Addressing the "Redress": A Discussion of the Status of the United States Postal Service's Transformative Mediation Program

LELA LOVE:  We have gathered to examine the Postal Service Mediation Program, a significant initiative in the history of mediation and conflict resolution. [see slide]


For the last decade, mediation has gotten hotter and hotter, more and more popular. The current popularity of mediation among lawyers and in courts poses a significant danger that the legal system will morph mediation into a process that resembles adjudicatory processes. Frank Sander, a prominent Harvard law professor, has described mediation as the 'sleeping giant' of the conflict resolution world because it is a real alternative to adjudication. What the Postal Service has done is to provide a large-scale testing ground for a type of mediation very different from the legalized settlement-oriented programs in many courts, a type of mediation more in keeping with the roots of mediation as a process focused on helping parties understand and resolve their own conflicts. The Postal Service process is based on "transformative mediation," a model introduced by Professors Robert A. Baruch Bush and Joseph Folger in their book, The Promise of Mediation (1994).


Baruch Bush is a law professor. I believe it took courage for him, based in a law school, to say that the core values of mediation are party empowerment and inter-party recognition – not settlement or problem-solving. It took courage because courts and lawyers are looking for settlement, for numbers – how many cases can be moved off the court dockets? It also took courage for the Postal Service – the largest employer in the country after WalMart – to adopt transformative mediation as their model. The Postal Service has a hierarchical system and aspects of its workplace environment have given rise to the term "going postal." One does not readily juxtapose "transformation" and the Postal Service. Their courageous adoption of the transformative model shows a commitment to another range of interest and values.


Pam Zuczek, the leader of the Postal Service Mediation Program in the New York area will describe the history, operation and aspirations of the program. Then Professor Lisa Bingham, a nationally renowned researcher, will share the results of her study of the program. Professor Bingham, with her sharp and critical eye, aided with vast numbers of exit surveys and questionnaires, will be able to give us at least preliminary answers to whether transformative mediation settles cases, whether it improves the workplace environment, whether it creates more understanding between workers and supervisors, whether it reduces the numbers of EEO complaints, and so on. If transformative mediation, empowerment and recognition, settles cases as successfully as settlement-oriented mediation, wouldn't that be interesting!


There is a light bulb riddle, which asks: 'How many mediators does it take to change a light bulb?' One answer is: 'None. Mediators empower light bulbs to change themselves.' If any mediation program has a claim to trying to do that, it is the Postal Service program. So Professor Bingham's observation and data about what's happening in that program are important.


Before proceeding, I would like to thank Abe Tawil, for initiating this symposium and spurring it on. I would also like to recognize the work of the Cardozo Online Journal of Conflict Resolution. The Journal pulled this program together in record time and has undertaken to disseminate what happens here today to a larger audience via the publication of these proceedings. They have done phenomenal work.


Now, let's have Pam Zuczek introduce us to the Postal Service Program.


PAM ZUCZEK:  Good morning, it's early on a Monday morning. My name is Pam Zuczek and I am the EEOC ADR Coordinator for the New York metro area. I am the Coordinator of the REDRESSTM Program.


We had spoken a little about ‘going postal' – and although we have the Califano Report – I don't know if any of you have heard about it, but the Postal Service contracted to do a study to find out about certain things like the workplace environment and check out this ‘going postal'. Although we have found out that this is a myth, what is true about the Postal Service is that we are the second largest employer in the country. We employ close to 800,000 employees – and when you have that many people working for you – you have workplace issues and it is also true that we may have been, or still somewhat continue to be, a hierarchical employer. We are very concerned about improving the workplace environment – that's our goal. We still have to deliver the mail every day and we want it to be on time – we want to service our customers – but we have internal customers and they are our employees – they are our greatest resource and we need to deal with our employees in a way that they feel they have a voice. If you are one of 800,000 employees and you don't feel you have a voice – that can make you sort of feel trapped. So we want our internal customers to have a voice the same way as our external customers have a voice.


All that being said, I am sure that you know that President Clinton had mandated ADR programs throughout the federal government and we have an internal equal employment opportunity – (EEO) department processing complaints of discrimination against the Postal Service. With that department we were mandated to have an alternative dispute program. We looked around and tried to see what we were going to do with dealing with our employees in the workplace – even before this mandate from President Clinton.


In 1994, there was a class action complaint in Florida that had been going on for 10 years – no one could get this class action resolved. There happened to be a wonderful lady and I just absolutely love her and I have to mention her name, who was the person who started the REDRESSTM Program for us in the Postal Service, and who connected with Professor Lisa Bingham. Her name is Cindy Hallberlin and she has since moved on from the Postal Service. But she was instrumental in getting this class action lawsuit into mediation and it was resolved quickly. Cindy had the brilliant idea that this would work for the Postal Service. She began looking more into mediation, talking with Lisa and the REDRESSTM Program was born. REDRESSTM, which stands for Resolve Employment Disputes Reach Equitable Solutions Swiftly, is the name of our ADR Mediation Program.


Our REDRESSTM mission is to provide mediation of EEO disputes by outside qualified mediators to all complainants who want it, in a timely and meaningful fashion that will address the immediate conflict and ultimately make a positive difference in the climate of the workplace – that was our REDRESSTM mission. Our goal is to improve workplace climate, to improve communications between supervisors and employees – those internal customers of the Postal Service – and to reduce our formal EEO complaints. Now I say that was a goal to reduce formal EEO complaints, but if you notice we didn't have in our goals, any numbers, we didn't say we wanted to reduce those formals by a certain percentage or a certain number. We didn't say we wanted to close so many EEO complaints. What we did was we said primarily we want to improve workplace climate. We set that as a goal: improving workplace relationships.  In using mediation to do that- and we'll get into that transformative model- we were hoping that it would have success on improving that workplace climate. What we found that came out of that was not only success, but a high success in our numbers.


Like I mentioned, in order to get into our REDRESSTM Program, you have to believe that you have been discriminated against and you would be a person who would be filing a complaint of discrimination in our EEO Department. When you file an EEO complaint with the US Postal Service, the first stage of that complaint process is what we call the informal stage, or the counseling stage, and once you enter into the counseling stage of our EEO complaint process, you have the option to elect mediation. You can proceed through that traditional process without mediation because mediation is voluntary on the part of the complainant. Or you can elect to go to mediation instead of going through traditional counseling. If you elect to go to REDRESSTM Mediation you bypass that counseling stage, and you meet face to face with the party you are in conflict with. Also, you are in mediation within 14 days of the time you filed your complaint as opposed to our traditional EEO process which would take 30 to 60 to 90 days before you would even hear from a counselor. So mediation we found is quick. You file a complaint, how do you feel? You want to resolve it. Mediation we found was fast. We use outside mediators for our REDRESSTM Mediation Program. Those mediators only use facilitative transformative mediation. They are facilitative rather than directive. And like I said before, the process is a voluntary process for the complainant, those things, the voluntariness, the use of the outside mediator, the quick response before people become hardened in their positions, and that face-to-face interaction, are all elements of what has brought REDRESSTM to success. Again our mediators use the transformative mediation process. Transformative mediation assumes that people have the capacity to solve their own problems given the right environment and support.


