transformative mediation at work: employee and supervisor perceptions on usps redress program

14

Click here to load reader

Upload: lisa-b

Post on 09-Feb-2017

214 views

Category:

Documents


1 download

TRANSCRIPT

Page 1: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

This article was downloaded by: [The Aga Khan University]On: 16 October 2014, At: 11:17Publisher: RoutledgeInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: MortimerHouse, 37-41 Mortimer Street, London W1T 3JH, UK

International Review of Public AdministrationPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/rrpa20

Transformative Mediation at Work: Employee andSupervisor Perceptions on USPS Redress ProgramYuseok Moona & Lisa B. Binghamb

a Kyungsung University, South Koreab Indiana University, USAPublished online: 25 Mar 2014.

To cite this article: Yuseok Moon & Lisa B. Bingham (2006) Transformative Mediation at Work: Employee andSupervisor Perceptions on USPS Redress Program, International Review of Public Administration, 11:2, 43-55, DOI:10.1080/12294659.2003.10805084

To link to this article: http://dx.doi.org/10.1080/12294659.2003.10805084

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) containedin the publications on our platform. However, Taylor & Francis, our agents, and our licensors make norepresentations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose ofthe Content. Any opinions and views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be reliedupon and should be independently verified with primary sources of information. Taylor and Francis shallnot be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and otherliabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Page 2: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

INTRODUCTION

Workplace conflicts are an unavoidable aspect oforganizational life. They can lead to job stress and orga-nizational damage. However the effects are not alwaysnegative, harmful, or dysfunctional. Commentators sug-gest that effective conflict management can result inbeneficial changes such as increased productivity,improved relationships, better communication, andincreased problem solving capacity. Traditionally, orga-nizations used litigation, a typical rights and power-based approach as the default means to resolve work-place conflicts. However, litigation can be a painstaking

and time-consuming process with high costs. Moreover,if not settled litigation often results in a winner-loserrelationship, which may terminate any direct dealingsbetween the parties. Increasingly, organizations areimplementing alternative dispute resolution (ADR)methods including arbitration and mediation. In the pasttwo decades mediation (one of the oldest forms of dis-pute resolution in human history) has become a pre-ferred ADR choice in various fields including familydisputes, labor-management negotiations, internationalrelations, community disputes, and environmental dis-putes. In particular, the use of mediation is growing inpublic and private sector employment (U.S. GAO 1995,

43© International Review of Public Administration2007, Vol. 11, No. 2

TRANSFORMATIVE MEDIATION AT WORK: EMPLOYEE AND SUPERVISOR PERCEPTIONS

ON USPS REDRESS PROGRAM

YUSEOK MOONKyungsung University, South Korea

LISA B. BINGHAMIndiana University, USA

The United States Postal Services (USPS) has implemented a nation-wide mediation pro-gram called REDRESS. The program uses the transformative model of mediation which pro-hibits the mediator from taking a directive or an evaluative approach in mediation, but insteadrequires that mediators seek to empower the parties and generate opportunities for recogni-tion of each others’ perspectives. This study examines the perceptions of participants aboutprocedural justice under the early implementation of the transformative mediation model byanalyzing exit surveys. This study finds that a great majority of both employees and supervi-sors are satisfied with the mediation process and the mediators. This result holds even foremployees whose disputes were not resolved at the mediation. In addition, this study finds thatthe mediation process and individual mediator performance are significant contributors tooutcome satisfaction. Based on these results, this study concludes that the transformativemodel of mediation, which focuses not on resolving the immediate issue but on the empower-ment and recognition of participants, is a promising alternative to traditional routes of disputeresolution in the public workplace.

Key Words: transformative mediation, dispute resolution, procedural justice

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 3: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

1997).Mediation is a process whereby a neutral third party

intervenes in a conflict to facilitate negotiation and deci-sion making between the two or more parties (Moore,1996). Unlike litigation where judges may impose asolution, a mediator has no power to decide the out-come. Instead mediators help two or more parties createa mutually agreeable solution through assisted negotia-tions. The mediation process is informal and the factsare presented not through cross-examination but in anarrative fashion. The parties may raise any issues andarguments without any legal constraints (Ury, Brett, andGoldberg, 1988). Proponents of mediation argue thatmediation leads to faster resolution, lower costs, flexibil-ity, creativity in settlements, the potential for preserva-tion or enhancement of relationships, and higher partici-pant satisfaction with the mediation than traditionaladjudicatory processes.1 Empirical research has reporteddifferent settlement rates, the satisfaction level of partici-pants, and other positive results of mediation in variousfields.2 Much of this research deals with how mediationworks in a specific setting with a relatively small sam-ple. None of this research examines the transformativemodel of mediation as described by Bush and Folger(1994). To understand how mediation works in a largeworkplace the field needs a more comprehensive empir-ical research with larger samples.

In July 1999, the United States Postal Services(USPS) completed national implementation of a trans-formative mediation program for disputes concerningdiscrimination in employment. This program is namedREDRESSTM (Resolve Employment Disputes, ReachEquitable Solutions Swiftly). In a pilot study about anearly facilitative mediation model in the REDRESSTM

program, Bingham (1997) found that mediation settledseventy-one percent (71%) of disputes in full or in part.The study also found that the satisfaction of participantswith the final outcome of mediation is determined byprocedural justice factors and the perception of partici-pants about the resolution of the dispute. Employees andsupervisors were equally satisfied with the overall medi-ation process and the mediators, but supervisors wereslightly more satisfied with outcome. The study lookedat a sample of REDRESSTM participants in the FloridaPanhandle using a facilitative model of mediation.

