Reinsurance Arbitration Reform: Is It Time For A Mediation Option And A Flexible Discovery Limit?

December 18, 2008

In recent years reinsurance arbitration has become a slower and more expensive process.  The authors report the findings of their own survey of reinsurance professionals on the arbitration process and review both the criticisms of the process and proposals for its reform that have been advanced by industry commentators.  Based on the success of reinsurance claims mediations in other countries and the results of their survey, the authors propose a formalized mediation option and a flexible discovery limit to improve the reinsurance arbitration process.

The insurance industry needs a system for resolving reinsurance claim disputes that is fair, cost-effective and efficient.  A review of the literature indicates that the current system of reinsurance arbitration often fails to realize these goals.  In a recent survey of reinsurance professionals conducted by the authors, industry participants cited the expense of arbitration as a serious issue and express a desire for tighter discovery limits to control the cost of arbitration.  This article lays out the problems and explores the potential solutions, including a formalized mediation option and a flexible discovery limit.

Is the Arbitration Process Broken?

Arbitration has been the standard mechanism for the resolution of reinsurance disputes for more than a century.  The  roots of reinsurance arbitration can be traced back to the 1790s.[1]  As one reinsurance executive has observed, the arbitration process worked successfully as a dispute resolution mechanism for many years:

In the ‘old days' when reinsurance was largely transacted within a small circle of industry professionals who knew each other and expected to have ongoing relationships, the less formal approach [of arbitration] made sense . . . . Arbitration could be conducted fairly quickly, privately and resolved [by experts] in a manner that allowed the parties to continue doing business with each other.[2]

Another set of commentators has observed that reinsurance arbitration became the norm in what might be called a "kinder, gentler era" when arbitration was viewed as the best way for business partners to obtain fast and cost-efficient resolution of their disputes.[3] 

In the "old days," reinsurance arbitration seldom involved lawyers, formal discovery, or a litigious atmosphere.  Claims executives on both sides exchanged file materials and met with a panel of experienced company executives who had themselves lived the "honorable engagement" standard.  The dispute was handled informally and privately.  Expert witnesses were seldom called because the arbitrators were viewed as the experts.  The panel either mediated the dispute to a settlement or decided it in a manner that provided a measure of justice and allowed the parties to proceed with their ongoing business relationship. 

Observers now seem to agree that the "old days" are gone.  "The primary difference now is the relationship of the parties . . . .  Today reinsurance disputes are almost always between parties who no longer have an ongoing business relationship."[4]  As the relationship between the parties has changed, reinsurance has become more like any other business.  As in broader corporate society, disputes have become more lawyer-driven and more contentious.  And satisfaction with the process is a far cry from the "old days":

The arbitration system, the predominant method of dispute resolution in the treaty reinsurance business, is broken . . . . It has deteriorated into a process in which few of the originally intended benefits [of speed, fairness and cost-effective resolution by industry experts] are realized.[5]

Proposals for changes in the process are beginning to emerge amid real concern that "[w]ithout changes, the current process is in danger of collapse."[6] 

The complaints about reinsurance arbitration follow a consistent pattern.  Participants, particularly ceding companies, complain about:

  • the increased length and the expense of arbitration which includes the fees of counsel, a party arbitrator, experts, the umpire, and rented hearing space often in major hotels;
  • the litigation-like antagonism that reinsurance arbitration generates;
  • the ease with which the party with the weaker position can delay the arbitration; and
  • the structural tendency toward compromise results even where one party is clearly right; and
  • the ability of a party to repeatedly take weak positions given the lack of published precedents in arbitration.[7]

One article summed up the frustration of participants with the increasing cost of arbitration by observing, "[n]ot surprisingly, the reinsurance arbitration process, as it is practiced today, is satisfying no one - except arguably those charging by the hour."[8] 

The inevitable charge that reinsurance arbitration is as slow, expensive and unsatisfying as litigation was made in a 2007 article proposing that mediation be added to the process as a prerequisite to reinsurance arbitration.  The article's authors concluded that at this point in its history reinsurance "arbitration can be - and increasingly proves to be - time-consuming, slow, expensive and even more uncertain than litigation."[9]  Suggesting that the delays and expense currently being experienced in reinsurance arbitrations cannot continue, the authors predicted "a significant change is about to happen . . . ."[10] 

The sentiments collectively expressed by the commentators may be in part driven by the industry's dissatisfaction with the cost and efficiency of the discovery phase of arbitration as managed by arbitration panels.  Robins, Kaplan, Miller & Ciresi L.L.P. recently surveyed cedents, reinsurers, attorneys and experts involved in the reinsurance arbitration process.  While the results should not be considered scientifically reliable, they do reflect a consensus that discovery is an important area of concern.

