Mary G. Marcus, Ph.D. and Walter Marcus conducted the first
statistically significant study of financial outcomes in Connecticut divorces.
They compared results of both mediated and adversarial cases across judicial
districts. No differences were found in percentage of family income
women received, in percentage of liabilities women received, in the likelihood
of receiving alimony or in the amount of alimony obtained in mediated versus
adversarial divorces. Similar percentages of couples, 62% in both formats,
selected joint legal/wife physical as the most popular choice of custody
arrangement and identical numbers of days per month were spent by children with
father, 9.5, and mother 20.5. Mediated divorces took less time than adversarial divorces
and were significantly less likely to result in post-judgment modification, thus
sparing couples and families added emotional and financial costs. The Marcuses presented their research results at meetings of
the Regional Bar Association Family Law Committee, Connecticut Bar
Association Family Law Section, the Connecticut Psychological Association, the Connecticut Council for Divorce
Mediation, the Arizona Dispute Resolution Association and the Massachusetts
Mediation Council. An article about the
study appeared in The New York Times and complete results were
published in Mediation Quarterly.
Mary G. Marcus, Ph.D. and Walter Marcus conducted the first statistically significant study of financial outcomes in Connecticut divorces. They compared results of both mediated and adversarial cases across judicial districts.
No differences were found in percentage of family income women received, in percentage of liabilities women received, in the likelihood of receiving alimony or in the amount of alimony obtained in mediated versus adversarial divorces. Similar percentages of couples, 62% in both formats, selected joint legal/wife physical as the most popular choice of custody arrangement and identical numbers of days per month were spent by children with father, 9.5, and mother 20.5.
Mediated divorces took less time than adversarial divorces and were significantly less likely to result in post-judgment modification, thus sparing couples and families added emotional and financial costs.
The Marcuses presented their research results at meetings of the Regional Bar Association Family Law Committee, Connecticut Bar Association Family Law Section, the Connecticut Psychological Association, the Connecticut Council for Divorce Mediation, the Arizona Dispute Resolution Association and the Massachusetts Mediation Council. An article about the study appeared in The New York Times and complete results were published in Mediation Quarterly.
The following is the study as published in Mediation Quarterly:
To Mediate or Not to Mediate: Financial Outcomes in Mediated versus Adversarial Divorces
Divorce Mediation has become a popular and frequently used alternative to adversarial divorce across the United States over the last 20 years. Mediation has been seen as a means of reducing court workloads, lowering the cost of divorce, enhancing compliance with court orders for child support, and most importantly, enhancing post-divorce adjustment of children (Keilitz, Daley, and Hanson; Pearson, 1993; Kelly, 1990; Meierding, 1993; Emery and Jackson, 1989). While the promise of mediation has not always been fully realized, the benefits have been substantial enough to have led 38 states and the District of Columbia to develop programs within their court systems to mediate custody, visitation, and child support (Keilitz, Daley, and Hanson, 1992).
In addition to these court-based or court-annexed programs, private divorce mediation is a growing alternative to adversarial divorce as evidenced by the more than tripling of membership in the past 10 years in organizations such as the Academy of Family Mediators (Costello, 1999). In contrast to many court-based programs, private mediation addresses all divorce issues including spousal support, division of assets, division of liabilities, child support, and parenting issues. Private mediation thus offers even greater potential to reduce court costs, benefit children, and reduce costs for participants.
