E-Discovery — Setting standards, not cover

The Sedona Conference has a slogan: “Moving the law forward in a reasoned and just way”. That slogan is exactly the way I have described this impressive organization and what do.

The Sedona Conference regularly drafts and disseminates commentary articles on various areas of the law and the practice of litigation. I do not always agree with everything promoted, but in every publication, something is important for consideration and much of it is worthy to set standards in some cases. The name of Sedona goes a long way with judges in the value of what the organization sets forth and how it influences standards.

Recently, the Sedona Conference authored and posted for public comment a publication titled, “Commentary on Defense of process: Principles for Developing and Implementing a Sound Discovery Process”. The publication would be expected to discuss and set forth criteria for the reasonable process for planning and evolving an orderly, fair discovery process for defendants and plaintiffs alike.

What the publication actually seems to be is an instruction book for how producing parties in a lawsuit could justify their actions or inactions in the discovery process, without regard for the fairness or appropriateness for the party’s actions.  It was the first publication in which I felt Sedona had adopted a particular party’s position and then tried to form rules and guidelines for the “defense” of the position without considering the correctness or reasonableness.

Last year the Advisory Committee on the Federal Rules of Civil Procedure redrafted various rules, proposed them to congress and got them rubber stamped for passage. These new rules propose many changes, some good, most bad, and all appearing to have been drafted by corporate in-house counsel. The problem with most changes is they restrict the breadth and effectiveness of a party’s ability to conduct discovery. As justification, the committee used e-discovery and its “monumental costs” and its complexity as excuses.

In this recent Sedona publication, “Commentary on Defense of process: Principles for Developing and Implementing a Sound Discovery Process”, it feels almost as if the Sedona Conference feels the need to, in one hand justify those rule changes and, in the other hand, provide cover for the producing parties trying to take advantage of the rules in avoiding justified discovery. The publication, in some areas, seems to provide cover for producing parties with judges by citing to this Sedona publication.

At pages 3 and 4 of “Commentary on Defense of process: Principles for Developing and Implementing a Sound Discovery Process”, Sedona summarizes the thirteen “Defense of Process” principles:

The Sedona Principles for Defense of Process

Principle 1. An e-discovery process is not required to be perfect, or even the best available, but it should be reasonable under the circumstances. When evaluating the reasonableness of an e-discovery process, parties and the court should consider issues of proportionality, including the benefits and burdens of a particular process.

Comment: Nothing is perfect and I have seen no discovery process that has been perfect. Rule 26 demands communication and disclosures between the litigants for very good reason. The discovery process today is not going to a series of filing cabinets, reviewing paper and producing documents. Today’s discovery is more complex because it includes elements of computer science to accomplish it. The first principle, however, says the discovery methods used need not be “even the best available” to satisfy the successful defense of a party’s production. What should be promoted is, through thorough exchange of information and expert input, the discovery process should be the best possible under the circumstances.

Principle 2. An e-discovery process should be developed and implemented by a responding party after reasonable due diligence, including consultation with persons with subject-matter expertise, and technical knowledge and competence.

Comment: And that “consultation with persons with subject-matter expertise, and technical knowledge and competence” should be shared by both parties at least in terms of substance.

Principle 3. Responding parties are best situated to evaluate and select the procedures, methodologies, and technologies for their e-discovery process.

Comment: What is the purpose Rule 26? Why are responding parties alone the best suited? Why is the best suited not a collaborative process?

Principle 4. Parties may reduce or eliminate the likelihood of formal discovery or expensive and time consuming motion practice about an e-discovery process by conferring and exchanging non-privileged information about that process.

Comment: This seems to conflict with the defensive mechanisms provided for in Principles 1 and 2.

Principle 5. When developing and implementing an e-discovery process, a responding party should consider how it would demonstrate the reasonableness of its process if required to do so. Documentation of significant decisions made during e-discovery may be helpful in demonstrating that the process was reasonable.

Comment: Shouldn’t responding parties focus more on achieving a reasonably efficient process and less on being able to defend their process? Maintaining notes and records related to your choices is important, but suggesting duck and cover techniques suggests that should be the focus.

Principle 6. An e-discovery process should include reasonable validation.

Comment: Agreed.

Principle 7. A reasonable e-discovery process may use search terms and other culling methods to remove ESI that is duplicative, cumulative, or not reasonably likely to contain information within the scope of discovery.

Comment: Other than relevance, shouldn’t deduplication be a decision the receiving party should make? Given present technology, deduplication can be handled just as easily on the receiving end.

Principle 8. A review process can be reasonable even if it does not include manual review of all potentially responsive ESI.

Comment: And, the best way to assure that reasonableness and reduce discovery disputes is through a transparent or, at worst, a translucent process.

Principle 9. Technology-assisted review should be held to the same standard of reasonableness as any other e-discovery process.

Comment: Other than human review and culling, the only thing we are left with is technology assisted review (TAR). TAR should be held to the same standard, but should not be subject to the same methods or verification. Using TAR allows for an expanded participation by all parties because verification can be done, at least in part, without requiring one human looking over the shoulder of another. There are benchmarks that can and should be shared between the producing and receiving parties, including:

• Sampling criteria
• Sampling size
• Prevalence
• Responsive vs non-responsive documents
• Recall statistics
• Precision statistics
• Depth of recall
• Confidence level
• Confidence interval

In addition, initial seed sets should be provided in terms of positive documents. All this data assists in having a level of confidence.

Principle 10. A party may use any reasonable process, including a technology-assisted process, to identify and withhold privileged or otherwise protected information. A party should not be required to use any process that does not adequately protect its rights to withhold privileged or otherwise protected information from production.

Comment: Agreed, to the extent detailed privilege logs are provided in a reasonable period immediately following the production.

Principle 11. Whenever possible, a dispute about an e-discovery process should be timely resolved through informal mechanisms, such as mediation between the parties and conferences with the court, rather than through formal motion practice and hearings.

Comment: Agreed.

Principle 12. A party should not be required to provide discovery about its e-discovery process without good cause.

Comment: This principle should be necessary in only the rarest of circumstances if basic transparency/translucency was followed in the collection, culling, reviewing and production process. If a producing party must only use a reasonable process that “is not required to be perfect, or even the best available” we are establishing a low standard and one-sided expectation.

Principle 13. The court should not decide a motion regarding the adequacy of an e-discovery process without a sufficient factual record. In many instances, such a motion may not be ripe for determination before there has been substantial or complete production.

Comment: Were you trying to protect a receiving party in these principles, one might argue that Principles 11 and 12 should be regularly and aggressively violated to protect the best interests of the receiving party’s client. If you promote the necessity of a “sufficient factual record” you would be compelled, in many instances, documenting through record actions. In addition, to have a sufficient factual record some amount of discovery about the producing party’s discovery process seems almost unavoidable in a fair world.

Judges continue to write decisions in which they promote the value of technology assisted review and, in particular, predictive coding as the answer to all major discovery projects. I believe that predictive coding has an important place in the many tools used in the ESI discovery process. But, as the man said, “Trust, but always verify.”

Predictive coding and TAR (sans keyword searching) are both relatively new technologies. No solid standards have been established, the software varies from provider to provider and some courts are only now realizing some of the complications of these approaches.

No longer is a human reviewing the discovery and presumably making ethical, though flawed decisions on the discovery of each. Instead we are giving over those decisions to software programs with varied programing and conceded margins of error. And, bundling that software by human review, potentially flawed, to create the standards (review sets) for the imperfect or, at least, non-standard software systems to build upon.

The goal should be to provide a reasonable, fair and balanced discovery process. Not simply a “defensible” discovery process.