Our mediators are helping our people at the table to feel empowered and to feel that the other person across the table recognizes them, recognizes their viewpoints. We have our employees sitting down together, our supervisors, their subordinates, talking to each other, trying to work hard with the facilitation of the mediator and understanding the other person's viewpoint. That's our goal, get them to the table, get them to sit down, get them to start talking to each other, get them to try to look at the other person's perspective, even if they don't agree, let us see if they can look at it. When these things start to happen, that's when the transformation begins to take place, as reported daily to me by our mediators, as reported daily by our supervisors and complainants. It's a wonderful, wonderful process.


I won't talk too much about numbers because Lisa will. I talked about four elements of the success of REDRESSTM. Well there is a fifth and that's Lisa Bingham, that's our research and evaluation. And she collects our data for us, and she's going to go into that today. But I will say, even though we didn't set a goal for reduction of formals we were hoping that success might manifest itself in a reduction of formals, well it did. In fiscal year 1999 we had 10,998 formal EEO complaints and by accounting period 12 of fiscal year 2000 we had 9,012 formal EEO complaints. That's a reduction of 1,986 formals in one year. That's quite a large figure, because each formal complaint could possibly go through to litigation and it's extremely costly – so that's a large number. We also found that our national closure rate – now we deal with 13 APs (Accounting Periods) in our fiscal year, so this was accounting period 12 near the end of our fiscal year 2000 – was 80%, that's looking at what we mediated in the fiscal year 2000 – approximately 10,369 cases across the country. I just think that's a phenomenal number that we know that across the country there were 10,369 different parties that sat down at a table and said, ‘Okay, we want to work this out together.' And 6,089 of those were resolved at the table; 1,674 did not continue to pursue their EEO complaints. They may not have signed a settlement at the table, they may have just gotten up and shaken hands and said 'I'm just not going to pursue that complaint.' And after the mediation there were 3%, or 272 people, who walked away, thought about it, came back and said, ‘Hey we can resolve this,' even though they didn't sign an agreement or settle it right there.  Those are the people who needed a little time to think about it, they walked away and came back and said, ‘You know what, we can work this out.' So those are pretty phenomenal numbers. To get 80% closure rate on a process, or on 10,000 plus complaints, when you didn't set that as a goal, that's a great by-product. We just have fantastic mediators that work for us, we do. I have to give Abe Tawil, Roger McKechnie, Paul Schlossbach, Jerry Hornblass and Sylvia Murzin, Lela are on the roster, and there are others of you.


I just want to say that the mediators that work with us truly, truly believe in the transformative model, and work that transformative model with the thought in mind that they want to help us improve our workplace environment. REDRESSTM has gotten national recognition. We have an individual from the headquarters REDRESSTM team that was invited to be on Vice President Gore's task force on reinventing the government; he worked with that task force for over a year.


We received from the American College of Civil Trial Mediators, the Institution Award of Excellence. That was in April of 2000. The Office of Personnel Management (‘OPM') gave us their first award for Outstanding Alternative Dispute Resolution Programs in 1999. Last year, we also received an award for an Outstanding Mediation Program. Our first year, we were recognized by OPM for our Research and Evaluation that Lisa did for us. Just last week, by the way, the CPR Institute recognized us with an award. So I can't tell you what importance Lisa does play in our REDRESSTM Program. I will say this: it's very hard to get money internally for a project when you work for the Post Office. It's very difficult and this was a new project, REDRESSTM was a new project, and Cindy was pushing to have mediation initiated across the country and she had to talk to the high-ups, to the Postmaster General and his VPs. She had to get money to establish this program. The way that Cindy was successful in doing that, the reason that I am able to stand here today and speak to you, is because of the research that Lisa did for the Postal Service. Because Lisa put it all together in a package for us to show us what each individual at the table was doing, what each individual at the table was reporting about their mediation experience. Lisa gathered thousands and thousands of bits of information and put them together in a nice package for us so that we could present information on how the program is going, as we need internal support. I just spoke about the awards that we are getting, we are now getting external support for our programs, so what Lisa is going to tell you today is really very important and something that continues to grow for us.


Thank you very much for inviting me here today and here's Professor Lisa Bingham.


LISA BINGHAM:  Thank you very much; it's a real pleasure to be here today. One of the reasons we are here today is because of the CPR Institute Award. Last week I didn't know about it, last week it was a big secret. It was very exciting to hear about it for the Postal Service. It's also a pleasure to be here today because Cardozo is my favorite cousin, Dana Davidson's alma mater, and I wish she could have been here this morning.


The research I'm going to talk about has been the most exciting intellectual experience of my life and whatever the Postal Service feels its benefits have been, they are dwarfed by the benefits that I personally have had working with just an amazing, imaginative bunch of people. The creativity and the guts it took to pull this project off continues to amaze me. We started the evaluation at the birth of the program. Cynthia Hallberlin and Karen Intrater, who is currently Managing Corporate Counsel of the USPS Law Department, came down to the Society for Professionals in Dispute Resolution (formerly SPIDR, now ACR, the Association for Conflict Resolution) meeting in October of 1994 to meet with me. They had built the mediation process into the consent decree that Pamela was talking about, and they knew that if they didn't get someone outside the organization to take a look at the program as a pilot in the Florida Panhandle – to see how it was working and be able to report objectively back to the organization – it would wither and die. There would be nowhere to go from there.


In 1994, I was studying employment arbitration, and I was paying to get my data – copies of arbitration awards. The USPS Law Department had no budget for evaluation. Over dinner at SPIDR, we discussed the possibility of my doing research on the new mediation program. Cindy and Karen turned to me and said, ‘Look, how much of a sample do you need?' And I said, ‘I don't know, if you could get me a sample of maybe 200 cases?' They both grinned, exchanged a meaningful look, and said, ‘We can get you thousands and thousands….' At that point I knew I was in deep trouble! We started the evaluation by setting up systems for continuous, contemporaneous data collection, specifically participant exit surveys completed at the end of the mediation session, and data tracking reports completed by the mediators.[1]


The evaluation now has three basic components: process, outcome, and quality assurance. [see slide] First, we look at the process of implementing the program, called process evaluation. As my colleague Charles Wise likes to say, you can't do an evaluation of the program unless you're sure there's a ‘there' there. Second, we look at the outcomes that are generated by the program. Finally, we try to monitor quality control. The program has grown from the pilot in the Florida Panhandle, to expanded pilots across 27 cities using different models, to the current nationwide model that uses transformative mediation. As it was rolled out nationally, we expanded data collection.