While mediation is being used extensively in the court system and ever more privately, criticism of divorce mediation is also prevalent in popular and scholarly forums. In an August 4, 1992 article, The Wall Street Journal, headlined, "Mediation Seen as Being Biased Against Women." The newspaper cited that women’s rights groups "point to at least 14 reports that found that women in mediation have sharply less bargaining power than men." But even more powerful than reports, is the tale quoted in the same The Wall Street Journal article of a California woman who participated in mediation and whose mediator "required her to give up her Northern California home and job to return to Southern California so that her ex-husband could visit their son more readily." In a professional family law publication, Family Law Quarterly, we also read horror stories about women who have been strong-armed by mediators into accepting settlements that leave them homeless and impoverished (Bryan, 1994). These articles encourage divorcing individuals to believe that adversarial divorce is superior in its process and outcomes. Most potential clients do not comprehensively review the literature as did Jessica Pearson, Ph.D., for the National Symposium on Court-Connected Dispute Resolution Research. She found "few differences in the reactions of men versus women to the mediation experience" and "where gender differences occur, they tend to favor women" (Pearson, 1993). In Keilitz, Daley, and Hanson’s 1992 study comparing participants’ satisfaction in mediation and adversarial court divorce programs, women were as likely as men to say that they felt they had participated in a fair process, that they had not been pressured to agree to something they didn’t want or pressured to agree quickly and to feel satisfied with the agreement. Ironically, in Keilitz, Daley, and Hanson’s research, men actually were more likely to perceive that the mediation process and the mediators were biased against them!
Criticism of divorce mediation by attorneys can lead clients to question whether divorce mediation will yield as optimal a result in their case as would an adversarial approach. More than half of a sample of divorce attorneys in New Jersey were unable to find any benefit in divorce mediation (Kressel and Hochberg, 1987). The attorneys felt that clients were too angry to benefit from mediation or too lacking in judgment to make use of a process in which the client controlled the negotiation. More recently, in the study state of Connecticut, a prominent divorce attorney was cited in The New York Times as saying that "mediation may work for the small number of cases in which there are no real property issues and no real problems regarding the children" (D’Agostino, 1999). Attitudes such as these lead clients to question whether divorce mediation will yield as optimal a result in their case as does an adversarial approach.
There have been few previous studies that have compared actual outcomes of mediated to adversarial divorces. Joan B. Kelly’s extensive Divorce and Mediation Project which was initiated in 1983 attempted to ascertain differences which existed between mediated and adversarial respondents’ perceptions of the process and outcomes of divorce. Among the variables examined were satisfaction with the financial agreements in both child support and non-child related aspects (Kelly, 1989). Kelly’s mediation group was as satisfied with the property agreement as her adversarial group. In both groups, child support was seen as inadequate by 38% of the women. Kelly’s study was a self-report study and focused on attitudes rather than hard data.
Pearson attempted to assess the equity of mediated divorce agreements by sending questionnaires to former mediation clients (Pearson, 1991). She divided this sample into clients who had resolved issues on their own, with attorney assistance, through a public mediation program, through a private mediation program and with a judge. She found no differences in the amount of child support based on the form of dispute resolution. Similarly, she found no differences in division of assets among the various dispute formats with women receiving an average of 52% of assets. Also, the percentage of alimony was similar among methods, with 26% of cases yielding alimony awards. Pearson’s study offers a limited comparison of mediated and adversarial methods, however, since all participants had been exposed to mediation during some aspect of their divorce.
Carol Bohmer and Marilyn Ray in another questionnaire study compared the financial outcomes in mediated, attorney-negotiated, and judicially assisted methods of divorce in New York and Georgia and found equivocal results (Bohmer and Ray, 1994). The New York study which was based on 124 couples and was done before child support guidelines had gone into effect indicated that a smaller percentage of mediated cases resulted in cash settlements. Bohmer and Ray also found that asset splits for mediated vs. attorney negotiated cases were approximately equal, with males getting a greater percentage of the assets, 56/44 in mediated cases, 57/43 in attorney-negotiated, and 68/32 in judicially-assisted. Additionally, women in mediated cases in which sole custody was awarded were significantly less likely to be awarded the family residence than they were in attorney-negotiated or judicially-assisted cases. In contrast, in Georgia where child support guidelines had been instituted, in a sample of 83 cases, Bohmer and Ray found that women received a similar percentage of maintenance settlement awards in mediated and attorney-negotiated cases. Regarding one time only cash settlement awards, women who mediated in Georgia were more likely to obtain these awards than women who used attorney negotiation. There were no significant differences in Georgia in division of assets among the divorce procedures used.
