COST: Analysis
By Digital Discovery CLE Teaching Fellow Alexi Maltas

The embrace of digital technology for the communication and storage of ideas and information has led to an unprecedented proliferation of recorded data. For example, consider email: it is not uncommon for the average employee at a company to receive more than 30 emails each day. For a company of 10,000 employees, this means around 2.1 million emails a week and around 110 million emails each year. More than 35 percent of corporate communications these days will never reach paper. Since email is so quick and easy to use, people tend to be more informal, and therefore less cautious, in email messages. Add to this other forms of electronic media such as computer hard-drives, other magnetic and optical storage media, voice mails, palm pilots, etc., and there is an enormous Pandora's box - or treasure chest - of information within each company. Furthermore, most of these files contain hidden, embedded information - such as when and by whom it was created, when it was printed, to whom it was sent, changes that were made, on which computer it was generated - all without the knowledge of the file's creator. Finally, electronic files that are "deleted" are rarely fully erased from a system's databases. (Much of this paragraph adapted from Jessen, John, "Special Issues Involving Electronic Discovery," 9 Kan. J.L. & Pub. Pol'y 425, 2000.)

In the electronic age, the first request for discovery can be a powerful tool in litigation, one that can quickly change the dynamic of the case. Digital discovery represents another significant step toward equalizing the relative power of litigants. Geoff Stewart, of Jones Day, says: "The ability to search by key word and by Boolean logic will become an equalizer in litigation. When I began to practice law in the 1970's, we associates at large firms chuckled at the superiority of our filings over those of small firms because of the better stenographic resources big firms had over small ones. Word processors eliminated that, just as Lexis/Nexis and WestLaw later eliminated the disparity between firms in library and IT resources. Now, the ability to use searches, data mining and artificial intelligence may allow even small firms to tackle large ones on an even footing when it comes to discovery."

A responding company that does not have a clear preservation policy in place may find itself facing a hugely expensive and time-consuming discovery request, potentially searching through thousands of back-up tapes. Nothing brings a settlement offer to the table faster than a discovery request that will cost as much as what the requestor is seeking. Requestors will inevitably seek as broad a discovery order as possible, one that will be sure to preserve everything even plausibly relevant, and that may stun the respondent with it size and scope. Respondents, who are likely unaware of the contents of their databases, will seek to narrow the scope and increase the specificity of the request in the hope of limiting the extent that the Pandora's box will be opened.

The biggest issue in cases of extensive discovery of backed-up data is the cost/benefit analysis. Reasonable search procedures (through, for instance, a statistical sampling method or a Boolean method similar to a Westlaw or Lexis search) may yield 95 percent of the relevant data, but tracking down that last 5 percent could raise costs and length of time exponentially. Should a court nevertheless order discovery on that last 5 percent? Do requestors have the right to find the potential "smoking gun" in that last 5 percent? Should respondents have to bear an enormous cost to find information of questionable relevance? How does a court determine whether the high cost of the request is due to the respondent's record-keeping procedures or the requestor's overbroad request?

Generally, courts have held that, when the responding company's own record-keeping scheme is the reason that discovery is burdensome, then the responding company will be expected to bear the cost. See, e.g., In re Brand Name Prescription Drugs Antitrust Litigation, 1995 WL 360526, N.D.Ill., 1995; Toledo Fair Hous. Ctr. v. Nationwide Mut. Ins. Co., 703 N.E.2d 340, Ohio Com.Pl., 1996. As the courts have defined it, this standard tends not to be a high one: included in the category of record-keeping schemes that lead to burdensome discovery are many standard practices, such as creating back-up tapes that are retained but are not catalogued. Courts consider retrieval of information an "ordinary and foreseeable risk" when a party chooses to store information electronically. See In re Brand Name Prescription Drugs Antitrust Litigation.

However, at some point, especially when the relevance of the information sought is marginal or speculative, if respondents are producing a substantial percentage and producing more would increase expenses dramatically, requestors may have to bear some or all of the cost. The exact location of this crossover point - the point that marks where requestors would start to pick up the tab because the burden on the respondent would be too great - has not been uniformly determined. The court's decision will likely be based on a balance of the relevance and importance of sought information, the cost and burden of production, and the reason for the costs. (See, e.g., Anti-Monopoly, Inc. v. Hasbro, Inc., 1995 WL 649934, 1995 U.S. Dist. LEXIS 16355, S.D.N.Y., 1995. See also, Grenig, Jay E. "Electronic discovery: Making your opponent's computer a vital part of your legal team." American Journal of Trial Advocacy 21 (Fall 1997), 293.)

The central impediments to the data-owner's ability to convince a judge to force sharing of discovery costs is that (1) traditionally, the owners of data have born the costs of searching it; (2) the owners of data chose and benefited from the information system; see, e.g., Linnen v. A.H. Robins Co, 10 Mass.L.Rptr. 189, *9 (Mass. Super. 1999) (3) sensible lines along which to split costs do not readily suggest themselves; and, (4) in order to make a judgment about sharing costs of discovery, a judge will have to get pretty deeply involved in the details of the discovery problems, a task from which judges have historically distanced themselves.

Ultimately, companies need to have a clear preservation policy in order to have a handle on their information. An easily searchable database containing a few weeks or months worth of back-ups is much more manageable and cost-efficient than one going back years or decades on thousands of back-up tapes in a distant warehouse. Such a policy will keep expenses down and minimize the otherwise very real possibility of surprise. It will also allow companies to have some degree of control over the information its employees create and disseminate. However, we should not underestimate the difficulty of implementing such policies where electronic data is concerned. Harry Baumgartner, from BASF, explains that every big company is looking for a system that will enable it to get control of its digital information, and, he says, no one is yet offering such a product on the market. Moreover, companies have a tough time implementing a drastically shortened retention time for back-ups. If they have any pending litigation, any destruction of back-ups could risk a spoliation sanction, as destroying back-ups would mean destroying material not-yet indexed and, therefore, potentially relevant to the litigation.