This study uses the same methodology as Bingham(1997) to examine the participant perceptions of thenational transformative mediation model. How effec-tive can transformative mediation be in resolving work-

place disputes? How do participants perceive transfor-mative mediation? How do the roles as employee-com-plainant and supervisor-respondent in the mediationprocess affect the perception of transformative media-tion by participants?

This paper will briefly describe the use of mediationat USPS and the theory behind the transformativemodel. Second, it will briefly review the literature on theprocedural justice theory that provides a framework forunderstanding the perception of mediation by partici-pants. Third, it will review the literature on the theory ofsocial accounts that provides another framework formediation. Fourth, it will review the research on the dif-ferences of perception about the mediation based on therelationship of the disputants, employees, and supervi-sors. Fifth, it will describe the research method and pre-sent results. Finally, it will discuss the implications ofthese results.

MEDIATION AT THE USPS

The USPS, one of the largest civilian employers inthe nation has a substantial history of workplace con-flicts (Baxter, 1994). The USPS employs approximatelythirty percent (30%) of the federal workforce but gener-ates about a half of all federal Equal EmploymentOpportunity (EEO) complaints. These are complaintsalleging discrimination on the basis of race, sex, age,disability, and other protected classes. The number offormal complaints of discrimination doubled from 7,000in 1990 to 14,000 in 1997.

The traditional route for discrimination complaints isthe administrative process of the Equal EmploymentOpportunity Commission (EEOC). In the EEO process,an employee may file an informal complaint, and thenan EEO counselor engages in counseling, which con-sists of informal efforts to resolve the dispute. In thisprocess, the EEO counselor proposes suggestions forsettlement to the parties and may or may not have a jointmeeting with the parties. If the employee is dissatisfiedwith the outcome of the counseling, they may file a for-mal complaint requiring a full investigation. The formalcomplaint process may result in an adjudicatory hearingbefore an administrative judge. If the employee is notsatisfied with a judicial decision they may take the issueto the federal courts.

The USPS began to explore alternatives to this tradi-tional complaint processing model in 1994. After a four-

44 Transformative Mediation at Work Vol. 11, No. 2

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 4: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

year experimental pilot program using facilitative medi-ation, the USPS implemented employment mediationusing a transformative mediation model nation-widestarting in March 1998. Both the original facilitativepilot program and the current national REDRESSTM

program use an outside neutral mediator, a third partywho is an independent contractor compensated by butnot an employee of the USPS, to facilitate a conversa-tion between disputing parties. Under both models,employees who file an informal EEO complaint mayrequest mediation in lieu of counseling. The USPSschedules mediation within two to three weeks of therequest by the employee. Mediation sessions last nomore than one day, and usually around four hours. Com-plaints may return to the traditional EEO process if themediation fails to resolve the dispute.

A unique feature of the current national REDRESSTM

program is the use of the transformative model advocat-ed by Bush and Folger (1994). The transformative modelfocuses on generating empowerment and recognitionamong parties. Empowerment is achieved when disput-ing parties experience a strengthened awareness of self-worth and the ability to deal with whatever difficultiesthey face regardless of external constraints (Bush andFolger, 1994). Recognition is achieved when disputingparties are given some degree of empowerment andexperience an expanded willingness to acknowledge andbe responsive to situations and common human qualitiesof other parties (Bush and Folger, 1994). While the set-tlement-driven mediation model allows for the mediatorto be directive to reach a settlement, the transformativemodel prohibits the mediator from taking a directive roleby suggesting solutions. It rather safeguards againstmediator pressure to obtain any particular result in medi-ation. The USPS adopted the transformative modelbecause of the potential for allowing parties to resolvethe immediate dispute as well as any underlying causesof conflict which otherwise would remain unresolved.

THEORETICAL FRAMEWORK AND HYPOTHESES

Procedural Justice Theory

Literature on procedural justice provides a theoreticalframework for explaining the perceptions of mediationby participants. Substantial research shows that the per-ception of fairness by personnel affects organizational

decisions, behavior, and attitudes. Thibaut and Walker(1975) found that procedural fairness judgments affectthe willingness by personnel to use various legal proce-dures. Lind et al. ( Earley and Lind, 1987; Lind, Kanfer,and Earley, 1990) found that the perception of procedur-al fairness by workers affects the acceptance of tasksand work goals. Lind and Tyler (1988) found that inmany different organizational decision-making contextsemployees decide whether or not to accede to authoritieson the basis of the perceptions of the fairness of from theauthorities. If the employee feels that the supervisor isacting fairly, they assume that the directive is legitimateand obeys; but if the employee feels that the supervisoris acting unfairly, they question the motives of the super-visors and disobey. Sheppard, Lewicki, and Minton(1992) found that both supervisors and employees areconcerned about procedural justice in the resolution ofwork and labor disputes.

Procedural justice theory predicts that participants inthe decision making process will be more satisfied withthe actual outcome of the process when they perceivethe process is fair (Lind and Tyler, 1988). According tothe theory, the perception of fairness by participants willincrease when they feel they have opportunities to pro-vide input to: decision-makers, a decision-makingauthority that is neutral and trustworthy, and treatmentthat is dignified and respectful. Lind, Kanfer, and Earley(1990) noted that the results of procedural justiceresearch can be best explained by the voice effect. Thevoice effects are either instrumental or non-instrumental.A group of scholars (Brett, 1986; Lind, et al. 1993)explored the instrumental aspect of voice: disputantsgiven the opportunity for exercising a voice to increasepotential influence over the outcome of the dispute. Thenon-instrumental aspect of voice is well explained bythe group-value model. The group-value model assumesthat people are concerned about long-term relationshipsand that they want to establish and maintain the socialbonds within the group or organization. Because of theconcern about group status and group membership, peo-ple care about neutrality, trust, or standing, instead of theactual outcome of the process (Tyler, 1989; Lind andTyler, 1988).