On the process evaluation, we look at implementation. [see slide] The USPS maintained much of this data itself. The implementation took place over an 18-month period. The program is now 100% implemented. Implementation included things like: training supervisors in what transformative mediation was and conflict resolution techniques; training key stakeholders like union leaders and managers in every location; having stand-up meetings with employees to let them know about the program; and getting videotapes out on an internal education network that the Postal Service uses. This was not a case of the inverse ‘Field of Dreams' effect (we build it but they do not come). The Postal Service focused on trying to get people to come during the implementation phase. One of the miraculous things that they did involved goal setting. Some federal agencies set as a target or goal a specific settlement rate. They actually put into their contracting with outside mediators that they had to generate an 80% settlement. The Postal Service did not do that. Instead, it set as an internal organization goal, a participation rate of 70%. The participation rate last spring was 72%. The organization made its goal. Kevin Hagan from USPS headquarters said last week that participation was up to 74% and even 82% in some areas. So the USPS built it and people are coming. This means there is a ‘there' there, a program to evaluate, so we can start to measure outcome.


However, before we measure outcome, we also need to have some assurance that the model that the USPS implemented was the model the USPS had designed. To do this, we did an electronic survey of the program, coordinators and specialists, the people in jobs – like Pam's job – to tell us what they had observed.[2] We asked them to give us examples of what mediators were doing that fostered empowerment, fostered recognition, interfered with empowerment and recognition, or were directive, or evaluative. We wanted to make sure that the examples that the specialists gave us mapped appropriately to what the transformative mediation model indicated should be happening.


We then looked at the key hallmarks of transformative practice and found that there was a close correspondence between the specialists' examples of what they had observed the mediators doing that was transformative or directive and what the model defined as transformative or directive. Since it is the specialist's job to screen the mediator roster to make sure that the mediators are practicing the model, this is an important step.[3] The initial roster was about 3,000 mediators and in the past year or so, it has been reduced by half to about 1,500 mediators. Choices about who stayed on the roster were based partly on specialists' observations working with mediators and determining who is comfortable practicing the model and who is not. During that initial period of time, the USPS had to give people a chance to see if they could practice this model or if they were really more comfortable being directive or evaluative. For those of you who are not familiar with the terms, a more directive or evaluative mediator tends to look at what the case is worth and try to give people an assessment of what would happen in litigation. The directive or evaluative mediator would then try to steer the disputants toward a settlement, one that the mediator thinks is an appropriate settlement.


This is not what we wanted happening in the transformative model. This is perhaps an appropriate time for me to digress as to why we did not want evaluative mediation practice happening at the USPS. The easy example is to think about the macro level. This is why this project is so intellectually exciting for me. You have 800,000 employees, you have tremendous case flows, and you have these case flows at the macro level turning into something that is an organizational level of response. So think about 800,000 employees at the time the program started filing approximately 28,000 (it is dropping now so maybe it is 25,000 or lower) informal EEO complaints a year. The Postal Service was prevailing in 90%-95% of all of its complaints. By prevailing, I mean that the case either went away or was won on the merits because it was not within the scope of discrimination laws or there was no legal cause of action. If the Postal Service had chosen an evaluative or a directive model, one where the mediator had the authority to give people an opinion in the case, then 90 to 95% of the time that mediator would be telling people, ‘You don't have a case.' Given the fact that the Postal Service was paying the mediators, establishing the roster, training the mediators, assigning them to cases, and that this was an entirely voluntary program for the complainant, just imagine what happens when the first few employees go back to the workroom floor after going through mediation. They would likely say something like, ‘That mediator is biased in favor of the Postal Service, I didn't have a chance, it's a worthless process, don't use it.' In which case, nobody comes. In which case, there's no ‘there' there to change the organization.


So, for entirely strategic and pragmatic reasons, you needed to have the transformative model in this setting. The Postal Service was providing the infrastructure for the mediators, but if the mediators told the complainants 90 to 95% of the time that there was no legal case, the program would look biased to employees even though it was absolutely fair. Even if we assume for the sake of argument the mediators were absolutely on target when they told the employees that they did not have a viable case, an evaluative mediation program most likely would not generate as high a level of participation as the transformative model has done.


I digressed. I won't do that again. Let's talk about who comes to mediation. By and large on the complainants' side, these are people who are craft employees. [see slide]  There are a few people who are temporary employees and not Postal Service employees at all, and there is a small group of people who are not in bargaining units, eight percent approximately, who come to mediation. On the supervisors' side, 97% of these participants are not bargaining unit people. [see slide] Again, another one of the exciting things about this program is that it coexists with strong labor relations and local collective bargaining units. An employee who comes to this program can use REDRESSTM and file a grievance under the collective bargaining agreement and be represented in both cases by a union representative if they so choose.


Let's move from process evaluation to evaluating outcomes. I am just going to focus on two of the outcomes that have been the source of publicity. [see slide] The first one you might have seen in a New York Times article last fall having to do with the formal complaint filing. Second, we are going to talk a little bit about the transformative indicators on the exit surveys. If you are mediators in the program, you have been distributing exit surveys to participants for the last couple of years. There are now about 55,000 exit surveys in our national database. The theory that we brought to bear in order to look at what organizational level outcomes would occur as a result of this program came from a variety of disciplines. For those of you studying dispute resolution, you probably already know that this is a highly interdisciplinary field. On the research side, it is all over the lot. We have, of course, a lot of wonderful scholarship in law journals including your own Online Journal, and we also have scholarship in the procedural justice literature from social psychology. We have communications literature and clinical psychology literature. We have people in anthropology and sociology looking at the programs. We have economists reducing everything to rational choice. Everybody is looking at the mediation process from a different perspective.


In this first outcome study on formal EEO complaint filing rates, we looked at literature coming out of the industrial relations discipline on employee voice systems and at studies of court use of ADR.[4] We also looked at relevant literature on grievance mediation, where parties to a collective bargaining agreement use mediation as a step in their grievance procedure, and at a new, developing dispute system design literature on the idea of intervening at a very early step in the life of conflict. One thing you need to understand about the REDRESSTM Program is how early the intervention occurs. The Postal Service designed the program; I had no part in designing the program. The designers within the organization had the concept of getting people to the table as a structural step of the program within two to three weeks of a complaint; this was a critically important design choice. In other words, mediation does not have to be fast, in and of itself. A number of programs in the court setting have mediation occurring relatively late in the life of the case. In contrast, the Postal Service chose to marshal all the organizational resources to get people to the table within two to three weeks; this is what makes resolution potentially faster, this choice in the dispute system design. We also have some literature on experiments in federal and state courts with dispute resolution in the court-annexed setting.


These literatures were a little bit conflicting, but when you put it all together, there was some evidence that we might have an increase in case filings as a result of a better employee voice system. There was some evidence that we might see a decrease in formal complaints because early intervention would get people to the table soon. On the other hand, there was also a study done in New Jersey on court-ordered arbitration – non-binding arbitration – that found that the time to resolution increased because the mandatory arbitration step was scheduled six or more months after complaint filing. Everybody gamed that system so as not to settle the case until the night before arbitration, and that actually prolonged the dockets. We were not sure what to expect. Our hypothesis was that the formal EEO complaint filings would decrease.


