Mediation in civil justice: international evidence review

Outcomes for Users and for the Civil Justice System

Uptake of Mediation

The evidence reviewed for these jurisdictions suggests that mediation uptake may be encouraged through some level of mandatory process and through use of low fee or free services. However, uptake – or 'levers' to encourage mediation – are a recognised gap in the literature and need further research and exploration.

The Civil Justice Council in England's 2018 report on ADR and a recent EU-wide report – including England and Wales and Ireland – have both recognised that voluntary mediation processes have low uptake.^[42] Therefore, some level of mandatory mediation process was seen to be the best way of increasing uptake in both reports.^[43] Key informants in Australia also felt that mandatory systems are the only way to increase uptake. There is some debate on this in the literature, with some authors arguing that mandatory mediation should be a short-term measure to increase uptake before a voluntary process is returned to. There is only limited guidance on how that process would work in practice.^[44] Experience in Alberta shows there may be some credence to this viewpoint though, as the mandatory system of JDR – sometimes referred to as mediation – in Alberta's Queen's Bench court (the higher court) had to be suspended due to judicial workload, but has left a legacy of reasonably strong voluntary requests for JDR due to increased knowledge and understanding of this system.^[45] Some Australian key informants felt mandatory mediation was leading to wider voluntary uptake also, though there were conflicting views here.

A popular model for mandatory mediation is the 'Notice to Mediate' procedure used in British Columbia, which mediators in the sector believe may be producing a wider culture of mediation.^[46] As this model is very popular in discussions of good practice around embedding mediation it will be briefly outlined here. If all parties agree voluntarily to mediate the Notice to Mediate process is not needed, but if one party wishes to compel another to mediate, this process will be used. Courts can also compel parties to mediation using the Notice to Mediate. There are some exceptions to cases that can be mediated, but it is generally expected that general civil cases will attempt mediation before going to court – British Columbia has an opt-out system^[47] – however the Supreme Court notes that, since 2001, they only have one request for exemption and this was from the Crown on the basis that mediation was not appropriate.^[48] A Notice to Mediate may be served 'no earlier than 60 days after the filing of the first response to civil claim in the action and no later than 120 days before the date of trial'.^[49] According to mediator opinion, the BC system is becoming increasingly informal as it becomes more normalized.^[50] This is put down in part to the fact the Notice to Mediate can be served without court interference and that this means parties take responsibility for mediation themselves and are then encouraged to use mediation in future.

As will be discussed below, mandatory processes of referral do not necessarily lead to better outcomes for users. The evidence only shows increased uptake. Quek also argues that mandatory processes need careful thinking to ensure parties' rights to the full justice spectrum are maintained and that there are careful guidelines for what cases can be referred and how opt-out works, recommending following Florida's process of providing criteria for opt-out, to reduce judge discretion.^[51]

Key informants also stressed the need for low cost or free mediation services to ensure uptake. If mediation is prohibitively expensive then the majority of users will not be able to make use of the process. Maryland, for example, does not allow mediation to be mandated if a party cannot pay – costs are factored into the case for whether mediation is appropriate or not. As noted in the operational section of the report, there are a variety of systems of funding and certain courts may be able to provide free services. Community mediation, in the United States, is a free service, while private mediation will require payment.

Beyond these levers to uptake there is a gap in the literature around what encourages engagement with mediation and then settlement; further research is required.^[52]

Access to Justice: Procedural Justice and Substantive Justice

A key message that emerged from the evidence across all jurisdictions was that mediation scores highly on procedural justice outcomes.

Mediation scores highly on procedural justice outcomes such as fairness of process, mediator lack of bias, feeling heard during the process, being able to tell one's story, and positivity about using mediation again. Wissler^[53] reviewed several studies from Ohio that investigated user outcomes of mediation. This study made it clear that mediation produced strongly positive procedural outcomes. Litigants were highly positive about the mediator and the process of mediation, with 72% stating that the process was fair and 84% that they had a chance to tell their own story. Ninety-seven percent noted that the mediator treated them with respect and 63% that they felt they had 'considerable input' into the outcome. Seventy-nine percent would recommend mediation. Attorneys were equally as positive, with 89% saying the process was fair, 97% stating that the mediator was impartial, and 88% that they were effective in engaging the parties in meaningful discussion. Eighty-five percent would recommend mediation to their colleagues for similar cases.