Proponents of alternative dispute resolution suggestthat it will increase the perception about fairness of theprocess by participants and in turn will increase overallsatisfaction because it gives participants more controlover the process, greater opportunity for voice, andmore control over the outcome. They also argue that

January 2007 Yuseok Moon and Lisa B. Bingham 45

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 5: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

alternative dispute resolution is more likely than tradi-tional procedures to increase opportunities for a settle-ment because it allows more room to consider theunderlying needs, concerns, and interests of the partici-pants (Fisher, Ury, and Patton, 1991).

Theory of Social Accounts

The theory of social accounts may also explain whyparticipants have positive views of mediation. A socialaccount is an explanation given to change an individualperception of “responsibility for an incident or action;motives for an action; or the unfavorability of an inci-dent or action” (Bies, 1987: 353). Bies suggested socialaccounts can mitigate the perception of an employeeabout the dispute if managers use accounts appropriate-ly. The perception about an incident or a situation by aparty is often more critical than the objective nature ofthe dispute itself (Sheppard, Lewiciki, and Minton,1992). The perception about an incident or a situation bya party is based on the assessment of the intention ormotives of the other party. In a case where a decision isharmful, the affected party may experience what Biesterms, moral outrage. The party that has caused harmmay mitigate the affected party’s feeling of moral out-rage by using social accounts including explanation,apology, or other justifications. The effective use ofsocial accounts can mitigate hostility, and recover com-mitment to the organization. The mediation processgives managers substantial opportunities for socialaccounts to employees who might feel themselves as thevictims of unjust actions. Under the traditional EEOprocess as implemented at the USPS, managers general-ly do not have the opportunity to give direct or face-to-face social accounts to the employees until the adminis-trative hearing. The availability of more opportunitiesfor social accounts under the mediation process is likelyto increase the satisfaction by both parties with theprocess and with the outcome of mediation.

Both the theory of social accounts and the groupvalue model can explain how mediation works by focus-ing on socio-psychological aspects of dispute. The fol-lowing two hypotheses of this study are drawn from theabove arguments and findings of procedural justice the-ory and the theory of social account:

H1: The satisfaction level on the process of media-tion will be positively related to the level of satisfactionwith the outcome.

Perceptional Differences based on the Role of Participants

Literature on organizational behavior and proceduraljustice theory predicts that an individual position willaffect the judgment of the process. Lind, Kulik,Ambrose, and Vera Park (1990) pointed out that defen-dants tend to be more satisfied with arbitration thanplaintiffs are. Lerner (1982) suggested that people aremore concerned with profit and loss issues when theyact as representatives of businesses than when acting asprivate individuals. In contrast, literature on agency the-ory (Jensen, 1983; Eisenhart, 1989) predicts that corpo-rate agents are concerned with outcome only when thereis a strong link between the organizational outcome andthe agent outcome. In general, individual disputantshave the greater personal stake in the outcome of litiga-tion or ADR than corporate officials including supervi-sors. Individual disputants having greater personal staketend to monitor actual outcome more closely than corpo-rate actors. In the light of the agent theory the actual out-come will have greater effects on individual satisfactionof disputants with the outcome than on the satisfactionof corporate officials with the outcome. Bingham (1997)found that supervisors are more satisfied with the out-come of mediation than employees are, but found nostatistically significant differences in the satisfactionwith the process and mediator indices between employ-ees and supervisors. Brett and Goldberg (1983) foundthat management representatives tend to be more satis-fied with grievance mediation than labor representatives.These prior findings suggest that the role of the partici-pants within organizations will affect the level of satis-faction with a given mediation model. This speculationleads to the second hypothesis of this study:

H2: Compared with counterparts, supervisors aremore likely to have higher levels of satisfaction with theoutcome of mediation.

METHOD

Sample

The sample to test the two hypotheses of this studywas drawn from exit surveys completed by participantsof the REDRESSTM program in eleven national areas(similar to regions) including more than eighty-five postal

46 Transformative Mediation at Work Vol. 11, No. 2

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 6: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

service performance clusters (similar to districts) betweenFebruary 1998 and October 2003. The participants wereasked to complete an Exit Survey at the end of the medi-ation session. The participants completed the entirelyanonymous surveys, sealed them in an envelope, andmailed them directly to the researchers at the IndianaConflict Resolution Institute. Because of concerns regard-ing the confidentiality of mediation and USPS employeeidentity, the exit survey did not include questions regard-ing the demographic variables of participants and therewas no follow-up mailing to the participants. A totaleighty-one thousand and nine hundred two (N=81,902)surveys were collected. As shown in Table 1, forty-twothousand and four hundred sixteen (N=42,416) com-plainants and thirty-nine thousand four hundred eightysix (N=39,486) supervisors completed the survey.Approximately eighty-four percent (N=35,698) of thecomplainants were craft level workers and approximatelyninety-five percent of the respondents (N=37,313) wereexecutive and administrative level supervisors. Approxi-mately sixty-six percent of complainants brought accom-panying representatives such as union representatives(44%, N=18,618), fellow employees (15%, N=6,218),and attorneys (3.6%, N= 1,433) to the mediation. Themajority of the respondents (53%, N=20,953) came tothe mediation by themselves. Approximately seventy-two percent of the complainants (N=30,4470) and thirty-nine percent of the respondents (N=15,206) reported that

they had filed either regular EEO complaints or attendedthe REDRESS mediation processes more than once,before attending the mediation respectively.

