Skip to main content

Kicking the Hornet’s Nest: Judge Francis on eDiscovery, Inherent Authority, and the Supreme Court

May 25, 2017  |  19 min read

Judge Francis

Over more than thirty years on the bench, Magistrate Judge James C. Francis IV has seen discovery evolve from associates sifting through stacks of paper to technologically sophisticated systems culling through vast amounts of electronic data. He’s helped shape that evolution along the way as well, writing influential cases on issues such as discovery cost shifting, ESI preservation orders, and more.

judge-Francis small.jpg

A widely recognized expert on eDiscovery, Judge Francis recently made waves with his CAT3 v. Black Lineage decision, where he asserted that courts’ inherent authority to bring spoliation sanctions survived the 2015 amendments to Rule 37(e), contrary to the Advisory Committee's indications. We recently sat down with Judge Francis to discuss discovery, technology, CAT3, and the implications of the Supreme Court’s recent Goodyear opinion. A transcript of that conversation follows, gently edited for concision and clarity.

Logikcull: I'm sure you're aware there's a very small and declining percentage of civil cases in the federal system that ever go to trial. Meaning, cases that aren't decided on motion are often resolved through the discovery process.  Do you think that the discovery process has become a sort of the de facto dispute resolution mechanism in our system? Is that a fair assessment? Following that up, to what extent is the discovery system as it exists today capable of ensuring just results?

Hon. James C. Francis: Well, I think in part that's correct. As you’ve indicated, a fair number of cases are resolved by motion -- whether it’s motion to dismiss or motion for summary judgment. Those obviously reach the merits; so, they're not per se resolved by the discovery process. Those that go to trial, which is a very small number, or those that are settled, which is a much greater number, are impacted heavily by the discovery process.

Certainly, discovery is critical in allowing the parties to evaluate the strengths and weaknesses of their case for purposes of settlement. So, in that respect, I think discovery can be a means of reaching a just outcome depending on three things: the discovery has to be balanced, sufficient, and the cost effective.

Let me say a little bit about what I mean on each of those. By balanced I mean each side has to be equally able to obtain information to support its claims or defenses -- that is, there can't be a tilt in favor of plaintiffs or defendants.

Secondly, there has to be sufficient discovery. Unless a party gets what it needs in order to evaluate its case and the other party's case, then the discovery doesn't move us toward a just resolution.

"Unless a party gets what it needs in order to evaluate its case and the other party's case, then the discovery doesn't move us toward a just resolution."

Finally, it has to be cost effective in the sense that you want the outcome to be driven by the content of the discovery rather than the costs of obtaining that discovery. Obviously, no discovery is costless. There's always going to be some cost, but it can't be disproportionate to the value of the content that is going to be obtained.

Logikcull: In terms of your first point, balance: Under two years ago Rule 26 was updated, emphasizing the importance of proportionality in the discovery process simply by moving the word a little bit higher up in the phrasing. Do you think that tweak has had an impact on the way judges deal with discovery issues and proportionality post-amendment?

Francis: I do. First of all, parties are more likely to raise proportionality issues. Second, judges are more likely to raise them sua sponte and to deal with them forcefully. I would point out that one of the problems that has arisen (although I think it's a problem that will dissipate as time goes on) is that by moving it up in the rules and making it more visible, it has also become yet another boilerplate objection. So, I think both the parties and the courts have to get beyond that and be able to treat proportionality more intelligently.

Logikcull: In terms of eDiscovery and proportionality, one of the situations that we think about here a lot, is courts grappling with the actual costs of discovery and the various technologies that are available. So, you have a judge facing a situation where a party is making claims about how much it can review and how much it can produce using a specific technology by a specific deadline, but a lot of how that's done is through “black box” proprietary algorithms. Most of the parties don't know what's going on inside the particular review technology. What, in those instances, can judges and courts do to ensure that the information they’re relying on is accurate, that everyone is informed and can make an informed decision in terms of proportionality?

Francis: Well, this is really like any other evidentiary issue. I have to evaluate the reliability of the evidence that is presented to me. For example, I'm going to give more weight to the affidavit of the expert, whether it's a vendor or someone in-house, than I am to the bald representations of the attorney.

