The Trust as a Necessary Condition but Not Sufficient for the Psychological Contract

The perception of organizational injustice may trigger off a questioning of the trust established between both partners (employer and employee), which may cause a breach of the psychological contract and encourage the emergence of problems within the organization (doubt, mistrust, antisocial behavior, etc.). We support the idea that social influence can affect the psychological contract, itself arising from a trust established yet between partners. In order to conduct our empirical research, we selected a financial institution, named SND bank, based in the capital Tunis. Data were collected from 250 questionnaires distributed to investigate the causal relationships between variables, using structural equation models. Results showed that only procedural justice affects positively the employees’ subscription to behaviors of confidence and the respect of the psychological contract. Regarding to interactional justices, it affects the employees’ subscription to behaviors of confidence. As to Social influence, it affects the psychological contract.


Introduction
The employment contract can be considered as a contract of adhesion.At the beginning of the relationship, the individuals committed with a little regard for multiple powers of the corporate manager.In fact, organizations tend to use all the means at their disposal to ensure the individual investment.It is in this sense that the psychological contract has come to counteract the old contract and combat unethical practices.However, the legal and psychological contracts are considered necessary to preserve the employment relationship of any organizational problems, but still not sufficient.Indeed, the perception of an organizational injustice may lead to challenge to the trust established between both partners, and thus the psychological contract.
Consequently, this can encourage therefore the emergence of problems within the organization (doubt, suspicion, antisocial behavior, etc.).We support the idea that social influence may be the cause of this breach and then it could influence the psychological contract, arising from a trust between employer and employee.In addition, starting from the premise that individual is always seeking to protect himself from social reprehension, and preferring to be accepted by the group to satisfy his gregarious and need to belong (Maslow, 1943), he will therefore, accept social influence and became « an instrument of the will of others » (Mamlouk & Gharbi, 2007).
Such contract employment is supposed to guarantee a win-win relationship.However, the employee may feel in some cases a sense of injustice leading to a potential failure of the psychological contract, weakened by the distrust and the social influence.Our research integrates simultaneously the main three dimensions which are (organizational justice, trust and psychological contract) and takes into the account the social influence.Such researches, we have not seen in previous works.
Therefore, the research question is: may the breach of the psychological contract be so determined by the social influence?In what follows, we present the concepts of our model.We also point out the assumptions of our research and their justifications, the methodology adopted and the empirical study.

Conceptual Framework
The Organizational Justice and the Importance of Social Relationships in Developing Justice Perceptions "Is that fair?"A question of a paramount importance that an individual will ask at least one time during his social or organizational life.This issue is due to some confusion that may trouble and lead individual to question some corporate relationships, adopted behaviors, granted trust, deployed professional awareness, etc.
Over the years, forms of organizational justice arose to depict the relationship between individual and his organization.We can mention three types (Deutsch, 1985;Greenberg, 1990, Tyler andLind, 1992): The distributive justice which refers to the basic principles of fairness: « Do I get what I deserve?» (Adams, 1965).The procedural justice is concerned with the fairness of procedures used to implement decisions and appropriate policies: « Are that rules and regulations treating me fairly?» (Thibaut and Walker, 1975) and finally, the interactional justice that focuses on the quality of treatment received from leaders and the importance that the decision-making procedures are properly represented: « Have they treated me fairly » (Bies and Moaga, 1986;Tyler and Bies, 1990).To sum up, we can say that organizational justice in all its forms, plays an important role in organizations.It affects, undoubtly, organizational behavior.Thus, it can leads individuals to more easily accept to trust their superiors and subordinate their short-term personal interests in favor of the interests of the organization.

Trust as the Basis for Organizational Relationships
The conflict inherent to the human nature and the organization, arrises whenever the interests diverge.Never, trust has been sought and provided in a desired configuration, governed by skepticism and disloyalty (the limits of agency theory, Jensen and Meckling, 1976).
For this reason, trust is supposed to be an essential solution to lessen the feelings of injustice felt by corporate employees, which may adversely affect the viability of the psychological contract.Trust traces the symbiosis salary and emphasizes on reciprocity.In general, this trust is awarded at first by the top manager to the subordinate through a grant of a scrap of power or of a voluntary delegation of responsibility, of an initiative-taking or of a decision making.Secondly, the subordinate has to prove that he deserves the trust of his superiors by illustrating an unequivocal loyalty, an emotional and a physical involvement, a true and an excessive devotion to work (Pfeffer , 1994(Pfeffer , , 1998)).

