will be reciprocated under the rules of social and ethical norms (Bernhard-Oettel et al., 2011;Schumacher et al., in press).Employees will compensate for the lack of fairness that they feel is exhibited by the organization by lowering their OCB.Thus, the perception of fairness and procedural fairness is strongly related to the stress experience in an organizational setting.
Thirdly, we aimed to evaluate the moderator effect that perceived supervisor support has on the relationship between procedural fairness and organizational citizenship behavior.The results for hypothesis 3 suggest that perceived supervisor support is a significant moderator on the relationship between procedural fairness and OCB.This relationship states that for a high perception of supervisor support, as the procedural fairness decreases, workers reduce their OCB.On the other hand, for a low perception of supervisor support, workers do not alter their OCB.
The results are consistent with the betrayal perspective (Elangovan & Shapiro, 1998;Zagenczyk et al., 2009), which states that workers who perceive a high supervisor support show a sharp fall in performance, as the job expectations about the organization decreases.According to this, the failure of the supervisor to prevent instances of decreased justice is interpreted as an act of betrayal by workers.Moreover, in situations of low supervisor support, the theory (Bal, De Lange, Jansen, & Van der Velde, 2008;Bal, et al., 2010) is consistent with our findings in terms that the relationship between procedural fairness and OCB is not significant, since the reduction of the procedural fairness is another sign that indicates the supervisor does not care about the worker as a member of the organization, does not value him and does not give him sufficient resources to cope with adverse conditions at work (Aselage & Einsenberger, 2003).

Theoretical implications
Our research is founded on theoretic and empirical evidence, and contributes to new knowledge about the effects that perceived job insecurity has on organization citizenship behavior in the context of social exchange theories (Blau, 1964).
In this sense, it deepens the understanding of the mediating and moderating mechanisms that underline this relationship.Hence, the results of this study support a model where procedural fairness mediates the relationship between perceived job insecurity and organizational citizenship behavior, given that the fairness of organizational procedural process has a negative relationship with OCB.Our study gives new empirical evidence to support this argument.
Additionally, this study widens the existing results about the moderating mechanisms of supervisor support in the relationship between procedural fairness and OCB.This research contributed to evaluate how the perception of supervisor support can moderate the size and severity of the procedural fairness, which is generated by the perception of insecurity among employees.By means of this, we contribute to the existing research on social exchanges by focusing on contextual factors as moderators in these relationships.This will allow a greater theoretical knowledge of the psychological dimensions that motivate the organizational performance.
The effect of procedural fairness and supervisor support in the relationship between job insecurity and organizational citizenship behavior

Practical implications
Our study has important conclusions from a management practice perspective.When faced with insecurity about their position, employees can manifest perceptions of unfairness related to individual consequences.In turn, OCB is negatively affected.For example, a decrease of OCB can generate losses in efficiency and effectiveness in organizational performance (Organ, 1997).Also, given that this behavior depends on the decisions of the employee, it does not constitute a job requirement and no formal sanctions exist for not committing to or participating in these practices (Podsakoff, Mackenzie, Lee, & Podsakoff, 2003).A decrease of OCB has a strong impact on organizational capital.This is an important reason that explains the increase of failure in high-impact organizational change such as restructuring or merging.These processes involve a large number of severances, as well as employees that remain in the company and have not been formally backed by the organization (Kammeyer-Mueller, Liao, & Arvey, 2001).
In the literature regarding this topic, the supported notion is that it is crucial that the organization is clear about what workers can actually expect from it (Coyle-Shapiro & Conway, 2005;Griffin et al., 2001).This means that clear communication is a means of deflecting negative perceptions of organizational fairness, therefore reducing the negative consequences when workers sense insecurity about their employment.This is by no means suggesting that negative outcomes can be avoided by merely communicating effectively.In this case, a company that cannot guarantee job security should clearly communicate that it offers flexibility and employability (Conway & Briner, 2005).
Also, supervisors have an important position in the relationship between procedural fairness and OCB.Our study provides support for training and development of these practices.However, at the same time that organizations promote a high quality worker-supervisor relationship, this can also be expected to create higher expectations among workers (Bordia et al., 2008;Elangovan & Shapiro, 1998;Zagenczyk et al., 2009).To minimize these negative results, management should investigate the commitments it has made with employees, making an effort to comply.

Limitations
It is important to note several limitations of our study, which could influence the interpretation of the results.Firstly, from the cross-sectional design, no causal conclusions can be drawn (Singleton & Straits, 2005).Earlier studies do conclude however that job insecurity and fairness have an effect on individual behavior (Bernhard-Oettel et al., 2011;Ouyang et al., 2015;Schumacher et al., in press;Zhang et al., 2014).This should be tested in longterm research.
Secondly, our sample fundamentally included employees that have remained in the company after processes of organizational restructuring, from the metal mechanic industry.Future studies should evaluate if our results can be generalized to other populations of workers.A fact that is noteworthy is that our sample included different operative levels, including the administration, professionals and supervisors, which represent an advance for the research of these variables.Typically, samples used in other studies are concentrated on recent MBA graduates from prestigious and high-paying positions, with a work experience usually between six months and two years (Kiewitz et al., 2009).Contrarily, the sample analyzed in our study had an average of 8 years of work experience.
Another limitation is that all of the measures used in this study come from selfreporting surveys.Although these surveys are an appropriate measure that reflects individual perceptions about job insecurity, procedural fairness and perceived supervisor support, some concerns have been expressed about the use of these instruments.Studies based on self-reporting surveys are potentially subject to common method bias, which can artificially inflate the association Sergio Andrés López Bohle / Antonio Ramón Mladinic Alonso between the measured constructs (Podsakoff et al., 2003).Still under debate is mono-method bias (CMV), and empirical evidence has been given which raises questions about the importance of CMV problems in research within the field of organizational behavior (Spector, 2006).To decrease the potential influence of common variance among the results, recommendations by Hinkin (1995) and Podsakoff et al. (2003) should be followed.
Finally, future discussions should focus on researching the performance of our model in different work environments; with the objective of finding out if workers are more influenced by personal relationships (employee-supervisor) or by employee-organizational relationships (employee-organization).In a collectivist context, individuals look forward to building confidence through the development of more fluid personal relationships (Hofstede, 2001;Rodríguez, 2001).This is one reason why the role of the supervisor can be more significant for employees than their formal relationship with the organization.It is also certain that within some contexts, there are enterprises with management systems that privilege independent employee profiles, creating a low power distance between a supervisor and their direct subordinates (Rodríguez, 2001).For this reason, the supervisor once more becomes a central part in moderating the relationship between procedural fairness and organizational citizenship behavior.

Conclusion
This paper showed the mediator effect that procedural fairness had on the relationship between job insecurity and OCB.This means that we found support to the proposed underlying mechanisms of procedural fairness that enables a deeper understanding of the conditions that translate perception of job insecurity into lower organizational citizenship behavior.Finally, the perceived supervisor support has an effect on the relationship between procedural fairness and organizational citizenship behavior.This relationship indicates that for a high perception of supervisor support, as the procedural fairness diminishes, workers reduce their OCB.On the other hand, under conditions of low perceived supervisor support, as the procedural fairness decreases, workers do not alter their OCB.

Table 1 Means, standard deviations and correlations of the variables in the study
Interaction effect of Perceived Supervisor Support on the relationship between Procedural Fairness and Organizational Citizenship Behavior (OCB).