Highlights from the Public Hearing on Proposed Amendments to the Federal Rules of Civil Procedure

On January 12, 2005 in San Francisco, the Civil Rules Advisory Committee heard testimony from 15 witnesses. This was the first of three public hearings on the proposed amendments to the Federal Rules of Civil Procedure relating to electronic discovery. The following are some highlights of the testimony. The complete testimony can be found here.

Mr. Greg McCurdy, on behalf of Microsoft Corporation:
-Argued in favor of two-tier discovery and the safe harbor provision.
-Regarding the increasing volume of electronically stored information and corresponding increase of electronic discovery with which Microsoft must contend:

oMicrosoft receives between three and four hundred million external and internal e-mails a month. 90 percent of that external e-mail is spam. (p4)
oIn our recent discovery, over 99 percent of all pages produced were produced in electronic form – less than 1 percent were produced on paper. (p4)
oComparing two lawsuits brought by the same competitor on roughly the same topics – one in ’98 and the other in 2003: the amount of e-mail and other documents that the two groups of employees had kept was seven times (or more than seven times) larger five years later. (p6)
oWhile the volume had increased, the percentage of e-mail that was actually responsive and useful in the litigation decreased significantly. In about ’98 or ’99, from hundreds of custodians about 15 percent of their documents were responsive to the request. Five years later less than 4 percent were responsive (about 3 and a half percent.) (p6)

-Case example: Microsoft acquired a small company that was sued in federal court. The company had 115 back-up tapes that were a few years old, for which plaintiffs brought a motion to compel. The magistrate ordered their production. Restoring these tapes (which required new systems, new servers, new software, and new technology) would cost a quarter of a million dollars. Once those 115 tapes had been restored and could be searched, the search process and getting ready for production would have cost another million dollars. They chose to settle based purely on economic concerns, even though they had a very strong defense. (p10-11)

Mr. Bruce Sewell, on behalf of Intel Corporation:
-“In Intel’s experience, discovery and other defense costs often exceed actual liability costs.” (p14)
-On the volume of electronically stored information: Intel has approximately 80,000 employees and nearly 300 different offices located on several continents. It possesses an enormous amount of electronic information, which resides on tens of thousands of notebook computers, desktop computers, and active servers located around the world. (p14)
-Regarding disaster recovery tapes: Intel has dozens of disaster recovery storage sites located around the world (p15). A fairly modern disaster recovery tape holds about 200 gigabytes of information – the equivalent of 90 million pages of data on just one tape. Intel uses 22,000 of these tapes every week in order to effect its disaster recovery system. (p15)
-Costs of discovery: For ordinary discovery (i.e., regular discovery out of an active file server) there is a general model of about one dollar per page. It would probably cost ten times as much for discovery from backup data. Today, our productions routinely cost many millions of dollars. We routinely produce between three and seven or eight million documents for each litigation. The cost requirements are huge and overshadows the potential liability in the vast majority of these cases. (p17)
-“We are absolutely in support of the two-tier approach. And … what we are particularly supportive of is the notion of a presumption that certain kinds of things are inaccessible, and that disaster recovery tapes should be among that category.” (p18)

Ms. Kathryn Burkett Dickson, on behalf of California Employment Lawyers Association:
-The California Employment Lawyers Association has about 550 members – primarily sole practitioners or small firm attorneys who represent primarily individuals in employment discrimination and labor litigation. (p20)
-“We obviously love the provision for authorizing the receiving party to specify the format for production. We have had instances in our office already where we have asked for production of e-mails. They did come from the company’s storage system. They did supply them to us. They were Unix based. We can’t read them at all. It would cost somewhere between 15 and 30 thousand dollars to acquire the hardware and software we need to read those.” (p25)
-“I will say that I have followed somewhat the development of the storage technology industry. And all you have to do is just go on the Internet and look up EMC, Veritas, Legato, Hitachi, Intel. Everyone is saying they are creating hardware and software right now to store and to allow searches of massive quantities of data. The justification that you’ll see from a lot of those companies is that they have to have this information for a long period of time and that they have to have it in a reasonably searchable form. The technology is coming, and it’s coming quickly. … I really think the notion that we’re not going to be able to search massive quantities of data is a temporary problem.” (p26)

Mr. Michael K. Brown
Regarding 2 tier discovery:
-[In litigation representing pharmaceutical companies] “we get a request every single time for back-up information, for dynamic database information.” (p30)
-“I’m involved in a lot of major pharmaceutical kinds of litigation. I’m not aware of anywhere the only place the information was ever found was on a back-up tape or some other disaster recovery.” (p30)
-“There is no bigger cost in litigation today than electronic discovery. … it’s already huge dealing with the active data. You then take it to another level — much less start going to every country in which we sell a product, which is a different issue — and then the cost becomes prohibitive.” (p30)

