Donald Rumsfeld and a Nineteenth Century Minister on Mediation

Mediation is essentially an exercise in assessing risk and making decisions based on that assessment. This process plays out most obviously and concretely in litigated cases in which the opposing parties have had an interaction in the past, as momentary as a car crash or prolonged as a business or marital relationship, and they have no interests in any continuing relationship after the litigation is concluded. These are the cases where settlement by agreement or verdict resolves the relationship. The risks of a bad mediation are definable and contained.

In these cases especially, I proceed with the assumption that the more information the parties have the better they can assess the risk of not settling then and there. If parties could have perfect and complete information about the outcome of the trial, that is, if they could predict the future, mediation would be short and simple. If, by use of a crystal ball or time machine, the parties could know what the judge or jury will in fact do in the matter, then cases which are only about an award of money would be relegated to an analysis of the time value of that money to the insurer and to the plaintiff, plus a potential discount for avoiding the emotional and financial costs of a trial. Cases where there is more than money at stake raise additional issues, but the analysis is the same: perfect information about the future outcome would narrow immeasurably the range of potential mediated results. Life, however, is not so simple.

In the absence of knowing what the fact finder will award, the attorneys rely on what they believe might be awarded and, therefore, create either a hard and fast result or a range of results. Depending on the experience of the lawyers or the carrier, either directly or through some assembled data, a range of results can be blocked out. The lawyers may have past experience with the presiding judge, the jury pool of a particular county or the subject matter of the litigation. The insurance carrier or company may have access to computer models of similar cases in this and other jurisdictions. These data then form the framework for each party’s initial analysis of likely outcomes and therefore their negotiating position.

But the reliability of those positions must necessarily depend on the accuracy and completeness of the information the parties have about their own and the opponent’s case. There are of course unknowable factors such as the composition of the jury, or how the judge might rule on preliminary motions. The parties generally have thought about this unknowable information and made sometimes educated guesses about it. That is probably the best that the parties can do on that data.

Donald Rumsfeld, former Secretary of Defense, is known for the somewhat baffling-on-first-hearing quote on knowledge:

There are known knowns. These are things we know that we know.

There are known unknowns. That is to say, there are things that we know we don't know.

But there are also unknown unknowns. There are things we don't know we don't know.

The information exchanged by the parties or otherwise known to them is the known knowns and generally forms the basis on which the case analysis is founded. My experience in mediations, however, is that both sides rarely have all the knowable information. These are the pieces which they know they don’t have (unanswered interrogatories, undeposed witnesses, incomplete medical records, etc.) as well as information they don’t know that they don’t know. Movement in mediation is very often the result of developing information in these last two categories. A productive mediation is often one in which one side discloses new information for the first time. It is this learning that there were things that one side didn’t know which can change the negotiating dynamic dramatically. This type of revelation can happen in several ways. I usually schedule a phone conference among counsel two weeks before mediation to make sure everyone has what they need to be productive. If there are unanswered interrogatories, as yet unproduced documents, and so forth, we work through these and I attempt to obtain assurances that the missing data be provided before the mediation date. In other words, I attempt to get the parties to produce all the information which they know that they do not yet know.

But on the last category, the unknown unknowns, this is where the crucible of mediation can produce dramatic developments. Sometimes it is because mediation finally forces an attorney to really think through his or her own case and put information on the table. This new data may be previously undisclosed or disclosed but unappreciated. Or perhaps new facts are learned because the mediator played an interrogational role to bring out new information which neither party had sought to discover.  Why does new information lead to a better likelihood of settlement at mediation?

Enter Bayes’ Theorem. Thomas Bayes, a nineteenth century philosopher, statistician and minister, designed a statistical tool which shows that the probability of a certain result is increased as more relevant evidence is added to the equation. Although we need not delve into its mathematical details, it is helpful as a principle in mediation. It supports the proposition that one can analyze the probability of certain results occurring at trial in light of new evidence or previously known but unappreciated evidence. More information about a case improves both parties’ ability to more clearly predict outcomes.

One of the advantages I have as an experienced trial attorney is knowing how particular facts or laws might be maximized to advantage at trial and having a more attuned sense of how a jury might react to certain facts. But that experience merely provides a helpful framework for the parties. So long as both parties continue to have the same knowledge base they had at the beginning of a mediation, that mediation will only be successful if their range of predicted potential outcomes at trial already overlap. This is rarely the case. The charge of the mediator is to expand the knowledge of the parties so as to change the boundaries of those ranges. This requires the mediator to be proactive in getting each party to explore its legal and factual theories in ways and with details not attempted before. The general session at the beginning of mediation is an optimal time to begin to develop this dynamic: all parties are present (I usually require the adjuster who appears by phone to be on the line for this session) and can assess each other’s reaction to the new data. By encouraging parties to put more and more information on the table, especially never disclosed information, each has a better chance to assess risk, have their ranges overlap, and get the case closed.