This study demonstrates why it is totally cool to work with the Postal Service. I have students who have gone on and are working in the U.S. General Accounting Office (GAO), for example, and helping evaluate the various federal agencies from there. Evaluators know that it is rare to have an organization that has amazing information systems organized by zip code and maintained the same way all over the country, organizations where all the information you need is in one place. The Postal Service can provide us with a printout with national data on EEO informal and formal complaint filings, and it updates the printout every four weeks, which is an accounting period. It is just glorious from a researcher's standpoint. For this study, we have 5 and one-half years' worth of data in the database by accounting period for every four weeks by zip code. For each zip code and each accounting period, we have data on how many employees were on staff at that point, how many informal EEO complaints were filed, and how many formal EEO complaints were filed. We knew from other information during precisely which accounting period in that zip code REDRESSTM was fully implemented, that is, 100% implemented. This means that we can control, in what is called a multiple regression, for what was happening before and after implementation of REDRESSTM in every zip code in the country. [see slide]


This next slide may not look like much to you but what it shows, with the lower line, the rate at which formal EEO complaints increase with number of employees. The lower line in the graph, the pink line is after REDRESSTM, the higher line is before REDRESSTM. What that demonstrates is there was a statistically significant drop in formal EEO complaints, it averaged out to be 1.88 for formal EEO complaints per zip code per accounting period nationally, after REDRESSTM. We first did this analysis in August of 1999. We documented a drop in formal complaints of about 2,000 and the slope was still going down so the complaint filings were still dropping. In 1999, we projected that based on full implementation there would be another drop of about 2,000 cases – that is the significance of the slide that Pam showed you. The formal complaint filings continued to drop. At some point you will get leveling off. You will end up with a base line of steady formal EEO complaint filings, and you need that because the complaints that have merit should be going forward. You do not want to eliminate all formal EEO filings because you are not trying to eliminate access to justice; you are trying to resolve conflict that can be resolved by people talking.


This next slide shows where we were when we first did the analysis in 1999 and what has now happened in 2000. [see slide] As Pam previously showed you, formal EEO complaint filing is down from over 20,000 to 9,000 annually. Formal EEO complaint filing was at a high in 1997. In 1997, REDRESSTM was only implemented in about 27 cities, and you had a high of 14,000 formal EEO complaints a year.


What happened in 1990? There were a series of changes in the Postal Service including increased automation, but also there was a statutory change that authorized punitive damages in 1992, punitive damages in the amount of $300,000 for employees that filed EEO complaints. (Is that correct Pam?).


PAM ZUCZEK:  In 1992, EEOC allowed people to recoup compensatory damages up to $300,000 per allegation if they were successful in their allegation of discrimination before the EEOC, so I think that's what generated more activity that you are looking at on the screen.


LISA BINGHAM:  We did not go back to 1990 in our analysis. We could go back and get data and control for the date of that statutory change and that would be interesting to look at. What our analysis showed was that the difference in formal complaint filing rates before and after the implementation of REDRESSTM correlated with the accounting period in which that zip code achieved full implementation of the REDRESSTM Program.


To continue with our second example of outcome evaluation, let's talk now about the transformative indicators on the participant exit surveys. Reducing formal EEO complaint filings was not an explicit goal of the program. There were other objectives. The primary objective was facilitating communications between employees and supervisors. The participant exit surveys are completed at the end of each mediation session. The surveys contain a number of items from the procedural justice literature of social psychology and items that we have tried to develop to look at transformative mediation in particular. We have about 55,000 in the database now.


In addition, there are a number that have not been entered into the database yet. I would say there are about 60,000 total at the Indiana Conflict Resolution Institute today. We generally will look at these surveys comparing the experience between employees and supervisors because we want to look at the relationship between the two groups. The table that you see on this slide shows the percentage of the participants that agree or strongly agree with certain statements on the exit survey and are divided into the employee column and the supervisor column. [see slide] One item that was important to us was, ‘The other person listened to my views.' The fact that the percentage of participants who agree or strongly agree is exactly the same number for employees and supervisors indicated to us that the mediators are facilitating a bilateral conversation. The employees and supervisors listened to each other. Interestingly, the numbers are also very close on the item, ‘The other person learned something about my point of view.' This suggests that employees and supervisors both feel they are getting something out of the conversation. The item, ‘The other person is learning something about what I think,' that is running at about 61 to 63%. The last one, ‘I learned something about the other person's point of view,' again employees and supervisors are comparable 59 to 65%. It is interesting that more of the supervisors, 65%, say that they are learning something about the employee's point of view. This would make a certain amount of sense, because one of the complaints was that people just were not talking things through when they happened on the workroom floor. The supervisors may not have time to sit down and listen to the employee when they are trying to get the mail out.


AUDIENCE QUESTION:  Is the 6% significant (between the 59% and 65% viewpoints)?


LISA BINGHAM:  Yes. The problem with the data set is that it is so big that almost any difference achieves statistical significance. Then the question is, despite the fact that the difference is statistically significant, is it really meaningful? What and how much meaning do you attribute to it? With 55,000 exit surveys, the power of the analysis is such that almost any measurable fluctuation is going to be statistically significant.


AUDIENCE QUESTION:  Do  you have any theory about why that third area only hits that bigger difference? I would conjecture that employees would learn as much from the supervisors and vice versa.


LISA BINGHAM:  I have a mediator nodding in the back of the room and I would rather invite someone that has mediated in the program to answer the question if that is all right – go ahead.


MEDIATOR IN AUDIENCE:  I think that employees have an opportunity routinely in service talks and presentations to hear very clearly where the supervisors are coming from. I think that mediation exists as a powerful opportunity for supervisors to hear a wide range, far beyond other issues.


LELA LOVE:  It is my experience that conversations are quite personal, so what the supervisor reveals about why he or she did something is interesting.


LISA BINGHAM:   We have some indicators that have to do with the things that the mediators themselves did and we asked people on the exit survey to what degree do they agree or disagree with these statements ‘The mediator helped me qualify what my goals were, what my choices were'; ‘The mediator helped me understand the other person's viewpoints'; ‘The mediator helped me understand my viewpoint'. These percentages report the proportion who agree or strongly agree. As you can see, between 79% and 83% of the people experiencing mediation, agree that the mediator helped them clarify their goals and choices, which is an objective of transformative practice. That is an aspect of empowerment for the disputants. The item, ‘The mediator helped me understand the other person's viewpoint,' is evidence of mediators facilitating recognition. This is another objective of transformative practice. Here, we have basically two-thirds of the participants agreeing that the event happened. The next item is, ‘The mediator helped the other person understand my viewpoint.' This shows again 60% or 61% of the employees and supervisors respectively agreed that this event happened.


On the recognition side – this slide is the most interesting and intriguing – the employees are reporting and agreeing that 30% of the time they received an apology, that the other person had apologized to them. [see slide] Now think about all the other models of mediation and what happens in the court setting. In the court setting, lawyers are more likely than mediators to steer clients away from apologizing, because it may represent an admission against interest. It may have all sorts of consequences that are not intended. However, in this model, an apology is a wonderful example of recognition, of one person giving recognition to the other party. Employees say the other person apologized about 30% of the time, and if you look on the diagonal, you have supervisors saying I apologized to the other person about an aspect of the problem about 32% of the time. Those numbers corroborate each other. It gives me reason to believe that it is really happening, since employees and supervisors report agreeing or strongly agreeing that it happened at about the same rate. Again, it is a very large sample size. On the ‘I gave an apology' side of things, you notice a statistically significant difference between employees apologizing to the supervisors and the supervisors hearing the apology. That is interesting. To get at why that is happening might take some interviews. I welcome some explanations for why that might be happening. That is about as far as I can take that right now. I don't see anyone raising their hand to give me an answer.


