Abstract
In this study, we examined creativity in court‐connected mediation. We analyzed 129 mediated agreements from civil cases in Norway and Denmark and compared the outcomes with the parties' original claims to determine whether the agreement addressed only the disputants' demands or contained other elements. If the mediated agreements contained elements in addition to the original claims, we considered them to be “creative.” We devised a creativity scale and found that approximately two thirds of the cases contained creative elements and one quarter of them contained more than five creative elements.
We then sought to determine which aspects of the mediation promoted creativity by looking at a variety of mediation characteristics (length of mediation, characteristics of the parties, etc.). We found that lengthier mediations tended to feature more creativity as did cases that involved two private individuals rather than businesses. Cases whose issues involved inheritance as well as the division of property following divorce seemed to foster the highest levels of creativity. Finally, we found that the amount of money at issue also seemed to be relevant: the highest levels of creativity were found in cases in which more money was at stake. In this article, we also discuss the implications of our findings for future research, practice, and training.
Introduction
Mediation in court settings is often promoted with the promise that negotiated agreements will address both parties' interests and needs rather than determining each party's rights. This is evident in both the scholarly (Ury, Brett, and Goldberg 1988; Vindeløv 2012) and promotional literature.1 Implicit in this promise is the assumption that in the mediation process, parties will have the opportunity to think “outside the box,” to discuss the issues that underlie their demands, and to customize agreements to address their needs (Kovach 2004). As Richard Walton and Robert McKersie (1965: 23) stated: “bargaining is not just a process of dividing up existing resources but is also a process sometimes used for creating additional value or mutual benefits.”
Research has demonstrated a strong correlation between the use of problem‐solving (integrative) strategies and outcomes that are of higher quality than those achieved through the use of distributive strategies (Pruitt and Lewis 1975; Rognes and Schei 2010). Negotiation theorists have also stressed the importance of a well‐managed process to achieve the best possible results (Malholtra and Bazerman 2007). A mediation allows for integrative processes, whereas adjudication is more often a distributive process. Court‐connected mediation is carried out in a judicial context, but with the explicit goal of providing comparatively more integrative processes and outcomes than typically would be achieved through adjudication.
In this study, we explored the characteristics of mediation outcomes in court‐connected mediations. Our main research question was: to what extent can court‐connected mediated agreements be characterized as creative? In addition to this, we have explored a possible relationship between creativity in agreements and particular characteristics of the case.
We have applied a “creative product perspective” (Carnevale 2006). This approach is founded on Walton and McKersie's notion of “additional value” and “mutual benefits” (1965), and Mary Parker Follett's (1940) writing about the integrative potential of conflicts, in which the aim of an integrative agreement is to combine the parties' interests in such a way that all parties' aspirations are met. According to this perspective, a creative product is defined as something that has interest, novelty, and value (Simon 2001). In this study, we have categorized agreements that contained one or more substantial elements that were not part of the demands in the court case as creative. Thus, creative elements can be present or absent, and more importantly creativity can be present in varying degrees.
Research on creativity in mediated agreements has been sparse, both in mediation settings in general and in court‐connected mediation specifically. As pointed out by James Wall and Timothy Dunne (2012), most reports have tended to focus on such disputant outcomes as settlement rates, such mediator outcomes as prestige, and such third‐party outcomes as the reduction of court dockets. Wall and Dunne have criticized such research for its lack of empirical data and lamented that the literature on outcomes is “rediscovering the windmill” (2012: 238) (a critique that Wall had raised a decade earlier in a review coauthored with John Stark and Rhetta Standifer in 2001). With this study, we seek to contribute to the empirical research on actual mediation outcomes in an effort to gain greater understanding of what mediation negotiated agreements can and cannot achieve.
Literature Review
Mediation outcomes have been subject to a variety of research methodologies over the past thirty years. Settlement rates and party satisfaction, in particular, have been widely reported on (see Wall and Dunne, 2012 for an overview of the most recent studies). Studies have reported high settlement rates and high satisfaction rates across mediation fields. More recently, studies have addressed the quality of mediated agreements, a dimension that involves not just parties' stated satisfaction with the outcome and the process but that also seeks to measure how well the outcome actually resolved the parties' differences (Poitras and Le Tareau 2009; Baitar et al. 2012).
