The Inadequacy of the Government’s Search Efforts
To prevail on a summary judgment motion in a FOIA case, the government has the burden to show that its search was adequate. Affidavits supplying facts indicating the same are sufficient, and the court affords a presumption of good faith. The requestors need not show bad faith to defeat the government’s motion; they need only show some tangible evidence that the government has not satisfied its burden of showing the searches were adequate.
Judge Scheindlin found that it was impossible for the plaintiffs to evaluate the adequacy of the government’s electronic searches without knowing what search terms the government used. The court further found that the government did not conduct the searches with any uniformity. For example, ICE’s Office of Enforcement and Removal Operations—responsible for implementing Secure Communities—did not describe the search terms that it used at all.
“Things have changed,” Judge Scheindlin remarked regarding the effect of technology on the ability to access discovery. Although acknowledging that “the search obligations under FOIA are not identical to those under the Federal Rules of Civil Procedure,” the court noted that there are emerging best practices for e-discovery in general and regarding custodial and keyword searches in particular.
“Things have changed, and they haven’t,” responds Ian H. Fisher, Chicago, cochair of the ABA Section of Litigation’s Committee on Trial Evidence. Particularly, Fisher notes that the nature of litigation has changed in that we “live in an evolving electronic world,” but the basic principles behind discovery obligations remain the same.
“The [discovery] standard has remained the same, but what you physically have to do to defend your search has changed because technology has advanced so much and parties can store warehouses of data for relatively little cost,” says John P. Hutchins, Atlanta, cochair of the Section of Litigation’s Technology for the Litigator Committee. “While it may cost $1 to store one gigabyte of data, it could cost upwards of $30,000 for a junior associate to review that same gigabyte of data,” as noted in the 2007 Sedona Conference, which has a goal to promote best practices in e-discovery.
Emerging Best Practices for E-Discovery Includes Communication
A primary best practice for e-discovery is communication with the adversary, according to Fisher and Hutchins. Initially, communication with the adversary is instrumental in understanding your opponent’s ESI universe so that you can design effective searches. If a level of trust exists between counsel, it is possible for them to discuss their parties’ ESI structure without resorting to motion practice.
The need for effective communication does not stop, however, with the discussion of the ESI universe. For example, “If you’re going to do keyword searches, you need some sort of agreement beforehand regarding what keywords and combination of keyword searches will be used,” says Hutchins. Moreover, “The Nat’l Day Laborers case provided strong evidence that it is more efficient to agree upon a set of keywords—or some other search method—at the outset than to have the producing party unilaterally conduct a search that may result in a prolonged and expensive dispute,” says Fisher.
Simple Keyword Searches May Be Inadequate
Simple keyword searching, however, often may be insufficient. Judge Scheindlin stated, “Even in the simplest case . . . there is no guarantee that using keywords will always prove sufficient.” “Even the most complex cases, perhaps involving millions of pages of documents, come down to a few really important documents—usually a handful," says Hutchins. Predictive coding is a tool that allows the parties to get to the heart of the dispute quicker. It involves using iterative processes to teach machines to discern what documents are responsive and/or privileged to reduce the domain of responsive documents to a manageable level. Alternative search tools are available, including fuzzy logic to capture variations on words, conceptual searches, and other machine-learning and text-mining tools that employ mathematical probabilities.
Ultimately, litigators may protect themselves from adverse discovery rulings by including specific search terms and methodology rather than generalizations in their affidavits, says Fisher. But whether using predictive coding or traditional discovery techniques, being able to justify your methods to search the documents will stand as a bulwark against any possible challenge, advises Hutchins.
Ian S. Clement is a contributing editor for Litigation News.