The goal of our research is to study the mechanisms by which resorting to mediation (or a third party facilitator) may re-establish trust between parties who are on opposite sides in a contentious context leading to a relational impasse. This research is based on a qualitative study (64 interviews) on ten organizations taking advantage of the “Support for Social Dialogue” program set up by the DRT (Direction Régionale du Travail) and the ANACT (Agence Nationale pour l’Amélioration des Conditions de Travail) in France. This resolution framework consists of leading mediation interventions so as to improve, even restore, dialogue between social partners when, outside of true crisis situations, relations which are difficult or based on mistrust have become present within an organization. One first observation from our study is the success of the mediators’ mission. In all cases, the mediators were able to restore dialogue between the parties and thus able to develop, to different degrees, a trust relationship. A quarter of the people questioned expressly experienced an improvement in the trust relationship, and the sustained effect over two years was observed in only one organization. One way of measuring the effect of mediation on trust was to observe if the behaviour of the parties had changed. The involved parties adopt, in all the mediation cases studied, a less aggressive tone and respectful attitudes, even if these improvements are of limited duration in time. The measurement of trust in itself depends on different mechanisms, which vary from one situation to another. Trust is a multi facetted concept, including various dimensions requiring a detailed analysis so as to determine points of action on which the mediators have an influence. Our study shows the positive role played by the mediator on the trust linked to knowledge (Lewicki and Bunker, 1995b), as well as the cognitive dimension described by McAllister (1995). In fact, when there is a relational impasse, the mediator is able to restart the exchange of information among the parties, and in this way, promote the return to this form of trust. Concretely, they also provide a structured framework which makes the facilitation of communication possible. Mediators also engender a certain transparency which is reflected in the fact of participants being able to talk in a more direct and open manner with each other. The active listening taught by mediators also allows the actors to take a step back from the situation and acquire maturity useful in an improved understanding of the other’s point of view. We also find a positive role for the mediation process on the calculated trust (according to the classification by typology proposed by Lewicki and Bunker, 1995b) and on the emotional dimension of trust (according to the typology of McAllister, 1995). This is promoted by mediators in particular through a feeling of warmth and calm. It is possible to observe the growth of mutual respect between the participating parties. Mediation runs into limits along its most fragile dimension: identity confidence. The reduction of the effects of mediation over time upon trust could be linked in particular to this difficulty of permanently influencing the identity component of trust. It is possible that a tripartite approach in terms of the “psychological contract”, between employer representatives, management representatives and mediators, could act upon the emotional and identity dimensions of trust, even if these aspects are definitely more difficult to reduce and to understand, and that in a certain number of cases, ideological limits are the source of a lasting deadlock in communication. As well, there is the problem in the follow up to mediation over time. Continuity on the part of mediators in their intervention over time would be recommended so as to avoid a trust breakdown. In effect, the identity and affective dimensions of trust are, beyond a doubt, the first dimensions to be affected; the ideological divisions always reappear when new difficulties arise due to the pressure of internal constraints (mandates, production, emergencies, etc.). An effect of contagion can then be felt along the other dimensions, thus explaining a new deterioration in the trust relationship. As a result, mediation can be seen as an invaluable tool aimed at having an effect in contexts where one observes a “relational impasse” between social partners so as to restore the confidence and improve the social climate in the organization. An essential question emerges from this study: How to know from which threshold the general concept of trust swings from the negative side towards the positive side? In what proportions, must each of the sub components of trust be satisfied so as to go from a situation of mistrust to a situation of trust? It is from this question that stems the complexity of analyzing and measuring this phenomenon. Our study would need to be extended through complementary inquiries so as to lead to teachings having a more general impact. In fact, the external validity of this impact is limited in the measure that it is based on only ten organizations case studies and thus necessarily remains contextual