Analysis and Variables

To test the first hypothesis (H1) the satisfaction levelwith each measure of process, mediator, and outcomeindex for participants was calculated and examined bydividing participants into three groups according to thedegree of resolution of the dispute. A multiple regres-sion with the process index, mediator index, role (adichotomous binominal variable, coded 0 for employeeand 1 for supervisor), and resolved (coded 3 for fullyresolved, 2 for partially resolved, 1 for not at all) wasused as a predictor and an outcome index as the depen-dent variable. The levels of satisfaction with the media-tion process, the individual mediator, and the outcomeof the mediation were measured by using multiple itemswith a five point, Likert-type scale ranging from verysatisfied (coded 5) to very dissatisfied (coded 1). Theprocess index was constructed by summing up scores ofseven items asking about the level of satisfaction withinformation about the process, control over the process,the opportunity to be heard, the overall fairness over theprocess, the participation in the process, understandingof it, and treatment in the process (Cronbach’s alpha=.90). The mediator index was constructed by summing

January 2007 Yuseok Moon and Lisa B. Bingham 47

Table 1. Composition of Complaints and Supervisors

Complaints (n=42,416) Supervisors (N=39,486)

N % N %

Craft 35,698 84.2 1.209 3.1

Executive and administrative service (EAS) 3,080 7.3 37,313 94.5

Level Postal career executive service (PCES) 99 0.2 435 1.1

Other 1,763 4.2 200 0.5

Missing 1,776 4.2 329 0.8

No representative 10,810 26.9 20,953 53.1

Fellow employee 6,218 14.7 6,501 16.5

Representative accompaniedAttorney 1,433 3.6 257 0.7

Union representative 18,618 43.9 538 1.4

Other 3,165 7.5 6,862 17.4

Missing 3.165 7.5 4,375 11.1

Yes 30,470 71.8 15,206 38.5

EEO experience No 11,460 27.0 23,993 60.8

Missing 486 1.1 287 0.7

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 7: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

up the scores of four items asking about the level of sat-isfaction by the participants with whether the mediatortreated one with respect to, the impartiality, the overallfairness, and the overall performance of the mediator(Cronbach’s alpha=.93). This index differed slightlyfrom that used in Bingham’s (1997) study, in that theknowledge of the mediator about the substance of thedispute is not relevant in the transformative model ofmediation, since the mediator is not permitted to evalu-ate the merits of a case. As a result that item was omittedfrom the mediator index. The outcome index was con-structed by summing up the scores of five items askingabout the level of satisfaction by participants with theactual outcome, the speed of reaching a settlement, theoutcome relative to what was expected, control over theoutcome, and the long-term benefit to the relationship(Cronbach’s alpha=.88).

To test whether there are significant differences inthe level of satisfaction with each measure of the threeindices based on role (H2), multiple t-tests were imple-mented using role as an independent variable and eachitem of the three indices as dependent variables. To seewhether there are significant differences in the level ofsatisfaction with each of the three indices based on role,

three separate analyses of variance (ANOVA) were per-formed by using role as independent variable and eachof the three indices as dependent variables.

Finally, to get additional information about thethoughts of participants on the mediation program, thisstudy analyzed a random sample of five hundredanswers to two open-ended questions asking about whatparticipants liked best and what they liked least aboutthe mediation program.

RESULTS

Approximately sixty-nine percent of employees(68.7%) and supervisors (69.3%) reported that disputeswere at least partially resolved. A high majority, gener-ally over ninety percent (90%) of the employees andsupervisors reported being satisfied or highly satisfiedwith the process of mediation, and the individual media-tors, regardless of settlement. These numbers decreaseslightly if the dispute is only partially resolved or notresolved. The satisfaction levels with almost all mea-sures of process of mediation and the individual media-tor remains high even if the disputes were not resolved

48 Transformative Mediation at Work Vol. 11, No. 2

Table 2. Reports of Being Satisfied or Highly Satisfied with Various Aspects of Mediation by Role and Resolved

Fully Resolved Partially Resolved Not Resolved

Employees Supervisors Employees Supervisors Employees Supervisors% Satisfied % Satisfied % Satisfied % Satisfied % Satisfied % Satisfied

Satisfaction with ProcessInformation 97.8 92.8 94.1 88.3 83.7 84.6Control 96.5 91.0 90.4 84.9 77.9 79.1Opportunity to present views 98.8 96.8 96.1 94.2 89.2 91.1Participation 98.7 97.2 96.9 94.4 90.5 91.7Understanding 99.0 98.3 96.9 96.4 90.6 94.0Fairness 98.1 96.2 92.6 92.4 77.5 87.8Treatment 98.5 97.7 95.5 95.5 85.1 92.0

Satisfaction with MediatorRespect 99.6 99.3 98.9 98.6 96.9 97.9Impartiality 98.8 98.2 97.1 96.7 92.9 95.3Fairness 99.2 98.6 97.8 97.2 93.6 95.7Performance 99.4 98.5 98.2 97.3 93.9 95.0

Satisfaction with OutcomeOutcome 97.1 93.8 82.7 82.2 18.8 33.4Speed 95.4 88.4 87.8 75.8 69.3 64.2Outcome relative to expectation 94.9 90.4 76.8 77.4 24.1 40.7Control 95.7 91.3 81.3 81.5 37.3 54.3Long-term effects 91.7 88.1 69.4 73.5 27.5 42.2

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 8: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

at all. These results support the first hypothesis.A majority of employees and supervisors reported

being satisfied or highly satisfied with the outcome ofthe mediation, unless the dispute is not resolved at all.Table 2 shows the portion of participants (by employeesand supervisors) who responded that they were eithersatisfied or highly satisfied with each of the measuresbelow according to the extent of the settlement.

