Recent amendments to federal discovery rules have renewed attention on the concept of "proportionality" -- a cost-benefit analysis at the heart of many judicial determinations about fairness.
But how do you make those cost-benefit decisions if the cost piece is missing? Is it even possible?
In part two of our interview with US Magistrate Judge Laurel Beeler, whose bonafides as one of the most accomplished members of her court were covered in part one, we pose those very questions. We also ask her how often she sees advanced search methodologies like predictive coding used in her court.
Spoiler alert: never.
Logikcull: I wanted to ask you specifically about e-discovery costs. I don’t know what you’re perception is on this, but from our perception, e-discovery costs in general are hard to come by. It’s hard to know how much people are supposed to be paying and what the market rates are for these various products and services. The only way that we come across them are in cost recovery orders and stuff like that.
Judge Beeler: Exactly. That’s the only place where I’ve had any transparency into it.
Logikcull: Okay, so I want to ask you, given that, how strong would you say your grasp is on how much these various e-discovery activities that occur in your cases are supposed to cost?
Judge Beeler: Not very strong. I don’t have a lot of insight into it, and when you say what they "should" cost, that’s normative. It's a judgement call. And the only insight I get tends to be at the end of a case when you see the cost motions.
It’s extremely anecdotal, but I have this perception, and it may not be 100% accurate, that e-discovery costs, or certainly storage costs, are coming down. I’ve seen that in cost bills with different companies.
You could probably set up a fee to subscribe through ECF to litigated cost motions to really look at what people have spent on e-discovery and do a survey of vendor costs that way.
It varies, and I’ve seen some big differences in cost, but it does seem that costs have come down over the past couple years. [Another thing to consider is that] some law firms and companies are processing [discovery data] in-house....
This is just a complete aside, but I think that advising companies on the business angle of maintaining your systems is a pretty interesting area because there have to be some efficiencies that can attach to having a plan for company recordkeeping, which e-discovery stems from.
But a short answer to your question is that I don’t have a lot of insight into the costs except to think, “Wow, that’s expensive,” or “Huh, that seems pretty reasonable compared to what I’ve seen." And I have this overall perception that costs have declined somewhat in the last 6 and a half years that I’ve been doing this.
"I don't have a lot of insight into [e-discovery] costs except to think, 'Wow, that's expensive...'"
Logikcull: What I'm getting at is, in the context of proportionality, you have to make decisions occasionally as to whether these costs are proportional to their outcomes and the importance of the issues, and so on. If you don’t have a lot of visibility into that and there’s a ton of cost variability across the board, does that hamper you in your ability to make those determinations?
Judge Beeler: So that’s a really interesting question. I’ve thought about this and where I think you should start, considering that most cases resolve by way of settlement, is by taking an iterative approach to discovery.
In my old job, I did internal investigations with the aid of large law firms, so I know how an investigation works. And because I know how an investigation works, having done them, I know that you can do preliminary forensic analysis, for example, that can yield a lot of information that’s useful for settlement.
So it’s not saying “no, [this is too expensive],” it’s saying “let’s start here.” I think email, which can yield great evidence, really risks having a lot of irrelevant information in it that’s going to be very, very, very spendy. So I sort of know the areas that are very, very, very expensive, even if I don’t know the actual costs.
I think [there are some ways to contain costs] -- for example, by conducting a preliminary forensic analysis. In a trade secrets case, for instance, one bucket is a bad actor’s media. You can take a hard look at that, and you can take a limited scrub of the defendant’s media and do some kind of production through a third party that’s produced only to attorneys. And those things are relatively easy to do, if you stage them. It’s not particularly expensive. The no-stone-left-unturned forensic landscape or investigation can be quite expensive, but that gives you a lot of information. So that’s one idea I have about investigations about managing costs.
The other is the email one that I mentioned, that is very expensive, and I think is best at the end of discovery when you really know a lot about what happened, and then you can take targeted email searches between known people, between known time periods, and you can really keep the costs down. So, sure, it would be nice to know how much everything costs. I assume it’s super expensive. Even if costs come down, it’s still expensive, so that’s part of it. And so I think that discovery can be managed effectively in aid of cost reduction. And if I don’t know, then you ask the lawyers, and if they don’t know, they can ask the tech people....
