State, local governments struggle to keep up with e-discovery requirements

City, county and state governments are under increasing pressure to improve their electronic records management capabilities. Besides producing e-mails and other documents under legislation such as the federal Freedom of Information Act (FOIA) and state open public records laws, they must be prepared to respond to e-discovery requests if their organization is involved in litigation. A typical e-discovery project includes several steps, from document identification and collection to legal hold, review and production. 

Yet agencies often lack the resources to deploy a sophisticated archiving and retrieval system, with analytics to help determine which documents are relevant to the search. Nevertheless, citing budget woes does not relieve them of the responsibility to produce documents in a timely fashion. This was made clear in an often cited 2007 legal case, Toussie v. County of Suffolk, in the U.S. District Court for the Eastern District of New York. A Lexis.com summary states that the plaintiffs filed a series of discovery motions following the county's production of only two e-mail documents. The court told county attorneys at a hearing that "you can't just throw up your hands and say we don't store [e-mails] in an accessible form and then expect everybody to walk away."

The county then hired a computer forensics firm to restore 417 backup tapes at an estimated cost ranging from $400,000 to $963,000. The restoration process produced 2,403 pages of responsive e-mail and attachments. 

KMWorld asked two experts, Matthew Nelson and Steve Marsh, who have helped public sector agencies with e-discovery implementations to discuss how government attorneys and IT leaders can work together to cope with the increasing volume of electronic data.

Matthew Nelson is e-discovery counsel at Symantec and a legal technology expert with more than a decade of experience helping organizations address electronic discovery and regulatory compliance. (Twenty-five state government customers have purchased Symantec's eDiscovery Platform powered by Clearwell.) Nelson is the author of the book Predictive Coding for Dummies.

Q  KMWorld: What are some of the biggest challenges that public sector organizations face in complying with e-discovery requirements? Is it getting their hands around all the types of content? 

A  Nelson: The biggest challenge in both private and public sectors is that there has been an explosion of information. You have information spread across multiple departments and agencies. 

Q  KMWorld: In the current investigation into the George Washington Bridge lane closures in Fort Lee, N.J., a legislative committee has subpoenaed a large number of documents and phone records from New Jersey Gov. Chris Christie's office and the Port Authority. That case makes me ask how many state agencies would have all that information in archives that could be searched for the specific matters and timeframes requested? 

A  Nelson: I think that is a good example. The reality is that a lot of agencies are in a situation where the information is scattered all over the place. This information growth has evolved over time. But they have realized that as government agencies, they must respond on a regular basis to public records requests as well as e-discovery, so what they are trying to do is address it both proactively as well as reactively.

On the proactive side, the idea is that with archiving you can centralize where you are keeping important records that you have an ongoing legal obligation to keep as a government agency. Historically, folks thought of archiving as a storage management tool, which it is because it allows you to reduce your storage footprint through de-duplication and compression technology. But the other huge value with archiving is you can use it to automate record retention policies. So the agency may want to keep only those records it has a legal or business need to retain. The way that helps from an e-discovery standpoint is that you are decreasing the amount of information you have to go through at every request because you are actively getting rid of stuff you don't need to keep, and therefore decreasing the size of the haystack that you have to sort through to find information.

Q  KMWorld: So archiving helps on the proactive side. What about on the reactive side?

A  Nelson: Organizations are still looking for ways to respond on a case-by-case basis. From an e-discovery standpoint, initially there were a bunch of point solutions available to manage various aspects of e-discovery. You could buy a product that helps manage the legal hold process, or another solution that will allow you to collect data throughout your network. Other solutions focus on the filtering and culling of the data, and still others allow you to review the data after it has been culled. The big move in the public sector is toward focusing on integrated solutions. You want a single solution that handles all those steps in the process. It is less expensive than paying for maintenance of multiple solutions and it reduces the risk of handing ESI [electronically stored information] from one platform to another. The more times you are transferring data, the more risk you face in terms of losing data or overlooking something.

Q  KMWorld: One of the latest trends with e-discovery is predictive coding, which, as I understand it, uses machine learning to "train" document review software to recognize relevant documents within an archive. Can you explain some of the latest developments with this technology?

A  Nelson: There is still a lot of confusion about predictive coding in the market. For instance, predictive coding and technology-assisted review are terms that are often used interchangeably, and they mean different things to different people. I think of technology-assisted review as an umbrella topic, and within it there are a variety of tools within a litigator's toolkit. Those tools might include keyword searching, concept searching, the ability to find similar documents and the ability to automatically group documents by domain.

With predictive coding, you can leverage that technology to review a fraction of the documents at a fraction of the cost and then let the technology rank those remaining documents by degree of responsiveness. That technology is very important, especially in the public sector, because you don't always have the resources to do an attorney review of the documents during discovery before they are produced to the other side.

Q  KMWorld: And is the use of this technology being accepted by the courts?

A  Nelson: We are seeing broader acceptance across the board in the judiciary in terms of using this technology. In 2012, Judge Peck in the Southern District of New York issued an opinion in the case Da Silva Moore v. Publicis Groupe, stating that computer-assisted review in e-discovery is "acceptable in appropriate cases." That is promising, but it isn't a bright-line statement. Machine-learning technology has been around for decades, but the problem is that this type of technology hasn't been used extensively in the legal community yet. But it has caught on, and agencies are really interested in using the technology because of the promise that it can be more accurate and reduce costs.