In a study by Charkoudian, Eisenberg, and Walter on ADR in Maryland (including, mediation and settlement conferences, with 88% of the study participants being referred to mediation) it was found that it, in comparison to those who went to trial without ADR, those who used ADR experienced immediate positive impacts, which were described as 'significant'. These impacts were experienced if parties settled their dispute through ADR or had to go to trial but had been through ADR beforehand – it seems the positive benefits are experienced simply by going through the process.^[54] The short-term impacts reported are as follows:

ADR participants were more likely than those who proceeded through the standard court process to indicate that: (1) they could express themselves, their thoughts, and their concerns; (2) all of the underlying issues came out; (3) the issues were resolved; (4) the issues were completely resolved rather than partially resolved; and (5) they acknowledged responsibility for the situation. Importantly, this was true for all ADR cases, including those that reached an agreement in ADR and those that did not settle.^[55]

This procedural positivity was replicated in other areas, with Ontario's mandatory mediation programme evaluation – conducted after 23 months and used to justify the continuation of the programme in Toronto and Ottawa, and then expanding it to Windsor – finding that 82% of Ottawa litigants and 65% of Toronto litigants were satisfied with the overall mandatory mediation experience.^[56] In Ottawa 61% of litigants agreed that justice was served by mediation, with a rather more mixed picture in Toronto, but nevertheless the largest category (39%) of litigants agreeing that justice was served. Eighty-eight percent of litigants in Ottawa would use mediation again and seventy-three percent in Toronto also, with similarly positive responses from lawyers (86% in Ottawa and 66% in Toronto).

The Community Justice Centre (CJC) of New South Wales also provides some useful satisfaction and perceptions of access to justice statistics from their work, showing positive feelings about mediation:

What clients say about CJC after attending mediation:

95% were satisfied with the assistance CJC provided in resolving their dispute
98% found the staff and mediators helpful and courteous
79% said mediation with CJC helped with their situation
94% said CJC mediators understood their concerns
98% stated they felt they were able to have their say in mediation
83% stated they were provided with useful information before the mediation
85% would use CJC again
88% would recommend CJC to others.^[57]

Rates may vary depending on the programme being run, but overall, in Australia, from the 'existing survey research, it seems that dispute resolution processes can provide greater satisfaction to disputants [than litigation] in many instances'.^[58] Throughout Mack's National Alternative Dispute Resolution Committee (NADRAC) report high satisfaction levels are reported from a number of studies and evaluations, ranging from around 60-80%.^[59] Although Sourdin reports studies with satisfaction rates ranging from 36-59%. This range depended on exactly what parties were asked: they were less satisfied with the mediation outcome, but were more satisfied with mediation as a process and overall.

In conversation with Australian academics, they argued that mediation can provide 'personal justice' and 'procedural justice'.^[60] Mediation can provide a solution to parties' disputes much more quickly than litigation, allowing parties to get on with their life and providing relief. It can be a process of more creative solutions, which allow parties to make agreements on areas and issues that the law cannot touch. This can mean a more satisfying, win-win scenario and that different kinds of conversation can be had, equivalent to a day in court, but maintaining relationships. This kind of comment on the potential positives of mediation was made by the majority of key informants across all jurisdictions, though with some notes of caution that will be outlined in this report.

Finally, looking at another jurisdiction, the Small Claims mediation service in England and Wales receives positive feedback, with 94.4% of clients whose cases settled stating they would use the service again, and 85.9% of those whose cases did not settle saying they would use the service again (these statistics reflect responses from 2007-2011).^[61]

However, the evidence does not present a purely positive picture: while satisfaction levels are high in terms of process, they are not necessarily as high in relation to the outcome. There is a lack of evidence around causality on this issue, so debate remains speculative. For example, in the same study above by Wissler – which was a review of several studies conducted in Ohio – though process was viewed very positively, with 79% stating they would recommend mediation to others, a rather smaller percentage (55%) felt satisfied with their outcome in the present case. Wissler is not fully clear why this drop in percentage occurs, as none of the studies reviewed investigate this. Those who settled were likely to be more positive about mediation that those who did not, showing that outcome does have an impact on how people view their general experience, however those who did settle were also more likely to report pressure to settle from the mediator or from others involved.

The Ontario Mandatory Mediation evaluation asked whether the settlement was fairer in mandatory mediation than it would have been going through another process of resolution (left unspecified), and the results were mixed. In Ottawa, 41% agreed, with 16% disagreeing and the rest unsure; this was improved when lawyers were asked, with 50% agreeing it was fairer than other processes, 18% disagreeing, and the rest unsure. Toronto litigants were unsure in 49% of cases, with 30% agreeing, and the rest disagreeing; lawyers provided similar results, with 48% unsure, 31% agreeing, and the rest disagreeing.^[62] It should be noted that Toronto as a city was less used to mediation than Ottawa at the time, which the evaluators note, may have an impact on the results.

