Lex Electronica

Revue électronique du Centre de recherche en droit public

Online Dispute Resolution: The Next Phase

Ethan KATSH (*)

Lex Electronica, vol. 7, n°2, Printemps / Spring 2002



Ce texte traite de l’évolution des modes alternatifs de résolution de conflit (MARC) en ligne. L’auteur dresse un historique des différents projets de médiation en ligne en passant par ses « débuts » en 1999 lorsque l’Online Ombuds Office (OOO) fut approché pour l’élaboration d’un système de médiation pour les clients d’eBay, par SquareTrade.com, par les règles de l’ICANN, par Cybersettle et Clicknsettle, etc. Il expose ensuite le courrant que prendra la cybermédiation et le cyberarbitrage dans les années à venir. Ainsi, le médium informatique devient lentement une « quatrième partie » aux discussions et peut venir en aide aux arbitres et aux médiateurs dans la gestion et la communication d’informations. De plus les fonctions d’affichages propres aux ordinateurs, c’est-à-dire la possibilité d’incorporer images, graphiques, plans, etc., devront être mis à l’œuvre par les systèmes de MARC en ligne si ceux-ci sont destinés à prendre de l’expansion dans d’autres domaines que ceux leur étant présentement réservés.


This paper sketches out the evolution of alternative dispute resolution (ADR) mediums online (ODR). The author traces the history of the different ODR projects which have come and gone over the years, from the Online Ombuds Office’s (OOO) eBay mediation platform in 1999, to SquareTrade.com, ICANN regulations, Cybersettle and Clicknsettle, etc. He then predicts what will happen to ODR in the coming years. The digital environment is fast becoming a « fourth party » to discussions and, is used properly, can support mediators and arbiters in managing information and communications. Furthermore, the medium’s qualities, such as the possibility to use images, graphics, blueprints, etc., will have to be exploited by new ODR platforms if they want to cross over to fields previously reserved to more traditional ADR forums such as those involving government agencies.

1. In early 1999, several colleagues and I were asked by the online auction site eBay[i] to design a pilot project to mediate disputes that arose out of transactions between bidders and sellers. eBay was, and is, one of the most successful of online businesses. On any day at the beginning of 1999, this site had over 1.7 million items for sale. Each week there were close to a million transactions. We did not know how many disputes we would encounter, but we decided to do a two-week pilot project to judge both demand and our ability to respond.

2. In the middle of March 1999, eBay placed a link to our project, the Online Ombuds Office (OOO) at the University of ­Massachusetts, on the eBay customer service page. Even though the customer service page was three levels down from the main eBay home page, and even though no announcement about the service was made to eBay users, over one hundred and fifty complaints were filed during the following two-week period. The value of the disputes ranged from one dollar to fifteen thousand dollars. Our online mediator was able to obtain participation from both parties in more than three-quarters of the disputes and successfully mediated about half. This was probably a lower success rate than would have been obtained in a traditional setting, but as a first large-scale effort, we were both satisfied and encouraged.

3. For us, the eBay project was a turning point in an effort that had begun three years earlier. We had established the Online Ombuds Office in May 1996 because we believed that, as the Internet grew, there would be a growing need for dispute resolution services. We also believed that dispute resolution services could be delivered over the Internet. We expected, in other words, that the Internet would eventually need to confront the issue of dispute resolution and that when it did, there would be opportunities to deal with online problems using online means.

4. Our initial interest in ODR grew out of our own online experiences. We are sensitive to environments in which conflict occurs, and we routinely look in new environments to see what steps have been taken to prevent conflict and to deal with disputes. Our belief that cyberspace would not be a harmonious place was based on a very simple observation. Cyberspace seemed to us to be too active, too entrepreneurial, too competitive, and too lucrative a place for it not to have many conflicts. Even in 1996 it was apparent that cyberspace involved too many people applying their creative energies and imaginations in new ways for there not to be a need for processes to settle disputes. Cyberspace was a place where the number of transactions could only grow, and, we thought, where transactions and relationships go, disputes will follow.

5. Our assumption that the Internet might become a kind of ­“dispute resolution space” and serve as a vehicle for resolving disputes was also based on a rather simple observation. This was that cyberspace was a place where powerful tools were being developed for communicating, storing, and processing information. We knew that these activities were also at the heart of dispute resolution. As capabilities for working with information and managing information online improved, we thought that opportunities for directing these capabilities at dispute resolution would also improve.