In reviewing the literature regarding the difference in financial outcomes in mediated versus adversarial divorce cases, it became apparent that there were many assertions about advantages or disadvantages of mediation but that the few actual studies yielded results that were varied, ambiguous, and sometimes conflicting. The current private study in Connecticut with samples large enough to draw meaningful conclusions, was an effort to determine objectively the benefits and detriments of mediating divorce cases.
In this Connecticut study, 199 privately mediated divorce cases were compared with 201 adversarial divorce cases. Since the State of Connecticut court forms have no indicators of whether a divorce has been mediated or not, we were unable to select the mediated cases randomly. We wanted to ensure that cases deemed to have been mediated were cases that had been mediated from start to finish, and so we contacted members of the Connecticut Council for Divorce Mediation and asked them to supply the researchers with docket numbers of cases they had mediated to completion. The Connecticut Council for Divorce Mediation is a statewide organization of mediators who have met training and experience criteria in mediation. Since we were concerned that we would not be able to identify 200 private mediation cases in the current year, we asked mediators to provide us cases that had gone to judgment in 1996, 1997, or 1998. We received responses from 20 mediators throughout the state, with the majority of cases from the Stamford, Hartford, Litchfield, and Tolland judicial districts. To find a comparison sample of adversarial cases, we were able to randomly select cases from the court files in the same judicial districts as the mediated cases. We limited adversarial cases to those in which both parties were represented by attorneys so that all adversarial parties would have had the benefit of legal representation. If mention of mediation was found in a potential adversarial file, such as Court Family Relations mediation of custody issues, that case was eliminated from the adversarial sample in an effort to keep the comparison groups as pure as possible.
To increase the likelihood that mediated and adversarial cases were comparable, we matched mediated and adversarial cases on length of marriage and whether or not the parties had children under 18, and as stated above, by judicial district. Adversarial cases were also limited to those that had gone to judgment in 1996, 1997, and 1998 as had the mediated cases. Again, to maximize comparability of samples, any adversarial cases in which the State of Connecticut had contributed to the maintenance of the parties were excluded from the sample since we had no mediated cases in which the state contributed maintenance benefits. Although we did not control for this, in our random sample of adversarial cases, only 3 of 201 cases had been tried to conclusion.
Dependent variables which we examined were: percentage of family income obtained by wife and husband; percentage of family assets obtained by wife and husband; percentage of family liabilities retained by wife and husband; amount of alimony awarded; duration of alimony awarded; and amount of child support awarded. Other variables examined were: any relationship between length of marriage and percentage of family income, assets and liabilities; length of time from initiation of divorce complaint to judgment; frequency of post-judgment actions in mediated vs. adversarial divorces; frequency of various custody arrangements in mediated vs. adversarial cases; any relationship between type of custody arrangement and percentage of family income, family assets and family liabilities; and the impact of the husband’s assumption of financial responsibility for children’s college education on the amount of periodic alimony the wife receives. Percentage of family income was determined by summing the wife’s earned and investment monthly income, any child support awarded, and any alimony greater than $1/year awarded, and dividing this sum by the total family monthly earned and investment income.
The aforementioned data were collected by the first two authors from examination of divorce judgments and financial affidavits in court files in the various judicial districts. For research design and data analysis, the last two authors, two researchers with strong statistical experience from the University of Connecticut Health Center, used the SPSS statistical program to analyze the data with statistical techniques including t-tests, chi-square tests, and analyses of variance, Pearson correlations, multiple regressions, and multivariate analysis of variance.
As indicated in Table 1, no significant differences were found between the percentage of family income that women received in mediated versus adversarial cases state-wide. Similarly, no significant difference was found between the percentage of liabilities that women in mediated cases retained versus women in adversarial cases (Table 1). However, again as reflected in Table 1, the percentage of family assets that women retained was found to be significantly different in mediated versus adversarial cases, with women in mediated cases retaining a greater percentage of family assets.