"For example, I'm going to give more weight to the affidavit of the expert, whether it's a vendor or someone in-house, than I am to the bald representations of the attorney."

Similarly, I'm going to credit specific representations as to the cost and burden of using a particular technology than I am to conclusory assertions about disproportionality. I'm no more likely to know anything about what's inside the “black box” than the counsel are, but I can evaluate what the parties or their representatives or experts tell me about what the cost and burden of using that particular technology is going to involve.

Logikcull: What tools do courts have available to reduce discovery burdens and do you think that they are using them sufficiently?

Francis: Well the two that leap to mind are phasing and sampling. By phasing, I mean choosing critical issues to take discovery on and setting aside the less critical issues, waiting to see whether the critical issue resolves the case or at least drives future discovery in a different direction. That way we can focus our resources on what makes a difference.

In sampling, I'm talking about taking whatever data is relevant to a particular issue and deciding how we can take a sample of that data in order to make inferences about the whole, so that we don't have to examine each and every data point. I don't necessarily mean a random sample in terms of statistics. It may simply be a sample sufficient to give the parties and the court comfort that this is a snapshot of the larger picture. I don't think that the parties are using these kinds of tools sufficiently. I think sometimes we spend a lot of time fighting over what the appropriate sample is and never get to the ultimate use of that kind of a tool.

"[I]f you're not going to take opportunity to make discovery more speedy and cost efficient by availing yourselves with these opportunities, I'm not going to cut you slack in terms of the schedule. You're going to live with what I give you."

Logikcull: If you're faced with a situation like that, how do you as a judge try to move things forward while also allowing the parties to bring their objections or concerns?

Francis: Well, to the extent that I can jawbone with them and try to get them to agree on a particular protocol, I'll do that. I think I am a little hamstrung because I don't think that in most instances I can impose upon them some of these tools that will move discovery along and make it more cost-efficient. I think, ultimately, the choice of those tools is party-driven. I can say that if you're not going to take opportunity to make discovery more speedy and cost efficient by availing yourselves with these opportunities, I'm not going to cut you slack in terms of the schedule. You're going to live with what I give you. So, there are some levers that I can use, but ultimately, I don't think I can impose these kinds of things on them.

Logikcull: You've been on the bench now thirty-two years, giving you a long view of how things have evolved. There was no eDiscovery in 1985, for example. So, how would you evaluate the level of technical competence in the lawyers before you over that long period? Do you think lawyers today are sufficiently equipped to handle the complex and often technological issues around evidence and discovery that are now more and more a staple of litigation?

Francis: I think, generally, the level of competence is getting better. I think the problem is that there's increasing disparity between the haves and the have-nots. There are some highly sophisticated, technically competent attorneys, but there are still folks who are really in the Stone Age. That creates problems because the asymmetry in competence engenders distrust and confusion and creates problems for the parties and for the court.

"[T]he asymmetry in competence engenders distrust and confusion and creates problems for the parties and for the court."

To put that in perspective, that disparity is really a function of the increasing sophistication of the technology, generally. Thirty-two years ago, everybody knew how to segregate the documents: the paper files that had to be preserved, how to preserve them by putting them in a separate room, how to go through them and cull them by hand, and how to review them for privilege by looking at each and every document. So, the competence necessary for doing that process was fairly low and fairly well understood.

But now you've got so many different kinds of data repositories and means of search that folks necessarily have a wide range of skills. Now, are they equipped to handle it? I think the answer is: some, yes and some, no. There are probably more folks who don't know how to handle it [than folks that do].

Logikcull: You are an expert on eDiscovery, along with a small handful of your fellow judges and also lawyers and technical consultants who know this stuff in and out. But as you indicate, there's a huge disparity -- not just in federal courts, but I'm thinking also of state litigation where you might have a practitioner or state judge who doesn't deal with these issues all that often. So, what do you think can be better done in terms of distributing knowledge and resources so that the profession overall is better educated about complex evidentiary issues?

Francis: I think to a large extent the training is available both to judges and to practitioners. What really needs to be done is to convince the attorneys and judges that they need to take the time to educate themselves in these areas. To some extent, folks view electronic discovery as somebody else's problem, but I don't think they would view civil procedure as somebody else's problem. Yet, they are very much parallel.