6. Our project with eBay confirmed for us that there would be a need for ODR at active online marketplaces. A few months after we completed our pilot project, eBay added dispute resolution to the array of “safe harbor” services that were available to consumers who were concerned about making a purchase.[ii] Any person who encounters a problem with a transaction can, at the click of a mouse, request assistance from the dispute resolution company SquareTrade.com.[iii] Between February, 2000 and February, 2002, SquareTrade handled over 200,000 disputes, most of which came from eBay.

7. The need for dispute resolution at a place like eBay is relatively easy to understand. eBay has huge numbers of items offered for sale (currently over six million), huge numbers of transactions (currently over two million per week), and eBay assumes little or no responsibility for the transaction. eBay is not really a party to the transaction and the actual sellers, many of whom are small businesses or individuals, are often without customer service departments. eBay does not wish to participate in the transaction but it does wish users to feel comfortable making bids and purchases. The array of “safe harbor” resources along with the use of credit card chargebacks has made eBay into an online marketplace where buyers are able to find almost anything they are interested in purchasing, and are also wiling to make bids and purchases without a high degree of risk.

8. Prior to our project with eBay, the question I was asked most often was how mediation could possibly occur without face-to-face interactions. I generally answered that there would always be disputes that might be difficult to handle without face to face meetings but that there were also many contexts in which even the tools we had available in 1999 would be helpful to parties and to an online mediator or arbitrator who was called in to work with the parties. As a result of what has occurred during the last two years with eBay and in other environments in which ODR has been employed, I no longer hear such skepticism about online dispute resolution. There are a great many questions and challenges that face ODR but its potential use and value has, I believe, generally been accepted both by persons outside the dispute resolution community and by those within.

9. During the last three years, two other ventures involving large-scale use of dispute resolution at a distance have contributed to the change in attitudes about the value of ODR. In late 1999, ICANN, the Internet Corporation for Assigned Names and Numbers,[iv] put in place a dispute resolution process for resolving disputes over domain names, the Internet addresses like ibm.com or disputes.org. These are disputes in which a trademark owner asserts that the registration and use of the domain name by someone else is infringing upon the rights of the trademark owner. The ICANN process has been the subject of some controversy[v] but there have been over five thousand decisions and it is relatively fast, efficient, and inexpensive. It may need modification but it is an example of a context in which face to face meetings might be costly and difficult to arrange and in which they are unlikely to be necessary.

10. The third online dispute resolution process is the fairly simple procedure of blind bidding. Blind bidding, employed by companies such as Cybersettle[vi] or Clicknsettle,[vii] consists of disputing parties submitting settlement offers to a computer. If the offers are within a certain range, the parties split the difference and settle the dispute. If they are not within range, negotiation continues in a traditional manner and neither party learns of the other’s offer. Blind bidding can be useful but only when there is a single issue in controversy and that issue involves money, or something that can be represented numerically.

11. These three examples have been enormously important in demonstrating that online conflict resolution tools are needed and can be delivered. There had been conferences about ODR prior to 1999[viii] and we had even resolved some cases online.[ix] Yet, it was the processing of large numbers of disputes in important venues that has allowed ODR to be recognized as a process that belongs in cyberspace and that is needed in cyberspace.

12. For the Internet, the years 1999 to the present have seen ecommerce expand, the number of online users increase, and the range of online activities continue to grow. Yet, many Internet start-ups have failed and the search for profitable business models has been difficult. There is not as much opportunity for subsidized experimentation as there was two to three years ago. Even in this environment, however, the Web continues to grow, both in numbers and in the range of processes that one can actually engage in online. The power of online applications continues to grow and this will be a catalyst for the further growth of ODR.

Moving ahead

13. The three examples of ODR discussed above represent starting points and the simple beginnings of ODR development and application. Ebay disputes can be considered simple since they involve single transactions rather than ongoing relationships. In addition, the vast majority of these disputes fall into predictable categories, such as an item which arrived broken, an item which appeared different from what was advertised, etc.

14. Domain name disputes fall at the less complex end of the continuum because the process employed is arbitration. These disputes are resolved on the basis of documents submitted by the parties. ODR is an online communications process, or set of processes, and mediation will generally involve a more complicated series of interactions than arbitration. In general, therefore, mediation will be more difficult to model online than arbitration.

15. The blind bidding process is simple because there is a single algorithm that is useful only in single issue disputes involving something that can be represented in numbers. Its simplicity is both its value and its problem. It can be employed easily and cheaply but only in a narrow category of disputes.

