House Judiciary Subcommittee On The Constitution: Hearing On The Costs And Burdens Of Civil Discovery

Wednesday, February 22, 2012 - 14:09
General Electric
Thomas H. Hill

Thomas H. Hill

Editor’s note: Thomas H. Hill, Associate General Counsel, Environmental Litigation & Legal Policy, General Electric Company, submitted the following testimony to the House Judiciary Subcommittee on the Constitution. The hearing on The Costs and Burdens of Civil Discovery took place in December 13, 2011. With the exception of brief introductory remarks, the Editor presents Mr. Hill’s testimony in its entirety.

We at GE are pleased to assist the Subcommittee as it examines the important issues related to the costs of discovery – and in particular the costs associated with preservation – that burden potential litigants in the U.S. courts.

I practiced law in Michigan before joining GE, and I am a member of the State Bar of Michigan, the U.S. Supreme Court Bar, and the American Bar Association. I have appeared before several of the Federal Circuit and District Courts. I also serve as a member of several state and national committees and boards, including the Civil Justice Reform Group, the George Mason University Law and Economics Institute, and Lawyers for Civil Justice. I have served as co-chair of the ABA Committee on Litigation Management, and I was an editor of the ABA report on litigation management, entitled "Radical Solutions to Litigation Management.”

In my two decades at GE, I have managed a number of the company’s most complex cases in the areas of product liability, toxic tort, environmental law and insurance, throughout the U.S. and internationally. Over the past ten years in particular, I have observed GE’s embrace of modern technology and the corresponding explosion of electronically stored information, or ESI. I have also experienced the adverse impact that these developments have had on discovery, litigation and dispute resolution and the costs associated with each.

Many stakeholders in the federal civil litigation process feel the need to reform the Federal Rules of Civil Procedure to address these and other challenges. Many issues need to be addressed, including the inter-relationship and scope of pleadings and discovery, developing standards for preserving documents to ensure a fair, just and balanced legal system, which I will address today, and addressing costs with a view toward how these issues impact our civil justice system. In short, I agree with the diverse spectrum of stakeholders who feel reform should be implemented now.

The Judicial Conference Rules Committee has chosen to begin its examination with the law governing the preservation of information in litigation. It’s a good start.

American Companies Must Significantly “Over-Preserve” Documents At Great Cost Because Their Discovery Obligations Are Vague And They Face The Threat Of Sanctions

I would like to speak with you today about tremendous costs that are affecting U.S. global competitiveness and job creation at this critical time. Because the Rules Committee has chosen to start with the issue of preservation of information, the costs I will be addressing today are litigation-related costs incurred by American companies with minimal discovery benefit to the courts or to either plaintiffs or defendants engaged in civil litigation – the costs of over-preservation of information in anticipation of litigation. Let me explain what I mean by the over-preservation of information. You may intuitively think that the costs of litigation begin when GE is served with a complaint in a new action, engages counsel, and begins the process of preparing for and moving towards trial. Under the current rules, the vast majority of the time, that assumption would be wrong. Instead, GE incurs considerable costs well before then.

GE, like all potential litigants, must preserve information in anticipation of litigation. But there are no clear rules or even consistent guidance regarding when we must preserve information or what should be preserved – or for how long. This uncertainty forces GE to make case-by-case determinations about when litigation may be anticipated and the potential scope of litigation that a plaintiff might bring some day. Essentially, GE must guess whether a case will ever be filed, guess as to what claims may ever be brought, and then do its best to preserve an unspecified amount of information for an indefinite period of time.

This uncertainty is created because the current discovery rules do not provide parties with adequate guidance as to when their obligations to preserve ESI begin (or even what ESI to preserve) in anticipation of litigation. Court decisions interpreting these obligations are inconsistent, and judges often evaluate whether a company adequately preserved ESI with 20/20 hindsight. As a result, parties who seek to comply with these rules must cast a wide – and very expensive – net. American companies, like GE, that face the potential for litigation are forced to preserve vast amounts of ESI that may never actually be required in the case. Certainly much of it will never be reviewed or produced and will never see the inside of a courtroom.

For example, let’s say that it is conceivable that a particular act or decision by management might possibly lead to a shareholder derivative suit against GE. GE must decide how likely it is that such a suit could be filed. If litigation may be “reasonably anticipated,” GE must begin preserving potentially relevant information. Because no suit has been filed, however, GE has literally no one – no opposing counsel, no plaintiff, no judge – with whom to negotiate to determine what “potentially relevant” ESI might reasonably be needed or what claims might be asserted. In the face of this uncertainty, GE must over-preserve – that is, GE must preserve vast amounts of information without any guidance from the other side, even though a case may never be filed, to avoid the risk of sanctions.