To further test the first hypothesis that satisfactionwith the process would contribute to satisfaction withthe outcome, a multiple regression was performed usingthe outcome index as the dependent variable and the

process index, the mediator index, resolved (fully coded3, partially coded 2, and not at all coded 1), and role(employees coded 0 and supervisors coded 1) as inde-pendent variables.3 Table 3 reports the results of theregression. The regression was significant at .001 signif-icance level, and explains that around fifty-six percent ofthe variance in the outcome index. All the independentvariables except role proved to be significant factorscontributing to higher satisfaction. The standardizedcoefficients report that the resolution of the dispute is thebiggest contributor to the satisfaction with the outcome.The process of the mediation is the second biggest con-

January 2007 Yuseok Moon and Lisa B. Bingham 49

Table 3. Multiple Regression for Outcome Index with Role, Resolved, Process Index, and Mediator Index

Variables Coeff. Std.Error Std.Coeff T P(2-tail)

Constant -2.797 0.131 -21.268 0.000***Role 0.820 0.025 0.082 32.368 0.000***

Resolved 2.894 0.016 0.487 184.990 0.000***Process 0.519 0.004 0.444 137.403 0.000***

Mediator 0.026 0.008 0.010 3.080 0.002**

No. of cases=68,642Adjusted R-square=0.562*** Significant at 0.001 level or better.** Significant at 0.01 level or better.

Table 4. Mean Satisfaction in Items of Measures by Role with Independent T-tests.

Employees Mean Supervisors Mean DF T-test P

Process Index Measures 31.49 31.79 78854 -9.84 0.000***Information 4.41 4.40 80772 7.33 0.000***Control 4.32 4.31 80621 2.13 0.033*Opportunity to present Views 4.58 4.59 80830 -2.86 0.004**Participation 4.60 4.60 80468 1.19 0.234Understanding 4.58 4.67 80856 -20.17 0.000***Fairness 4.39 4.55 80263 -26.65 0.000***Treatment 4.55 4.68 80838 -23.67 0.000***

Mediator Index Measures 19.20 19.31 80252 -8.39 0.000**Respect 4.86 4.88 81005 -5.64 0.000***Impartiality 4.76 4.81 80798 -11.93 0.000***Fairness 4.78 4.83 80867 -10.73 0.000***Performance 4.79 4.80 80588 -2.74 0.006**

Outcome Index Measures 18.40 19.69 72807 -34.93 0.00***Outcome 3.40 3.81 80463 -42.73 0.000***Speed 4.16 4.02 75540 16.81 0.000***Outcome relevant to expectation 3.50 3.87 80298 -41.29 0.000***Control 3.78 4.09 79897 -37.39 0.000***Long-term effects 3.58 3.90 79524 -38.06 0.000***

* Significant at 0.05 level.** Significant at 0.01 level.

*** Significant at 0.001 level or better.

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 9: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

tributor to the satisfaction with the outcome by a slightmargin. The mediator has smaller but significant effectson the satisfaction with the outcome. These results alsosupport the first hypothesis.

To examine whether there were significant differ-ences in the satisfaction between employees and super-visors, multiple T-tests were implemented using role asan independent variable and each of the measured itemsas dependent variables. Table 4 reports mean satisfaction(5=very satisfied, 1=very dissatisfied) and T-test results.

The mean differences are statistically significant onvarious measures of process index: information, control,opportunity to present views, understanding, fairness,and treatment. In the measures of information and con-trol the mean satisfaction by employees is only margin-ally higher than that of supervisors. In contrast, in themeasures of understanding, fairness, and treatment, themean satisfaction by employees is slightly less thanthose of supervisors. The differences on specific indica-tors are small but significant, partly as a function oflarge sample size.

In each item of the mediator index, the mean satis-faction of supervisors is slightly higher than that ofemployees, and the mean differences are statisticallysignificant on all of the measures. These differences in

all the individual indicators again are small but signifi-cant partly as a function of large sample size.

In all of the outcome measures except speed, themean satisfaction of supervisors is slightly higher thanthat of employees, and these mean differences are statis-tically significant.

To examine mean satisfaction differences on thethree indices, three separate one-way analyses of vari-ance were performed using each index as a dependentvariable and role as an independent variable. Table 5reports the results. The ANOVA on all of the threeindices are statistically significant. What is noteworthyis that the substantially large sample size makes even theslight mean difference between the employee andrespondent groups statistically significant. Overall, themean values of both the employee and respondentgroups are very high and the differences on all of thethree indices are very small. This finding means thatover all, supervisors and employees are equally satisfiedwith the mediation process, the individual mediators,and the outcome of mediation. As shown in Table 4, themean values of the supervisor groups on all three indicesare slightly larger than those of counterparts and theemployee groups. This result indicates that overallsupervisors have slightly higher level of satisfaction

50 Transformative Mediation at Work Vol. 11, No. 2

Table 5. Three One-Way-Analyses of Variance using Role as Independent Variable and Process Index, Mediator Index, and Outcome Index as Dependent Variables

Dependent Variables Sum of Squares between Groups DF F-ratioii P

Process Index 1826.07 1 96.85 0.000***Mediator Index 270.46 1 70.41 0.000***Outcome Index 30297.68 1 1220.34 0.000***

* Significant at 0.05 level.