Ms. Joan Feldman, on behalf of Computer Forensics, Inc.:
-“This is a key fact in American businesses, that for each document that exists, there are probably a minimum of five exact copies of that document located throughout the system, active and online, not including things that might be on back-up tapes or on off-site storage.” (p39)
-“Another issue we may have to take on that may be more difficult is more near duplicates.” (p39)
-“I watched one of my clients spend over three and a half million dollars and three months of a special magistrate’s time arguing on what the difference between what’s a field in a database and what’s a record in a database. I am not making that up. What I’m saying is that they chewed through millions of dollars for attorneys’ motion for practice, our time, the court’s time, because early on when the discussion came up about producing databases, they weren’t clear on their terms.” (p40)

Mr. Thomas Y. Allman
Regarding preservation:
-“I would say that in probably 75 to 85 percent of your cases, it is not in fact a problem. The great majority of cases are tried — the information is collected and preserved without a lot of disputes over preservation.” (p44)

Mr. Jeffrey Judd:
-Argued against specific reference to preservation in the list of items to be discussed as part of the Rule 26(f) meet and confer process. (p53)
-Endorsed the proposed two-tier approach to discovery, but expressed concern that the focus on accessible vs. inaccessible in the note should be accompanied by a reminder that a proportionality test still applies when determining whether information which is accessible need be produced. (p54-55)

Mr. Gerson Smoger:
-Argued generally against changing the Rules to implement two-tier discovery, or to provide a safe harbor against sanctions. He argued that courts already can and do treat accessible and inaccessible information differently, that the imposition of sanctions is currently a very infrequent measure, and that these additions would only serve to make the system too rigid. (See testimony starting on p55)
-“… in dealing with electronic discovery, we are really dealing with all discovery. We have to be cognizant of the fact that more than 90% of all information is electronic now. … So these are not separate rules for electronic discovery. In reality, these are rules for all discovery, with a small exception.” (p62)

Ms. Jocelyn Larkin on behalf of The Impact Fund:
-” … we are currently lead counsel in the Dukes vs. Walmart litigation, which is a gender discrimination class action on behalf of 1.5 million female employees. I will say we’ve gotten an awful lot of electronic discovery in that case. We filed it four years ago, and last I checked, Walmart is still in business. They seem to be doing fine, despite having to deal with us and many other large cases against them.” (p63)
-In support of the proposed requirement that parties discuss any issues relating to preserving discoverable information: “… putting into the rules the requirement that essentially the parties work on that at the front end is going to be very important.” (p64)
-JUDGE SCHEINDLIN: In the typical case when you make this request to defense counsel, do you for example ask them to retain all back-up tapes?
MS. LARKIN: Often we do, because we have no idea what’s there. At the outset, as we’ve described, we don’t necessarily know.
JUDGE SCHEINDLIN: And do you ask them to stop recycling?
MS. LARKIN: No, no, we do not. (p65)
-On the challenges facing nonprofits who are working with limited resources: “… for us, every time we think about whether we’re going to district court, we have to think about the time it takes, how much it costs us, and really whether we’re ultimately going to have … a friendly reception from the district court. … And so one of the reasons that I have a good deal of difficulty with the two-tiered system I think is that it increases the likelihood of litigated discovery disputes.” (p65)

Mr. Frank Hunger:
-Recommending a provision in Rule 26(b) that if the court “decides that good cause has been shown for inaccessible information, that there arise a presumption that the party who is requesting the information have to pay, and this presumption can be overcome by clear and convincing evidence that an injustice will be done if that is in fact made a part of the proceeding.” (p69)

Mr. David Dukes:
-“in a recent case, one of my clients searched between 400 and 600 million electronic documents. That search led to 8 million electronic documents that were deemed to be potentially responsive. Now, this was significant national litigation. And my point here is not that they had to search too much. But what they did search, we would consider to be reasonably accessible information or active data. And the point that I make here is, I think this is an example that, even if we limit this to two-tiered discovery where clients are searching for reasonably accessible information, we are not going to eliminate a lot of things that would potentially lead to discoverable evidence. I think you’re going to find that you’re getting the bulk of the material through what is actual, active data.” (p74)
-“Another example is one of my clients in the last several years had seen their IT staff, which is devoted strictly to complying with electronic discovery requests in litigation, increase over 50 percent.” (p74)
-“One of the questions you’ve asked is, will technology solve this issue of accessible information versus inaccessible information? I have some personal history for that. In 1990 I was national counsel for a software company. I spent three years on discovery of electronic information in the Bay Area, spent two and a half months before Judge Claudia Wilkin over in the Oakland courthouse on a software case. And at that time consultants and software vendors were telling us, we have products that will make everything in a corporation accessible. What we found was there was a disconnect between the marketing arm of these consultants and the technology arm of these consultants. They couldn’t do it then. … we need to be careful not to be lulled into the belief that the marketing pitch about technology addressing these issues of accessible information is going to really solve it. So I would ask you, if you’re seriously considering that argument, make people show you how technology will solve this accessible versus inaccessible issue. Please don’t just rely on the marketing pitch as to how it would be done. It wasn’t done ten years ago, and based on what I’ve seen, there still does not exist technology that can make everything in a corporation that people are requesting be produced accessible.” (p75)