Few studies, however, have examined the content of agreements. In 1979–1980, Craig McEwen and Richard Maiman studied nearly four hundred mediated and adjudicated small claims cases in Maine by both interviewing parties and observing hearings (McEwen and Maiman 1981). They found that when parties reached agreement via mediation, only 12 percent of the agreements involved conditions other than payment, although respondents in 40 percent of the cases indicated that other issues were present (McEwen and Maiman 1981: 253). This finding contradicted their hypothesis that mediated settlements were likely to be creative in nature.
In another study, an evaluation of child protection mediation found different results, suggesting that the type of dispute matters. In an extensive investigation conducted in five California courts (involving more than eight hundred court files), researchers found that mediated agreements often addressed communication problems and other underlying issues that had not been part of the legal dispute (Thoennes 1997). Although the mediated agreements were similar to judgments in many respects, they included a number of issues beyond the legal dispute. Similarly, in two other studies of court‐deferred mediation in family disputes, researchers found that agreements addressed issues such as, for example, restrictions on contact, responsibility for transport, and location of child exchange, in addition to custody and visitation issues (Beck, Walsh and Weston 2009; Putz et al. 2012).
In a different study of legal disputes embedded in prior relationships, Dwight Golann (2002) reported similar results after examining fifty disputes that were either already being adjudicated or in which there was a threat of adjudication and in which the parties had a prior relationship. The study examined both the effects of mediation on the relationship as well as the integrative result of the agreement.2 Golann found that more than half the agreements reached included elements outside the legal claim.
Several studies have investigated settlement creativity in mediated civil cases with inconclusive results. In a study of court‐connected civil case mediations in nine Ohio courts, Roselle Wissler (2002) found low levels of creativity. The settlement included nonmonetary, or a combination of monetary and nonmonetary, provisions in less than 20 percent of the cases, while the remainder included monetary provisions only (2002: 666). Wissler argues that the high percentage of exclusively monetary provisions reflects the significant number of personal injury cases that are typically settled with monetary provisions only. Contract cases are different, however. Wissler found that nearly 65 percent of the agreements in contract cases included nonmonetary provisions (2002: 667).
Ralph Peeples, Catherine Harris, and Thomas Metzloff (2007) observed a lack of creativity in a small study of medical malpractice cases. And in a more recent study of court‐connected mediation in Finland, Kaijus Ervasti (2011) found that only 20 percent of the mediated agreements contained elements that went beyond the legal claims in the court case.
We have conducted two other studies, however, that have come to quite different results. In a study of 139 mediated agreements in court‐connected mediation, Solfrid Mykland and her colleagues (2009) compared the number of elements in the original claims with the number in the mediated agreements. They found that both the complaint and the answer to the complaint each contained an average of 2.38 demands and the mediated agreements contained an average of 4.53 elements, indicating that issues in addition to the legal claims were addressed in the mediations and incorporated in the agreement. In a study of forty‐two court‐connected mediations, Lin Adrian (2012) found that about half the agreements contained elements beyond what had been identified in the original complaint and answer. She confirmed this finding in an observational study where she sat in on and observed twenty court‐connected mediations over a period of nine months with the permission of parties and mediators (Adrian 2012). These studies suggest that the mediation goal of engaging in a different kind of conversation than would occur in court, resulting in qualitatively different outcomes, seems to have been realized.
Few studies have used the agreements themselves as data for analysis. Most have been based on interviews, questionnaires, and observational notes, etc. — methods that might not adequately capture outcomes. Using the content of the agreements has at least two advantages. First, doing so makes it possible to judge creativity with a higher degree of certainty than would be the case, for example, when relying on self‐reported observations by mediators or parties. Second, it enables the researcher to glean concrete information about the extent and nature of the creative elements themselves, in addition to determining whether or not creativity is present.
For the purpose of this study, we aggregated data from the Mykland study (Mykland et al. 2009) and the Adrian study (2012), developed new codes, recoded the material, and conducted a refined analysis. The purpose of this aggregated study was to accrue more data for analyses, to address in‐depth some unanswered questions in our previous individual data processing and analysis, and to develop a more detailed picture of creativity in court‐connected civil mediation. We have sought to examine which factors in particular seem to contribute to creativity in court‐connected mediation, focusing on four variables: type of disputant, length of mediation, type of case, and the amount of money in dispute.
Type of Disputant
The central actors in a mediation process are the mediator and the disputing parties. The parties' effect on the process has been partially addressed in previous research. In 1955, Henry Landsberger found that labor mediators adopted a pressing style with intransigent parties. James Wall and Suzanne Chan‐Serafin (2009) also found that mediators applied more pressing techniques when the disputants revealed high aspirations.