Even if a case were to be filed, however, GE’s situation may not improve much, and the tremendous costs of over-preservation continue to compound. The plaintiff’s counsel has no incentive to narrow or to minimize GE’s obligation to preserve ESI. Rather, the opposite is often true – that is to say, because the preservation of ESI is expensive and the costs of preservation are seldom shifted from the preserving party to the requesting party, the plaintiff’s counsel may well use those costs as leverage to encourage the defendant to settle the case or as a weapon in his arsenal during pre-trial negotiations. This is not what our system of justice should be about. There is certainly no incentive for the plaintiff’s counsel to narrow the categories of ESI that must be stored or to minimize the disproportionate burden on GE that is created by the uncertainty of the rules.

Without more clarity, companies have no choice but to over-preserve ESI to protect their reputations in light of the possibility of sanctions. GE and most companies take this obligation seriously. GE has been in business for over 130 years and is the only original company on the New York Stock Exchange. We are consistently ranked by Fortune magazine as one of the world’s most admired companies and have been ranked by Ethisphere magazine as one of the world’s most ethical companies four years in a row. GE’s reputation is its most valued asset, and the key to protecting that asset is maintaining the highest standards of integrity and compliance. The threat of sanctions – the risk that a court would impose sanctions on GE for failure to preserve documents – poses a disproportionate reputational risk too great for most companies to bear. Uncertainty forces companies like GE to incur exorbitant costs to protect their reputations by preserving ESI that only rarely benefits either plaintiffs or defendants. With clearer rules defining our obligations, we could substantially reduce these wasteful litigation costs without impacting the actual merits of the litigation. This is an important issue – in this difficult economic period, the significant resources currently wasted on over-preservation would be better allocated to creating jobs and growing our economy.

GE, Like Other Large American Companies, Creates An Astounding Amount Of ESI

At GE, I have observed many of the changes that have occurred as the company transitioned from what was largely a paper communication world to one that relies on instantaneous communication over a variety of media. The volume of information created today exponentially exceeds that which was created even four years ago and the challenges for determining what and how to preserve such information are daunting, even in a company as large and sophisticated as GE. These challenges are even greater for smaller companies. It is quite common in today’s electronic age for small businesses to use an iPhone, iPad and/or a laptop computer. These devices can store the equivalent of millions of pages of material. As the number of employees of a small business grows, the data created can increase exponentially. The current rules require that such a party make sure that all entries and information are properly preserved, not overwritten or otherwise altered. In addition, the smaller the company the greater the likelihood that all employees’ data may be subject to holds, production and review. This impacts their ability even to function, and in the event that they become the subject of preservation and production, the effort may impose costs that dwarf the financial resources of the small business.

Rules that made sense when discovery involved boxes of documents that measured in the thousands of pages are not helpful in determining our obligations regarding the preservation and production of terabytes[1] worth of documents. A single terabyte is the equivalent of about 500 million pages. If GE preserved and produced only a single terabyte of data, the production would be equivalent to roughly 24,666 banker boxes full of documents or 11,000 copies of the U.S. Code or 22,727 copies of the Oxford English Dictionary. That is an astounding amount of data to preserve at great expense.

GE has approximately 290,000 employees. We have normal turnover, retirements and transfers of nearly 35,000 employees each year. In addition, GE is involved in the acquisition or divestiture of approximately 1,000 smaller entities each year. These routine personnel changes create the need to keep track of close to 500,000 individual data sources who could be subject to preservation obligations depending on the nature and breadth of any claim. The amount of data that our employees create on any given day is beyond comprehension.

The complexity of this obligation to preserve data is daunting. Recently we have begun to create a database of preservation and litigation holds. While this database is still under construction, currently, we have about 10,000 employees who are on litigation holds, which means those employees have been sent communications prohibiting the destruction of information on their computers or other electronic devices or in their files. Many employees have multiple holds. For example, I have no reason to disagree with Microsoft’s estimate that for preservation purposes alone, it collects 17.5 GB from each custodian in litigation (the equivalent of over 430 banker boxes of documents per custodian), which is up from an average of 7 GB per custodian just three years ago. From what I have seen, I believe Microsoft’s experience is consistent with that of many other companies. It is inconceivable that any litigator could review, select, and use as evidence at trial even a small fraction of the information that GE preserves. We are engaged in preservation for the sake of preservation.

Magnifying this problem is the complexity created when you consider the number of different media and communication systems and back-up systems in use today. Companies preserve duplicative information and preserve metadata and other forms of ESI which are rarely the best or only sources of information that the court, juries or the parties will use to resolve a dispute. Companies preserve astronomically more data than is ever reviewed, let alone used, by the parties in litigation.