Ms. Jean Lawler on behalf of Federation of Defense & Corporate Counsel:
-“The FDCC supports both the two-tiered system for discovery of electronic information and the safe harbor provision. We agree with the reasonably accessible standard but believe that it should refer only to data used in the actual course of business, not the disaster recovery systems.” (p78)
-“In terms of safe harbor for sanctions, we believe that there should be safe harbor where information is unavailable due to routine computer operations, and we do believe that there should be some willfulness factor there.” (p79)

Mr. Kenneth Conour:
-“In my practice, I get a lot of requests for data, but I also get requests for actual databases. I’m not sure exactly what this means. When you’re talking about dynamic databases, you’re talking about something like on an Oracle or Sequel or another platform. And typically these databases are large relational databases. The use of the database is made through an enterprise application that a company licenses at a cost of several hundred thousand dollars. The databases often contain dozens of tables, with each table containing multiple fields and sometimes tens of thousands or hundreds of thousands or even millions of records.” (p84)
-“But it’s not just the ease of getting it. It’s the ease of producing it that I think is really driving this. … In pharmaceutical litigation, we are always asked to produce safety databases. These are the databases we use to track adverse event reports. Everyday the company is entering new information into the database, generating reports for the FDA, generating reports for its own safety surveillance. These databases typically include hundreds of tables. Each table will have multiple fields with sometimes millions of records. Only some of those records will be of interest to the litigants, because they will involve patients using the particular drug in the litigation. To go through that database and pare down the records to those that might be relevant to the litigation requires you to go through each table and extract out that information which isn’t relevant to the litigation, figure out how the tables are linked, figure out what the codes mean, and go ahead and do that. To do that, you have to generate queries to isolate that data. And these queries are different than what you normally use in business. In business you usually use a query that helps you come up with some small subset of data. But for litigation purposes, you have to design a query that goes across all of the tables and pulls out all of the data that might be relevant. So it takes a substantial amount of time to do that. Once you’ve done that, you’ve narrowed it down, you now have to export that data into some other fashion. And in exporting that data, you have to be careful that you haven’t somehow impaired the data itself or the data structures. Because when you’re dealing with that large volume of data, you’re going to end up with those types of problems. Even then, before you can hand it over to the other side, you have to then go through all of the records and redact out that information that’s required to be protected by law, and those can be in millions of freeform narrative records where you have to go through each and every one and redact out the information. So I would submit that in that context, that type of the database is not reasonably available for production.” (p87-88)

Mr. Henry Noyes:
-“Essentially, I agree that it is helpful for the parties to meet and confer at the outset about preservation of discoverable evidence and issues that might arise with respect to discovery of electronic information. But I think that the insertion of this new phrase, ‘electronically stored information,’ is not necessarily good.” (p95)
-“I however don’t think that it’s beneficial to insert into the rules a specific requirement that the parties meet and confer regarding whether on agreement of the parties the court should enter an order protecting the right to assert privilege after production of privileged information.” (p95)
-Argued in favor of distinguishing between two tiers of information (p97), but not requiring the party to identify the inaccessible information (p98), and argued that a rule regarding treatment of what is reasonably accessible versus not reasonably accessible “ought to apply to both categories of information, electronic and otherwise.” (p99)
-Argued against proposed Rule 26(b)(5)(B) regarding inadvertent production of privileged information. (beginning at p100)

Mr. Charles R. Ragan:
-Regarding the problems presented by dynamic databases: “that phenomenon is probably the one aspect of the developing technology that may not be adequately reflected in the notes as you have them today. … In my experience, the proliferation of databases that are proprietary, customized, and unique to a particular company is the single most difficult issue. … To take a snapshot of that is an impossible situation. If that were the rule, it would freeze every entity that has computers, in my mind.” (p106)
-Argued that additional guidance from the federal rules is important because it will raise the CLE bar, which in turn causes the overall understanding and standard of care to rise with it. (p107)
-Argued that the standard for safe harbor should not be negligence, but something higher. (p109)