Wall and Dunne (2012) have argued the need for a greater focus on the dynamics between the parties' behavior and the mediator's, and how such dynamic interactions influence the outcome. To explore this, we would need data about the process itself, which are not available in this archival study. We do, however, have information about the parties and their relationships to each other, and are thus able to explore the relationship between the type of parties involved in the dispute and creativity in mediated outcomes.
Time and Outcome
Integrative processes in negotiation are typically time‐consuming because of the need to share information and to develop mutually acceptable agreements for all parties involved (Malholtra and Bazerman 2007). Researchers have found that greater time pressures in negotiation lead to less effective information processing, more reliance on heuristics in the process, and generally produce less integrative outcomes (e.g., De Dreu 2003; Harinck and De Dreu 2004). Previous research thus clearly indicates that time spent in negotiation has an effect on outcome. Because mediation can be understood as assisted negotiations, it seems likely that there is a relation between time spent in mediation and creativity in outcome.
Type of Case
Wall and Dunne (2012) highlighted the importance of conflict type in the mediation process. Conflict type affects the mediators' strategies and presumably outcomes. Nancy Thoennes (1997) found that mediations concerning children featured greater creativity, and Wissler (2002) found the converse to be true for personal injury cases. Hence, it seems likely that case type affects the degree of creativity.
Amount in Dispute
The size (monetary value) of cases handled in mediation can vary considerably. In 1981, McEwen and Maiman found low degrees of creativity in mediated small claims outcomes, but Wissler (2002) found that 64 percent of contract cases included nonmonetary terms in their settlements. If we assume that contract cases typically involve disputes over a greater amount of money than would be at stake in a small claims case, it is possible that some of the differences in creativity of mediated outcomes could link to case size.
Study Context
The Danish system of court‐connected mediation is, in large part, modeled after the Norwegian system, so the two are very similar.3 Both countries have more than ten years of experience with court‐connected mediation in civil matters — first as pilot programs in selected courts and since 2008 nationwide.4 Courts offer mediation services at both the trial court and appeals level, and mediators are primarily either judges or attorneys, although generally the mediator may not be involved as either counsel or judge should the case proceed to trial. In contrast to most other national systems of court‐connected mediation, judge mediators are active and not retired judges.
Some level of mediation training is a prerequisite for serving as a mediator and mediators are appointed by the court. In Denmark, all mediations are provided free of charge, but in Norway there is a charge for mediations carried out by attorney mediators.5 Consequently, in Denmark, cases are divided almost evenly between those mediated by judge mediators and those mediated by attorney mediators, but in Norway most mediations are conducted by judge mediators.6
Mediation usually takes place early in the litigation process, after the complaint and its answer have been filed. Parties' attorneys may attend mediation sessions if the parties wish, although their presence is not required. Finally, if the case is resolved in mediation, the agreement can be made enforceable.
In terms of cultural context, Denmark and Norway share many features and they rank quite similarly on all five of Geert Hofstede's (2001) cultural dimensions, including power distance and long‐term/short‐term orientation, which are factors that could potentially affect creativity in mediation.7 Power distance refers to the level to which inequalities in a given society are accepted, and both Denmark and Norway have fairly low scores on this dimension. Long‐term/short‐term orientation refers to the importance attached to the future as opposed to the past and the present. Both countries have long‐term orientations attaching value to actions that are important for the future and investment in things to come.
Both Norway and Denmark are Scandinavian welfare states, and each has approximately five million inhabitants. The Danish and Norwegian languages are also similar and each can be readily understood by speakers of the other, which made it possible to work with the data in our study without translation. The justice systems are also similar, and Scandinavia in general has had a tradition of regional legislative cooperation and sharing of ideas — court‐connected mediation is one example. We ran a number of statistical analyses to check for potential differences in the data explained by nationality, and none revealed any significant differences.8
Methodology
The data are merged from two studies conducted in Norway and Denmark, respectively (see Mykland et al. 2009; Adrian 2012), and comprise the written mediated agreement as well as the complaint and answer from 129 civil cases resolved in court‐connected mediation.9 About two thirds of the cases (eighty‐eight) originated from Norway and were collected in 2008, and approximately one third of the cases (forty‐one) are Danish and were collected in 2008–2009.10 We did not include cases involving child custody and visitation in our sample because these cases have different characteristics, both procedurally and substantively, than the typical civil case.
As noted above, combining the Norwegian and Danish data is methodologically sound because the data were produced in similar contexts, and our tests revealed no significant national differences. Our data collection in both settings applied purposive sampling (Frankfort‐Nachmias and Nachmias 1996) because of great variations in mediation activity across courts.