Examples Of Over-Preservation At GE

I would now like to go beyond hypotheticals and provide the Subcommittee with real-world examples of these exorbitant costs and describe how American business interests will benefit from increased clarity in this area. We have shared these examples with the Judicial Conference in Dallas. The costs expressed here are very conservative and do not quantify additional costs incurred from business interruption, legal expense and other stressors.

Case number one: How long does “pre-litigation” last when no claim is brought?

Under the current ill-defined preservation standards, GE must preserve documents when it “reasonably anticipates” litigation. This vague standard has no meaningful scope or time limits. To comply with an amorphous “reasonably anticipates” standard, GE has been preserving documents for several years in a specific matter where no litigation is pending. Indeed, no case may ever be filed. Because no court has jurisdiction and there is no opposing counsel, GE cannot negotiate or seek direction to confirm or otherwise adjust the scope of the documents it is preserving. Thus, GE is incurring significant costs for extended periods preserving data that no one may ever even ask to see. Unfortunately, this example is far from unique.

The example I just described involves a single business at GE and is confined to GE’s U.S. operations. Thus, the breadth of the hold in this example is relatively narrow – there are 96 custodians. In a company the size of GE, it would not be out of the ordinary for hundreds or even thousands of individuals to be involved in a subject matter that becomes the subject of litigation that lasts for many years (which is typical for most of our matters). Even in this relatively narrow example, however, 3,800,000 documents have been identified to date, which total 16,000,000 pages of data. This data only has been coded and scanned for preservation. There has been no substantive review involved. We are simply “preserving” this data. However, the custodians generate approximately 500,000 new documents every six months which must also be preserved, coded and scanned. This ever-growing amount of material costs approximately $100,000 per month to preserve in just this one case. To date, excluding legal review, GE has incurred $5.4 million in fees for preserving this data. And because the rules do not define an  end point,  the meter is still running.

Case number two: The status quo does not incentivize proportionality.

I would also like to discuss a second impact of over-preservation, which is the disproportionate nature of discovery costs once litigation is filed. GE is involved in a smaller matter where we valued the liquidated damages at less than $4 million. In order to comply with its preservation and discovery obligations, GE has collected, preserved and produced material generated by 57 custodians. These individuals have created approximately 3,100,000 documents, requiring review and production at significant and disproportionate expense. In this matter, GE has incurred $1.5 million in fees for coding and quality control review of its production and $4 million in fees for document review, the creation of privilege logs, and the subsequent production of the material. In a case valued at less than $4 million, GE has spent, to date, $5.9 million simply organizing and producing documents to be reviewed. Few of these documents will ever be used in the actual discovery process.

Because the current rules impose obligations on us in the absence of any agreement, often opposing counsel is not interested in  meet and confer  negotiations to reduce this burden. Indeed, many even fail to actually review the vast majority of the documents produced. Rather, the norm is to have the party who created the data be responsible for all expenses associated with locating, preserving, collecting and producing the material. Only after that expense do the recipients engage in some sort of winnowing process. Rarely do courts consider cost allocation, which can incentivize an efficient focus on information necessary to prove the case. Without such economic issues fairly addressed, ESI production becomes leverage to skew resolution of matters not on the merits but rather on the economics of the case.

The outcome in this example was typical. When faced with the decision between reducing the scope of production and the plaintiff’s objection and/or shifting costs to the requesting party, the court simply defaulted to a determination that GE was better able to bear the cost of the production and had the resources to do so. Under that test, costs are never shifted.

What is most important is that it is money wasted. Resources that could be put to better, more productive economic uses to create new products are forever lost.

We Need Clearer Rules As To Our Discovery Obligations

As the above examples demonstrate, American companies are incurring significant, needless costs because they lack clear guidance as to their discovery obligations. I appreciate this opportunity to illustrate for this Subcommittee the real-life challenges faced by an American company burdened with the uncertainty of data preservation under the current rules. With clearer guidance and rules, companies can expend the resources necessary to comply fully with the document preservation rules and use the savings to create jobs and invest in their future and the U.S. economy. We will continue to work with the Judicial Conference Rules Committee in its effort to develop amendments to the Rules of Civil Procedure that will help solve some of these preservation problems as well as other systemic problems. We applaud the efforts of this Subcommittee in exercising its oversight role under the Rules Enabling Act.

[1] One terabyte is 1,000,000,000,000 bytes, or 1 trillion (short scale) bytes, or 1024 gigabytes. The U.S. Library of Congress Web Capture team claims that “[a]s of September 2011, the Library has collected about 254 terabytes of data . . . The archives grow at a rate of about 5 terabytes a month.” See Library of Congress, Web Archiving FAQs (2011),