I don’t know U.S. Magistrate Judge James L. Cott of the Southern District of New York. I couldn’t even find a photo of him save the one at left singing with the Harvard Krokodiloes circa 1978. But Judge Cott has my admiration and sympathy for a good decision that’s under attack by well-funded business interests who’ve picked the wrong case to rally their troops.
The truth is, in Pippins v. KPMG, it was KPMG’s poor presentation of the issues and evidence (ironic, as they’re one of the biggest e-discovery vendors in the trade) that left Judge Cott no option but to rule as he did–and surely the smart folks hired to condemn the decision know that. But that’s not stopping the rush to pillory the magistrate. One big firm commentator even called the result “stupid.” That’s just mean.
What really ticks off the big firm bar and Chamber of Commerce lobby is the fact the judge didn’t simply rule and slink away. No, His Honor had the temerity to explain why he couldn’t just incant “proportionality” and miraculously gloss over the glaring absence of factual and legal justification to support KPMG’s request to discard evidence.
Pippins is a federal Fair Labor Standards Act and New York State Labor Law class action brought on behalf of current and former KPMG audit associates allegedly deprived of overtime. Before class certification, KPMG elected to amass 2,500 hard drives from computers used by members of the putative class. No one ordered KPMG to do so. It was KPMG’s decision, and probably a sound one. Neither was KPMG ordered to preserve relevant material by retaining the entire contents of the hard drives. They could have made a defensible choice to preserve just the relevant contents and toss the drives. Again, holding the data as complete hard drives was a cautious and cost-effective choice, but it was KPMG’s choice, not something foisted upon them.
Having pursued a broad whole drive preservation methodology, KPMG thought better of it and asked the court for a protective order allowing it to abandon its preservation strategy and destroy the drives it held. Of course, KPMG doesn’t need the court’s permission to do this, as it wasn’t acting pursuant to a preservation order. What KPMG wants is the court to anticipatorily shield it from a claim of spoliation should it later turn out that the drives KPMG discarded held relevant, material and non-duplicative information. Courts aren’t in the habit of giving such rulings, but Federal Rule of Civil Procedure 26(c) affords judges broad discretion to issue protective orders “for good cause … to protect a party or person from annoyance, embarrassment, oppression, or undue burden or expense.” As movant, KPMG bore the burden of proving its need for protection.
You don’t have to be a computer forensic expert to know that these drives will shed light on the audit associates’ work done using the computer and the timing and duration of that work. But will it be unique and material evidence when compared to information available from more accessible sources? Hard to say without knowing what you’ve got.
But if you happened to be a computer forensic expert, you’d also know that the information on the drives demonstrating the hours worked by the users and the nature of that work isn’t going to be found by running a few artful keyword searches across the drives. Instead, it will require analysis of usage by contextual examination of metadata and system files like logs, links and registry entries. The drive will serve up a detailed picture of its usage, but it won’t be neatly found in a file called, “Hours I Worked for Which I Wasn’t Paid According to Law.doc.” Or maybe it will be in files like that, as well.
Which is the point: You can’t say what’s on the drive without having a look!
And that’s exactly why Judge Cott had no choice but to deny KPMG’s request for a dispensation to destroy the hard drives. In what seems to me to be a monumentally obtuse (some might say “stupid”) strategy, KPMG took no steps to look at the contents of any of the 2,500 hard drives it preserved as potentially responsive evidence, not even the drives of the three named plaintiffs. Of course KPMG didn’t have to look at all 2,500 drives it deemed important enough to save, but maybe just a few, a sprinkling, a smidgen, a sample!
Hmm, sampling, where have I heard that before? Oh, wait, everywhere! Everyone in e-discovery has been talking about the importance of sampling for some time–everyone including KPMG’s own very able e-discovery leadership. KPMG has outstanding e-discovery and forensic technology services teams, so you have to wonder: Do the cobbler’s children have no shoes?
When you hold the data, you don’t need the other side’s or the judge’s permission to sample your collection. You just do it, because then, when you make claims about the contents of the drives, you’re not just talking out of your ass! You have evidence. I grant no one actually tries cases anymore, but has everyone forgotten that really nifty stuff called “evidence?!?”
So, on one side, you’ve got plaintiffs—KPMG employees and former employees—who’d actually used the computers, knew something of their contents and asserted there would be relevant data on the drives (i.e., “usage information, log-in/log-out information, application records, forms, and other contents stored on the laptops in the course of the business day, which would be helpful in showing both the range of hours during which they worked and the substance of their work,” (slip opinion at p. 4, quoting the plaintiffs). On the other side is a defendant who held the drives at great expense ($1.5 million and mounting), but hasn’t looked at a single one of them and won’t let the plaintiffs look.
It’s worth mentioning that discovery was stayed pending class certification, leaving the plaintiffs without the arsenal of requests for production, interrogatories and 30(b)(6) depositions to otherwise secure hard evidence about the contents of their former hard drives. KPMG held all the cards, I mean, drives.
What’s a judge to do?
Did Judge Cott order KPMG to undertake a costly forensic examination of the drives?