"To some extent, folks view electronic discovery as somebody else's problem, but I don't think they would view civil procedure as somebody else's problem. Yet, they are very much parallel."

You need eDiscovery knowledge almost to the same extent now that you need civil procedure knowledge. Convincing them that it is on the same plane is one very important thing.

I think, frankly, that the problem of disparity is going to become greater and more evident in the near future. The reason for that is there are many more data sources now coming online than there used to be, when you think of things like the “Internet of Things” and all of the different handheld devices that exist.

This means that evidence will become relevant in cases where electronic evidence has never been relevant before, i.e. matrimonial cases, personal injury cases, where counsel aren't as sophisticated about the technology. So, as more practitioners become involved in these cases, more of them will have to come up to speed.

Logikcull: It seems like the “Internet of Things” and the growth of data collection is becoming more of a concern among practitioners and certainly among some judges. We spoke to Judge Scheindlin recently and she brought up the same concern: You have more data being generated, more data being collected. [That raises issues] not only just in terms of competency in using that data, but dealing with these massive amounts of data that are going to be collected by everything from medical devices to cars to fitness trackers, electric sensors, whatever it might be. That seems like it's going to pose a pretty big challenge in the future.

Francis: It undoubtedly is going to be a challenge. I think that technology will help, new search tools and that kind of thing will be helpful. I think that at least in the short run, the increasing volume and the increasing diffusion of that kind of information is going to outstrip the technology. It's going to mean higher eDiscovery costs.


Logikcull: I want to switch now to inherent authority. Your CAT3 decision definitely generated a lot of buzz -- at least among the people who care about this sort of stuff. Were you expecting that response? Were you surprised by the response?

Francis: I knew I was kicking the hornet's nest; so, I was not surprised. I was fully aware of the Committee note and of the controversies that took place on the Advisory Committee that led to that note. They were focused to a large extent on what they viewed as the Second Circuit's going off on a frolic and detour in Residential Funding and they were concerned that a court might resurrect Residential Funding via inherent power.

Logikcull: In fact, some courts are still using Residential Funding.  Your colleague in the Eastern District of New York, Judge Tomlinson, in an opinion just a week or two ago, citing CAT3, talked about how impractical it would be to apply the new rules under a case that was briefed under the previous ones. There seems to be a handful of cases that are still going to be dealing with the pre-amendment rule; so in that way, I guess, they haven't been totally excised. Do you have any thoughts about that?

Francis: That's really just the tail I think. At some point, there's no longer an argument that the pre-2015 rules could apply because the case will have been filed after December of 2015. So, I don't think that's going to be a major problem going forward. I think that the Committee's greater concern was in the long run, would a court just try to gut the rules by exercising its inherent authority.

Logikcull: That brings us to the Goodyear decision, the Supreme Court’s first decision on both discovery and inherent authority since the last decision on the topic 26 years ago in Chambers. It hasn't gotten a lot of attention, from what I’ve seen. Do you think that's because the situation isn't going to be frequently reoccurring? There's the particular facts of Goodyear that were sort of unique in that the case had already settled and the court was somewhat at a loss with how to deal with this bad-faith violation. Or, do you think it is going to have an impact? Will it be a meaningful case in terms of discovery and sanctions?

Francis: I think its impact will be extremely limited. I think you could say that the rumors that Goodyear would be the demise of inherent authority were premature. It really is limited in in two respects. One is, as you’ve suggested, the facts were so extreme that the district judge really felt compelled to exercise this pretty draconian attorney’s fees sanction.

"I think you could say that the rumors that Goodyear would be the demise of inherent authority were premature."

More importantly is how narrow the decision itself is. It did not say that you don't have inherent authority to impose punitive attorney’s fees. All it said was if you're going to do that, you have to provide the party with due process as you would in any contempt proceeding.

So, it doesn't even limit the circumstances under which in inherent authority can be exercised. All it does is impose due process requirements -- quite properly, I'd say.

Logikcull: And it discussed, in tandem, both sanctions under the explicit rules and under inherent authority. Do you see, then, Goodyear as supporting your dual authority theory?