16. It was quite predictable that early uses of ODR would occur at the less complex end of the complexity spectrum. More complex disputes will require not only third parties who are comfortable with the online environment but tools and resources that are more powerful than what is currently available. Any form of ODR will require a structured interaction between the parties and the third party that involves the network and the machines connected to the network. In a recent book,[x] my colleague, Janet Rifkin, and I have referred to technology and the Net as being a “fourth party”, something that is an influence on the process of communication and negotiation, something that adds value to the third party, something that typically does not replace the third party but can displace her, in the sense that the third party operates with an ally or assistant alongside.

17. The “fourth party” is embedded in software and can take many different forms. In the early large scale examples of ODR mentioned above, the added value of the Net as a “fourth party” was quite minimal. Yet, the power and availability of Internet-based information processing applications can only grow. As this occurs, the value of the “fourth party” will also grow and new uses will be found that facilitate resolution in more complex disputes.

18. The “fourth party” is a metaphor for applications that enhance the expertise of the third party and thus do more than simply deliver the expertise of the human third party across the network. Figure 1 identifies a few areas in which traditional tasks performed by third parties can be enhanced by increasingly mature online applications.

19. The informational activities in Figure 1 are common but, in the past, it has only been when the parties were physically together that they were performed with any degree of efficiency. For example, scheduling meetings with several parties can be done fairly quickly when all are in the same room looking at calendars but scheduling using telephone or paper becomes cumbersome as the number of participants grows. A mediator meeting face to face with a few parties can survey opinions and may even be able to evaluate whether consensus exists by looking at facial expressions. When the parties are not together, however, ascertaining how parties feel again grows difficult as the number of parties involved increases.

20. Many inefficiencies caused by the parties being apart might be labeled “tolerated inefficiencies”. We are accustomed to them and have accepted many of them as inevitable. The network, however, changes significantly our ability to overcome these “tolerated inefficiencies”. As we grow more comfortable with the network, we realize that certain parts of how third parties traditionally handle disputes should be reevaluated. For example, we have new tools for communicating with parties in between face to face sessions. Should we do so? For this and other instances of “tolerated inefficiencies”, we need to decide whether what we are accustomed to is still appropriate given the new tools we are acquiring that allow us to change how long interactions with parties might take and where they might take place.

21. As another example, the process of monitoring performance or enforcing the terms of any agreement. Currently, it is the responsibility of the parties to make known any problems that might occur. New options, however, become available when channels of communications stay open after the agreement has been reached and various communication tasks can begin to be handled by the “fourth party”.

22. Monitoring performance has never been a very efficient process. If a scheduled payment has not arrived, for example, do I need to call the other party? Should I alert the mediator? Should I have the mediator call the other party?  One aspect of an agreement in the future will be to employ a Web site to monitor enforcement. If payments are required at intervals, automatic reminders can be sent. Boxes can be clicked on to indicate that a payment has been sent or some other action has been done. Obviously, the complexity of monitoring tools will have to be appropriate to the complexity of the performance required. Certainly, however, a “legal watchman” or early warning system of non-performance will be quite useful.

23. The medium's visual capabilities provide a whole range of intriguing possibilities for alerting us to change and to the direction of change. Images and numbers can be employed to show change in ways that are not possible with print. Increases and decreases can be demonstrated visually through changes in size, shape or color. In the monitoring context, for example, lack of performance might send a red flag to the attorney for one of the parties. This could be an actual image of a red flag and the red flag, if ignored, could grow larger over time, something that would be both meaningful and attention getting.

24. Symbols like a red flag remind us that text itself can fall into the category of a “tolerated inefficiency”. Text is often inefficient compared to the spoken word and text can be inefficient compared to various forms of visual communication. There are many images online but we are still not very effective in employing modes of visual communication. As bandwidth increases, however, it can be expected that new applications employing both image and sound will become available.

25. Visual communication should not be equated with pictures. Nor should visual communication be equated with graphs and charts. Or with maps and icons. What we are presented with by creative use of the screen is a range of visual communicative opportunities. This should, and ultimately will be, quite liberating, as we learn new ways to combine text and image for purposes of persuasion and explanation. As with many new opportunities, however, we will also be challenged to determine how to use forms of expression that we have probably received little or no training in before.