So those are different strategies that I think can be brought to bear on managing costs, even if you don’t know the exact dollar figure that attaches to them, because I do know... the relative costs that are attached to it. Or maybe I’m just making myself feel more comfortable about not having perfect insight into the costs that are attached to it. But I have enough insight to know it’s expensive. And I have enough insight to know that any litigation can be expensive. Consider the costs to hire lawyers. It’s just expensive.
"There are different strategies that I think can be brought to bear on managing costs, even if you don't know the exact dollar figure that attaches to them... Or maybe I'm just making myself feel more comfortable about not having perfect insight into costs."
Logikcull: Do you think that more can be done to increase visibility into e-discovery costs? You mentioned earlier that something as simple as having making it easier for parties to have access to Pacer feeds that mention orders with cost recovery. Is there anything else like that you think would be helpful for you to make these decisions?
Judge Beeler: Sure, I would love to have more insight into it. I think that it’s good to ask lawyers along the way. I do think that it’s always important to manage discovery toward the dollar value that’s at stake in the litigation. What’s reasonable e-discovery in the context of patent litigation with core technology is different than what might be relevant for an employment case -- it’s still consequential to the parties, but with a relatively low dollar value for the ordinary spend of federal litigation. So, I think it’s good to have insight. I do ask the lawyers the questions, and it would be great to have more insight. We get a nice little list of all the BNA reporters that comes out. I read that to keep up with what’s being said, but beyond that, I don’t have any great ideas about how to increase insight.
Logikcull: Some people have talked about proportionality in terms of actual percentages that you can put on discovery costs or attorney fees or whatever, that are, for instance, X-amount of either the amount awarded or the total amount in dispute or something like that. Is that a realistic way to look at it? If you were to say, for example, it would be disproportional for e-discovery costs to make up more than 35% of the total value of the case? Is thinking about it in terms like that useful or realistic?
Judge Beeler: Well, I think you don’t want someone to spend more pursuing a case than they might recover from it. So it’s always a good idea to know what things cost and what the total value is of the case at the end. I don’t know that quantifying it by reference to fixed percentages is exact enough to really work in all cases. I think people do use rules of thumb, that are tethered to the ultimate dollar value of the litigation. It seems pretty mechanical to just have a formula that you would apply.
"You don't want someone to spend more pursuing a case than they might recover from it."
Logikcull: Before I let you go, I’m curious, a lot of people have been talking about predictive coding over the last four or five years, and other advanced search technologies that have emerged. Have you seen predictive coding or other newer methods being used more in your court?
Judge Beeler: I have not seen it. Have you seen it? I mean, I’m curious whether you’ve seen it. I’ve certainly read the articles about it.... There was an article in the New York Times about Target, about how Target knows you’re pregnant. There’s so much information that you’re able to glean from people’s behaviors that it seems to me that this idea of predictive coding, this sort of iterative approach to identify what’s more likely to yield more useful information, seems great.
I remember reading about it in the Chinese Drywall litigation, or at least there was a lot of talk about a potential utility of doing large scale mining of information to figure out where you are likely to get relevant information. I think the interplay of that and investigations seems pretty interesting. But I haven’t seen it at all. And I haven’t heard anybody talk about it in the lunchroom in this district, and maybe it’s not something that grips people’s attention. But I think it’s very interesting and I haven’t seen it. Maybe I just don’t have that insight into it because it would be super big litigation that everyone has spent the money on it anyway. I don’t know, what’s your thought about it?
"This sort of iterative approach to identify what's more likely to yield more useful information seems great. But I haven't seen it at all. And I haven't heard anybody talk about it in the lunchroom in this district."
Logikcull: I mean, I have not seen it as much either, but I don't have the visibility into it that you might, so it’s hard for me to say one way or the other. But I do get the sense that it is being used in very complex cases with a lot of documents, as you alluded to, and I also get the feeling that it’s being used in at least some fraction of cases where the other side doesn’t know that it’s being used and the court doesn’t know that it’s being used either. But that would just be a guess.
Okay, well we are going to end it there I think, but this was a super interesting conversation. I really appreciate you joining me.
Judge Beeler: That’s nice of you to say, and it was nice to chat with you too.
Part one of this interview is available here. To learn more about the evolution of judicial thinking on predictive coding, download our Guide to the TAR Solar System below.