Does Mediation Work? (Footnote Version)

Researchers have conducted more than 50 studies since divorce mediation first appeared more than two decades ago. As one researcher puts it, the easiest research has already been done, i.e., comprehensive research on the outcomes of mediation. Enough data has been collected and enough analysis conducted to begin drawing clear conclusions about whether mediation works. Along several key axes, the answer is an unequivocal yes.

A.               Settlement Rates

Mediation produces agreement in 50 to 80 percent of cases.[1]  This is the case whether the mediation is court-referred or privately placed, whether mediation is voluntary or mandatory, and whether the mediating couples had a history of domestic violence or intense marital conflict. There is some evidence that settlement rates of more than 85 percent suggest a more coercive style of mediation. This is particularly likely to be a factor if settlement rates are the only criterion for judging success of mediation. In general, single-issue mediations have a lower settlement rates than mediations involving multiple issues.[2]

B.               Client Satisfaction

1.                  Overall.

Couples who mediate the issues of their divorce are significantly more likely to be satisfied with the experience of their divorce when compared with couples who have finished an adversarial divorce. At final divorce, according to one study, 69 percent of mediation respondents were somewhat to very satisfied, compared to only 47 percent of adversarial men and women.[3]  Interestingly, this preference for mediation was not uniform. On several issues, such as the warmth and sensitivity of the professional, adequacy of information, and satisfaction with child support, there were no statistically significant differences between the perceptions of mediating and adversarial couples. On most issues, however, such as the perceived skill level of the professional, the creativity of the professional, the effectiveness of the professional in helping clients deal with anger, the professional’s success avoiding imposing his or her viewpoint on the client, impact on the spousal relationship, satisfaction with the property settlement, satisfaction with arrangements around spousal support, satisfaction with parenting schedules and arrangements, and understanding children’s needs and issues, mediating couples reported significantly higher levels of satisfaction.[4]

2.                  Among Women.

In general, the difference between men and women in their level of satisfaction with mediation is not statistically significant. This is in contrast to adversarial divorce, where men are significantly more dissatisfied than women with the process and outcome.[5]  There has been some discussion of findings that women are disadvantaged in mediation, but that initial research has been discredited. A recent review of the literature spells out the chronology:

In a much hailed experimental study, (two researchers) found that women in litigation felt that they had won more and lost less than their counterparts in mediation, whereas the reverse was true of men. More recent work by (these same researchers and their associates) has since qualified their original results. These recent studies showed that in the litigation group, women’s satisfaction rose through time whereas men’s satisfaction declined. In contrast, in the mediation group, women’s satisfaction, already very high on entry, changed very little, whereas men’s satisfaction increased through time. On the whole, then, women in mediation expressed greater satisfaction with both process and outcomes than did their litigation counterparts.[6]

C.               Effect on Terms of Agreement

In general, mediated agreements tend to be more comprehensive than settlements reached either voluntarily or involuntarily in adversarial court.[7]  In general, mediation results in more joint legal custody compared to adversarial processes,[8] but not necessarily a different parenting schedule.[9]  Researchers have not noted a statistical difference in the treatment of child support payments, although mediating fathers are more likely to agree to pay for “extras” for their children and are more likely to agree to help with college expenses.[10]

D.               Long-term Mental Health

Researchers agree that mediation does not seem to have any long-term statistically significant effect on the psychological adjustment of either divorcing couples or their children, whether the mediation is custody only or comprehensive.[11]

E.               Cost in Time and Money

Mediating couples tend to resolve the issues in their divorce in substantially less time than that taken by their counterparts in litigation.[12]  They also tend to spend significantly less money.[13]  In one study, couples in the adversarial sample reported spending 134% more (more than twice as much) for their divorces than those in the mediation sample. [14]  Most reports tend to find less dramatic differences, however, in the 30-40% range.[15]

F.                Compliance and Relitigation

Researchers generally report higher rates of compliance with mediated agreements, when compared to agreements reached in the adversarial process. This includes parenting schedules, payment of child support and spousal support, and completing the final division of property.[16]  Relitigation rates are low in general among mediated samples and are lower than in adversarial samples.[17]

G.              Process Studies

What remains is the more difficult and expensive task of exploring the process of mediation to learn what styles work well, and why. The little bit of research that has been completed appears to show that the mediators who are more effective in helping their clients reach agreement are more active in structuring the process, focus more on problem solving than on reconciling positions, discuss options and solutions rather than facts, and maintain flexible over the process of mediation that varies depending on the characteristics of their clients.[18]