In a study conducted in Queensland's Community Dispute Resolution Centres, 1339 users took part in a survey. Over the studied period of 1999-2003, 81.5% reached a written or verbal settlement during mediation, with a further 18.6% reaching partial settlement, on average across the 5 years.^[63] There is a mixed picture for satisfaction with outcomes, with participants in the study being asked whether their agreement had addressed the concerns they raised in mediation – only in results for the year 1999 did more people respond 'yes' than 'no'. Results for the other four years of study were not so positive – between 57.7% and 70.4% reported that their agreement had not addressed these concerns. Nevertheless, all years under study show a strongly positive response to the question 'Did you feel heard [in mediation]?', with between 78.4% and 83.6% stating that they did so. This suggests that the process of mediation was a positive one, but that the final outcome may not have felt as satisfactory. These agreements have nevertheless, in the majority, been sustained.^[64] These results chime with findings from across the jurisdictions.

Though not one of our specific state jurisdictions, one of the most significant studies of the impact of mediation on court processes ever undertaken – and therefore important to cite here – looked to explore a variety of case management principles that were put forward by the Civil Justice Reform Act in California in 1990. These principles required more action on ADR in federal district courts, intended to save costs and time. These studies were divided up into a number of different themes and results published accordingly.^[65] Participants in this study – lawyers and litigants alike – appeared to like ADR processes and the pilot raised awareness of mediation and other ADR amongst lawyers and the judiciary. However, there are more mixed outcomes for users and the system otherwise, with positive outcomes more focused on case management by judges than use of a specific form of ADR:

This report discusses the effects of the CJRA [Civil Justice Reform Act] case management principles on time to disposition, costs, and participants' satisfaction and views of fairness. The study found that the CJRA's package of case management policies, as it was implemented, had little effect on any of these outcomes. However, what judges do to manage cases does matter. A package of procedures containing early judicial management, early setting of a trial date, and shorter time to discovery cutoff could reduce time to disposition by 30 percent, with no change in litigation costs, satisfaction, or perceived fairness.^[66]

Though our review is not systematic, these results suggest that mediation provides strong procedural justice, but may not provide as strong satisfaction around the settlement itself.

Settlement does not necessarily reflect a purely positive outcome – with the potential for a feeling of being pressured to settle – and should be reviewed in relation to other outcomes. However, as much of the empirical research on mediation does focus on settlement, this data will now be examined.

Settlement Rates

Settlement rates varied across our jurisdictions. It was difficult to pin down exact reasons for this variation, but some broad enablers and barriers can be identified.

The evidence on settlement rates extends across a varied type and quality of evidence, and this should be borne in mind. Settlement rates in mediation are highly variable, with the majority in this review falling between 40 and 90%, with some rates below or above. The range may relate to the context-specific nature of programmes and pilots, but without strong evidence about what encourages or prevents settlement this remains speculative. There was no jurisdiction with higher or lower rates than any other, but variety across jurisdictions and between programmes.

As an example of this variability, in a large review of US studies on mediation (general civil cases), Wissler found most studies reported between 27 and 63% settlement rates, but two reported a lower rates of between 13 and 22% and two a higher rate of 71-80%.^[67] However, to illustrate the importance of looking at settlement in connection with other outcomes, this major review of US data presents a mixed picture of mediation outcomes overall, with it being unclear whether mediation relative to trial improves settlement rate, increases compliance, reduces time to disposition, or actually reduces trial rates; cost savings and perceptions of savings are also varied.^[68] It is clear mediation settles cases and people feel the process is fair, but in comparison to litigation the picture is less clear.^[69] The same review found no conclusive evidence around whether the method of referral made any difference to outcome either.

However, Maryland specific research contradicts this on some measures, finding ADR (the majority of participants went to mediation) was popular with parties in terms of fairness, saved resources, and increased confidence in the courts just by going through the process, in the short and longer term, whether or not settlement was reached and litigation required.^[70] This research noted that mediation can provide more creative solutions to disputes, and have wider outcomes such as self-empowerment, improved relationships, and increased skills to handle disputes in the community. Both the Maryland based research and Wissler's review found improved relationships between parties overall. There are, therefore, important outcomes that require consideration alongside settlement.

As a Supreme Court contact in Florida stated, although they have an experiential sense of positive settlement outcomes, such simple, quantified outcomes without other types of information may not be useful in the long-term:

We do not require the court ADR programs to report the settlement rates of mediations or any type of consumer satisfaction data. One reason for not requiring them to report settlement rates is that we do not want the success of mediation to be evaluated only on the settlement rates. Our experience is that of the cases mediated about 50% of appellate and circuit court cases settle, 60-75% of family law cases settle, and 70-75% of county and juvenile dependency cases settle.^[71]

Similarly, Noce et al. argue that qualitative measures are better for an understanding of mediation than quantitative measures. They argue that qualitative questions can truly get at what is 'good' about mediation for all parties and quantitative data is rarely of useful quality.^[72] The current evidence base can provide few certainties about mediation, given the complexity of carrying out meaningful, quality, empirical research.