The Psychological Contract to Better Understand and Manage the Employment Relationship
The employment relationship requires the appeal of a framework that defines the rights and obligations of each party in the relationship.This framework has long been embodied by a written and formalized legal employment contract.Governed by a set of various laws and conventions, it tends to cover the more material aspects of the employment relationship, concerning the manner in which the work is organized, governed, evaluated and rewarded (Kallenberg & Reve, 1993).The psychological contract is a concept that allows to relieve ethical failures of formalized approaches of contracts.Today, we are becoming to speak more and more about this type of contract containing a series of unwritten mutual expectations favoring « an implicit understanding » between the stakeholders (Argyris, 1960: 97).More sensitive to the rule of the linkage between staff and his organization, Levinson and al. (1962) and Schein (1965) highlighted the phrase "mutual expectations" to emphasize on the reciprocal relationship between the employer and employee in a swap contract.

The Social Influence as an Omnipresent Organizational Reality
The herd instinct compels the individual to be submitted to the corporate culture.This influence can be so deep that the individual becomes an instrument of the dominant thought.Social influence, driven by shared values and subjective representations, directed individuals and collective behaviors.It can, thus, weaken the psychological contract linking the person to his organization.«Other people play in the life of the individuals the role of a model, an object, a partner or an opponent, and the individual psychology appears as a social psychology » (Freud, quoted by Maisonneuve, -1950 -).The notion of influence refers to the idea of a mysterious power, considered more effective than invisible.Therefore, influence refers to act on others.Insofar as it is almost obvious to note that many things can influence us (such as a situation, a group, one person, a context, a policy, a tone, a set of terms, a certain way to behave in a way, a charisma etc..).Moreover, Tarde (1973) describes the desire of individuals to undergo intentionally to social influence in order to avoid social disapproval.In this sense, social influence is likely to condition the perceptions of employees and the organizational justice, to govern the establishment and sustainability of the psychological contract and cause changes in behavior.Links that we wish to test are shown in Figure (1) below.

Fig.1: Conceptual Model The Legitimacy of the Conceptual Model
Certainly, distributive justice focuses on the allocation of resources.However, it refers to the perception of fairness or unfairness of remunerations or rewards which may result from it, a sense of justice or injustice felt by the employee in respect of its organization or supervisor.Furthermore, procedural justice is oriented more towards the account and the justness of the rules of processes (Folger, 1987, Lerner 1977, Leventhal et al., 1980).Interactional justice focuses on the employee in his sociological dimension and justice in terms of identification, recognition and interpersonal exchange.Authors of both meta-analysis, Charash Cohen and Spector (2001), Colquitt and coll. (2001) showed a significant relationship between trust and distributive, procedural and interactional justices, and stipulated that trust in the organization can be predicted by the three types of justice.Thus, our first three assumptions are as follow: H.1.Distributive justice affects positively the employees subscription to the behaviors of trust.

H.3. Interactional justice affects positively the employees subscription to the behaviors of trust.
Organizational justice and psychological contract have joint shares inherent to the social exchange theory of Blau (1964).We rely on the research works of Konovsky & Pugh (1994) -who found that organizational justice (distributive, procedural and interactional justices) is positively related to the organizational citizenship behaviors (Materson, Lewis, Goldman & Taylor, 2000 ) and to the organizational commitment (McFarlin & Sweeney, 1992), where a reciprocal relationship [within the meaning of Cropanzano, Rupp, Mohler and Schmike (2001)] is of paramount importance to sustain the viability of the psychological contract and to assess the relationship between procedural justice and the adoption of organizational citizenship behavior.Therefore, in this case we speak about the establishment of a psychological contract.
Moreover, according to the theory of cognitive references of Folger (1986), in a situation of distributive injustice, employees who have fair procedures as appropriate and conclusive clarification with regard to the reward, perceived as unjust distributive injustice; will hide it and will not give it excessive proportions.Indeed, we believe that the lack of procedural justice leads to a reconsideration of the psychological contract.This encourages employees to increase the unjust distribution, even if it is a priori a trivial injustice almost negligible.In this sense we state our fourth hypothesis: H.4. Procedural justice affects positively the respect of the psychological contract.Moisson and Peretti (2006) suggest that the relationship between trust and psychological contract is « an essential basis for the development of this type of contract particulary when the employee feels that his contract has been violated, the perception of trust vis-à-vis the other party is shaken » (2006: 7).Moreover, according to the empirical study of Robinson (1996), firms who are installing and maintaining relationships of trust with their employees, can evade negative surprises related to the breaches of psychological contracts.In this regard, he has proved empirically that trust influences perceptions and interpretations of a breach of contract among employees.Thus, we present our fifth hypothesis:

H.5. Trust affects positively the respect of psychological contract.
Certainly, social influence is a pervasive organizational reality that can change through subjective norms, behaviors adopted by all employees operating in a group setting.Moreover, the archives of history are full of experiments conducted by researchers in companies, whose mission is to clarify the concept of social influence and its impact on individual perceptions.Specifically, Taylor and Todd (1995) have discerned a significant relationship between subjective norms and intentions of behavior.For Bogozzi (2000), the effect of subjective norms on behavioral intentions is significant for the case of social behaviors expressing themselves in the presence of others.From another side, social influence has not been searched as a variable that can affect the psychological contract.Therefore, our sixth hypothesis takes the following structure:

H.6.
Social influence affects the psychological contract.

Methodology
It is clear that the role of a questionnaire is to reproduce as possible the most reliable measure of the phenomenon (Igalens & Roussel, 1998).Following the collection of data by the research questionnaire, we used a double analysis.The first one is exploratory and leads to assess the quality of the measurement scales in order to purify the research questionnaire.For this, we used a Principal Component Analysis (PCA) with an analysis of internal consistency.
The second type, is confirmatory and implies the validation of the selected dimensions resulted from the exploratory analysis.In addition, this analysis will undertake to investigate the causal relationships between variables by the Structural Equations Models (SEM).The sample devoted to the exploratory study consists of 150 employees belonging to a financial institution which is in our case the Tunisian « SND 1 » bank, all grade 5 Journal of Human Resources Management Research and gender combined.The main sample characteristics are as follow:  Price & Mueller (1986) allow to measure distributive justice within the organization.Among the six items of the measurement scale, one was eliminated because it has a very low quality of representation and a factorial contribution.

• Procedural Justice
Factor analysis shows that the six items of procedural justice are all connected to the same factor.Indeed, the unidimensionality of the variable is confirmed with the identification of a single component representing 47.758% of the total variance explained.The indicator KMO for the data corresponding to this construct is 0.781, which authenticates the appropriateness of factor analysis on these data.The reliability analysis reports a Cronbach alpha of 0.778, a level quite acceptable and very close to the result obtained by Niehoff & Moorman (1993), which reached a Cronbach's alpha of 0.92.

• Interactional Justice
Factor analysis shows that the nine items comprising the scale of Niehoff & Moorman (1993) on interactional justice are all connected to the same factor.The unidimensionality of the variable is confirmed with the identification of a single component representing more than 61.014% of the total variance explained.After checking the KMO (0.904), we conducted a reliability analysis that reveals a Cronbach's alpha of 0.920 and a respectable level that precedes the one obtained by the authors, namely a value of 0.90.

• Trust towards the Superior
Factor analysis shows that the eight items adapted from Tyler and Degoey (1996) measuring the trust in the supervisor are all connected to the same factor.Indeed, the unidimensionality of trust felt towards the superior is confirmed with the identification of a single component representing more than 69.132% of the explained total variance.The KMO, disclosing a coefficient of 0.876, shows that data measuring this construct are well suited to factor analysis.The reliability analysis reports a Cronbach's alpha of 0.936, a satisfactory threshold very close to the result obtained by the authors mentioned above, which shows a value of 0.93.

• The Social Influence
It must be remembered that we adapted the scale for measuring social influence in our research.To do this, Ajzen (1991) identified only three items.In addition, Hom and Hulin (1981) have adopted a version of five items.As recommended by Ajzen (1991), we have taken the liberty to retain six items to measure social influence within the firm.Factor analysis shows that the six items are all related to the same factor.Moreover, the unidimensionality of social influence is confirmed with the identification of a single component representing more than 53.111% of the explained total variance.The indicator shows a KMO coefficient of 0.832, which justifies the factor analysis of this construct.The reliability analysis reports a Cronbach's alpha of 0.818, a satisfactory threshold very close to the result obtained by Hom and Hulin (1981) who gave off a Cronbach's alpha of 0.93.