26. These new opportunities for expression made possible by the “fourth party” will pose challenges for some third parties. New media can be both helpful and disruptive at the same time. The third party of the future will require the same level of sensitivity to pictures, images, icons, charts, figures, graphs, scales, tables, diagrams, maps, sketches, blueprints and colorful and animated graphics, as there is to textual or spoken communication. This does not mean that one needs to be able to actually draw any of these items. The ease with which numbers are, even today, transformed into charts illustrates the assistance that the “fourth party” will provide to us. But there will still be a need to learn how to use visual tools in the most effective way, just as we continually face the challenge of expressing ourselves clearly and forcefully as we speak and write.

27. The best example of the enabling power of the online environment for graphical communication comes from an extremely complex international negotiation. This was the negotiation between the Bosnian Serb, Croat, and Muslim factions that resulted in the Dayton Peace Accords in 1995.

28. The negotiations in Dayton were the first to employ digital maps, maps that not only had detailed boundaries that could be adjusted, but that could “visualize” terrain. The negotiators were experienced in the use of maps but were also “used to paper maps, the crisp appearance of printed detail, and the flexibility of drawing on their map copy where and when they wished”. The digital maps provided computer generated images of terrain and allowed the negotiators to try out different boundary lines, and contributed significantly to the successful completion of the Dayton negotiation.

29. What might be admired most about the use of maps in Dayton is not how they were used but that the negotiators were willing to learn how to use them. The most challenging aspect of employing ODR in an offline context will be to recognize that the manner in which we have been resolving disputes, even if we have been quite successful, can be enhanced. Mediators and arbitrators need to recognize that however complete their toolkit was a year ago, it is more complete today, because the array of available tools is growing.

The next phase

30. The first phase of ODR confirmed that there would be demand for it and that ODR was quite feasible in contexts which required fairly simple and widely available online tools. The second phase can be expected to involve extending what is currently being done and growing in several different directions:

31. Phase one began with great concern that ODR would not succeed because the richness of offline communication would be missing. As phase one ended, several arenas had been identified in which some type of ODR proved to be of considerable value. We will, during phase two, undoubtedly find additional contexts in which fairly simple online interventions or interactions will of considerable value. As is true of the availability of offline mediation, however, use will be limited by cost and economic considerations. As with ADR, ODR will be employed when the cost of dispute resolution is less than the value of the dispute in question. Some online tools, such as blind bidding, can be delivered at low cost, and some technological alternatives may be less expensive than human ones, but we have also learned that while ODR can bring many kinds of efficiencies, ODR is not inevitably a low cost option. Indeed, ODR ventures face the same challenges of delivering a service at a profit that are faced by Internet start-ups generally.

32. One arena in which the calculation of costs and benefits may work in favor of experimentation and use is likely to involve government. Dispute resolution is a core function of many, if not most, government agencies. In such agencies, dispute resolution serves a variety of purposes. In some federal agencies, such the Internal Revenue Service, dispute resolution is part of the enforcement role. In other agencies, such as the Equal Employment Opportunity Commission, formal systems exist in order to obtain factual information and determine whether standards have been violated. In still other agencies, such as the Environmental Protection Agency, dispute resolution occurs as a part of rule-making and other processes that are designed to foster consensus in about some action. In general, government agencies enact and/or enforce regulations and standards, and dispute resolution processes of some kind are inevitably necessary.

33. Many of the most numerous types of disputes, such as those involving citizens and the Internal Revenue Service, can be resolved merely by the efficient exchange of information. Many such disputes are, even now, not handled through interactions with humans, and the use of the network will be an increasingly appealing and, in all likelihood, economical option. Other contexts will involve more complex disputes but online options will be appealing and pilot projects can be expected to demonstrate what is possible both in the short term and in the long term.

34. Disputes handled by government agencies originate in relatively traditional arenas and we will learn much from the hybrid ADR/ODR systems that will be employed. ODR adds to the choices that will have to be made in applying dispute resolution resources. For example, difficult choices will have to be made concerning use of synchronous v. asynchronous modes of communication, how to employ differing combinations of text, images and video, whether to rely totally on the Web or to also use email, and what the right mix of online and offline “meetings” should be.