H.               What We Don’t Know

Research is scant to nonexistent on the following issues: Should the mediator tell clients about legal issues or simply direct them to their attorneys? Should lawyers be present or absent from actual mediation sessions? Should the mediator welcome and deal with emotional issues or push them aside? Should mediators be directive or passive? Should mediators work individually or in teams? Should sessions be sequential, or should mediation be conducted in a marathon format? Should the parties meet together or in separate caucus rooms? Should mediators be in the same racial group as their clients, or does it matter? Should pure mediation be replaced with a med/arb system that gives the clients assurance of a resolution? All these and other questions remain to be answered as researchers explore the effectiveness and the process of divorce mediation.


[1] Benjamin, M. and Irving, H. H., “Research in Family Mediation, Review and Implications,” Mediation Quarterly,1995, 13(1) 53-82, at 57; Kelly, J., “A Decade of Divorce Mediation Research,” Family and Conciliation Courts Review, 1996, 34(3) 373-385, at 375.

[2] Walker, J., McCarthy, P., & Timms, N. (1994). Mediation:  The Making and Remaking of Cooperative Relationships. Newcastle Upon Tyne, UK:  University of Newcastle Upon Tyne, Relate Centre for Family Studies.

[3] Kelly, J. (ed.) “Empirical Research in Divorce and Family Mediation.”  Mediation Quarterly, 1989 24(1). The couples who mediated and the couples who completed the adversarial process had similar levels of marital conflict and tension reported in the two years prior to separation. They also measured at comparable levels for difficulties with marital communication, feelings regarding the divorce, degree of unilateral decision making about the divorce, and overall cooperation. The groups were similar in the amount of reported divorce-specific anger at the spouse and in hostility toward the spouse.

[4] Id.

[5] Emery, R. (1994). Renegotiating Family Relationships:  Divorce, Child Custody, and Mediation. New York:  Guilford.

[6] Benjamin & Irving (1995), at 59.

[7] (1) Kelly, J. (1993). “Developing and Implementing Post-Divorce Parenting Plans:  Does the Forum Make a Difference?” In C.E. Depner & J.H. Bray (Eds.), Non-residential Parenting:  New Vistas in Family Living  (pp. 136-155). Newbury Park, CA:  Sage;  (2) Pearson, J., & Thoennes, N. (1989). “Reflections on a Decade of Research.”  In K. Kressel, D. Pruitt, & Associates (Eds.), Mediation Research:  The Process and Effectiveness of Third-Party Intervention. San Francisco:  Jossey-Bass.

[8] Emery (1994); Pearson, J. (1991) “The Equity of Mediated Divorce Agreements,” Mediation Quarterly, 7, 347-363.

[9] Emery (1994).

[10] Kelly, J (1990). Mediated and Adversarial Divorce Resolution Process:  An Analysis of Post-Divorce Outcomes. Fund for Research in Dispute Resolution, Northern California Mediation Center).

[11] Emery (1994); Walker, et al. (1994).

[12] Emery (1994) (less than half the time); Benjamin & Irving (1995) (fewer sessions and fewer service hours than litigation).

[13] Pearson, J. & Toennes, N. “A Preliminary Portrait of Client Reactions to Three Court Mediation Programs.” Conciliation Courts Review, 1985, 12(1), 1-14.

[14] Kelly, J. (1991) “Mediated and Adversarial Divorce Resolution Processes:  An Analysis of Post-Divorce Outcomes.”  Fund for Research in Dispute Resolution.

[15] Pearson, J. & Theonnes, N. “A Preliminary Portrait of Client Reactions to Three Court Mediation Programs” Conciliation Courts Review, 1985, 23(1), 1-14.

[16] Emery (1994); Irving, H. H. & Benjamin, M., “An Evaluation of Process and Outcome in a Private Family Mediation Service.” Mediation Quarterly, 1992, 10(11), 35-55; Kelly (1990); Pearson & Thoennes (1989).

[17] Irving & Benjamin (1992); Pearson & Theonnes (1989).

[18] Slaikeu, K., Culler, R., Pearson, J. & Thoennes, N. (1985) “Process and Outcome in Divorce Mediation.” Mediation Quarterly, 10.