• The Psychological Contract
The measure of sampling precision discloses a KMO of 0.918, a factor that confirms the data for measuring this construct are well suited to factor analysis.
As for the reliability analysis, it released a Cronbach's alpha of 0.962, well ahead of the results found by Kickul (2001a), namely a value of 0.70.

The Application of Structural Equation Models
The validation phase of the measurement scales was performed on a sample of 250 employees belonging to the same bank.We present below a table that shows the main characteristics of our sample.• The Distibutive Justice Regarding the measurement model of distributive justice, we note that it presents a good fit to the data.Indeed, the chisquare is low "1.04" and x2/ddl = 0.347 < 5, the RMR = 0.0063 and the RMSEA = 0.00 based on the residuals are close to 0. GFI = 0.998, AGFI = 0.992 and CFI = 1, are close to 1.The CAIC model is much lower than the CAIC of the saturated model, indicating good parsimony.The Rho of Jöreskog is equal to 0.854.The scale therefore has a good reliability.

• The Procedural Justice
According to the measurement model of procedural justice, we notice that it presents a good fit to the data.x2/ddl = 2.898 < 5, RMR = 0.0325 and RMSEA = 0.087, are acceptable.GFI = 0.977, AGFI = 0.920 and CFI = 0980, are close to 1.The CAIC model is lower than the CAIC of the saturated model, indicating good parsimony of the model.The Rho of Jöreskog is equal to 0.827.Therefore, the scale has a good reliability.

• The Interactional Justice
While the vertical interactional justice model displays a RMSEA = 0.08 deemed not acceptable, as long as it slightly exceeds the limit set by Browne and Cudeck (1993).
Nevertheless, the goodness of fit of this model is quite good.Indeed, the GFI, NFI, NNFI and CFI are greater than "0.9".AFM = 0.839, meanwhile, is close to 0.9, viewed as acceptable by Didellon & Valette-Florence (1996).x2/ddl = 4.326 is satisfactory.In fact, according to Jöreskog and Sörbom (1989), the ratio is considered satisfactory when it is less than 5.The CAIC and ECVI models tested are respectively lower than saturated models, indicating a good parsimony.The Rho of Jöreskog is equal to 0.933.Therefore, the scale has a good reliability.

• Trust towards the Superior
The model of trust showed RMSEA = 0.054 considered as satisfactory index (Browne and Cudeck, 1993;Didellon & Valette-Florence, 1996).The GFI, AGFI, NFI, NNFI and CFI are all above "0.9"and therefore very close to 1, (Didellon & Valette-Florence, 1996).The square root of the residuals average adjusted to the square is 0.016, close to 0. Indeed, furthermore this value is low, the more the the model fit is of quality.x2/ddl = 1.71 is very satisfactory.
The CAIC and ECVI indicate a good parsimony.The Rho of Jöreskog is equal to 0.963.Therefore, the scale has a good reliability.

• The Social Influence
Regarding the measurement model of social influence, we remark that it reveals a very good fit to the data.Indeed, the chisquare is low "4.38" and x2/ddl = 1.095 < 5; RMR = 0.0153.RMSEA = 0.02.GFI, AGFI, NFI, NNFI and CFI are close to 1.The AIC and CAIC of the models are well below the AIC and CAIC saturated models, indicating a good parsimony of the model.The Rho of Jöreskog is equal to 0.912, the scale has a good reliability.The model has a good fit.

The Psychological Contract
The model of psychological contract discloses a RMSEA > 0,10 deemed not acceptable.Nevertheless, the fit indexes are very good for the majority.Indeed, the incremental indexes are above "0.9".GFI and AGFI, meanwhile, remain poor but are close to 0.9.They are respectively the order of "0870" and "0770", deemed acceptable by Didellon & Valette-Florence (1996) as long as these indexes are close to "0.9".x2/ddl = 4.73 is acceptable.In this regard, according to Jöreskog and Sörbom (1989), the ratio is considered satisfactory when it is less than 5.The CAIC model tested is lower than the CAIC of the saturated model, indicating good parsimony of the model.The values of the Rho of Jöreskog are respectively 0.946 for the first dimension and 0.980 for the second.So, both subscales have good reliability.

Model and Discussion of Results
It is noteworthy that several dimensions postulated in the literature are not emerged in response to quantitative analysis, so that new dimensions have emerged.The components of each concept of our model have significantly changed the outcome of exploratory and confirmatory factor analysis.Indeed, throughout the construction phase and purification of measuring devices, we were led to remove some items that had abnormalities.In what follows, we will outline the overall quality of fit of the model by sketching the absolute, incremental and parsimony indexes, and then in a second step, we will present the final model with regression coefficients; and finally, in a last step, we will try to discuss the obtained results.