We sought to include courts that had extensive experience with the mediation process, had several mediators participating, and a larger portfolio of mediated cases to choose from. In Norway, we discussed the selection of courts with the Norwegian council for court‐connected mediation and chose the four courts that best matched our criteria. They are situated in different regions, range from small to large, and have average size caseloads. We asked the court to submit the complaint and the answer to the complaint in the thirty‐five most recently mediated cases, as well as the agreement if the case had settled.11
In Denmark, we chose the five courts (four trial courts and one appellate court) that had participated in the national pilot mediation project because they were the only courts that had sufficient experience with mediation at the time of the study.12 The selected courts are among the largest courts in Denmark in terms of caseload, but they resemble other courts in all other parameters. For each court, we sought to obtain the complaint, answer, and mediated agreement for the ten most recently mediated cases and received them for forty‐three cases. The pleading and the answer to complaint, as well as the mediated agreements, were obtained from court records and inquiries to parties and their attorneys.13 Our sample comprises all types of civil cases, and the distribution of cases seems to be representative of the way case types are distributed more generally in court‐connected mediation in both countries.14
A central concept in our study is “creativity.” We have operationalized creativity in a way that corresponds with the creative product perspective (Carnevale 2006); that is, agreements that contain one or more substantial elements that were not part of the original demands filed with the court case are considered creative. We argue that these substantial elements are new and add value to the agreement.
The data set is unusual because it enables comparison between the parties' stated legal demands (their positions) from the court filings with the mediation outcome as represented by the agreement. Thus, we are able to determine whether the parties' agreements match their original demands or contain additional elements that would not typically be included in a judgment.
To determine the degree of creativity, we coded and tallied each element that added value beyond the claims. Examples of such elements are delivering work or services beyond those noted in the claim, paying interest at unusually high or low rates, apologizing, and including contingencies (e.g., if the work is not fulfilled within a specific date, new conditions are set).
Examples of elements that we do not consider to be creative are those that were explicitly or implicitly included in the legal claim or are automatically included in a judgment, for example, payment of interest at standard rates, usual division of legal costs, regular payment deadlines, and “full and final settlement” formulations.
We developed the following scale, which provides a rich descriptive picture of the amount and distribution of creativity in the data set:
Code One: only one party's claim was met in the agreement.
Code Two: the parties' agreement fulfilled neither party's claim, but fell somewhere in between.
Code Three: the agreement contained one element beyond the claims included in the case.
Code Four: the agreement contained two to four elements beyond the claims included in the case.
Code Five: the agreement contained five elements or more beyond the claims included in the case.
Applying our coding system required familiarity with the judicial systems, specifically the definition of a legal claim, which we have. Understanding the implicit aspects of a legal claim, for example, rules and conventions surrounding interest, is particularly important.
We began by coding twenty randomly chosen settlements from our data set. To assure interpersonal rater reliability, we compared and discussed our criteria and agreed on how to code the different elements. Based on this system, we each evaluated half the data set and noted any codings we were unsure about, which we subsequently discussed.
The variables' categorical character, both in the independent and dependent variables, affects their suitability for different statistical analyses. Cross‐tabulation analyses are the most common for such variables, but logistic regression analyses are also possible. Therefore, we used both methods in our analyses.15 The number of categories for the independent variables (case type, types of party, case size, and length of mediation) combined with the number of codes for the dependent variable (creativity) made the total number of observations in each cell limited when using the creativity scale with five levels. When we analyzed creativity and type of party, we had a total of fifteen cells in the cross‐tabulation (three categories of parties and five levels of creativity), which made it difficult to gain five or more observations in each cell, which is the statistical criteria for using cross‐tabulation analysis. Because our 129 observations are not evenly distributed over the fifteen cells, these criteria were difficult to meet. Hence, to facilitate cross‐tabulation analyses, we developed a second scale that divided the cases into two groups. The first group (now labeled Code One) comprises settlements that showed no or very limited creativity (agreements containing one or fewer extra elements, Codes One to Three in the original scale). The second group (now labeled Code Two) comprises settlements that show greater creativity (agreements containing two or more extra elements, Codes Four to Five in the original scale).16 We argue that this reduction still takes into account the constructive validity, although the codes are not so fine‐tuned.
Findings
Applying our first, five‐category scale of creativity, 65.1 percent of the agreements contained creative elements, as shown in Table One.