Barriers to and Enablers of Settlement

Predicting settlement in mediation is very complex and may come down to differences on a case by case basis.

Barriers to Settlement

As Genn argues, predicting settlement in mediation is very complex, and may come down to 'personalities, depth of grievance, degree of conflict, willingness to negotiate and compromise'.^[73] These factors are very difficult to quantify and change with each and every case. There is therefore only a limited amount of evidence that explores barriers and enablers to settlement, but this will be explored here: it should be kept in mind that these are broad brush comments and cannot be guaranteed to help or hinder mediation settlement in every case.

Looking at Small Claims examples, where mediation tends to do well in terms of settlement, a study in British Columbia – which made use of available court data on 4136 mediations of Small Claims cases that took place between 1998 and 2004 – found that:

56% were fully settled at mediation, 2% were partially settled at mediation, and 42% were not settled at mediation.^[74]

In this study the barriers to settlement were being investigated from the mediators' perspective. Mediators rated 'unrealistic expectations' on the part of parties as the most significant barrier (69%), with lack of legal representation coming second (39%). These results will reflect the sample bias, but do chime with other research across the jurisdictions that stresses the need for education around mediation and the complexities that can arise for users who are self-represented and may need advice to achieve settlement.

This author goes on to state that case type might have had an impact also:

In this study, real estate cases had the lowest settlement rates overall, and mediators' barrier ratings provided some insight as to why this might be the case. Unsettled real estate mediations were rated substantially higher than other dispute types on the expertise/authority barrier scale and on all three items that compose that scale (i.e., 'lack of authority to settle', 'lack of representation/legal advice', and 'missing a necessary participant').^[75]

Genn notes that active coercion to mediate tends to be a barrier to engagement and settlement and that 'facilitation [and] encouragement… with selective and appropriate pressure' are better methods of increasing both willingness to engage and to finding a solution.^[76]

Enablers of Settlement

As argued in the last section, case type may make a difference to settlement. The evidence is not always clear on what case types are being referred to when settlement is discussed, but there is some evidence that divides civil cases up clearly, which will be drawn on here to discuss enablers to settlement. Small Claims cases appear to be particularly suited to mediation, with good settlement rates across jurisdictions.

The Small Claims mediation service for England and Wales, based in Northampton, sets its settlement targets at a minimum of 62%: this would be seen as a good rate of settlement.^[77] In general, they find that cases with 2 companies, rather than 2 consumers or 1 consumer and 1 company, are more likely to settle, as are those with smaller amounts in dispute.^[78] The staff here feel that most of their clients want resolution and that mediation saves the cost, time, and stress of a court case, allowing parties to move on with a resolution that suits both sides.^[79] A Manchester County Court pilot saw 90% settlement rates for a similar telephone mediation service, so there is potential for strong settlement rates via a distance service, which reduces certain costs.^[80] Gill et al.^[81] note that there has been some debate about whether distance mediation is 'true' mediation, as it takes place as a relay set of phone conversations between parties – meaning parties do not meet and discuss together. There are therefore both positive and negative voices around these kinds of schemes.

In Alberta, the Provincial Court – dealing with claims of less than $50,000 – has seen a 67% settlement rate, according to one local expert, which is deemed to be very successful.^[82] In Ontario, Small Claims cases that were mediated also have good settlement rates. Vidmar's study^[83] looked at reasons for compliance in these cases, focusing on the level of admitted liability. In full liability cases, there was an 89% settlement rate and in no liability a 43% settlement rate, with 69% for partial liability. Compliance was between 93 and 100%. However, Wissler has disputed Vidmar's claims^[84], findings that admitted liability made no difference to settlement. However, parties saw mediation as fair and as more fair than trial (for those who did ultimately go to trial).

In Ontario, settlement rates also vary by case type, with wrongful dismissal, negligence, Simplified Rules, and real property cases having the highest settlement rates (50-54%) and contract/commercial, collection and trust and fiduciary duties cases having the lowest (21-36%).^[85] This evidence therefore suggests some case types may work better with mediation than others, though provides no specific detail on why case type might make this difference.