Table 3: The Overall Quality of Model Fit
The final model provides a RMSEA = 0.037 viewed as satisfactory by Didellon & Valette-Florence (1996).In addition, Browne and Cudeck (1993) certified that a model with a RMSEA ≤ 0.08 is considered as a good model.The good fit of the model comes with a good quality adjustment, as long as GFI = 0.888, AFM = 0.865, NFI = 0.877, NNFI = 0.960" and IFC = 0.965, they are all greater than "0.8" and very close to 1, (Didellon & Valette-Florence, 1996).The square root of the residuals average adjusted to square is "0.085" near zero.Indeed, furthermore the value of the RMR is close to 0, the better the model fit improves.x2/ddl = 1.346 is a suitable index.As recommended by Jöreskog and Sörbom (1989), the ratio is considered satisfactory when its value is less than 5.The AIC = 634.690,CAIC = 973.799and ECVI = 2.549 of the tested models, are respectively lower than AIC = 870.000,CAIC = 2836.835and ECVI 3.494 of the saturated models, indicating a good parsimony.In summary, we can conclude that the goodness of fit of our model is satisfactory.

Distributive Justice, a Vulnerable Criteria to Build Trust Hypothesis 1: Distributive justice affects positively the employees subscription to the behaviors of trust." (hypothesis reversed).
It should be remembered that the distributive justice refers to a direct calculation of allowances of goods received by an employee (Folger & Cropanzano, 1998).It is noteworthy that we cannot talk about distributive justice without mentioning a priori the theory of relative deprivation cited by Stouffer and al (1949).This theory states that the reaction of individuals to financial stimuli is highly dependent on subjective comparisons made with other individuals or groups considered analogous.The theory of the relative deprivation shows the impact of comparisons on the development of judgments about the assets obtained, and consequently the behaviors and attitudes (psychological contract, trust) to adopt following the comparisons.
We support the idea that the distributive justice, even if it exists within the firm, does not necessarily encourage employees to establish and maintain a degree of trust in management, to establish and maintain a psychological contract with the organization, to behave commitment vis-àvis their work related to organizational citizenship and to develop and maintain a professional awareness.Trusting is not easy, furthermore less whether it is inherent in a relationship purely pecuniary.Probably, it will not last.Although they found a significant relationship between trust and distributive, procedural and

Indexes
Values Chi-deux/dl 1.346 RMSEA 0.037 interactional justices, Cohen and Spector (2001), Colquitt and al. (2001) showed that the trust placed in the superior is best predicted by procedural, interactional and distributive justices.relying on other basis, trust is no longer synonymous, in a configuration of procedural and interactional justices to cognition, motivation and affection than it is in a configuration of distributive justice that is purely material.Trust is not as interest, it is also « emotive » (Servet, 1994, p. 39).In this sense, the first assumption is invalidated.

The Procedural and Interactional Justice, Creative Elements of Trust
Hypothesis 2: Procedural justice affects positively the employees subscription to the behaviors of trust.(Hypothesis confirmed).

Hypothesis 3: The interactional justice affects positively the employees subscription to the behaviors of trust. (Hypothesis confirmed).
Our quantitative results are an empirical confirmation of both meta-analysis done by Cohen-Charash and Spector (2001) and Colquitt and al. (2001).According to the literature, procedural and interactional justices affect trust.Hence, we can distinguish between distributive justice and procedural justice, and thereby highlight the impact of each on the employees subscription to the behaviors of trust.Indeed, in a distributive injustice, the employee can see the potential injustice by seeing the results, that is to say, when having collected the reward.Following his efforts to achieve the desired objective, the employee will find that the distribution is not consistent with its expectations.In this case, the belief that he was injured in terms of distribution, the employee may choose not to subscribe to the behavior of trust.By cons, in a procedural injustice, the employee is in possession of flexibility allowing him to transmit to his superiors what is lacking in terms of procedural injustice.Thus, he can anticipate the result and remedy the procedural deficiencies to restore the potential equilibrium between the efforts he is deploying and the result.
We argue this nuance with the notion of « process control » set by Thibaut & Walker (1975).Furthermore, interactional justice is concerned mainly with aspects of the process of developing communication between the superior and the employee.E.g politeness, honesty, empathy, respect, etc.The employee with the perception of having been wronged, rejected his disapproval of the individual: a source of injustice rather than the procedure itself.While who is liable for decision-making can be exemplary in his application of all formal procedures, blameless as to the accuracy of his decisions, and absolutely impartial vis-à-vis his employees during the stage of the allocation of rewards.However, it is nevertheless true that we are never safe from a perception of the lack of consideration, respect or attention to employees.Sometimes, erroneously, a tiny collection of failure in the recognition of employees can be transformed into a real attack on their person.The lack of responsiveness to current and future needs of employees can be the cause of the appearance of pathological relationships within the organization and a questioning of trust.Indeed, this perception of negligence may give rise to malfunctions that may have an adverse impact on the work flows, such as: the filing of complaints (Bies, Shapiro & Cummings, 1988), present-absent behavior (Thevenet, 1992), robbery (Greenberg, 1993), withdrawal behaviors undertaken by employees (Folger & Cropanzano, 1998), deviations of production and property (Robinson and Bennett, 1995), Organizational Misbehavior (Vardi and Wiener, 1996), rewards' pirates (Peretti , 2004), etc.