** Significant at 0.01 level.

Table 6. Participant Responses (Frequency) to What they liked Best about Mediation

Employee Supervisor

Timely resolution or immediate fact finding 160 128Open face to face communication 120 159Opportunity to defend oneself 83 56Opportunity to be heard 32 36Process itself 28 15Impartiality of mediator 27 20Fairness of process 20 21Opportunity to solve problems together 8 19Improved Understanding 5 3Settlement of complaints at the lowest level 4 7Others (flexibility, informality, no stress, non confrontation) 13 34

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 10: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

with the outcome of mediation than employees. Theseresults support the second hypothesis.

To get additional information about the perceptionsof participants about the mediation program, this studyexamined answers to the two open-ended questions ask-ing about what they liked best and least about the medi-ation program from each of 500 employees and 500supervisors. This sub-sample was taken from the first500 in each group within the larger sample. Table 6 rep-resents a frequency distribution of the responses of par-ticipants to the question asking about what was likedbest about the mediation program.

A majority of employees and supervisors pointed totimely resolution or immediate fact-finding and face-to-face communication as the best aspect of the mediationprogram. The opportunity to defend oneself or fightback, the opportunity to be heard, the process itself, theimpartiality of the mediator, and the fairness of theprocess were also mentioned as the best aspects of themediation program, although the frequency distributionsare slightly different between employees and supervi-sors. Table 7 represents a frequency distribution of theresponses of participants to the question asking aboutwhat they liked least about the mediation program.

Around one hundred twenty employees and one hun-dred fifteen supervisors answered the question with pos-itive comments about mediation and identified no weak-nesses in the program at all. While a majority ofemployees (N=148) identified the failure to resolve thedispute as the worst aspect of the mediation program, amajority of supervisors (N=129) felt that the time con-suming nature of mediation was the worst aspect of the

mediation program. Fifty-nine employees and forty-fivesupervisors identified the presence of uncooperativebehavior from the other party as the worst problem inmediation. A relatively larger number of employees(N=34) than the counterparts (N=3) identified the lackof authority to make decision or to implement a settle-ment as the worst aspects of the mediation program.Twenty-five supervisors pointed out the lack of informa-tion given beforehand about the issue and the other partyas the worst aspect of the program.

DISCUSSION

The results of this study partially replicate Bing-ham’s (1997) evaluation study about the USPS pilot pro-gram of facilitative mediation in Panhandle Florida. Theoverall pattern of results here is generally similar tothose in the pilot program as reported in Bingham(1997). The reported combined sixty-nine percent (69%)settlement rate is slightly lower than that of the pilot pro-gram in which seventy-one percent (71%) of the dis-putes were settled in full or in part. This settlement rateis significantly higher than a mediation model usinginside neutral or USPS employees as mediators (Bing-ham, et al. 2000). It is a high rate of settlement for amodel of mediation that does not have settlement as itsprimary goal. Even though supervisors were found tohave slightly higher levels of satisfaction on the process,mediator, and outcome of the mediation, employees alsohave high levels of satisfaction in all of the three indices.These earliest results from the transformative model

January 2007 Yuseok Moon and Lisa B. Bingham 51

Table 7. Participant Responses (Frequency) to What they liked Least about Mediation

Employee Supervisor

No actual outcome 148 76Uncooperative behavior of the other party 59 45Time consuming 52 129Lack of authority to make a decision or to implement agreement 34 3Inability to confront the person 11 2Improperness of meeting place 8 4Lack of information about future 7 2Impartiality of mediator 6 5Improperness of time of the meeting 6 7Over/under representation 3 7Lack of beforehand information about the issue and the other party 2 25Inappropriateness of the issue to REDRESSTM 0 11Answered with some positive comments-No weakness in program 120 115Others4 44 69

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 11: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

indicate that employees and supervisors alike perceive itas a relatively well-balanced and fair program for deal-ing with workplace disputes in the USPS. This is animportant achievement in a workplace with a history ofwell-publicized employment relations problems.

The frequency distribution of responses by partici-pants to what they liked best (Table 6) shows that partic-ipants value mutual communication, opportunity to beheard, fair process, and opportunity to solve problemtogether as well as solutions to the dispute at issue.These results show that people care about not only actu-al outcomes, but also about the mediation process whenaddressing workplace disputes. This too supports theconclusion that the transformative model of mediationcan prove as effective in the workplace as traditionalfacilitative mediation in terms of participant satisfaction.

The results of multiple regression show that eventhough the resolution of the dispute was the biggest con-tributor to the satisfaction by participants with outcome,process, and mediator indices are also significant factorsof outcome satisfaction as is expected in the proceduraljustice theory, and consistent with numerous reportedstudies. This was true for both parties; employees andsupervisors alike. The differences in outcome satisfac-tion are explained largely by the underlying differencesin procedural justice indicators and settlement. Thesefindings suggest that an organization will not sacrificethe overall satisfaction of either one or both parties withthe mediation program by using a transformative media-tion model.

As shown in Table 4, satisfaction level by participantswith the outcome was found to be lower than that of theprocess and the mediator index. The mean satisfaction ofthe two groups is slightly but statistically significantlydifferent on multiple measures on the three indices.Overall with the outcome index, the mean satisfaction ofsupervisors is significantly higher than that of employ-ees. Similar patterns were found in a labor-managementcontext, where management was found to be more satis-fied with mediation than labor (Brett and Goldberg1983). This pattern emerged in perceptions of the civiljustice system by tort litigants (Lind et al., 1990). Thiscan be attributed to the differences on expectations fromboth groups about the mediation process and outcome.An employee dissatisfied with a certain issue that wasbrought into the mediation is likely to expect more thanthe supervisor does. While the employee tends to expectthe best, the supervisor expects the worst. Significant dif-ferences based on role are shown in Table 4 in satisfac-

tion with outcome relative to what the parties expectedsupport this analysis. Another reason for the differencesbetween the two parties may be traceable to the increasedopportunities for social accounts. In mediation, bothemployees and supervisors have more chances toaccount for the stories to each other. However, it is usual-ly the supervisor who has taken some action that formsthe basis for the EEO complaint. As respondents, super-visors may feel more satisfaction with opportunities toexplain any actions to the employees.