However, Community Justice Centre (New South Wales) data from 2017-2018 show that they arranged 1061 mediations and though there was some variation in settlement rate based on case type and referral method, they argue the range reflects a good rate of settlement across all cases types.^[86] Similarly, key informants were generally agreed that there is the potential to mediate any case, but that professional judgement should always be used to consider this on a case-by-case basis. Though this review did not focus on family cases, domestic abuse (DA) was the one case type most people were willing to accept may not be suitable for mediation, with some systems based on explicit exemption of DA cases from mediation and some systems based on judge discretion and precedence.^[87] There were dissenting voices here also though, with, for example, Justice Beaudoin from Ontario suggesting it is possible to mediate DA cases if the mediator has the correct training and by using a shuttle system.^[88]

With good evidence around Small Claims, but conflicting evidence around case type more generally, evidence on broader enablers to settlement in mediation were explored. Hahn and Barr found that across case type, the experience of the mediator and the ability of parties to choose their own mediator influenced the possibility of being able to settle in cases in Ontario, with more experience of mediation by the mediator and greater choice of mediator for the parties meaning greater likelihood of settlement. Non-roster mediators were as likely as roster mediators to completely solve disputes 'but roster-led mediations were more likely than non-roster mediations to resolve some (but not all) the issues'.^[89] If 6 or more plaintiffs and defendants are involved the case becomes much less likely to settle.^[90] The experience of mediators is defined as significant to settlement and satisfaction rates and the need for training is emphasised; the authors also stress the need for education about mediation more broadly.^[91]

Wissler's review of Ohio mediation projects also explored broad characteristics of those cases more likely to settle, to focus on which cases could be quickly, cost-effectively, and beneficially brought to conclusion using mediation.^[92] The factor found to have the most impact on likelihood of settlement was disparity between parties' initial positions – the closer to one another the more likely they were to settle. Other characteristics

were associated with an increase in the rate of settlement of more than fifteen percentage points: greater cooperation between the attorneys during the session, the mediator recommending a particular settlement, and the mediator assisting the parties in evaluating the value of the case.

There were a further set of statistically significant factors increasingly likelihood of settlement, but of less significance than those above:

• Case entered mediation by judge referral or party request rather than by random assignment
• Mediator previously had mediated a larger number of cases
• Parties had more preparation for mediation by their attorneys
• Fewer months between case filing and mediation session
• Mediator evaluated merits of case for parties
• Mediator did not keep views of case silent
• Motion to dismiss or for summary judgment was not pending
• Issues in the case were less complex
• Non-dispositive motions were not pending
• Liability was less strongly contested
• Attorneys spent more rather than less time talking during mediation
• Parties spent more rather than less time talking during mediation

Furthermore, if parties were better prepared for mediation they assessed it more positively afterwards.

With settlement a significant measure of success in the mediation literature, but with its usefulness debated, other outcomes for users are also worth exploring to examine the wider impact of mediation in civil justice.

Compliance

Rates of compliance across these jurisdictions were high, with empirical evidence strongest in the United States.

From the evidence reviewed here, compliance with mediation agreements is strong. In an oft-cited overview of US studies of mediation by Wissler, compliance was between 62 and 90%, with those studies that included a comparison group generally finding higher rates of compliance for mediated settlements than in other processes overall – though with one study finding no difference. Whether referral was voluntary or mandatory was not found to make a difference to outcomes overall.^[93]

In a study by Charkoudian, Eisenberg, and Walter they cite a range of literature from America that found compliance with ADR settlements (not exclusively mediation, though predominantly) as between 59 and 93%.^[94]

Key informants across jurisdictions spoke of compliance being strong with mediation settlements because of the ownership of the process and settlement that mediation entails.

Voluntary versus Mandatory Systems: is one approach better than the other for user outcomes?

The evidence in this review was conflicting about whether voluntary or mandatory approaches to referral to mediation provided better user outcomes; overall, there is not enough evidence to suggest that one approach works better than the other.

The discussion of whether mediation should be a voluntary or mandatory process was divided within and between jurisdictions and clearly has ideological overtones, depending on how mediation is defined and what its ultimate goal is seen to be. The difficulty of definitively defining a system of mediation as voluntary or mandatory is dealt with in the operational features section of the report.

In terms of outcomes for users, there was no conclusive evidence that favoured mandatory systems more than voluntary or vice versa: the evidence is conflicting, with some studies favouring one system, some another, and some stating that there is limited or no difference in terms of outcomes between voluntary or mandatory systems. For example, Wissler^[95], examining the difference between voluntary and mandatory schemes of mediation in small claims and common pleas courts in the USA, found mandatory mediation had lower rates of settlement than voluntary mediation, but neither group felt a greater pressure to settle. Fewer mandatory participants felt mediation to be fair or were satisfied with the outcome than voluntary participants. However, in a review across US evidence Wissler concludes that voluntary and mandatory mediation provide no higher rate of settlement than one another; compulsory mediation does not compel settlement.^[96] However, Mandell and Marshall found that mandatory mediation improved settlement rates overall. They also found that it was important to mediate early in the life of the case to achieve settlement.^[97] This literature shows the range and conflicting nature of conclusions on this issue, as well as the conflicting comments about their impact on outcomes for users.