LELA LOVE:  Perhaps supervisors are more trained as managers and their apologies are more articulate. This is a wild guess, maybe when an employee apologizes it's not as clear.


LISA BINGHAM:  Either it's not clear or the supervisors are not hearing. I see another hand in the back.


ANSWER IN AUDIENCE:  I suggest that the numbers should be even higher because at the end of the mediation, after the surveys are filled out, there is a lot of small talk. I often hear at that point, ‘I'm really sorry, etc.'


LISA BINGHAM:  Isn't that interesting. I see all the other mediators nodding.


On the mutual understanding piece, we have again comparable numbers of employees and supervisors agreeing that ‘The other person acknowledged their perspective, their views.' They are saying they received recognition from the other person. [see slide] That is a fairly high number, and that is further evidence that the mediators are doing what they are supposed to be doing. Also, we have somewhat higher numbers of people saying, ‘I acknowledged the other person.' I think in this case the difference is explained by the notion that it is harder to hear somebody acknowledging you than to feel that you have done it. Again, we find a similar pattern and an intriguing difference.


We also have some indicators in the exit survey to see if the participants' perceptions are that the mediators are doing things in the model that the mediators should not be doing. [see slide] Most of this data is from the period of implementation when we still had 3,000 mediators on the roster, before the specialists screened and reduced the active roster. There is some evidence here that there has been some digression from the model during implementation. For example, we have some participants agreeing that ‘The mediator told me who would win this case if we were in court.' This should not be happening at all, should not be happening ever, in the transformative model. Nevertheless, eight to nine percent of participants, both employees and supervisors, agree that it is happening.


A second item on the survey also tests for directive mediators – ‘The mediator pressed me to accept a solution I was uncomfortable with.' Again, this should not be happening. Both employees and supervisors agree it is happening. The employees say it is happening 14% of the time; the supervisors say it is happening about 10% of the time. That is a statistically significant difference. One explanation concerns what might be happening in the caucus. Recall that 90% to 95% of most EEO complaints get disposed of without litigation, that is, complainants usually do not prevail on the merits. You can very easily visualize a mediator who is disposed to lean on people, leaning more heavily on the complainant in caucus. A settlement-oriented mediator most often would conclude that the complainant does not have as strong a case as the respondent, at least from the standpoint of discrimination law, and thus, for that mediator, it makes sense to lean on the complainant. If it is happening in caucus, you could understand why complainant employees would report it more frequently than respondent supervisors. Again, it is digression from the model, and it should not be happening, so it is something we're continuing to track. It is a pattern that we hope to see change now that the roster has been further screened.


The next group of slides has to do with quality assurance and the third piece of the evaluation program. In terms of quality assurance, I just wanted to talk to you about three studies that are ongoing: The first one you have probably seen results of before – which is the overall satisfaction of the parties with the process, the mediator, and the outcome; the second looks at what happens when people bring representatives to mediation sessions and the different kinds of representatives that they can bring to the table; and the third one compares the inside mediator model that was one of the pilot models during the 27-city phase of experimentation, with the outside transformative mediator model implemented nationally, in terms of participants' perceptions of procedural justice and satisfaction. [see slide]


Let's talk about procedural justice, an amazingly resilient, well-established social psychology theory. A number of researchers have worked on this. Tyler and Lind[5] assert that basically people value a process based on what it shows about their role in the social group, called group value theory. This theory suggests that the important components of a process to settle disputes include things like the opportunity for voice or participation in the process and the participants' control over the process, among other elements. You can readily see this aspect of procedural justice and group value theory is directly related to the transformative model of mediation and its notion of empowerment of the parties. Another important element of procedural justice theory is the respect with which participants are treated in the process. Again, you can readily see the connection between this notion of respect and recognition in the transformative model. So, there is really a close correspondence between procedural justice theory and some of the elements of the transformative model.


In the exit surveys, we have all the traditional procedural justice indicators, and we have grouped them for analysis purposes into three big indices: one on the process, one on the mediators and one on the outcome. This slide represents all the process indicators: how satisfied are you with the information you got about the process, your control over the process, with the fairness of the process, your participation in it, your understanding of it and your treatment in the process? [see slide] These bars show the percentage of each category of participants who report they are satisfied or very satisfied with this aspect of the process. You have three bars in this chart. The green bar is supervisors, the blue bar is employees, and the red bar is employee representatives. These representatives can be attorneys; they can be union representatives. In fact, many people bring no representative and come to the table pro se, and then some people bring friends, relatives. People have brought their pets. They can bring anybody to the table that they feel the need to have there.


What is interesting here is how high the numbers are. You will see they range basically from 87% to the high nineties. The lowest number on the chart is the supervisors' satisfaction as to control over the process; 87% of them or higher are satisfied with their control. That number is pretty miraculous considering the supervisors have no choice but to come to the table. I mean the supervisors are mandated to come. For the employee complainants, mediation is voluntary, but the supervisors do not get a choice because they are the responding authority for the Postal Service. They must come to the table, but they need not agree to anything. Therefore, the fact that that number for control is relatively high is a very good sign. It is indirect evidence that the mediators are succeeding in fostering empowerment of even those parties for whom mediation is mandatory. For everybody else, we are looking at 90% to 98% satisfied or highly satisfied with all of these indicators.


This is not that different from satisfaction rates with mediation in many other settings, so it is not a particularly dramatic result from that standpoint. What is really cool about the numbers is that we can break them down by zip code. We can watch what happens in each district and it is so neat, the power of statistics. As soon as you have a reasonable sample size, you may have a little variation in the first couple of samples because they are small, but as soon as you have 100 exit surveys in that zip code, you have this pattern and it is just consistent.


I get this question all the time from the mediators: ‘why can't I or when can I see my own numbers?' It is true that it is conceivable we could analyze the database by individual mediator. However, that was not the purpose for which the national database was established. The national database was established as a program evaluation tool to make sure that this zip code looks like that zip code, and they both look like the next zip code. As long as the pattern is relatively the same by zip code, then we have consistent implementation nationally and consistent results nationally. That means there is no reason to go into that zip code and break down the print out by mediator. If we had a big disaster, if some particular zip code region tanked in terms of its numbers, then in theory we could go in and break it down by mediator and see if we had a problem mediator. However, I think that the mediators should find it oddly comforting that we are not doing that – that big brother is not actually out there looking at your individual numbers. Now, another related question mediators have asked is: ‘can we get them privately?' In other words, could we do printouts for the mediators but not for the Postal Service. This research has been done on a shoestring. We have major support from the William and Flora Hewlett Foundation, and we have support for the direct costs of data collection from the Postal Service itself. We have students entering the exit surveys. Each time we do a national analysis, it takes 40 hours just to do the printouts, so to do an individual mediator's run would be very time consuming.


