The Yale Law Journal


Recycling Electrons, Undermining Justice

01 Sep 2006

Discovery is the linchpin of our fact-based justice system. Because social policy in the United States is often enforced by civil suits, discovery is a cornerstone of civil rights, environmental, product liability, fraud, and antitrust law. Often the only way to meet the high standard of proof required in these areas of the law is to present evidence demonstrating clear patterns of behavior and showing who knew what and when. The recent revisions to the discovery rules missed the opportunity to create rules that would encourage the use of technology to make the justice system more transparent and more just.

In January 2005, representatives of Microsoft appeared before a federal committee and made a surprising argument: the blessings of modern technology may actually make discovery—the process by which litigants request from opposing parties documents that they believe will help win their case—more cumbersome and more expensive than in the days of paper documents. Microsoft was appearing before the federal committee tasked with updating the rules of discovery. For over five years, the committee—made up of distinguished judges, attorneys, and academics chosen by the late Chief Justice William Rehnquist—had been engaged in an effort to overhaul the rules of discovery for the digital age. It might seem odd that Microsoft would disparage the technology it helped pioneer, but the company wasn’t appearing before the committee in its capacity as an innovator of software. It was appearing as an often-sued defendant.

Information technology should make discovery a more effective tool of justice. Anyone who has stored a lifetime of work on a single CD knows that the storage and exchange of digital information is vastly less expensive than paper-based alternatives. What’s more, many conversations that previously took place by telephone or in the privacy of an office are now written and saved in e-mails. The result is that more and more information—some of it potential evidence—is being saved, and on media that’s becoming ever easier to search.

While the technology continues to improve, satisfying discovery requests for electronic data can still be costly: some types of storage media are difficult to search, older data formats may be hard to recover, and salvaging “deleted” data that still exists on a hard drive can be pricey. And the vastly greater volume of documents preserved today can also increase the cost of discovery. Before producing documents to an opponent, parties must determine whether these documents are covered by attorney-client privilege or otherwise protected from discovery. Eventually technology will make searching these growing mounds of material easier and cheaper, but in the meantime, more documents can mean more time and money spent on document review.

With these considerations in mind, the federal committee began the process of updating the rules. The committee discussed the issues for five years, reviewed 180 written comments from the public, and listened to testimony from seventy-four witnesses, myself included. The committee’s recommendations were ultimately approved without comment by the Supreme Court and will go into effect this December if Congress takes no action. Though the rules of discovery have been revised often—five times since 1980, an unusually high frequency compared with other rules of civil procedure—the committee’s proposed rules will comprise one of the most important changes to the rules of discovery in the last two decades.

Many of the committee’s recommendations have been welcomed—or at least accepted—by both those who sue and those who get sued. The recommendations with broad support include provisions requiring parties to discuss electronic discovery issues early in the litigation process and reach compromises without the costs of going to court, and a requirement that information be produced in a usable form.

But not all of the committee’s proposals pleased everyone. Some of the proposed amendments drew opposition from plaintiffs’ lawyers, public-interest groups, several large bar associations, and the Federal Magistrate Judges Association. They argued that the amendments would erode litigants’ ability to uncover evidence. The corporate defense bar and their clients, on the other hand, did not share these reservations. Alfred Cortese, a top corporate lobbyist in Washington, D.C., wrote in an April 2005 report that the wording of one of the most controversial amendments “is almost identical to the proposal [we] advocated.”

One of the biggest changes is also the most disappointing. A proposed rule limits judges’ power to impose sanctions on parties for failing to preserve evidence that might be subject to discovery in a lawsuit. When the U.S. government sued tobacco giant Phillip Morris, for example, the court ordered the company to preserve e-mail messages, but the company continued to delete them for over two more years. As a result, the judge barred the testimony of individuals who had not complied with the order and fined Phillip Morris $2.75 million.

The proposed amendment creates a new “safe harbor” against such repercussions for destroying data. According to the amendment, except under exceptional circumstances judges may not sanction litigants who destroy documents by using software that routinely deletes data, as long as the system is operated in “good faith.”

By excusing the loss of records due to the “routine” operation of electronic systems, the rule encourages corporations to use software that routinely deletes more data than business concerns might otherwise justify. This is the amendment that the corporate lobbyist Alfred Cortese identified as almost identical to one of his early proposals. Joy Cunningham, the president of the Chicago Bar Association, wrote that allowing a safe harbor “is akin to encouraging the breeding of dogs to eat homework assignments.”

Litigants could still be sanctioned for destruction of evidence done in “bad faith,” but that leaves considerable wiggle room. There are a host of legitimate business reasons for routinely destroying data, among them freeing up hard drive space on central servers and maintaining trade secrets. How will judges determine whether data was deleted in good faith or to reduce the probability of liability?

Supporters of the new provision claim that it is a practical response to the nature of software. In one report accompanying the amendment proposals, the chair of the advisory committee, District Court Judge Lee Rosenthal, explained that the rule was a response to a “distinctive feature of electronic information systems,” which routinely delete information. The deletion often cannot be stopped, she wrote, “without creating problems for the larger system.”

It is true that many current systems routinely delete data. Many database systems are designed to purge old records after a certain expiration period or to erase original date stamps when records are updated. What the committee did not fully appreciate is that deletion is the result of a deliberate software design choice, not an inevitable feature of electronic systems. If the rules and the judges who implement them encouraged or required retention, they would create incentives to produce software that does not routinely delete data, a feature that would usually be feasible to implement. This is especially true for large corporations, which often invest in custom-made data-management systems.

Take again the example of Phillip Morris. In the wake of the tobacco litigation, the company spent millions of dollars to design new software to work with its Microsoft Outlook e-mail system, making it easier to save documents and e-mails relevant to litigation. If the current incentives to retain data remained in place, other corporations would likely adopt similar innovations. Phillip Morris might even sell its adaptive technology, recouping expenditures and providing these tools more cheaply to companies that would otherwise have to design them from scratch. Or perhaps a retention provision would have created a new market for Microsoft. But this virtuous circle of innovation is discouraged by the proposed new rules, which assume the inevitability of deletion-happy software.

Microsoft’s testimony before the discovery committee attesting to the arduousness of complying with the rules of discovery in the digital age identified a number of credible reasons why responding to discovery requests for electronic documents has grown more costly in the last few years. But by accepting at face value these statements about the limits of our current technology—and in particular by underestimating its ability to adapt—the discovery committee has drafted rules that could be outdated before they are approved. The committee assumed the limitations of current and past technologies to be fixed, rather than encouraging both plaintiffs and defendants to find ways of using technology to improve the justice system. The committee also chose to make changes limiting the scope of e-discovery without conducting a rigorous empirical study that measures how e-discovery costs compare to the costs of paper discovery—or that weighs those costs against the benefits of improved transparency.

In the short run, the proposed amendments may reduce legal costs for companies like Microsoft. In the long run, however, they do a disservice to our tradition of transparency and reduce the chances that we will benefit from the technological innovations that Microsoft and its peers have helped engineer.

David Tannenbaum is a third-year student at Yale Law School and a Student Fellow of the Yale Information Society Project.

Preferred Citation: David Tannenbaum, Recycling Electrons, Undermining Justice, 115 Yale L.J. Pocket Part 155 (2006),