No significant differences were found in the amount of periodic alimony awarded in mediated versus adversarial cases (Table 2). Only cases in which periodic alimony was greater than $1/year were included so that the data was not skewed by those cases in which the wife was awarded $1/year for the purposes of being able to reopen the judgment should circumstances change or if the husband did not repay debts as agreed. In 30% of mediated and 30% of adversarial cases periodic alimony greater than $1/year was awarded to women. There was no significant difference between the 19.60% of mediated cases and 15.42% of adversarial cases in which $1/year alimony was awarded. Lump sum alimony was so infrequent an event that only five women in mediated and five women in adversarial cases were awarded it and thus no conclusions could be drawn about any differences in amounts of lump sum alimony. A significant difference was found in the years of periodic alimony awarded to women in mediated cases versus adversarial cases, with women in mediated cases receiving alimony for 7.67 years and women in adversarial cases receiving alimony for 6.04 years (Table 3).
Child support awards were found to be significantly different in mediated versus adversarial cases (Table 4). Women in mediated cases received significantly more child support than women in adversarial cases. A similar proportion, 62%, of couples in mediated and adversarial cases had children under 18, and a similar percentage of couples with children had no awards of child support, 15% of mediated cases and 17% of adversarial cases. Couples who had mediated had similar numbers of children as well. There were no significant differences between the number of days per month with each parent in mediated versus adversarial cases, with the mean of mediated cases being 20.42 days with the mother and 9.58 days with the father and in adversarial cases, 20.46 days with the mother and 9.54 days with the father. However, there was a significant difference in percentage of unspecified visitation arrangements, with significantly more, 27% vs. 16%, in adversarial than mediated cases.
Couples in both mediated and adversarial cases most frequently chose Joint Legal/Wife Physical as the custody arrangement with 86 of 125 mediating couples making this choice and 84 of 120 adversarial couples. There was a significant difference, however, in choice of Joint Legal/Joint Physical and Wife Legal/Wife Physical, with more mediating couples choosing Joint Legal/Joint Physical (21% mediated vs. 12% adversarial) and more adversarial couples choosing Wife Legal/Wife Physical custody (3% mediated vs. 12.5% adversarial). Multivariate analysis of the relationship between type of custody arrangement and the percentage of family income women receive yielded only one significant difference, that is between couples with Joint Legal/Joint Physical Custody and those with Joint Legal/Wife Physical Custody, in which women receive a greater percentage of family income if they have Physical Custody in both mediated and adversarial cases (Table 5). There was no impact of type of custody on percentage of family assets or percentage of family liabilities.
No relationship was found between who was the plaintiff and percentage of family income, percentage of family assets, and percentage of family liabilities awarded. In 63% of mediated cases and in 67% of adversarial cases, women were the plaintiffs.
A significantly greater percentage of parents in mediated cases, 26%, chose to provide for their children’s college education than the 10% of parents in adversarial cases. Husbands’ agreement to assume financial responsibility for their children’s college education was found to have no impact on the amount of periodic alimony greater than $1/year that wives received in mediated and adversarial cases nor on the percentage of family assets wives obtained in either divorce format. Length of marriage was a determinant of whether wives received alimony or not (Table 6). Discriminant function analysis indicated that only length of marriage determined whether women received alimony and that whether the divorce format was mediated or adversarial was irrelevant. Women married longer than five years were significantly more likely to receive alimony than those married five years or less. Similarly, women married longer were significantly more likely to receive alimony for more years than those married fewer years in both mediated and adversarial cases. However, length of marriage did not have a relationship to percentage of family income, assets or liabilities obtained by women.