The frequency distributions of responses from partic-ipants to what they liked best or least about mediation(Table 6 and Table 7) also show that positional differ-ences affect the perception by participants about themediation program. While more employees indicatetimely resolution or immediate fact finding as the bestaspects of mediation programs, more supervisorsanswered that open communication is the best aspect ofthe program. A larger number of employees than super-visors view mediation as an opportunity for self-defenseor for fighting back. Table 7 shows that perceptionsabout the weaknesses of mediation programs are mostlydetermined in terms of outcome. This tendency is moreprominent among employees than supervisors. Approxi-mately two hundred employees pointed out that no actu-al outcome or the lack of authority to make a decision orto implement an agreement were the worst aspects ofthe mediation program. It is important to note that evenin cases where the parties reached no resolution,employees, and supervisors were satisfied with theprocess and mediators.

One limitation of this study is that such factors as thenature of the dispute, the characteristics of the media-tors, and interactions of the parties were not controlledin the multiple regression. Controlling for these factorsmight increase internal validity. Some of this informa-tion was unobtainable due to concerns regarding theconfidentiality of the mediation and anonymity of par-ticipants. The interaction among such factors and theimpact on participant perceptions of the process, media-tor, and outcome is needed to be studied further. Thesufficiently large sample size also addresses this con-cern. It is reasonable to conclude that a normal distribu-tion of discrimination case types and mediators is repre-sented in the sample, particularly because the overallparticipation rate in the program ranges between 70 and75% of all EEO complainants.

Another limitation on this study concerns the natureof the model. This study does not directly compare facil-

52 Transformative Mediation at Work Vol. 11, No. 2

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 12: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

itative and transformative models, nor does it find thatparticipant satisfaction is higher overall in one modelthan the other one. It does find that the two models pro-duce comparable results in terms of procedural justice,participant satisfaction, and settlement. Some mightargue that the model studied here does not necessarilyreflect transformative mediation as described by Bushand Folger (1994), because the data address participantperceptions, not observations of mediator behavior.However, other studies suggest that the model beingimplemented does correspond to the ten hallmarks oftransformative mediation practice at least in terms of cri-teria being used by USPS EEO ADR specialists andcoordinators to screen and evaluate mediators on thenational roster (Nabatchi and Bingham, 2000).

CONCLUSION

Mediation is not a panacea that can or should resolveall workplace disputes. However, it is possible to imple-ment a transformative mediation program that bothemployees and supervisors will view as fair and useful.A mediation program that focuses not on settlement, butrather on generating empowerment and recognition canmake a constructive contribution to the workplace. Fur-ther research should examine whether the transformativemodel of mediation allows for participants not only toresolve workplace conflict, but also to improve thecapacity to resolve conflicts in the future. This in turncould lead to beneficial changes in both individualmembers and the organization.

NOTES

1. For comprehensive reviews of the literature, seeCarnevale and Pruitt, 1992; Wall and Lynn, 1993.

2. For settlement rates, see Brett, Barsness, and Goldberg1996; Brett and Goldberg, 1983; Gregory and Rooney,1980; Bowers, Seeber, and Stallworth, 1982; Skratek,1990; Roberts, Wolters, Holley, and Field, 1990; Walland Lynn, 1993.

3. To examine the presence of multi-collinearity amongindependent variables, this study conducted correlationanalyses with the four independent variables before con-ducting the OLS regression for the hypothesis testing.The highest correlation coefficient was .606, whichimplies there are no severe multi-collinearity problemsamong the independent variables.

4. Others for supervisor include such comments as listeningto falsehood, too much stress, chilling effect, moneyspent, lack of knowledge about the issue by the mediator,overlapping with other programs, and too impersonal.Others for employee include such comments as driving tothe session, not focused, anxiety, falsification of docu-mentation, hearing negative things, not enough timeallowed, inability to express the issue, no witnesses, hurtfeelings, waiting time to the date of mediation, and fear ofretaliation.

REFERENCES

Anderson, J. F., and L. B. Bingham. 1997. UpstreamEffects From Mediation of Workplace Disputes:Some Preliminary Evidence from the USPS. LaborLaw Journal48: 601-615.

Baxter, V. 1994. Labor and Politics in the U.S. PostalService. New York: Plenum Press.

Bowers, M.H., Seeber, R.I. and Stallworth, L.E. 1982.Grievance Mediation: A Route to Resolution forthe Cost -Conscious 1980s. Labor Law Journal33:459-466.

Bingham, Lisa B. 1997. Mediating Employment Dis-putes: Perceptions of Redress at the United StatesPostal Service. Review of Public Personnel Admin-istration 17(2): 20-30.

Bingham, Lisa B., Chesmore, G., Moon, Y., and Napoli,L. M. 2000. Mediating Employment Disputes at theUnited States Postal Service: A Comparison of In-house and Outside Neutral Mediator Models.Review of Public Personnel Administration20(1):5-19.