Distribution of Creativity within Five Categories
Creative Element . | Frequency . | Percentage . |
---|---|---|
One party's claim is met | 5 | 3.9 |
Between claims | 40 | 31.0 |
One extra element | 17 | 13.2 |
2–4 extra elements | 34 | 26.3 |
5+ elements | 33 | 25.6 |
Total | 129 | 100.0 |
Creative Element . | Frequency . | Percentage . |
---|---|---|
One party's claim is met | 5 | 3.9 |
Between claims | 40 | 31.0 |
One extra element | 17 | 13.2 |
2–4 extra elements | 34 | 26.3 |
5+ elements | 33 | 25.6 |
Total | 129 | 100.0 |
The level of creativity varied significantly. Just over 13 percent of the cases contained just one element beyond the legal claim, whereas about half the cases (52 percent) had two elements or more. Roughly one quarter of the cases (26 percent) contained five or more creative elements and can be thus categorized as highly creative. More traditional compromises, in which the parties settled on an amount that fell between their initial claims, constituted just under one third (31 percent) of the cases. Finally, in 3.9 percent of the cases, the settlement agreement fulfilled the claim of only one of the parties.
Having established that court‐connected mediation does indeed produce creative outcomes in many instances, our next step was to identify which factors could account for the presence of creativity in some mediations but not in others. We then examined the following four independent variables: the type of party, the duration of the mediation, the type of case, and the size of the claim.
For these analyses, we used our second (dichotomous) scale to generate more statistically reliable results. The distribution of the mediated agreements on this scale is shown in Table Two.
Type of Party and Creativity
The parties to the mediations we analyzed appeared in three different combinations: private individual versus private individual, business entity versus business entity, or private individual versus business entity (mixed cases). In a cross‐tabulation that analyzed creativity according to the types of party involved, we found a statistically significant pattern17 (see Table Three).
Type of Party and Creativity
Type of Party . | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Private to private | 14 (35.9%) | 25 (64.1%) | 39 (100%) |
Mix | 32 (47.8%) | 35 (52.2%) | 67 (100%) |
Business‐to‐business | 16 (69.6%) | 7 (30.4%) | 23 (100%) |
Total | 62 | 67 | 129 |
Type of Party . | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Private to private | 14 (35.9%) | 25 (64.1%) | 39 (100%) |
Mix | 32 (47.8%) | 35 (52.2%) | 67 (100%) |
Business‐to‐business | 16 (69.6%) | 7 (30.4%) | 23 (100%) |
Total | 62 | 67 | 129 |
As shown in Table Three, nearly two thirds (64 percent) of the settlements in the private versus private category included two or more extra elements in the settlements, compared with only just under one third (30 percent) of the business versus business disputes. The mixed party mediations attained this level of creativity in slightly more than half (52 percent) of the cases.
Duration of Mediation and Creativity
The average duration of the mediations in our sample was three hours. To examine possible relationships between time spent in mediation and creativity, we grouped the cases into those with durations shorter and longer than the average. The mediation sessions that lasted more than three hours were more creative than the sessions that lasted less than three hours, and our cross‐tabulation analysis revealed this difference to be statistically significant18 (see Table Four).
Duration of Mediation and Creativity
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Less than three hours | 27 (65.9%) | 14 (34.1%) | 41 (100%) |
Three hours or more | 26 (35.1%) | 48 (64.9%) | 74 (100%) |
Total | 53 | 62 | 115* |
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Less than three hours | 27 (65.9%) | 14 (34.1%) | 41 (100%) |
Three hours or more | 26 (35.1%) | 48 (64.9%) | 74 (100%) |
Total | 53 | 62 | 115* |
*N = 115 because we did not have information on the duration of mediation in the remaining fourteen cases in our sample.
As is shown in Table Four, nearly two thirds (65 percent) of the mediations that lasted more than three hours included two or more extra elements in their agreements, but this was true for just over one third (34 percent) of the mediations that lasted less than three hours. For agreements containing no or one extra dimension, the pattern was reversed: two thirds (66 percent) of these lasted less than three hours, while just over one third (35 percent) lasted longer than three hours.