Francis: I guess it would go too far to say I see it supporting it. I think the way I would characterize it is I don't see Goodyear undermining that view at all.

Logikcull: I want to talk a little bit about your Sedona Conference Journal article with Eric Mandel. You end it saying: “Going forward, the issue will be not whether the federal courts retain inherent authority to issue spoliation sanctions, but under what circumstances and to what extent they may exercise that authority.” In what circumstances and situations do you anticipate courts turning to inherit authority?

Francis: Well, let me identify a couple. One is the circumstance where a party intentionally destroys information, but it destroys it in the context of an imminent government investigation. In other words, the FTC or the SEC or somebody knocks on the door and issues a subpoena and in response, the party destroys information. Now, by its terms there's not yet any litigation and perhaps we can say the spoliator hasn't yet anticipated litigation; they've anticipated this investigation. But then in subsequent litigation, would that spoliator be subject to sanctions under 37(e) or would it have to be under inherent authority because the spoliation did not take place in connection with the litigation? So that’s one example.

Another example is the incompetent spoliator. Under 37(e), there's a remedy only if there was information lost. If someone attempts to destroy information but is ultimately unsuccessful, then probably 37(e) doesn't apply but they may have created problems for their adversary that deserve some kind of remedy. That might be a remedy under inherent authority.

Logikcull: You also mentioned metadata is one area that inherent authority might give you grounds for sanctions in terms of spoliation. Why treat metadata different than normal ESI?

Francis: Well, I think the answer is you treat it differently sometimes. If the metadata that you're talking about is substantively relevant, for example, track changes in a draft of a contract that's at issue in the case, then I don't think you treat it differently. That is relevant discoverable evidence and would be subject to 37(e).

There's other metadata that may be useful in the case but not relevant in the evidentiary sense. For example, it might facilitate automated searching of data. In those circumstances, I think it's more doubtful that its loss would be covered by 37(e); so, you may have to resort to inherent authority for any remedy.

Logikcull: The Sedona Conference Journal article could be given a narrow reading, that you're arguing that where Rule 37(e) doesn't apply, courts can turn to their inherent authority to address these issues. But in the CAT3 case you seem to make a broader claim stating that, Where exercise of inherent power is necessary to remedy abuse of the judicial process, it matters not whether there might be another source of authority that could address the same issue.” Are these two things at odds or am I reading in a conflict in that doesn't exist there?

Francis: They’re just two different circumstances. The point is, if there is a gap in the rule, then the exercise of inherent power is appropriate, even if it's not required to preserve judicial power.

But if some kind of exercise of authority is required to preserve judicial power, then it's appropriate even if it's at odds with a rule or a statute. This is really a separation of powers argument -- that is, a statute or a rule can't withdraw judicial power to the point where it undermines the ultimate authority of the courts.

"[I]f some kind of exercise of authority is required to preserve judicial power, then it's appropriate even if it's at odds with a rule or a statute."

Logikcull: You survey both theories, the narrow and broader views of inherent authority, within the Sedona Conference Journal article. I'm assuming you're falling on the broad side personally?

Francis: I guess I'm withholding opinion to some extent on the larger issue. On the narrower issue, I think there really is no question that after Chambers and that line of authority, inherent authority is available as a gap-filler.

I think there is a lot of academic debate about the broader question of when and under what circumstances inherent judicial authority can fly in the face of a statute and I think without seeing a particular case or set of facts, I'd be reluctant to offer a more general opinion on that.

Logikcull: Can you tell us where you see this issue going in the future? Is this something that courts are going to explore, or will it remain largely an academic, theoretical issue?

Francis: I think the broader issue will be largely an academic one. The question of where are the gaps in Rule 37(e) and can they be filled by an inherent authority is something that will come up. It may come up in circumstances where a court will say, “well it's probably within the scope of 37(e), but even if it's not, I’ll assert my inherent authority” so that they'll use a belt and suspenders. That maybe the way the courts address that most of the time.

This post was authored by Casey C. Sullivan, Esq., who leads education and awareness efforts at Logikcull. You can reach him at or on Twitter at @caseycsull.