Brett, Jeanne M., Zoe I. Barsness, and Stephen B. Gold-berg. 1996. The Effectiveness of Mediation: AnIndependent Analysis of Cases Handled By FourMajor Service Providers. Negotiation Journal12(3): 259-269.

Brett, Jeanne M., and Stephen B. Goldberg. 1983.Grievance Mediation in the Coal Industry: A FieldExperiment. Industrial and Labor RelationsReview37(1):49-67.

Bush, R. A. B., and J. Folger. 1994. The Promise ofMediation: Responding to Conflict throughEmpowerment and Recognition. San Francisco:

January 2007 Yuseok Moon and Lisa B. Bingham 53

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 13: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

Jossey-Bass Publishers.

Carnevale, P. J., and D. G. Pruitt. 1992. Negotiation andMediation. Annual Review of Psychology43:521-582.

Conlon, D. E. & Ross, W.H. 1993. “The Effects of Parti-san Third Parties on Negotiator Behavior and Out-come Perceptions.” Journal of Applied Psychology78:280-290.

Consensus40. 1998. The MIT-Harvard Public DisputesProgram.

Daus, M. W. 1995. Mediating Claims of Discrimination.Dispute Resolution Journal50(4):51-54.

Feuille,P. 1992. Why Does Grievance MediationResolve Grievances? Negotiation Journal8(2):131-145.

Feuille, P., and D. M. Kolb. 1994. Waiting in the Wings:Mediation’s Role in Grievance Resolution. Negoti-ation Journal10(3):249-264.

Fisher, R., W. Ury, and B. Patton. 1991. Getting to Yes(2nd ed.). New York: Penguin Books.

Folger, R. 1977. Distributive and Procedural Justice:Combined Impact of ‘Voice’ and Improvement onExperience Inequity. Journal of Personality andSocial Psychology35:108-119.

Gregory G.A. and Rooney, R.E., Jr. 1980. GrievanceMediation: A Trend in the Cost -conscious Eighties.Labor and Law Journal31:502-508.

Karambayya, R. & Brett, J. 1989. Managers HandlingDisputes: Third-Party Roles and Perceptions ofFairness. Academy of Management Journal32:687-704.

Karambayya, R., Brett, J., and Lytle, A. 1992. Effects ofFormal Authority and Experience on Third-PartyRoles, Outcomes, and Perceptions of Fairness.Academy of Management Journal35(2): 426-438.

Lewicki, R.J., McAllister, D.J. & Bies, R.J. 1998. Trustand Distrust: New Relationships and Realities.Academy of Management Review23:438-458.

Lind, E. A., R. J. Maccoun, P. A. Ebener, W. L. F. Fel-stiner, D. R. Hensler, J. Resnick, and T. R. Tyler.1990. In the Eye of the Beholder: Tort Litigants’Evaluations of Their Experiences in the Civil Jus-

tice System. Law & Society Review24(4): 953-996.

Lind, E. A., and T. R. Tyler. 1988. The Social Psycholo-gy of Procedural Justice. New York: Plenum Press.

Moore, Christopher. 1996. The Mediation Process(2dEd.). San Francisco, CA: Jossey-Bass Publishers,Inc.

Roberts, M. T., Wolters, R. S., Holley, W. H. and Field,H. S. 1990. Grievance Mediation: A ManagementPerspective. Arbitration Journal 45:15-23.

Skratek, S. P. 1990. Grievance Mediation: Does it ReallyWork? Negotiation Journal6:269-280.

Shapiro, Debra L., and Jeanne M. Brett. 1993. Compar-ing Three Processes Underlying Judgments of Pro-cedural Justice: A Field Study of Mediation andArbitration. Journal of Personality and SocialPsychology65(6):1167-1177.

Sheppard, B. H., R. J. Lewicki, and J. W. Minton. 1992.Organizational Justice: The Search for Fairnessin the Workplace. New York: Lexington Books.

Ury, William, Jeanne Brett, and Stephen Goldberg.1989. Getting Disputes Resolved: Designing Sys-tems to Cut the Cost of Conflict. San Francisco:Jossey-Bass Publishers.

U.S. General Accounting Office. 1995. EmploymentDiscrimination: Most Private-Sector EmployersUse Alternative Dispute Resolution. Washington,DC: United States General Accounting Office.

_______. 1997. Alternative Dispute Resolution: Employ-ers’ Experiences with ADR in the Workplace. Wash-ington, DC: United States General AccountingOffice.

Wall, James A., and A. Lynn. 1993. “Mediation: A Cur-rent Review.” Journal of Conflict Resolution37(1):160-194.

Yuseok Moon received Ph.D. from School of Publicand Environmental Affairs, Indiana University and is aFull-time Instructor at Kyungsung University, SouthKorea. His Research interests include innovation diffu-sion, conflict resolution, and community policing.

54 Transformative Mediation at Work Vol. 11, No. 2

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014

Page 14: Transformative Mediation at Work: Employee and Supervisor Perceptions on USPS Redress Program

Lisa Bingham was formerly an attorney practicinglabor and employment law. She was named the 2002recipient of the Willoughby Abner Award from theAssociation for Conflict Resolution. Bingham, the Keller-Runden Professor of Public Service, serves as the direc-tor of the Indiana Conflict Resolution Institute, which

she founded at Indiana University in 1997. The instituteconducts research and program evaluation on mediation,arbitration, and other forms of dispute resolution.

Received: August 22, 2006

Accepted after one revision: January 1, 2007

January 2007 Yuseok Moon and Lisa B. Bingham 55

Dow

nloa

ded

by [

The

Aga

Kha

n U

nive

rsity

] at

11:

17 1

6 O

ctob

er 2

014