Type of Case and Creativity
To examine whether the type of dispute was connected to the level of creativity in the mediated outcome, we divided the cases into four dispute categories: workplace disputes, disputes regarding inheritance or division of property after divorce, contractual disputes, and property disputes.19 We found statistically significant differences between the case types, as shown in Table Five.20
Type of Dispute and Creativity
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Workplace disputes | 7 (33.3%) | 14 (66.7%) | 21 (100%) |
Inheritance/divorce disputes | 5 (27.8%) | 13 (72.2%) | 18 (100%) |
Contractual disputes | 37 (62.7%) | 22 (37.3%) | 59 (100%) |
Property disputes | 8 (34.8%) | 15 (65.2%) | 23 (100%) |
Total | 57 | 64 | 121* |
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
Workplace disputes | 7 (33.3%) | 14 (66.7%) | 21 (100%) |
Inheritance/divorce disputes | 5 (27.8%) | 13 (72.2%) | 18 (100%) |
Contractual disputes | 37 (62.7%) | 22 (37.3%) | 59 (100%) |
Property disputes | 8 (34.8%) | 15 (65.2%) | 23 (100%) |
Total | 57 | 64 | 121* |
*Our sample size here is smaller (N = 121) because we left out the personal injury and libel cases.
Agreements settling cases involving inheritance and divorce contained the highest level of creativity — nearly three quarters (72 percent) of these contained two or more extra elements. Workplace (slightly more than two thirds or 67 percent) and property (nearly two thirds or 65 percent) disputes closely followed. Agreements to contractual disputes contained the least amount of creativity, with just over one third (37 percent) containing two or more extra elements.
Amount and Creativity
We also looked for a relationship between the amount of money in dispute and creativity. To do so, we divided the cases into three different financial categories: those involving amounts up to 50,000 kroner (roughly $9,000), 50,000–500,000 kroner (roughly between $9,000 and $91,000), and above 500,000 kroner. (Danish and Norwegian kroner have almost a one‐to‐one value.)
The results shown in Table Six suggest that when the amount at issue in a dispute increases, so does the likelihood that mediation agreements will demonstrate greater creativity. The most creative settlements were attained in the largest value cases. In this category, just over half (56 percent) of the settlements included more than two extra elements. The fewest number of creative agreements were found in the cases with the lowest amount of money in dispute, with just under two thirds (65 percent) of those agreements containing zero to one extra element. Our statistical analysis failed to confirm any statistically significant differences in this category, however.21
Amount in Dispute
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
0–50,000 kr. (∼$0–$9,000) | 11 (64.7%) | 6 (35.3%) | 17 (100%) |
50,001–500,000 kr. (∼$9,000–$91,000) | 33 (56.9%) | 25 (43.1%) | 58 (100%) |
500,001 kr.+ (∼$91,000+) | 12 (44.4%) | 15 (55.6%) | 27 (100%) |
Total | 56 | 46 | 102* |
. | No or One Extra Element . | Two or More Extra Elements . | Total . |
---|---|---|---|
0–50,000 kr. (∼$0–$9,000) | 11 (64.7%) | 6 (35.3%) | 17 (100%) |
50,001–500,000 kr. (∼$9,000–$91,000) | 33 (56.9%) | 25 (43.1%) | 58 (100%) |
500,001 kr.+ (∼$91,000+) | 12 (44.4%) | 15 (55.6%) | 27 (100%) |
Total | 56 | 46 | 102* |
*Our sample size here is smaller (N = 102) because in twenty‐seven cases the claim did not involve money.
Additional Analyses
Our cross‐tabulation analyses confirmed bivariate correlations between all but one of the independent variables and the dependent variable. The only independent variable for which we did not find a significant effect was the amount of money at issue in the dispute. We also ran a logistic regression analysis to explore the independent variables' direct effects on the dependent variable (creativity), but found that additional statistical information beyond the correlations already gained from the cross‐tabulations was limited.22 Our regression analysis did confirm that more time spent in mediation resulted in a higher degree of creativity, and that two different case types — divorce and inheritance disputes and property disputes — were more likely to produce mediation agreements with higher degrees of creativity.23
Discussion
This study contributes to the rather limited body of empirical research that has examined the content of mediated agreements and expands our knowledge of mediated outcomes and, in particular, factors related to creativity in mediation outcomes. The results suggest that creativity has an important role to play in mediation. In our sample, almost two thirds of the mediated agreements included elements that were not included in the parties' legal claims; hence, by our definition, they are creative.
Although we define any agreement with elements beyond those identified in the court filings as creative, it is clear that those agreements with just one extra element, such as a specially tailored payment plan, differ from those that include a variety of multifaceted solutions to a problem, such as providing alternative access to a property, making specific plans for achieving this access, and dividing the costs associated with the plan. Complex agreements with five extra elements or more occurred in just over a quarter (26 percent) of our sample, and these were often quite detailed and intricate agreements that are significantly different in character from judgments. Because these agreements embody the creative solutions that many mediators strive for, we believe our results have important implications for both practice and further research.