PAM ZUCZEK:  I would like to point out one thing about this particular slide. [see slide] Lisa told you that our supervisors and managers are mandated to come to the table. This process is voluntary for the complainants, the persons who are filing the EEO complaint and requesting mediation, but for the respondent, the supervisor, and the manager – it is mandated for them to be at the table because that is part of their managerial responsibilities, to show up at a mediation when they are part of the conflict. What is remarkable about this is that it speaks to the fact that our mediators truly are using transformative mediation, because imagine a supervisor being mandated to come to the table and then filling out a directive, then the mediator is saying you win, you lose, you're right, or you're wrong. Our supervisors are reporting that 80% of our supervisors are satisfied or very satisfied with the amount of control they have over this mediation process, although they're mandated to go there, saying I feel I have control in this process. Also 95% of our supervisors are reporting that this mediation is a fair process – it is voluntary for the complainant and they knew they were mandated but something happens in that room that 95% of our supervisors can report that this is a fair process, so we look at those numbers and we know what is happening in those rooms. I just wanted to point that out.


LISA BINGHAM:  The next slide is satisfaction with the mediator and again, the numbers are very high, running in the high nineties generally speaking. [see slide] This result is stable, absolutely stable, nationally. The result is stable not only for the employee but also for the employee representatives, when about 40% of them are union representatives. This next slide shows our outcome indicators, and you will notice that the numbers are lower. [see slide] We have numbers ranging from a low of about 65%, up to a high of 79% on various indicators of satisfaction with the outcome of mediation. The fact that these numbers are lower than the numbers we have just seen for the process and for the mediator is absolutely consistent with other procedural justice studies on grievance mediation in labor relations and dispute resolution in the court-annexed setting. People, generally speaking, consistently rate their satisfaction with the process and their mediators higher than their satisfaction with the ultimate outcome. That said, these are still good numbers given that we are talking about 2/3 of both sets of disputants being satisfied or highly satisfied. Now, there are differences between the employees and supervisors, and these differences are statistically significant. You can see there is satisfaction with outcome overall among 71% of supervisors while 69% of employees report being satisfied. As to speed, 71% of supervisors and 73% of employees are satisfied. We have 79% of supervisors and 73% of employees reporting satisfaction with their control over mediation outcome. In another outcome indicator, 71% of supervisors and 65% of employees report satisfaction with the long-term effects of mediation, and this difference is also statistically significant.


The result is that, as to the outcome of mediation, the employees are reporting satisfaction at a slightly lower rate than supervisors, but that the differences are statistically significant. This pattern as to the outcome is consistent also with all of the procedural justice studies on grievance mediation and court annexed dispute resolution. An explanation has to do with expectations. Generally speaking, in all these processes and other studies, the moving party or plaintiff is slightly less satisfied with the outcome than the responding party or defendant. This is probably because the moving party is the one going into the process trying to get something, and generally what they are getting is not what they hoped to get. Meanwhile, the responding party is the party whose expectations are depressed because someone is suing them, so the fact that they are getting out of this dispute at all is a bonus. For the responding party, it means that things are turning out much better than they expected them to. That is pretty much the consistent explanation that procedural justice scholars have given for this result.


AUDIENCE QUESTION:  Has anyone ever compared mediation with arbitration in terms of participant satisfaction?


LISA BINGHAM:  That comparison has been done by Deborah Shapiro and Jeanne Brett in an article from 1993.[6] It is a wonderful article doing the comparison in the labor relations setting. My recollection is that there was grievance mediation and arbitration, and the participants were more satisfied with mediation than arbitration.


AUDIENCE QUESTION:  With respect to the 14-day turn-around period, why wouldn't the number be high under the speed category?


LISA BINGHAM:  Part of that may have to due with the backlog of the cases before the program started. (Pam what do you think?)


PAM ZUCZEK:  Our goal is to get everyone to the table. Some take a little longer, some take 21 days; and then there are cases where somebody is on vacation or sick and it might take a little bit longer. I can only suppose that why we are getting a 70% reading on that is that when you are in a conflict, you want it now. We actually have some people calling us on the phone wanting to be in mediation and wanting to resolve their conflict that afternoon. So when we make you wait 14 days, that is difficult – especially when you are working with that person every day, it doesn't seem like 14 days, it seems like forever because the person you are in conflict with is still your supervisor – you need to get that resolved. I think that's why they're not saying this is the best thing that has happened. We think that is why it is 74% and 73% reporting that they are happy to us internally. This is a good number because if we ask them about the timeliness of the EEO complaint processing itself, where it could take a year to accomplish what they have accomplished here in 14 days, we would not be looking at a percentage in the mid 70s. So, it's sort of relative.


AUDIENCE REMARK:  Looking at the employee representative category, I'm glad to see that there is more importance found in the satisfaction of the outcome. There may be a variety of reasons for this; the representative may be more objective to things and the representation presents new opportunities.


LISA BINGHAM:  There is a possible two bites of the apple in terms of a collective bargaining agreement, but there has to be a cause of action under the collective bargaining agreement, for the grievance procedure to legitimately be invoked. So for that employee where there really is a communications issue that is outside the scope of the language in the collective bargaining agreement, that person may typically go to the EEO process. Now the union representative can represent them in mediation and provide additional services.


Now let's turn to the representative profile, who employees bring to the mediation table to represent them. Notice the biggest category is Union Representative 43%, then No Representatives 33% of the time, Fellow Employee 13%, Others (like mom, dad, friends) 13% of time, and attorneys 3% of the time. [see slide] Now actually, the fact that only 3% of the cases involved attorneys is consistent with the notion that 90% to 95% of these cases are not strong on the merits. It stands to reason that someone could only get a lawyer to represent them if the lawyer decided the case was strong enough that he or she could recover a contingency fee.


This table, ugly as it is, shows the breakdown of resolution rate by representation category. [see slide] The top line, the red bar, shows the percentage of cases reported as totally and completely resolved on all issues. The blue bar underneath shows the percentage of participants reporting that the case was partially resolved. Just to clarify, these percentages again represent the reports of the participants as to whether they felt the case was fully or partially resolved. You should know that the supervisors more frequently report that the case was fully resolved, and employees more frequently report that it was partially resolved. There is a difference in perceptions sometimes. This breakdown does not reflect employees and supervisors separately; it is the total pool of actual complainants and respondents, with the representatives taken out of the sample. What is interesting to me is that the rate of partial settlement is about the same throughout the representation categories, running between 32% and 35% – all those little blue bars pretty much line up. However, if you look at the attorneys at the bottom of the list, you will see that a full settlement with an attorney is happening at about half the rate of full settlement with union representatives or with no representative, and those last two categories are very close. In other words, attorneys get a full settlement maybe 15% or 16% of the time, while in the union representative or no representative categories, you get a full settlement 29% to 30% of the time. Now, this result makes sense if you assume that the cases with attorneys are the cases that are the strong cases. If they are the strong cases, then from a justice perspective, you want them moving through the process, and you want them to get addressed on the merits because they might not get addressed fully in mediation, depending on what the parties' needs were. It is very interesting that there is a difference.