Did he require KPMG to turn drives over to the Plaintiffs?
Did Judge Cott inject his own notions of what the drives might contain in the absence of evidence?
Uh-uh. No way.
Judge Cott did what smart, fair judges do. He preserved the status quo and sent the parties away with guidance that will either enable them to work it out or assist KPMG in returning with sufficient evidence to allow the judge to weigh the relief sought.
As Judge Cott succinctly put it: “[A]t this point it is not entirely clear what the hard drives contain, in part because of KPMG’s own efforts to keep that information at bay…. Additionally, KPMG’s ongoing burden is self-inflicted to a large extent. It is suffering from the effects of its own reluctance to work with Plaintiffs to generate a reasonable sample that may well be less burdensome to maintain.” (slip opinion at p. 11).
Put another way: Don’t argue disproportionate balance with your fingers on the scale.
But I’m not selling KPMG short here. KPMG did propose using sampling, except in a way that it had to know makes little sense. KPMG proposed to keep 100 of the hard drives it preserved, chosen as a random sample. KPMG would discard the other 2,400 drives, search the sample using keywords, preserve only the material responsive to the keywords, then discard the sampled drives, too. And they wanted to be protected from sanctions if it turned out the discarded drives held material, non-duplicative evidence. What’s wrong with this picture?
As I said earlier, using keywords to find usage data is like trying to pick up plastic paper clips with a magnet. It’s the wrong process, a point plaintiffs made with the court. If the search technique won’t find what you’re looking for, the size or quality of the sample doesn’t matter.
I also thought KPMG’s burden argument unappealing. Per the court, KPMG wanted “the costs that may be incurred if these hard drives remain available and subject to discovery in other litigation” to be weighed as a proportionality factor. There’s something vaguely repugnant in a litigant asking to destroy evidence in one case for fear that it might be probative in others. Sure, you worry about that, but don’t expect judges (who see too many claims of spoliation, as it is) to be swayed by such an odoriferous argument. At least when we’re on the record, let’s rise above our partisan interests and pretend that probative evidence is a good thing because it gets us closer to the truth and a just result.
KPMG was wise to seek a protective order limiting its preservation obligations. Nothing ventured, nothing gained. But it unwisely employed the old e-discovery playbook: Trotting out fabulous cost projections on flawed assumptions expecting to persuade the judge to rule by the seat of his pants instead of based on competent evidence. Judges today aren’t splitting the baby just because they don’t like e-discovery or think it costs too much. They understand that ESI you don’t preserve is ESI that’s gone forever (or must be painstakingly reconstructed at enormous cost).
Proportionality has its place in e-discovery…and in assessing the adequacy of a party’s preservation effort. No fair-minded person wants litigants forced to spend more preserving evidence for a case than the conceivable financial or societal value of the case. But Pippins isn’t the right test case for proportionality, and I’m betting the Chamber of Commerce wasted its members’ money meddling in magistrate’s court.
So, attaboy and illegitimi non carborundum, Judge Cott!
Afterword: The amici should count themselves fortunate that Judge Cott didn’t take too close a look at KPMG’s efforts to engraft the proportionality principles of Federal Rule of Civil Procedure 26(b)(2)(C) onto the preservation obligation, or the judge might have pointed out some truths that business interests really don’t want to see in print.
The duty to preserve evidence doesn’t flow from the rules of procedure. It’s a common law duty. Discovery, by contrast, is a creature of statute. It’s a right afforded and bounded by the rules of procedure. This is no trivial distinction. The Federal Rules of Civil Procedure may not abridge substantive common law rights without a congressional mandate. Thus, it’s improper to extend the reach of the Rules to limit substantive rights absent a clear statement of intent by Congress.
The proportionality language in Rule 26(b)(2)(C) is a limitation on discovery “otherwise allowed under these rules or by local rules,” not an abridgement of common law obligations. Further, the language of the Rule is expressly geared to the Court reining in requests for discovery in pending litigation where the interests of all parties can be presented and weighed. It’s a rule designed to protect one party from the oppressive conduct of another. Nothing in the Rule suggests that it was intended to limit pre-suit obligations or to insulate a party from the consequences of the party’s own failure to preserve relevant evidence.
Clearly, Rule 26(b)(2)(C) can’t be pressed into service as a safe harbor for botched preservation, but it can prove instructive to courts weighing sanctions for failure to preserve relevant evidence and considering motions for protective orders. Though the standards for imposition of spoliation sanctions vary across the circuits, all assess the reasonableness of the respondent’s conduct and the harm flowing from the alleged spoliation. In arguing the reasonableness of their conduct, a party may seek to demonstrate that the burden or expense of the preservation outweighed the likely benefit of the data not preserved considering the needs of the case, the amount in controversy, the parties’ resources, the importance of the issues at stake in the action and the importance of the discovery in resolving the issues. In seeking to mitigate the alleged harm, litigants can show that the evidence not preserved can be obtained from a source more convenient, less burdensome, or less expensive than the source not preserved. KPMG tried the latter approach. They just didn’t have the evidence to meet their burden.