Our logistic regression analysis indicates that time spent in mediation affects the level of creativity in the agreement. This makes intuitive sense. Obtaining creative outcomes in mediation is likely to involve thoughtful consideration and discussion of parties' underlying interests and needs, which can be a time‐consuming process. It also takes time to develop the kind of trust that allows the parties to talk openly about their interests and needs and other concerns they may have.24 Finally, it takes time to craft alternative solutions and consider all their implications.
Our finding of a relationship between types of parties and the presence of creativity in the agreement also makes sense. We found a higher incidence of creativity for cases involving private parties than those between businesses. For a private party, more is often at stake in a dispute than for a business. Private parties are usually one‐time players in court, and the stress of litigation may serve as an incentive to work harder to settle. Businesses, in turn, are more often repeat players in court, and involvement in disputes is often considered an inevitable side effect of doing business. In addition, businesses can often write off settlement amounts as a loss, whereas private parties must pay out of their own pocket. Hence, private parties may be more risk‐averse than businesses (Mnookin 1993), which could offer stronger incentives for settlement and for expending greater effort to find creative solutions.
Framing theory (e.g., Dewulf et al. 2011; Druckman and Druckman 2011) may also explain the higher incidence of creativity in agreements between private parties. How disputants and mediators frame the conflict can affect the scope of the conversation and possible solutions (Adrian 2012). The type of party (individual or business) could affect the framing. Private parties may have a greater range of interests and needs at stake than businesses, and consequently have a greater interest in developing alternative and possibly nonmonetary solutions. The mediator may consequently focus greater attention on these more creative options.
Framing may help explain our results with respect to the type of cases as well. In the cross‐tabulation analyses, we found that agreements resolving inheritance and divorce disputes, property disputes, and workplace disputes featured greater creativity, but our logistic regression analysis found a direct effect only for inheritance and divorce disputes, and property disputes. Mediators and parties may apply a wider frame to these kinds of disputes because disputants in these cases are more likely than those in contractual disputes to have interests in addition to money.
We also identified a pattern (although not a statistically significant one) with regard to the monetary amount in dispute: the greater the amount, the higher the level of creativity. More is at stake for parties involved in cases involving larger sums of money — especially for the defendant — and this may stimulate creativity.
In about one third of the cases in our study, we found no evidence of creativity in the outcome. In such cases, the agreement most commonly constituted a compromise between the demands of the two parties, although in some rare cases just one of the parties prevailed. We cannot know whether this reflects an actual absence of creative potential, or whether the mediator and the parties were simply unsuccessful in releasing potential creativity. In previous studies of mediations in the same setting (Mykland 2011; Adrian 2012), however, we found untapped creative potential and substantial variation between the mediators' efforts to both explore the parties' interests and to develop creative solutions, which suggests that creative potential may have existed even in those cases whose agreements included little or no creativity.
Research Limitations
Our research had several limitations, including our small sample size. Our reduced two‐step creativity scale helped capture interesting relationships between creativity and our four independent variables, but a larger sample would have made it possible to use our five‐step scale, which would have provided additional interesting information. In our analyses, this would specifically apply to the amount of money variable because the sample size for that variable (102) was particularly small.
In addition, our analyses suffer from omitted variable bias. Because of the small sample size and the limited number of variables, we have been unable to establish a control for how our variables affect each other and for other intervening variables. As noted, our logistic regression analysis indicates direct effects only for two of our variables, length of the mediation and type of mediation. We have shown an association between the other independent variables and creativity, but further studies with larger samples would enable us to understand more fully the effect of all these mediation variables on creativity in outcomes.
Future Research
Our findings suggest directions for future research. As noted above, a larger‐scale study could examine additional independent variables with greater precision.
In addition, the effects of creativity deserve attention. What is the relationship between creativity and parties' satisfaction with the process and the outcome? Is there a relationship between creativity and compliance with the agreement?
Studying creativity by analyzing agreements (creative product perspective) has obvious limitations. A focus on outcomes does not address the creative process (Carnevale 2006). It fails to ask and answer the question: what exactly did the mediator do to generate creativity in the mediation? Which specific mediator and party behaviors promoted or hindered creativity? Did it for example, make a difference whether the mediator had a facilitative or evaluative approach (Riskin 1996) and did the problem definition matter (Riskin and Welsh 2008)? Finally, gaining a complete picture would require focus on the participants' perspectives: which personal characteristics of mediators, parties, and attorneys might predict creativity in the course of the mediation and in the outcome (e.g., Kolb 1997).