We asked participants which of several options would best deal with the problem of the escalating cost of reinsurance arbitration.  Our question and the breakdown of responses by percentage from the 179 respondents appears in Figure 1 below:

Figure 1

Note:  Traditional mediation refers to the mediation process as typically conducted in a litigation setting.  Mediator's recommendation refers to traditional mediation with a mediator's settlement recommendation provided verbally to the parties in the event of impasse. Combination mediation/arbitration refers to a process in which, upon the parties failing to settle in mediation, the mediator delivers a binding arbitration award.  The phrase tighter discover limits is self-explanatory.

Because of the trends towards expansive discovery and multi-week hearings, many respondents expressed concerns about the cost of arbitration.  A surprising 37.8% reported that reinsurance arbitration in their experience actually costs the same or more than litigation.  Less than 10% reported that they had found reinsurance arbitration to be much less expensive than litigation:

Figure 2

These survey results mirror the frustration with the cost, delay and inefficiency of discovery in reinsurance arbitration that is expressed in the literature.  Some critics have even pointed to the disincentive for arbitration panels to meaningfully limit discovery:

Arbitrators have been heard to say that they are duty-bound to give the proceedings their full and complete attention, particularly in light of the fact that the parties are not likely to be able to appeal the decision.  Thus arbitrators often resolve doubts in favor of expanding the proceedings, allowing additional discovery and spending more of the participants' money . . .  parties to reinsurance arbitration have been heard to question whether arbitrators who are paid by the hour have appropriate incentives to resolve cases in an efficient manner.[11] 

Amid the extensive discovery and the related discovery disputes that now arise in arbitration proceedings, one set of commentators claims that reinsurance arbitrations now "resemble trench warfare [and] routinely take years to come to a hearing."[12]       

The burdens of discovery in reinsurance arbitrations may fall disproportionately on cedents.  Cedents possess much of the information about the claim adjustment and/or contract formation that a reinsurer may wish to discover and so a cedent may find itself consuming valuable time and resources producing that information.   And cedents argue that they suffer the most if recovery on a meritorious reinsurance claim is delayed for years by discovery and discovery disputes.  One reinsurance executive recently acknowledged that there has developed "a view on the part of ceding companies that certain reinsurers (typically having gone into run-off and therefore having no ongoing business relationships to preserve) manipulate the arbitration process to simply delay payment, make collection more difficult and stimulate settlement on terms more favorable for the reinsurer."[13] 

It seems then that reinsurance arbitration, if not broken, needs some reform aimed at expedition and cost-control.  There is no shortage of ideas in the literature.  What changes to the arbitration process have been proposed?  Given eroded relationships, the present tendency toward trench warfare, and the legitimate need for discovery and a full hearing on the merits, what changes will do the most good?

What Reforms Have Been Suggested?

Dakim-Grimm's and Cloutier's 2003 article proposed a sea change in the American system of reinsurance arbitration including requirements that all arbitrators be neutrals trained and certified in dispute resolution and the rules of evidence and bound to an ethical code enforced by the certification entity; discovery limits and time limits; a requirement of written, reasoned decisions; reduced confidentiality; and a policy favoring awards of fees and costs as a means to reduce abuses.[14]   There have been gains in certification and training procedures in recent years, but five years later many of the visionary elements of this proposal remain unrealized.   

A year later, in 2004, Winn and Davis advanced a proposal to create a system of administered mediation with procedural rules defining time limits, an ethical code, discovery limits, and presumptive responses to attempts to delay or derail arbitration.[15]  Winn and Davis also proposed increased use of mediation.[16]  While less ambitious than the proposal advanced by Dakim-Grimm and Cloutier, Winn's and Davis's proposal has also spurred limited formal movement.