There was some concern that if parties feel compelled to mediate they may also feel compelled to settle: Moore, discussing the recent legislation in Ireland, explores the conflict between this concern and the increasing recognition that mandatory mediation processes may be needed to improve uptake.^[98] Furthermore, key informants were concerned that the authority of judges may have an impact on the pressure to settle, as parties are concerned to do what they think the judge will want.^[99] Ensuring parties are not pressured to take part in mediation and settle should they not wish to is therefore an important consideration. Part of ensuring this may be access to legal advice.

Legal Representation and Legal Education

Legal representation is a topic of debate and concern within the mediation literature of these jurisdictions as lack of knowledge of the law may lead parties to give away important rights – this may then have an impact on settlement, satisfaction, and access to justice outcomes.

If parties have limited knowledge of their rights, both taking part in mediation and settling can become problematic – they may give away important rights or not know of requests and demands they could make – and mediators may not be able to help such parties, either because of their own lack of legal knowledge or because they are restricted by rules or legislation as to what advice they can offer.^[100] This then has an impact on the settlement, sense of satisfaction with the process, and access to justice. The jurisdictions under study here do not mandate for legal representation in mediation, but mediators are allowed to give parties time to ask for legal advice if needed, speak to court advice services if available, or to allow lawyers to attend mediation. Lawyers' fees obviously add costs to mediation for users. Certain key informants felt that, nevertheless, mediation with lawyer advice was the best form of mediation. Others noted that it forms a place for lawyers within ADR, alongside the potential for them to act as lawyer-mediators. Most felt that parties were best served with the full support of legal advice in mediation, wherever this came from (i.e. mediators require legal knowledge to support self-represented parties or lawyers should attend). Research in Maryland found that those with representation were more likely to feel they had been heard throughout the process.^[101]

In a US study by Engler^[102] about the unrepresented facing a represented adversary in mediation, the unrepresented party will be faced with the fact no one in the mediation setting is meant to provide legal advice. This can mean unfair settlement, unequal settlement, or the signing away of rights. This article cites a case in Florida where a mediator asked whether they could advise an unrepresented party that they knew to be unaware of certain rights they had and were therefore not getting a full and fair settlement – the courts said no. This article argues that this is not impartiality, but a favouring of the represented. This article points out that mediation can become harmful in this situation and represent negative outcomes for users.

There is recognition in the literature and from key informants that lawyer culture may need to (further) change before discussing mediation with clients and negotiating within the mediation setting will be second nature, even where legislation exists to mandate lawyer-client discussion – more information and education is needed, both formally in law degrees and informally through discussion within the legal profession and beyond.^[103] Such awareness will also help to ensure mediation is used appropriately, with an understanding of the power dynamics involved.

Power Dynamics

The issue of power dynamics in mediation, across case types, is an important issue to pay attention to and consider in detail.

As Boulle and Field note^[104], there are contexts in which power imbalances within mediation will be significant and there are understandable and important questions to ask and be aware of here: these imbalances may not be mitigated and unfair settlement can result. They argue that such inequality can be found in many places in the justice system, but that judges and arbiters should be using objective standards to make decisions, minimising subjectivity.^[105] Thinking carefully about the appropriate use of mediation may help to alleviate some of these concerns, as well as proper training for mediators to help them spot potential problems between parties. Mediators also need to try to recognise and minimise their own power within mediation, especially when their life experience may be very different to that of their clients.^[106]

Mediators are generally not empowered to impose 'fair' settlements, even when they can recognise power imbalance.^[107] The above example from Engler^[108] of the Florida mediator shows how difficult mediators can find this and how process may prevent them from assisting those with less power during mediation. They may be able to stop the mediation, but again, only if they are trained well enough to recognise a difficult situation.^[109] The impartiality and neutrality of mediators become significant issues – they can lead to actually favouring a more powerful party.^[110] There may be space for mediators' evaluation of disputes or at least ensuring fairness, though this may be a significant ask of mediators who may not have proper legal training. It may also make the case for the involvement of lawyers or other legal advice for parties throughout.

It is also worth stating that power is not straightforward and there are nuances in how it works, shifts, and changes in any given setting. Boulle and Field, following Astor, note that 'power is not a commodity that a mediating party has or does not have, but a situated phenomenon involving subtlety and contradictions in its exercise'.^[111]

Cost and Time Savings for Users and Civil Justice System

There is potential for both cost and time savings for users and the civil justice system by making use of mediation, but the evidence shows this cannot be guaranteed. Empirical evidence is conflicting on both cost and time savings and there are difficulties in measuring cost savings accurately.

