Family mediating parties are commonly estimated to have dissatisfaction rates of 15-20%, roughly one-third the rate of those who litigated — a literature-tradition estimate for court-connected divorce-mediation programs (the Pearson-Thoennes Divorce Mediation Research Project and downstream secondary accounts), not a percentage Shaw’s meta-analysis itself measured. Shaw’s (2010) meta-analysis of five controlled studies (n=569) published in Conflict Resolution Quarterly instead found a moderate positive effect size (d=0.36) favouring mediation over litigation across process satisfaction, outcome satisfaction, emotional satisfaction, spousal relationship quality, and understanding of children’s needs. Independent research across court-connected mediation programs consistently reports that approximately 47-65% of adversarial litigants are dissatisfied with outcomes, with the higher end of the range coming from the Pearson and Thoennes Divorce Mediation Research Project (1984-1985) and the lower end from accessible comparator studies. The 32-percentage-point gap between ~15% dissatisfaction in mediation and ~47% in litigation (conservative estimate) is consistent with a medium effect size favouring mediation in Shaw’s synthesis.
Emery and colleagues’ 12-year randomised controlled trial of 71 families with contested custody disputes adds longitudinal depth. Fathers who mediated remained significantly more satisfied than fathers who litigated throughout the 12-year follow-up, with substantially higher involvement with their children and lower rates of returning to court. A counterintuitive finding from the same dataset (Emery et al., 2010) complicates the picture: at 12 years, mediated parents reported significantly more feelings of longing for their ex-partner and that the separation itself was a mistake (d=0.39, p < .05) compared to litigated parents. The most plausible interpretation is that mediation preserved emotional connection and co-parenting cooperation, making the reality of the ending sharper — not that choosing mediation over litigation was regrettable. Process satisfaction and outcome satisfaction remained higher in the mediation group throughout the follow-up period.
Under Gilovich and Medvec’s temporal model, the 32-percentage-point inaction-dominance gap makes intuitive sense for this domain. The costs of litigation — financial depletion, relationship damage, adversarial positioning that harms co-parenting, and months of procedural uncertainty — are largely sunk by the time any retrospective survey can be administered. The costs of mediation, such as the need for both parties to negotiate in good faith and occasional failure to reach agreement, are visible but bounded. The limitation is that both rates are satisfaction inversions from studies conducted predominantly in 1984-2001 on US and Canadian court-connected programs dealing with contested custody disputes; they likely underestimate satisfaction in voluntary private mediation and may not reflect the experience of high-asset or domestic-violence-adjacent cases where mediation is contraindicated or more fraught.