More recently, Foster has suggested from the reinsurer's perspective that the industry would be best "served by exerting more vigorous control to keep the arbitration process focused on the business goals served by the reinsurance relationship."[17]  The goal is laudable but the question must be asked whether the parties have the ability to control an ongoing arbitration.  As has been observed, "like a runaway train, once arbitration has begun, it is often difficult to stop."[18] 

In 2007, Phillips' and Moss's article made a focused proposal for the inclusion of a mediation provision in facultative certificates and treaty arbitration clauses:

reinsurers would be well advised to consider an industry-based approach to formulating and adopting key mediation clauses, protocols and practices . . . mediation is ripe for take-up, with potential for significant cost savings.[19]

Phillips and Moss praise the International Reinsurance Industry Dispute Resolution Protocol promulgated by the International Institute for Conflict Prevention and Resolution in cooperation with the Lloyd's market.  The protocol, which may be incorporated by reference in certificates or treaties, suggests a "comprehensive method of identifying reinsurance claim disputes early on, agreeing on a rigorous . . . method of exchanging adequate information . . . and engaging in structured negotiation (and, if necessary, mediation)."[20]  A copy of the protocol may be found at  The protocol seeks to avoid delay by calling for "notice and exchange of information within 30 days, negotiation commenced a fortnight thereafter, and private confidential and non-binding mediation brought on if the matter cannot be resolved within another 15 days."[21]  Some London-based reinsurance agreements have begun to include mediation clauses but the trend is nascent.

Is Mediation Coming?

Given the issues with reinsurance arbitration today, the idea of mediation has some appeal.  Mediation, where successful,  provides the parties a measure of justice in a less contentious, less expensive setting than full-blown arbitration.[22]  Aside from  strong interest in London, mediation seems to be working in at least one large national market:  Australia.  "Despite the popularity of arbitration clauses in reinsurance contracts, it is not uncommon in Australia for these to be replaced with clauses allowing for . . . mediation or expert determination-or a clause allowing for resolution via arbitration only after other [ADR forms] have failed."[23]  In Australia, some 30% of non-life reinsurance contracts reportedly contain mediation or expert determination clauses.[24] 

The United Kingdom and Australian industry participants may be on to something.  A combined 63.7% of our survey's respondents cited combination mediation arbitration or tighter discovery limits as the ideas holding the most promise for reducing the cost of arbitration.  Excluding respondents with no experience mediating a reinsurance dispute, a similar percentage of respondents (63.0%) reported that mediations have been somewhat effective or very effective in resolving reinsurance disputes.  Of those with an opinion, 49.6% of respondents reported that they have found mediation most effective in resolving reinsurance disputes at the pre-hearing stage.  A smaller but still significant percentage (28.2%) of respondents reported that mediation had been most effective in resolving disputes at the pre-demand stage, which is consistent with the IICPR protocol approach advanced by members of the London market.

Based on the views of industry participants as well as the trends in the United Kingdom and Australia, the addition of a mediation option would likely improve the current system.  The authors recommend that mediation be added as an option and not as a mandatory requirement applicable to all disputes because there are classes of cases and parties ill-suited to mediation.  Requiring mediation of all disputes would only doom parties dealing with disputes that are inappropriate for mediation to additional delay and expense without much hope of resolving the disputes.     

The trigger for a mediation option need not be more formal than a provision in the agreement or in a rule of an organization such as ARIAS (the AIDA Reinsurance and Arbitration Society), to the effect that "Any party to this reinsurance agreement may demand mediation of a dispute and nominate five persons with experience as umpires or mediators in reinsurance disputes, and the party receiving the demand shall, absent a statement of good cause in writing, select one of the five nominees and the matter shall proceed to mediation before that person within sixty (60) days."[25]  Alternatively, the provision could designate the umpire as a pre-hearing mediator, satisfying the interest of a substantial number of industry participants in combination mediation-arbitration, but that alternative restricts the timing of mediation and raises obvious conflict issues which would have to be fully considered. 