All jurisdictions were interested in the efficiency agenda, to a greater or lesser extent. For some areas this was described by key informants as the major reason for using mediation – particularly mandatory mediation – as something dramatic needed to be done to clear the court backlog. In others a more mixed picture was presented, with a sense that efficiency was balanced with wider aims, such as increasing access to justice (though this is in part connected with efficiency), empowerment, and self-determination. In the American states in particular, community justice values around mediation are frequently seen as the first narrative around mediation, with the efficiency agenda coming only later or at least very much hand-in-hand.^[112]

Overall, this review found very limited cost effectiveness evidence that was underpinned by strong empirical evidence. There are therefore very limited statements that can be made about this, however, it is worth noting all jurisdictions in this review have, to some extent, moved forward with mediation despite this lack of empirical evidence.

The Law Reform Commission of Ireland sums up the general viewpoint in the mediation literature around cost and time savings and is therefore quoted at length here:

Research on the efficiency of ADR processes (some based on Irish experience) indicates that mediation and conciliation processes often provide a speedy resolution to a specific dispute. That research also indicates that there is – to put it simply – no such thing as a free conflict resolution process, alternative or otherwise. Where the resolution process is provided through, for example, the courts or the Family Mediation Service, most or all of the financial cost is carried by the State. Where the resolution process involves private mediation, the cost is often shared by the parties involved. The Commission accepts, of course, that the additional financial costs involved in an individual case that goes through an unsuccessful mediation and must then be resolved in litigation has to be balanced against the possible savings where a complex case is successfully mediated. The Commission nonetheless considers it is important not to regard ADR as a patently cheaper alternative to litigation costs; in some instances, it may be, but where a mediation or conciliation is not successful it obviously involves additional expense. On the whole, the Commission accepts that careful and appropriate use of ADR processes is likely to reduce the overall financial costs of resolving disputes.

In addition, the other aspect of efficiency – timeliness – may be of great value to the parties. The Commission is also conscious of other values associated with ADR processes, including party autonomy and respect for confidentiality. The point of noting the narrow issue of financial cost is primarily to indicate that the available research strongly supports the view that ADR assists timely resolution of disputes, but is less clear that direct financial costs savings may arise for the parties.^[113]

There is a lack of clear evidence on cost savings, though there is a common sense feeling that, in the correct circumstances – that cannot be guaranteed – mediation can save costs, either for litigants, the civil justice system, or both. Time may also be saved, though the wider literature debates this also. These potential positives of mediation should, as the Law Reform Commission of Ireland note, be reviewed in relation to other values of mediation and ADR. Some of the wider literature will be highlighted here to give a sense of this uncertainty around cost and time.

Macfarlane conducted an evaluation of the mandatory mediation system in Ontario in the 1990s, and found that parties and lawyers felt mediation saved them processing time – this was then confirmed by examining court records. Lawyers estimated cost savings for their clients of between $1000 and $5000. Overall the experience of mediation was positive, in terms of cost and time savings and sense of fairness of process. If a party settled, they were more likely to be satisfied with mediation overall – confirming above findings.^[114] Similarly for Florida, Shultz's work found that processing time decreased (cases were looked at from time of referral, rather than filing). As few cases were mediated in this study, judicial workload was not significantly reduced, but parties perceived a cost saving and felt that they had greater access to justice than those whose cases were adjudicated.^[115]

Although not of the strongest quality, a synthesis of cost effectiveness evidence in mediation by Mediate British Columbia is useful as an overview of what is known about this particular aspect of the outcomes of mediation for users and civil justice systems.^[116] This report suggests that mediation can save costs in a number of ways, noting that it has the potential to resolve cases before reaching court or earlier in the process; that mediated agreements are complied with more than court-imposed outcomes, hence preventing further trial or appeal; and that court proceedings, when they do occur after mediation, can be shorter and cost less due to issues being narrowed, clarified, or settled etc.. Court resources can therefore be reallocated. Also, legal and court fees are less and disagreement and conflict are lessened, in workplaces, for example, so cost savings are made through this greater harmony. The report also states that 'Mediation can reduce the cost of civil litigation in which government and/or crown corporations are involved.'^[117] This report does note that there is evidence that suggests that mediation is not cost effective, but argues that this evidence shows that mediation needs to be used appropriately – for example appellate cases may be too complex for mediation – and that there is always the possibility for mediation to save money, even if it cannot be guaranteed in every case. The report suggests empirical research into which circumstances mediation and other forms of ADR work best, including around cost effectiveness.^[118]

