The unexamined consequences of taxing eDiscovery costs against losers

The unexamined consequences of taxing eDiscovery costs against losers

Image courtesy of

Earlier this year, a federal judge in Denver defied a growing and increasingly authoritative body of case law in awarding a victorious defendant in a discrimination suit about $57,000 in eDiscovery costs, forcing a not-for-profit halfway home for recovering opium addicts to foot the bill of the party it sued -- in this case, a Colorado state health regulator.

The decision, which reimbursed the prevailing defendant for costs it incurred to perform a broad range of discovery activities (including ESI "retrieval"), marked a significant departure from some amount of judicial consensus that had seemingly congealed around the opinion that only eDiscovery costs that could be construed as "copying" or "duplicating" information -- about 10% of all eDiscovery costs, by some estimates -- are reimbursable to winning parties.

In sum, this particular federal judge said that, under the federal cost-taxing statute and Federal Rule of Civil Procedure 54, which allows costs "other than attorneys' fees" to litigants who prevail, eDiscovery costs should be paid by the loser. Period. Not some eDiscovery costs. Not costs that can be interpreted as falling within the narrow limits of "exemplification," as the cost-taxing statute outlines.

All eDiscovery costs.

The ruling was hailed as a restoration of sanity. Of course eDiscovery costs should be awarded to prevailing parties, the commentariat said, as if such a thing is common sense. The cost-taxing statute, 28 USC § 1920(4), was written for a paper world -- when out-of-pocket costs for making actual physical copies of documents for use in cases was to be taxed.

Now that litigation is waged with arsenals of electronic data, the statute should be interpreted to include costs incurred for that, advocates of broad cost awards said.

But that line of thinking, while compelling, sets aside or ignores completely very serious access to justice issues that further tilt the playing field against less-monied parties, who are already at a disadvantage in federal courts. Broad allowance of eDiscovery costs also serves to reinforce the EDRM-driven patchwork of disconnected vendor services and eDiscovery processes, incentivizes problematic billing practices, and potentially penalizes parties who utilize outside help or technology that is either "soup-to-nuts," priced at a flat rate, or both.

The gatekeepers

There are generally two gatekeepers a prevailing party must bypass to recoup eDiscovery costs, and it is worth pausing here to consider whether either is qualified to adequately evaluate the legitimacy of the bills thrust before them.

The first is the clerk of the court, who has the responsibility to comb through the invoices submitted for cost recovery purposes to evaluate a) whether those expenses were legitimately (i.e. "reasonably") incurred -- and not, for instance, merely for the convenience of counsel and b) whether those expenses fall within the parameters of what can be recovered according to the language of the cost-taxing statute, precedence in the jurisdiction and other factors.

If all eDiscovery costs were awarded in every situation (i.e. if every court in the U.S. became Denver), the latter determination becomes moot. But a clerk still must rifle through potentially highly technical invoices to discern whether the costs incurred for the activities to which those bills speak were "necessarily incurred" to fulfill discovery requests.

Generally, the application by the prevailing party to recoup costs is accompanied by affidavits by the lawyers and the service providers who performed or supervised eDiscovery in the case -- affidavits that include technical jargon about the activities and data at issue. (e.g. "Non-standard acquisition of SharePoint files.")

Should those costs pass muster with the clerk, and if a costs challenge arises from the losing party (they always do if the amount in question is significant), the determination of awards goes to the final (pre-appeal) arbiter -- the district court judge. This person is also in a position of handicap, having had limited interaction with the discovery issues in the case -- those generally being the purview of magistrates. It would be unfair to ask a spectator on the arena's mezzanine level to properly score a boxing match. This is hardly different.

To be sure, judicial officers by and large aren't asked or expected to be experts in every matter that comes before them. The key point here is that -- as the lack of coherence among cost-taxing decisions bears out -- the judicial awarding of eDiscovery costs is a crapshoot, and thus it is hard for parties to make an informed calculation about what they stand to lose or gain should a case go to trial or a dispositive ruling be handed down. It is clear from the Colorado decision that some judges are completely unaware that there is even a debate over whether electronic discovery costs should be taxed.

The chilling impact of eDiscovery cost awards

Setting aside the qualifications of the auditors, which in itself is a consideration that has the potential to impact the appetite of parties to engage in broad discovery, there is a potential chilling effect to taxing eDiscovery costs against losing parties -- a chill that grows as litigants become more aware of the prospect of paying (or recouping) vast post-judgment sums.

The issue is fairly straightforward: Taxing eDiscovery costs essentially asks parties seeking information to push more of their chips into the center of the table. When parties request broad electronic discovery -- when they request any electronic data at all -- they run the risk of having to pay whatever costs their adversaries incur to produce that information should they lose.