Procedural Justice, a Means to Perpetuate the Psychological Contract Hypothesis 4: Procedural justice affects positively the respect of the psychological contract. (Hypothesis confirmed).
The fourth hypothesis refers to the theory of cognitive references of Folger (1986).It is concerned with procedures which act as a support for distributive justice by conveying to employees the appropriate explanations of the distributive justice Journal of Human Resources Management Research 10 while allowing them to engage effectively in their work (establishing psychological contract), to mobilize and make them adhere to the common objectives of the firm.
It is clear that the theory of cognitive references emphasizes on the importance of procedures as a means to support a favorable perceptions distribution without focus, necessarily, on the character of correctness of the relevant distribution.Employees who received convincing explanations, adequate and correct in relation to the compensation they have had, will feel motivated to maintain the viability of the psychological contract that binds them to their organization or their supervisor.Besides the implementation of formal procedures, the supervisor's ethical duty to his employees will determine the scope and the how of the employees behavior.Indeed, the absence of relational and interactional side between employer and employees leads to a violation or a breach of the psychological contract.This will encourage them to amplify the unjust distribution by assigning proportions, even if it is a priori a small injustice.Moreover, our empirical results confirm the work of Tekleab, Takeuchi & Taloyr (2005) who found that procedural justice is a precedent for the breach of the psychological contract.

The Trust, a Necessary Condition but Not Sufficient for Psychological Contract Hypothesis 5: Trust affects positively the respect of the psychological contract. (Hypothesis reversed).
It is quite possible that an employee is in a context where his contract was fulfilled and respected.Starting from this premise, we learn nothing when discussing what Robinson and Rousseau (1994) have demonstrated through their longitudinal study of MBA students, indicating that the perception of violation is negatively correlated with trust.Simply, the perception of an employee's noncompliance of obligations is the precursor to a loss of trust.However, this same employee may be judged as unfair even when he refers, then, to another referent that has kept his promises.That is to say, paradoxically as it may seem, he can perceive an unfair treatment when he confronts his contributions in the context of the relationship that binds him to his employer, with what he received in return, despite the perception that the employer has kept his part of the contract and fulfilled his obligations and promises.This confirms the work of modern biology that provide a different truth: « Individuals of many species learn much about their environment through the others » (Moscovici, 1994: 120).
Trust is usually mentioned as a relational concept that leads to cooperation in a configuration for building new forms of organization (Harrison and Laplante 1994).The confusion has been raised by Derbel and Mamlouk (2003) in their article entitled « The Dilemma of trust and cooperation: interdependence of actors and supremacy of organizational system ».The results that the two authors led through a real case, can consolidate the reversal of our fifth hypothesis.Derbel & Mamlouk (2003) start from the distinction between two types of trusts in this case « a trust expressed in terms of interest (calculated) » and « trust founded on an institutional arrangement (organization) » and confront two types of value systems or mutual expectations namely « integration » or « no integration ».By values system, authors intend ways of being or acting as an actor or a group appraises as ideal.
The dilemma arises when the actor is skeptical due to the uncertainty that characterizes the organizational context.He voluntarily chooses to not integrate and therefore to not share the same values that the party with which he is in relation, by managing the interdependence by the simulated forms of trust.To reinforce this point of view, we refer to the quotation of Derbel and Mamlouk who stipulated that « none of the parties wants to participate in a game that ignores the rules.The situation is unclear.There's a Dilemma » (2003: 151).This explains the unexpected result that we have reached stipulating that the trust does not necessarily affect the psychological contract, this is especially true in cases where both parties do not share the same values and use the organizational blur to implement their individual strategy.