The other difference that we are starting to see – that I do not have a slide on – is that the employees are statistically significantly less satisfied with both their participation in mediation, and their control over the mediation process when they bring a lawyer to the table. The result makes sense, because the lawyers are going to be taking control of the case, doing the talking, and making sure their client does not make any admission that the lawyer does not want the client to make. These are all the things that lawyers do unless they're trained in mediation advocacy, and trained in particular in mediation advocacy in a transformative model. I think it is a fairly rare training for lawyers to have, although maybe Cardozo is doing it.


AUDIENCE QUESTION:  Those rates are reflected by whom?


LISA BINGHAM:  These are the rates at which the parties report in exit surveys that the case is fully or partially resolved. We have a separate data set to corroborate the parties' reports; this data set consists of mediator reports. Mediators complete a data tracking form at the conclusion of each mediation case, to let us know how many exit surveys they distributed and what the outcome of mediation was. We can also tabulate these mediator reports in terms of representation category, so we can do a double-check. There is a little fluctuation that may have to do with response rates. We do not have a 100% response rate on the exit surveys; the response rate on the exit surveys is running over 70%, which is very good. We have close to a100% response rate on the mediator reports, because mediators must complete it as a condition of getting paid, so there are differences. We can keep track of our response rate on the mediator reports by checking them against an independent data set at headquarters from the specialists.


I would like to move on to our third quality assurance study, comparing inside neutral mediators and outside neutral mediators.[7] [see slide] An inside neutral mediator is someone who is an employee of the Postal Service, on the payroll of the Postal Service, potentially in a bargaining unit, who has been trained in mediation. The outside neutral mediators are independent contractors, as many of you know, who are not employees of the Postal Service. The inside model was used in Rochester, New York. This data came from the Rochester, New York area, and it was Mary Trabert's program. She was one of the first specialists – and some of you may know her – she was just terrific. The inside neutral mediators, I have to say for the record, were terrific people, and did a great job. The inside neutral mediators screened cases; they picked the cases they felt were more amenable to mediation, cases that might be more likely to settle. They also brought co-mediators when they felt it was necessary in terms of equalizing people's perceptions of their personal empowerment at the table. They would bring in both a bargaining unit and non-bargaining unit mediator.


Notwithstanding all these efforts, they were not able to compare favorably to the outside neutral mediators. What we found was that the satisfaction with mediator impartiality and settlement rates were lower in the inside model, and the differences are statistically significant. These are not huge samples; these are samples sizes of between 100 and 200. In satisfaction with outcome, there is a bigger difference, about a 15-percentage point difference between 61% of the parties reported satisfaction with the outcome with inside neutrals, versus 76% as to outside neutrals. Again, for participants reporting satisfaction with long-term effects, there is a statistically significant difference of 61% versus 72%. This last one was the killer – people reporting that the case was resolved in full or in part. In the inside neutral model only 56% of the time did parties report that the dispute was resolved in full or in part, whereas in the outside neutral model 75% of the participants reported full or partial settlement. This was despite substantial selection bias that would have predisposed the inside model to do better. What I mean by that is in the inside neutral model, the program administrators were intentionally identifying cases to go into mediation that they thought would settle. In the outside neutral model, there is no screening except for very rare categories of cases involving violence or certain kinds of egregious sexual harassment, am I right Pam?


PAM ZUCZEK:  Correct.


LISA BINGHAM:  In other words, everything – a much wider variety of cases – is coming into the outside model. You would figure, with this selection bias, there would be a lower reported resolution rate in the outside model; however, it was dramatically higher. This was part of the evidence that headquarters used to decide to go with the outside neutral model nationally. Notwithstanding the fact that, in theory, the outside neutral model was more expensive, in practice, I am not sure; because if you get such dramatically different results, the outside neutral model may pay for itself. We are at the end. I will review just the highlights. We have fewer formal complaints after the program. [see slide] We have more effective outside neutral mediators than inside neutral mediators. We have higher resolution rates with union representatives. In the private sector, a lot of employers are using an employment dispute resolution program with mandatory arbitration as a way of forestalling union representation. This data suggests that employment dispute resolution may coexist with union representation in a constructive way. It's just interesting to me that you can have a very effective role for the union representative in the employment dispute resolution program, and that employment dispute resolution need not be conceived as an anti-labor relations concept. Then we have what we think is the beginning of evidence that REDRESSTM is improving communication.


There are other studies that are ongoing that include interviews before and after implementation of the program. We did a random sample of employees in the production facility in New York before the program was implemented and then again last spring, and that data is being coded and analyzed as we speak. We just sent out a survey nationally to the 3,000 mediators on the national roster to get feedback about their experience in the program. All the studies I just described based on exit surveys and participant satisfaction are continuing. We plan to work longitudinally to determine how the program performs over time, because there is some literature that suggests that a new program gets a honeymoon and then performance drops off after a period of time. We have not seen that with REDRESSTM, and we have been finding stable results in different places for six years now since late '94, when the pilot program was first implemented in the Florida Panhandle. So I will be happy now to open the floor to questions.


AUDIENCE QUESTION: Before the program was implemented, how did the union reps feel about it and do they feel the same way now?


LISA BINGHAM: Very interesting question, constantly in flux – depends on where we are. First, how the union representatives feel about it depends on whether we ask the question on the local level or at the national level. Second, it depends on what bargaining unit you are asking. Since this is a political question, I am going to have Pam answer it.


PAM ZUCZEK:  Ask the question again.


AUDIENCE QUESTION:  I wanted to know how the union reps felt about the program before it was implemented and if it has changed since.


PAM ZUCZEK:  The one thing that I can say is there was not a lot of support from our unions, and I believe there are several reasons why in the beginning. Probably one is encroachment, because if you are the only game in town that sometimes feels good. A second is a lack of information and we had to do educating. We had to partner with them; we had to let them know that the collective bargaining agreement is their arena. They are there to protect that agreement; we had to let them know that we were not encroaching on their territory and that they typically were representatives in the EEO complaint process even before we were talking about REDRESSTM. And we had to let them know that their part in playing representatives in the mediating process would not change; it might just be enhanced. Let me just say that there was skepticism. At this point, the figures that you saw were high positive figures for representatives, and most of those reps were union reps reporting those figures. So you can say from a skeptical union or unions and even from unions that are saying to their people "do not participate in REDRESSTM," they are going to the table and reporting high levels of satisfaction with the process, they are partnering with us. They are telling their people to come to REDRESSTM. They're selling REDRESSTM for us and we think that is the biggest thing that I can say. We have union reps selling REDRESSTM because they are satisfied with the process.


AUDIENCE REMARK:  Maybe because they are looking more effective to the employees, like sometimes when they go to arbitration they are not quite sure which way to go because the arbitrator may make decisions that may go completely against the employee's decision.


PAM ZUCZEK:  So in other words, your saying that they feel more control in this mediation process than in going to arbitration and they well should feel that way because they are part of the decision-making process.