This is an important and under-examined issue, not just because it stands to sway millions upon millions of dollars in post-judgment allowances each year, but because, if the ability for winners to recover vast sums for eDiscovery expenses, thereby defying the "American Rule," ultimately prevails, it could further stand to chill parties with worthy claims from fully litigating discovery, seeing that period through, and pushing on to more substantive matters. Given that large producing parties in federal cases are generally well-funded defendants, one could reasonably make the argument that this chilling effect disproportionately impacts smaller plaintiffs.

In other words, plaintiffs now stand to lose harder and more spectacularly should they have the fortitude and stamina to make it to trial, or otherwise obtain a dispositive ruling on the merits.

There are, to be sure, potentially positive consequences here -- namely, that discovery demands may be tempered for fear that the requestor will ultimately foot the bill. This may give pause, for example, to the proverbial patent "troll," who may think twice before forcing the hand of a data-rich Silicon Valley company (and, in fact, the 9th Circuit, the playground to many such lawsuits, has traditionally been generous in awarding eDiscovery costs).

Indeed, one person's chilling effect may be another person's proportionality -- which is to say, overly broad discovery requests are less likely to be made if the requestor is leery of having to pay the costs incurred to meet the request. And, in general, eDiscovery costs are more likely to be in alignment with the value and importance of the case.

Dear vendor, you are now in the 'copying' business.

Congress amended the language of the federal cost-taxing statute in 2008 to include costs, other than attorneys' fees, for “exemplification and the costs of making copies of any materials where the copies are necessarily obtained for use in the case.” What that word-stew means is anyone's guess, but "any materials" can be presumed to cover more costs than “copies of papers," which is what the statute previously covered.

"The definition of 'copying' as well as 'non-attorney's fees' has been broadening in some district and circuits, and narrowing in others. So we have a vary inconsistent view of what's recoverable across the nation," Bill Gallivan, a founder and executive of the eDiscovery vendor Digital WarRoom, said on a recent ACEDS webcast. "The circuits are all over the map on not just what is the definition of 'recoverable cost,' but how much and how many of those costs you can recover."

As Gallivan's quote emphasizes, the tremendously ambiguous language of the statute -- to what discovery activities do "exemplification" and "copies" refer? Data conversion? OCR? Forensic collection? Production? -- is the primary reason for why there is no consensus on how to tax costs, or what, actually, is taxable.

But the ambiguity also opens the door for shrewd parties and their vendors to craft the language of their eDiscovery bills and receipts in such a way that they are more likely to be awarded by the court based on whatever reimbursement winds are prevailing in that particular jurisdiction. This is not to say that vendors are writing bills that describe work that is different from what was performed (although some may be) just so the costs may be deemed recoverable, only that there is an incentive to draft invoices and affidavits in such a way that is more mindful of what may be reimbursed, and less forthcoming about the work actually done. (Consider, too, whether the aforementioned gatekeepers are in a place to make a call one way or the other.)

As a sidebar, it is also interesting that, given the inclination for some courts to resist awarding costs incurred for intellectual efforts (project management, for instance), vendors are potentially encouraged on invoices to reposition high-value work as lower value work as a means for their client to recover those expenses.

Reinforcing the status quo

Cost-taxing has other implications, perhaps lesser, but still worthy of mention here. First, the way in which bills are examined for categorization into different eDiscovery process "buckets" (e.g. forensic copying vs. processing vs. culling vs. analysis, etc.) reinforces that process fragmentation and may encourage piecemeal billing that can be easily fit into those buckets. This potentially discourages use of alternative fee arrangements, principally all-in flat-fee billing where bills are not easily analyzed by the recoverable/non-recoverable binomial. How, for example, would a judge distinguish allowable costs versus non-allowable cost incurred for the use of a tool that performs multiple functions across the discovery process, but is billed at one lump sum, or on a subscription basis? (consider, too, that such a tool is perhaps more likely to be used by a party of lesser means.) Is reimbursement of costs in this instance an all or none proposition?

Secondly, taxing eDiscovery costs also, for good or bad, encourages outsourcing of eDiscovery services. For all the uncertainty, the one constant in the cost-taxing equation is that expenses are generally only allowed to parties who utilize outside help in one way or the other. This phenomenon, again, perhaps disproportionately impacts parties of lesser means.

Still, there are pluses to be taken from the debate over cost-taxing that should be lauded for as long as this dialogue continues. For one, that the true costs of eDiscovery see the light of day at all -- and are available for public consumption -- is a side effect whose democratic merit is hard to argue.

Robert Hilson is a director at He can be reached at

Want to see Logikcull in action? 

Let us show you how to make Logikcull can help you save thousands in discovery.

Want to see Logikcull in action? Let's chat.

Our team of product specialists will show you how to make Logikcull work for your specific needs and help you save thousands in records requests, subpoenas, and general discovery.