Practice and Training Implications
Our study has implications for mediation practice, especially for mediation in legal settings. It indicates that creative agreements are possible, which could help encourage mediators to work more deliberately to achieve creative resolutions. The exact steps that mediators can take to promote creativity within mediation remain largely undefined, although this research suggests that time is a critical element.
For trainers, the study suggests that creativity is a topic that deserves attention in mediation training. Although further research is required to determine exactly which mediator practices and techniques promote creativity, the texts of highly creative agreements have obvious pedagogical value, and their close examination deserves a place in the mediation training curricula, perhaps accompanied by commentary provided by the mediators who helped write them.
Conclusion
Our study supports the assertion that mediation can produce creative solutions to disputes that are not available through adjudication. It also confirms how creatively dynamic mediation outcomes can be, with one quarter of the agreements we examined containing five or more creative elements. Our identification of those factors associated with creativity offers insight to mediators as they seek to assess a dispute's creative potential. The challenge for future research is to determine what exactly mediators can do to exploit the potential for creativity that each resolution may or may not contain.
NOTES
A result was coded as integrative if the agreement included at least one item beyond payment, release of liability, and confidentiality, and if the mediator estimated that the agreement had significant value to one or more of the parties (Golann 2002: 308).
Court‐connected mediation is regulated through the two countries' respective Administration of Justice Acts.
The parties must pay the initial court fee before mediation is offered.
Mail from the National Court Administration, May 7, 2014, on file with author.
See Exhibit A5.1 in Hofstede (2001). The dimensions are power distance, uncertainty avoidance, individualism/collectivism, masculinity/femininity, and long‐term/short‐term orientation.
A chi‐squared test using a p‐value of 0.10 found no significant difference between the Norwegian and the Danish part of the sample on the dependent variable, creativity.
Originally, we had 139 cases, but 10 cases were incomplete and did not have the necessary information available for the purpose of this study (e.g., missing pleading or answer). The incomplete cases apparently did not deviate in any other way from the study's sample.
There has been no change in rules and practices of court‐connected mediation in Norway or Denmark since the data were collected.
Within the three courts that submitted their most recent mediated cases, the settlement rate was 73 percent. One court submitted their 35 most recent settled cases by mistake, but as the purpose of this study is to analyze the content of the agreements we included these in the sample.
We compared agreements from the trial courts with agreements from the appellate court and found no difference. Hence, we include data from both settings in the study.
In the remaining seven cases, obtaining some or all of the necessary documents was unsuccessful either because the parties did not want to participate, the attorneys did not want the researcher to contact the parties, or it was not possible to get in touch with relevant parties. These cases resemble the other cases in the study, and there is nothing to suggest a selection bias. Two cases were subsequently excluded from the study as the agreements were entered a year or more after the mediation meeting, and thus could not conclusively be considered a result of mediation.
We have to rely on the evaluation reports of the pilot projects in the two countries because there is no other statistical information available.
The regression analyses produced only marginal additional information; therefore, most of the article is based on the cross‐tabulation analyses.
The same categorization of the dependent variable was used in the logistic regression analysis.
A chi‐squared test using a p‐value of 0.05 showed statistical significance; chi‐square = 0.037, df = 2 and value = 6.575.
A chi‐squared test using a p‐value of 0.05 showed statistical significance: chi‐square = 0.002, df = 1, value = 10.020.
We left out cases that appear infrequently in the study for statistical purposes. These were personal injury cases (seven) and a libel damages case (one).
A chi‐squared test using a p‐value of 0.05 showed statistical significance: chi‐square = 0.009, df = 3, value = 11.466.
A chi‐squared test using a p‐value of 0.05 showed chi‐square = 0.378, df = 2, value = 1.946.
We computed dummy variables for all the independent categorical variables.
The duration of mediation showed significant effect on creativity at the p‐level of 0.05, the case types, inheritance and divorce disputes, and property disputes both showed effect on creativity at p‐level of 0.10. The chi‐square for the model was 19.77; the Cox and Snell R2 = 0.158 and Nagelkerke R2 = 0.211.
Anecdotal evidence from a district court in Norway shows that when the mediators decided to continue all mediations until after lunch, the rate of agreements rose. Likewise, it is plausible that the rate of creativity rose simultaneously.