The Social Influence, an Important Aspect Affecting the Psychological Contract
Hypothesis 6: Social influence affects the psychological contract.
The empirical research of Taylor and Todd (1995) found a significant relationship between subjective norms and intentions of behavior.For his part, Bogozzi (2000) showed that the effect of subjective norms on behavioral intentions is significant in case of social behaviors acting out in the presence of others.We can explain the result we have reached by the conformism that means the prescription of an individual to the hegemony of the norms of a group, causing a change in his behavior.Moscovici (1998) reinforces the idea that « compliance occurs when an individual changes his behavior or attitude to bring it more in harmony with the behavior or attitude of a group » (1998: 26).This behavioral change must go hand in hand with the integration standards usually required by the group vis-à-vis those who intend to join it.In consonance with this reasoning, Fischer (1987) outlines the three pillars of the process of compliance.The first relates to a situation of tension and ambivalence felt by the individual, marked by an antagonism between cognitive and his ways of thinking prior to the pressures to which he is compelled to submit.The second pillar was especially interesting to show the weight and impact of social influence on the individual.An evidence that is illustrated by his obvious convenience over everything that is recommended.The last pillar, states the result of the transformation justified by a certain self-fulfillment felt by the individual, and this by his alignment with new behaviors and attitudes, but also and especially by a voluntary renunciation or unconscious on his part to certain behaviors adopted and undertaken previously (questioning the sustainability of the psychological contract).
Moreover, according to Paicheler (1985) « The individual is delivered to the judgments of his similars.A two reasons: he fears a negative ruling and seeks to generate positive opinions; he relies on others to establish his point of view, in concert with them » (1985: 101).For his part, Milgram (1974) explains the phenomenon of social influence by obedience which corresponds to a change in behavior to obey the order from a legitimate or perceived authority.These theoretical contributions explain our empirical finding stating that social influence affects the psychological contract.In this sense, the sixth hypothesis is verified.

Conclusion
We believe that the contributions of this paper resides in the non-significant relationships emerged from the research model.Empirically, our model showed nonsignificant relationship between distributive justice and trust.However, the work of Cohen and Spector (2001), Colquitt and al. (2001) confirms the opposite.Indeed, trust is not simple, maybe it isn't inherent in a purely monetary relationship.Probably, it will not last.Trust is not as interest, it is also « emotive » (Servet, 1994, p. 39).
We could also raise a relationship presented by the theory as obvious, « Trust affects positively the respect of the psychological contract.
(Hypothesis reversed) ».A relationship not verified by our model.Indeed, according to the theoretical point of view, our paper focused on psychosocial variables previously little studied in the management sciences, in this case the social influence.A phenomenon which happens to be in strict relation to the non verification of the relationship that binds the trust to the respect of the psychological contract.A complex and pervasive social influence in an organizational setup, is able to change understanding of the terms of the psychological contract.Moderated by bad intentions and referred manipulatives, social influence is likely to impair the functioning of work, to undermine the trust already established between the partners in the relationship, to make organizational relationship vulnerable and to cause the breach of the psychological contract.
It is up to the initiative of the supervisors of « SND » bank to meet the promises they give to their human resources by creating a climate of trust within the meaning of Whitener and al. (1998).In the same vein of ideas, Tremblay (2006) evokes the skills of some employers who can establish a climate of social exchange based on reciprocity.A model that is built through respect of certain psychological benefits that an employer provides to its human resources as interactional justice, trust, moral support, recognition.In general, these psychological benefits are dependent on a set of organizational practices (practices of human resource management, work organization, executive leadership, vision, mission, values, goals).Feedback from employees will result in commitment and motivation to work higher, and by mobilizing collectively more.It is the duty of top management of the « SND » bank to insist, among others, the consolidation of a state of balance of trade involving the two parties allowing the anchoring of the aspect of reciprocity in a logic of a win-win.« SND » bank, by spreading stable characteristics, constitutes among its employees collective perceptions that are forged under the attention that it shows them through a few dimensions as it respects such as autonomy, trust, recognition, integrity and justice.