If brokers and underwriters respond to the dissatisfaction of their customers, as they inevitably must, it seems to the authors that some form of mediation is coming.  One alternative worth considering here is qualitative mediation, sometimes referred to as evaluative mediation.  This process includes all the features of more traditional mediation with the bonus of a mediator's settlement recommendation supported by detailed reasoning in the event mediation fails.   As an alternative to full blown arbitration, qualitative mediation offers real potential advantages:

  • Early engagement of decision makers;
  • Reduced delay
  • Two avenues to dispute resolution in one process;
  • One day process vs. weeks or months of hearings;
  • Relatively low cost especially if done pre-discovery;
  • Opportunity to air issues in evaluative setting while minimizing confrontation; and
  • Potential for written evaluation supported by detailed reasoning


The qualitative mediation process can work particularly well if a respected umpire serves as the mediator.  Parties may feel they had their "day in court" years earlier and at much less expense than in a full blown arbitration.  It is an alternative that may meet the goal of returning to a fair and cost-effective process in which disputes are resolved informally with the assistance and input of an industry expert familiar with the custom and practice of the industry.  If some form of mediation is coming, qualitative mediation may be worth special consideration because of its potential to be a good fit with the industry's culture and circumstances.

Is Mediation Alone Enough To Save Reinsurance Arbitration?

Mediation or negotiation will work in some cases, but the remaining cases will proceed to hearing where cost remains an important issue.  As the survey results reflect, substantial sentiment exists in favor of tighter discovery limits as the preferred method to cut costs.  Among respondents, an impressive 42.5% cited tighter limits on discovery as holding the most promise for reducing the cost of reinsurance disputes.  This is consistent with the concern, as Foster reports, that some parties use discovery as a tactical weapon to delay or force favorable settlement of claims.

The authors acknowledge that limiting discovery or types of evidence across the board in a dispute resolution venue from which there is only a limited right of appeal is difficult. Discovery limitations are inherently unattractive when viewed from the perspective of the party with the least knowledge and greatest need for discovery.  There is always the risk that a blanket limitation will trigger the vacating of an award on the ground that the panel "refus[ed] to hear evidence pertinent and material to the controversy."[26]

The industry has clearly placed need for cost-effective management of discovery at the top of its list of concerns.   The issue has not been tackled head on, but there have been efforts to simplify proceedings over all for smaller disputes.  ARIAS offers in its rules an option for a streamlined arbitration process with no discovery and a decision by a single arbitrator, but the option is limited to disputes under $250,000.  The short-form arbitration idea has made limited headway with some agreements now including clauses calling for short-form arbitration of disputes involving less than $1,000,000.[27] Foster suggests limiting such short-form procedures to insurers in run-off,[28] but it remains to be seen how much support could be garnered for a system having one set of rules for active reinsurers and a different set for run-off reinsurers.

One reform more directly aimed at discovery that may be worth considering is a system that limits discovery in the first instance, but allows a party to exceed the limits subject to potential cost-shifting.  Some federal courts have instituted local rules with presumptive limits on the number and length of depositions, and those rules inspire this proposal as follows.  Each party to an arbitration might be limited by the reinsurance agreement to discovery of documents arising in the course of the claim or the account and not more than five (5) depositions.  A party could still take relevant discovery that goes beyond the limitations, but the party taking the additional discovery would be liable for the reasonable fees and costs of incurred by the party providing the additional discovery if the party providing the additional discovery prevails on the claims to which the additional discovery relates.[29]  We refer to this idea as a flexible discovery limit.  The powers of arbitration panels to limit discovery through protective orders would be unchanged.  The concept would improve the current system by creating a measured disincentive to unjustified discovery, while not actually limiting the right of any party to take discovery that is needed in pursuit of a meritorious claim or defense.

In sum, an easy-to-use mediation option would improve the current arbitration process, if only to quickly resolve the subset of those cases where the parties (if given a forum) would rather work out their differences than pay for a full arbitration.  For the remaining cases, a new disincentive is needed that does not unduly burden everyone's interest in fair and just results, but does present some reasonable quantum of risk for a party whose discovery requests are more extensive than the norm.  We propose a flexible discovery limit as one alternative to address the problem. 

The survey results suggest to us that the first steps along the road to a better reinsurance arbitration process are a mediation component easily triggered by either party and a system that presents any party who wishes to pursue broader discovery with some potential economic downside.  The system should also provide some level of compensation for the party with a meritorious position that must nevertheless spend extra time and money responding to outsized discovery requests.  A two-level discovery rubric that allows some potential for cost-shifting meets this goal.

Whether or not the ideas discussed in this article are the best ones to "fix" reinsurance arbitration is an unknown, but we do know that the price of doing nothing is high.  New ideas are worth considering as a means of restoring a measure of efficiency balanced with procedural fairness to reinsurance arbitration.  Taking stock of the current system, we can probably do better.