AUDIENCE QUESTION:  Earlier you mentioned a 30% to 40% settlement rate and the latest slide showed a 70% satisfaction rate with outside mediators, why the difference?


LISA BINGHAM:  You need to understand that there are different data sets and different ways of reporting. The 56% to 75% numbers that you saw on the slide by comparing inside and outside neutrals, were calculated by adding together people who reported resolution in part or resolution in full, so that was the combination of part and full and that was exit survey reports from a sample. The closure rate that Pam reported is completely closing out the case at different stages of the process. This number is based not on the participants' perception and their reports of closure, but on the receipt of documentation closing out the EEO complaint by somebody in Pam's office. So that is more objective, less perceptual data. Then the slide that showed the breakdown in settlement by category of representative was an analysis done on something like 10,000 of the exit surveys. That sample put together employees and supervisors; it consisted of the reports of employees and supervisors deciding subjectively whether they think that mediation resolved the case in full or in part. Again, there is a little bit of disconnect. There is a tendency for the supervisors to report it is completely resolved, and the employees more frequently report it is partly resolved, so you're going to get fluctuation in the numbers there.


AUDIENCE QUESTION:  So if you were asked what is the settlement rate for cases, what would you say?


LISA BINGHAM:  I guess I would go with the closure rate, the 80% closure rate reported to you.


PAM ZUCZEK:  That's at all levels. Remember that that 80% closure rate includes closure at the table. Also after mediation we have people that are coming back and saying ‘I thought about that and yes, I want that as a settlement' or people who are even talking to each other just before they file a formal. There are different things once that mediation process begins. It actually begins even before people get to the table. It begins when people elect mediation and we start talking about REDRESSTM and we start talking to people about the control that they have with the process and people that have already been to mediation once sometimes are back in the system. They may get a reminder of what they have experienced in mediation and we have been finding closure rates before people are getting to the table because of the REDRESSTM program they already have that experience so it's at different levels.


LISA BINGHAM:  On that point – the Postal Service people who administer the program coined the term ‘premediation' for people getting together before the mediation process is scheduled and working it out bilaterally. They started tracking that internally, and I think the last number I heard from headquarters was that there were something like 1,200 ‘premediated' resolutions among people who had requested mediation and then resolved the dispute themselves before they got to the table. That is a great number. I want to find a way of capturing that number nationally, because you would expect that number to go up if people are learning. We have already seen another trend that I want to capture in the analysis. You would expect an initial increase in informal EEO complaint filings, because people want access to the new program. However, after that initial increase, you would expect, over time, informal EEO complaint filings to drop if people are starting to work things out themselves. That result has started to happen; it has started to happen. Actually, there has been a drop in informal EEO complaint filings, but we have not yet done the analysis to see how it correlates with the mediation program.


LELA LOVE:  This is sort of a personal question. I imagine you want to get certain results – positive results so that people like us and others in this room are happy and also it may be other reasons, so I just wanted to know how you can comment with a sharp cold eye.


LISA BINGHAM:  Well it is relatively easy when you have huge data sets and analysis where the numbers speak for themselves. I make it a point of submitting much of the research to refereed journals, like the Review of Public Personnel Administration. You notice I have not published these findings in law reviews and there is a reason for that. That reason is that I want somebody outside my shop to scrutinize the sample, the way we have analyzed the data, the method we have used, the theory we have used and the claims we are making from the data. By having the research subjected to peer review, I feel that it keeps me honest and a lot of evaluation research never goes there. Evaluations often get published in these monographs that are privately circulated, or they get published in law journals. Law journals are wonderful; you know I publish in law journals too, but just not this stuff. The problem with law journals is that there is a perception outside the legal academic community that because these journals are student-edited, there is insufficient critical perspective brought to bear on at least the statistical analysis in the papers. So, peer review through refereed journals is one mechanism I use to keep that clear, cold eye.


I also try to maintain other research. There was a story; I think a USPS manager told it to me. It was about a business professor who had been invited to do some lectures for the Postal Service and he initially declined, but when he was told, ‘You will have an audience of fifty managers, and each of these managers is the economic equivalent of the CEO of a Fortune 500 company,' and so he agreed to take the speaking engagement. He was given more speaking engagements, and finally he said, ‘I'm not going to take any more speaking engagements.' And the manager asked him, ‘Why won't you do any more speeches for Postal Service managers?' The business professor said, ‘I had this dream and in the dream I died and was looking at my gravestone and it said: He worked for the Post Office.'


It is a great story, because it captures this notion of getting totally absorbed by an organization that is really huge, demanding, and exciting. It has its own subculture. I can't describe what it is like to interview postal workers moving the mail at 3:00 AM in Florida, or in Cleveland, or in New York. It's amazing to me that there is this consistent pattern among the employees and supervisors and managers, which is that there is a very high public service motivation, a high level of dedication, and an abiding sense of service to the public, which on another level liberates people to criticize the organization because they feel it is actually in service of this higher motivation. The people who work at the Postal Service are bright, interesting, motivated, and very diverse. It is among the most if not the most diverse employer in the country. It is just a fascinating institution, so it is very easy to get absorbed by it, which is the moral of his story. Part of how I do not do that – get absorbed – is that I do very few speaking engagements at the Postal Service, and I do other research and publication. I try to keep balance, I just try to keep myself honest by submitting all the numbers to peer review, but there is no question that it is difficult.


LELA LOVE:  So maybe we'll just close it down. Thank you very much.



[1]. For a description of the history of the program, see Lisa B. Bingham & Lisa Marie Napoli, Employment Dispute Resolution and Workplace Culture: The REDRESSTM Program at the United States Postal Service, The Federal Administrative Dispute Resolution Deskbook 507-526 (Marshall J. Breger et al. eds., 2001).


[2]. Tina Nabatchi & Lisa B. Bingham, Transformative Mediation in the USPS REDRESS, 18 Hofstra Lab. & Emp. L.J. (forthcoming Spring 2001).


[3]. For a more complete description of roster management, see Traci G. Gann & Cynthia J. Hallberlin, Recruiting and Training Outside Neutrals, The Federal Administrative Dispute Resolution Deskbook 623-629 (Marshall J. Breger et al. eds., 2001).


[4]. Lisa B. Bingham & M. Cristina Novac, Mediation's Impact on Formal Discrimination Complaint Filing: Before and After the REDRESS, Rev. Pub. Personnel Admin. (forthcoming Fall 2001).


[5]. E. Allan Lind and Tom R. Tyler, The Social Psychology of Procedural Justice. New York: Plenum Press (1988).


[6]. Debra L. Shapiro & Jeanne M. Brett, Comparing Three Processes Underlying Judgments of Procedural Justice: A Field Study of Mediation and Arbitration, 65 J. Pers. & Soc. Psychol. 1167-1177 (1993).


[7]. Lisa B. Bingham, et al., Mediating Employment Disputes at the United States Postal Service: A Comparison of In-house and Outside Neutral Mediators, 20 Rev. Pub. Personnel Admin. 5 (2000).