35. Much of phase two will concern “what happens when the digital world merges with the physical world”.[xi] Bringing online tools and approaches into offline practice can be expected to occur gradually as tools are developed, found to be appropriate, and become a routine part of the mediator or arbitrator’s toolkit

36. SquareTrade, the firm that handles eBay disputes, currently also works with the California Board of Realtors and handles disputes arising out of real estate transactions. These two clients, eBay and the California Association of Realtors, symbolize two extremes, one a company that exists only in cyberspace and one a group focused on the most tangible of transactions. What SquareTrade is learning is that online options developed for the exclusively online environment of eBay can be of value offline as well. SquareTrade began its online dispute resolution effort by requiring its mediators to have traditional qualifications and experience. One perhaps unexpected consequence is that the tools and approaches learned by online mediators can now be applied in more traditional contexts.

37. The earliest uses of ODR were in relatively simple transactions and disputes and ODR can be expected to be employed similarly in the offline environment in different niches where there are large numbers of recurring disputes. ODR will also find a place offline, and perhaps find a place quite rapidly, in offline disputes that are of high value or involve many parties or issues. In such contexts, parties may be able to afford the use of technology, be sensitive to the value of using online resources, and the efficiencies and benefits of supplementing traditional approaches will be clear.

38. Perhaps the most important but also least noticed change of the past three years has been in the growing sophistication of ODR software.For ICANN domain name disputes, eResolution used a totally online process that was both powerful and easy to use. Our eBay pilot project in 1999 relied on email, a resource that is widely available, inexpensive to use, and allows the quick exchange of messages, but has no significant information processing features. In the Web-based dispute resolution process that SquareTrade has designed, the parties are asked to engage in a period of “direct negotiation” before a mediator is appointed to work with them. SquareTrade has found, quite remarkably, that even without the intervention of a human mediator, approximately eighty percent of the disputes filed are resolved through direct negotiation. This is an impressive success rate, particularly since almost all the parties had engaged in some form of email negotiation before the complaint was filed. The filing of the complaint is generally a sign that one party, at least, believes that there is little or no value in any further negotiation via email.

39. While a mediator does not participate in the “direct negotiation” process, SquareTrade participates. What does it mean to say that “SquareTrade participates”? The process is called “direct negotiation” but the reality is that it is “mediated negotiation”. It is negotiation that takes place mostly through screens designed by SquareTrade, information supplied by SquareTrade, and a process structured by SquareTrade. Figure 2 contains the first screen someone with a dispute would see. This is not a complaint form but a screen with a bit of information that might be expected to provide encouragement and optimistic expectations for the complainant. The information is a suggestion that 80% of the parties are pleased with the process. This, along with the phrase “Building Trust in Transactions”, contributes to trust. The statement that “the case filing should take 15 minutes” contributes to feelings of convenience. The links on the right side of the screen suggest expertise and experience.


SquareTrade Case Filing Screen

40. Cyberspace is, increasingly, a place where there are processes available to users as well as information. This should not be surprising since processes are sets of informational transactions and exchanges. What makes building processes out of informational transactions difficult is structuring and regulating the flow of information and the numerous informational exchanges among the parties. Tools that have made it possible to create online auctions, stores, casinos, and other online processes will, however, over time also allow civic institutions such as courts and dispute resolution systems to function online.

41. The building of new software tools is what will make the “fourth party” an increasingly robust resource. The more informational tasks that can be done online, the more useable ODR will be in more complex disputes, disputes where there are many issues, many parties or both. Software development occurs less rapidly than hardware development and adapting and learning to apply to new features also takes time. We cannot predict how quickly various tools will appear but we can be certain that a process has begun that will inevitably lead to their development.


* Professor of Legal Studies and Director of the Center for Information Technology and Dispute Resolution at the University of Massachusetts. Email: Katsh@legal.umass.edu.

[i]  <http://www.ebay.com>

[ii]  <http://pages.ebay.com/help/community/index.html>

[iii] <http://www.squaretrade.com>

[iv]  <http://www.icann.org>

[v]  Michael Geist, "Fundamentally Fair.com? An Update on Bias Allegations an the ICANN UDRP" <http://www.udrpinfo.com/home.php>

[vi]  <http://www.cybersettle.com>

[vii]  <http://www.clicknsettle.com>

[viii]  <http://mantle.sbs.umass.edu/vmag/disres.html>

[ix]  <http://www.ombuds.org/casetran.html>

[x]  Ethan Katsh and Janet Rifkin, Online Dispute Resolution (2001)

[xi] Neil Gershenfeld, When Things Start to Think (1999)




© copyright 1995-2008 Lex Electronica Tous droits réservés / All Rights Reserved ISSN 1480-1787