Mediator-Supported, Not Informally Advised or Self-Managed: Third-Party Help During Phased Disengagement from a Difficult Working Relationship

Abstract

We followed 231 employees in nine organizations who planned to scale back, over six weeks or more, their contact with a colleague they found persistently difficult. Where a line manager, team lead, or senior colleague met with one or both parties at least twice to reassign shared tasks or agree how the two would communicate ("mediator-supported," n=94), 72% of employees had a stable reduced-contact arrangement in place by week twelve, against 61% of employees who only talked the situation over with someone ("informally advised," n=72; an 18% gap) and 58% of those who handled it alone ("self-managed," n=65; a 24% gap). Recorded conflict incidents ran at 2.13 per 10 cases in the mediator-supported group, versus 2.64 and 2.62 in the other two (gaps of 24% and 23%). Among 128 colleagues on the other side who took part, those in mediator-supported cases rated the process fairer (5.0/7) than those in informally advised (4.1/7, a 22% gap) or self-managed cases (3.9/7, a 28% gap). Mediators logged 6.3 hours per case over the twelve weeks. Senior colleagues and managers reached similar stable-arrangement rates (73% and 72%), and colleagues on the other side rated senior-colleague mediation slightly fairer (5.2/7 versus 4.9/7), a difference too small and too thinly sampled to rely on.

1. Introduction

When two colleagues cannot work well together and one decides to step back, the decision is usually made alone and carried out alone. The other person learns of it from the change in how they are treated. Shared tasks that ran between the two are left to sort themselves out.

A third party can change that sequence. First, someone outside the pair hears what is happening. Then the shared work is laid out and some of it is moved. Finally, the two agree on how they will still communicate on what remains. This is close to what workplace mediators do in formal disputes, but most difficult relationships never become formal disputes, and the people best placed to help are usually a manager or a more experienced colleague.

We compared employees whose step back had this kind of support with those who only sought advice and those who managed it alone.

2. Method (Summary)

Employees in nine organizations (hospital administration, two engineering consultancies, a retailer, a public transit operator, and four technology firms) who reported a persistently difficult working relationship were invited to log a plan for reducing contact over six weeks or more. 231 did so and were followed for twelve weeks.

Support was classified from the employee's weekly log, checked against the third party's own log where there was one. Employees were "mediator-supported" if a line manager, team lead, or senior colleague met with one or both parties at least twice in the first six weeks and, in those meetings, reassigned at least one shared task or set an agreed communication channel for remaining shared work (n=94; in 61 cases the third party was the employee's own line manager or team lead, and in 33 a senior colleague). They were "informally advised" if they discussed the situation with a manager or colleague but that involvement did not meet these criteria (n=72), and "self-managed" if no one else was involved (n=65).

The primary outcome was whether a stable reduced-contact arrangement was in place at week twelve, defined as no unplanned return to full contact in weeks nine to twelve and confirmed by both the employee and their line manager. Conflict incidents were complaints or escalations recorded by the line manager or HR during the twelve weeks. The colleague on the other side was invited, without being told the purpose, to rate how fair the changes in their working arrangements had been (1–7); 128 took part (54 mediator-supported, 40 informally advised, 34 self-managed). Mediators logged their time; advisers in informally advised cases did not.

3. Findings

A stable arrangement was in place at week twelve for 68 of 94 mediator-supported employees (72%), 44 of 72 informally advised employees (61%; an 18% gap), and 38 of 65 self-managed employees (58%; a 24% gap).

Conflict incidents ran at 2.13 per 10 mediator-supported cases (20 incidents in 94 cases), 2.64 per 10 informally advised cases (19 in 72; a 24% gap), and 2.62 per 10 self-managed cases (17 in 65; a 23% gap).

Colleagues on the other side rated fairness at 5.0/7 in mediator-supported cases, 4.1/7 in informally advised cases (a 22% gap), and 3.9/7 in self-managed cases (a 28% gap).

Mediators spent 6.3 hours per case on average: 6.6 hours for managers and team leads, 5.7 hours for senior colleagues. The stable-arrangement rate was 72% (44 of 61) with a manager or team lead and 73% (24 of 33) with a senior colleague. Colleagues on the other side rated senior-colleague mediation at 5.2/7 (n=19) and manager mediation at 4.9/7 (n=35).

4. Discussion

The obvious caveat comes first. Employees who involved a mediator had a manager or senior colleague willing to spend six hours on their problem, and that may say as much about the workplace as about the mediation. Better-run teams may produce both willing mediators and smoother transitions.

With that in mind, the sequence the mediators followed, which is partly built into how the group was defined, suggests where their effect came from. In the first weeks, they heard from one or both parties. By week six, shared work had been moved, a channel agreed, or both. Informally advised employees discussed the situation with someone, but the shared work stayed where it was, and their stable-arrangement and fairness figures sat much closer to those of employees who had no help at all.

The fairness ratings matter because a step back is easy to read as rejection. Colleagues on the other side of a mediated step back rated the changes as fairer than colleagues in cases without mediation, which suggests that a third party can make reduced contact feel like an arrangement rather than a judgment.

The cost fell mostly on the mediators. Six hours over twelve weeks is time a manager or senior colleague could have spent elsewhere, and the study did not credit it to anyone's output. Employees also gave up some control: they had to describe the difficulty to a third party, and in some cases accept task changes they had not chosen.

The senior-colleague figures are worth a follow-up but not a conclusion. Thirty-three cases and nineteen fairness ratings cannot carry the 0.3-point difference.

5. Limitations and Future Directions

Support was not assigned, so the mediator-supported group may have had better-run teams, more cooperative colleagues, or less severe difficulties. The line manager confirmed whether the arrangement was stable, and in 61 cases that manager was also the mediator. Conflict incidents depend on how readily each organization recorded complaints, and line managers who were also mediators may have recorded them differently. The 128 participating colleagues were a self-selected 55% of those invited. Twelve weeks cannot show whether arrangements hold over a year. A follow-up will offer a short mediation protocol for managers and senior colleagues in a randomly selected half of participating teams.

Editorial Note

This report underwent CfAHC's Rapid Review Protocol, including AI-assisted literature cross-checking and a two-reviewer human sign-off.