The mean time in days from service of the complaint to the divorce judgment was found to be significantly longer in adversarial, 325 days, than mediated, 245 days, cases. The frequency of post-judgment modifications was also significantly higher in adversarial cases, 22%, than in mediated, 9%, cases (Table 7)
Our findings are contrary to predictions made about the results of mediation for women by those who argue that mediation favors men. Not only do women obtain similar percentages of family income and family liabilities in mediated versus adversarial cases, but they also obtain similar amounts of periodic alimony and similar percentages of $1/year alimony. In a non-financial arena, our study indicates there also is no difference in the number of days children spend with each parent in the differing formats nor in the type of custody arrangement most frequently chosen. Surprisingly, women in mediated cases obtain a higher percentage of family assets, receive periodic alimony for more years than their adversarial counterparts, and obtain greater amounts of child support. While women obtain superior outcomes via mediation in several financial areas, men who mediate were more likely than those using the adversarial process to obtain joint legal and joint physical custody.
That women who participate in divorce mediation obtain similar percentages of family income and family liabilities as women who use the adversarial process is consistent with Pearson’s results as well Bohmer and Ray’s Georgia results (Pearson, 1991; Bohmer and Ray, 1994). That women who participate in mediation get a larger percentage of family assets than women in adversarial cases is an unexpected finding not predicted by Pearson’s or Bohmer and Ray’s results. While Pearson found that there were no differences in percentage of assets awarded to women in mediated and adversarial formats, she found that the percentage was similar to that in our study, approximately 52%. Bohmer and Ray noted that women received fewer assets than men although divorce format was unimportant.
Results in our study that suggest mediation is actually advantageous for not only women but couples are that mediation leads to significantly fewer post-judgment modifications, that mediation is a speedier process, and that children’s college education is more likely to be included in the agreement when couples mediate. Reduction of post-judgment actions and faster process suggest mediation would be less costly to couples and less disruptive of their lives as well as advantageous to the court system. Pearson and Anhalt found that unspecified visitation orders are responsible for 40% of the post-judgment litigation and so our finding that in mediated cases there are significantly fewer unspecified visitation orders again suggests mediation is advantageous to families (Pearson,1993). However, whether mediated cases ultimately yield fewer long-term post-judgment modifications remains unclear (Pearson, 1993). Our finding that whether husband or wife is the plaintiff is irrelevant to financial outcomes should alleviate some acrimony and anxiety on the part of spouses whether in mediated or adversarial divorces.
In summary, our results indicate that most financial outcomes of mediated versus adversarial divorce formats are not significantly different, and that three major financial outcomes, percentage of family assets obtained, amount of child support, and years of periodic alimony actually favor women who mediate. Since adversarial divorce is generally demonstrated to cost more than mediated divorce (Pearson, 1993; Kelly, 1990), and as our study shows, to take longer and to lead to more frequent post-judgment modification, divorce mediation is an option worthy of serious consideration by most divorce clients.
Pearson, J. The Equity of Mediated Divorce Agreements. Mediation Quarterly, 1991, 9(2), 179-197.
Pearson, J. Ten Myths About Family Law. Family Law Quarterly, 1993, 2, 279-299.
Mary G. Marcus is with the Center for Divorce Mediation & Alternative Dispute Resolution, Inc. in Norwalk, Connecticut.
Walter Marcus is with the Center for Divorce Mediation & Alternative Dispute Resolution, Inc. in Norwalk, Connecticut.
Nancy A. Stilwell is Assistant Professor, Department of Behavioral Sciences, University of Connecticut Health Center in Farmington, Connecticut.
Neville Doherty is Professor Emeritus, Department of Behavioral Sciences, University of Connecticut Health Center in Farmington, Connecticut.
Table 1a. Wife’s Percentage Family Income
Table 1b. Wife’s Percentage Family Assets
Table 1c. Wife’s Percentage Family Liabilities
Table 2. Wife’s Periodic Alimony Per Month
Table 3. Wife’s Years of Periodic Alimony
Table 4. Wife’s Amount of Child Support Per Month
Table 5. Relationship of Type of Custody to Wife’s Percentage of Family Income
a. Wife’s % of Family Income by Med/Adv. Type of Custody
b. All effects entered simultaneously
Table 6. Likelihood of Wife Receiving Alimony by Length of Marriage
Table 7. Time in Days Between Complaint and Judgment