[1] Winn and Davis, Arbitration of Reinsurance Disputes: Is There A Better Way?, Dispute Resolution Journal (October 2004), p. 1.  
[2] Foster, The Brave New World of Reinsurance Contracts, 18-10 Mealey's Litigation Report Reinsurance 10 (2007), p. 1.  
[3] Dakin-Grimm and Valerio, A Case Against Reinsurance Arbitration?, National Underwriter Property & Casualty & Ben. Mgt., Vol. 106, Issue 35 (Sept. 2002), p. 1.
[4] Winn and Davis, Arbitration of Reinsurance Disputes: Is There A Better Way?, p. 3.
[5] Dakin-Grimm and Cloutier, Coming of Age: Arbitration Needs An Update, Vol. 104, Issue 5 (September 2003), p. 1.  
[6] Winn and Davis, Arbitration of Reinsurance Disputes:  Is There A Better Way?, p. 3.
[7] Foster, The Brave New World of Reinsurance Contracts, pp. 1-2; Dakim-Grimm and Valerio, A Case Against Reinsurance Arbitration, pp. 2-3.  
[8] Dakim-Grimm and Cloutier, Coming of Age: Arbitration Needs An Update, p. 3.
[9] Phillips and Moss, How To Stop Costly Litigation From Ruining Your Results, Reinsurance Magazine (February 2007), p. 1.  
[10] Id.
[11] Dakim-Grimm and Valerio, A Case Against Reinsurance Arbitration, p. 2.  
[12] Dakim-Grimm and Cloutier, Coming of Age: Arbitration Needs An Update, p. 2.
[13] Foster, The Brave New World of Reinsurance Contracts, p. 2.
[14] Dakim-Grimm and Cloutier, Coming of Age:  Arbitration Needs An Update, pp. 4-5.
[15] Winn and Davis, The Arbitration of Reinsurance Disputes:  Is There A Better Way?, pp. 2-3.
[16] Id. at 3.
[17] Foster, The Brave New World of Reinsurance Contracts, p. 7. 
[18] Dakim-Grimm and Valero, A Case Against Reinsurance Arbitration?, p. 2.,/span>
[19] Phillips and Moss, How to Stop Costly Litigation From Ruining Your Results, p. 3.  
[20] Id.
[21] Id. at 3-4.  
[22] Baron, Role of Mediation In Reinsurance, AIRROC Matters (Spring 2006), pp. 13-14, 23.
[23] Mann and Giblett, Mediation-A Rival To Arbitration In Reinsurance Disputes?, Mondaq-Goliath (Oct. 2006), p. 1.  
[24] Id. at 2.
[25] The authors acknowledge that, depending on contract wording, the introduction of mediation wording may require corresponding adjustments in other clauses, e.g. clauses triggering penalties for overdue payments.
[26] Federal Arbitration Act, 9 U.S.C. § 10(a)(3).
[27] Foster, Reinsurance Arbitration Clauses Through the Looking Glass, ARIAS Quarterly, Vol. 15, No. 3 (Third Quarter 2008), p.6
[28i] Foster, The Brave New World of Reinsurance Contracts, p. 7.
[29] An analogous proposal might limit each party to the use of one expert witness in the hearing.  Parties would be free to call more experts but would face cost-shifting if they call additional expert witnesses (increasing costs for both sides) in pursuit of claims and defenses on which they ultimately do not prevail.

The articles on our website include some of the publications and papers authored by our attorneys, both before and after they joined our firm. The content of these articles should not be taken as legal advice. The views and opinions expressed in this article are those of the author(s) and do not necessarily reflect the views or official position of Robins Kaplan LLP.


Related Publications

March 14, 2024
How Many Cases Have You Tried to a Verdict?
Gabriel Berg, Lauren Coppola - New York Law Journal
January 2024
Crack the Code: Evaluating Cyber Claim Exposure
Taylore Karpa Schollard - The Robins Kaplan Insurance Insight
January 2024
Don’t Shoot The Messenger
Michele Detherage - The Robins Kaplan Insurance Insight
January 2024
The Weight of Words
Melissa M. D’Alelio - The Robins Kaplan Insurance Insight
Back to Top