Charkoudian studied the benefits to the police in Baltimore City, Maryland, in referring callouts to personal conflicts to community mediation.^[119] This type of mediation can prevent disputes escalating to reach the civil courts. These services were provided free of charge. Such calls were allowed to accumulate for several months before referral, to test whether the dispute could be resolved by the parties without intervention; there were on average 7.56 calls before referral. The results from this in time and cost savings were:

Mediation saved 273 minutes per situation per six-month period, or 4 hours and 33 minutes of patrol time. Assuming two officers to a patrol car, this is 9 hours and 6 minutes of personnel time. Financial savings resulting from this can be assessed using the officer's salary. The upper bound of the cost of conflict-related calls is $193.35 (found by dividing the police budget by the number of calls), and the lower bound is $24.38. With these figures, the financial savings to the police department in a six-month period from one mediation lies somewhere between $1,649.27 and $208.00.^[120]

Overall, Charkoudian argues that time savings are probably more significant than cost savings, but that cost savings do exist.

In the Court of Appeals in Maryland, according to a report prepared by Malhotra-Ortiz, there are around 1300 cases a year, and most will be screened for ADR, including mediation; a large proportion will go to mediation and other ADR.^[121] This same report states that mediation helps to minimise the return to court of the same disputes and reduces time and cost of appeals, though without quantifying these savings.

A more negative study, looking across the USA, is Kuhner's study of court-mandated mediation, which cites settlement rates of between 21 and 60%. Linked to these settlement rates, Kuhner notes there is no saving of time or money for parties who settle and there was in fact some increase in lawyer work in some States, though this was balanced by a decrease in work in other areas.^[122] Kuhner argues that court backlog is reduced by using mediation, however mediation can add money and time costs that exhaust the capacities of parties (particularly as mediated cases may be highly complex and require litigation anyway, at least of some points), meaning they give up on their case: he notes this is a serious justice concern.^[123]

Kuhner goes on to argue that courts save money by using mediation (as opposed to parties), as mediation costs a fraction of the cost of litigation (he suggests between 3 and 10% of the cost).^[124] He states that even if only a small percentage of cases go to mediation and settle, courts will save money; however, he does not find that parties necessarily save time or money.^[125] He is concerned that judges may push for settlement, adding pressure to parties and mediators, and also that mediators may have no incentive for quick settlement due to fee structures. In some cases then, mediation may be drawn out unnecessarily and in others pushed towards a conclusion too quickly.^[126] He is therefore sceptical of the use of mediation as a means of access to justice, seeing it as more a means to save court money and reduce case backlog. However, Kuhner does note that lack of settlement or rising costs for parties does not mean they did not get something important from mediation in terms of understanding their opponent better or narrowing the focus of their concerns, but that these are not definite outcomes of mediation. The courts can have juxtaposed interests to parties who bring cases.^[127]

McEwen's study of corporate mediation in America, and views on cost and time savings, makes clear that large corporations can save money through mediation rather than litigation, and that ADR should be standard.^[128] However, this article also shows corporations are concerned about the costs of discovery and feel this is where much of the cost lies, though they think discovery is important to negotiating well. There may also be structural, cultural, and reputational incentives not to settle (concern with being seen as weak if suggesting ADR, for example) and certainly not to settle early and a new approach is required to challenge these issues.^[129] Corporate and commercial mediation may therefore present a specific set of issues that do not pertain to wider civil cases with smaller claims and/or individual parties.

Mediation: not a panacea

Mediation should not be seen as a panacea, but used where appropriate, giving users access to a range of dispute resolution processes including trial.

The importance of viewing mediation as a tool that people influence – parties, mediators, lawyers – is significant to understanding its success. McEwen argues this clearly:

Once said, the importance of lawyers and parties to mediation seems pretty obvious. Certainly, it was readily apparent to one of the corporate counsel we interviewed… As he put it: 'Mediation in and of itself is not going to reduce costs or time. You could go back and forth for years and then go into mediation. Or you could go into mediation immediately. It's everything that happens around mediation that makes it more or less expensive.... Mediation itself, sitting in a room with a so-called neutral third party, is no panacea. In this view, mediation is a tool. Its effects depend on its uses and on the skills, goals, and orientations of its users'.^[130]

Though there are some voices who believe any case may be mediated, with appropriate judge or other professional discretion used in decision-making, most comment and evidence from this review agrees that the right dispute resolution process is needed for the right case. As a 2004 report for NADRAC states^[131], the most appropriate resolution tool/approach for each dispute is needed, rather than a blanket response. Trial may sometimes be the right response, as may another form of ADR. The situation of litigants – self-represented, having suffered abuse from the other party, etc. – needs to be taken into account when deciding this, as much as the case type and other characteristics of the dispute.