This week’s episode looks at the latest decision in the In re Diisocyanates Antitrust Litigation. At issue this time is whether the production of search term hits for text messages and calendar entries was adequate to fulfill the defendant’s obligations under the Federal Rules of Civil Procedure (FRCP). Join us as we explore the Court’s ruling and the implications it may have for future cases involving ediscovery.
Welcome to Episode 97 of our Case of the Week series publishing partnership with ACEDS. My name is Kelly Twigger. I am the CEO and founder of eDiscovery Assistant and the principal at ESI Attorneys, and I’m so happy to be here with you today. We are quickly approaching our 100th episode. Thanks so much for sticking with me on this journey of learning ediscovery through case law.
Welcome to those of you and a huge thank you who participated in the University of Florida Levin College of Law eDiscovery Conference last week, especially to the planning committee, of which I was very privileged to be a part, and to Bill Hamilton and Maribel Rivera, who never ceased to amaze me with their talent, leadership and generosity of spirit.
If you registered but were not able to view all of the sessions, they will remain available for viewing for three months in the Whova application, so you can log in there and be able to view those sessions.
I also want to specifically thank Judge McCoy of the Middle District of Florida and Ryan Tilot of Gunster for joining me on the case law panel. We had fun discussing some of the most important takeaways from Case law in 2022. If you missed the conference this year, be sure to bookmark it on your calendar for next year (dates to be announced).
We had about 3000 attendees online and more than 200 in person this year, and we’d love for you to be a part of it with us next year. I’ll keep you posted on when we have an official date for that next conference.
All right, this week’s decision comes to us from a case that we’ve covered a couple of other decisions in the In re Diisocyanates Antitrust Litig. This is the decision from January 26th of 2023 from United States District Judge Scott Hardy. As always, we tag the issues for each decision in our eDiscovery Assistant database, and this decision’s issues include search terms, text messages, ESI protocol, and failure to produce.
As we’ve discussed in our previous Case of the Week sessions on this matter, this is a multi-district litigation regarding an alleged conspiracy to reduce the supply and increase prices for ingredients in the manufacture of polyurethane foam and thermoplastic polyurethanes. There has been extensive discovery in this case, as evidenced by all of the decisions in our eDiscovery Assistant database in this matter. And we are currently before the Court on a motion to compel production of custodial calendars and text messages from three of the defendants that are part of the alleged or alleged to be part of the conspiracy in this case. Those defendants are Huntsman, WCA, and BASF.
Now, prior to this motion, the defendants had already produced many, many documents responsive to requests for calendar entries and text messages. And those productions were made following the application of search terms in accordance with the ESI Protocol that the parties agreed to in the case:
- Huntsman produced approximately 27,000 calendar entries and 1500 text message transcripts,
- WCA produced approximately 5500 calendar entries and 3400 text message transcripts, and
- BASF produced almost 20,000 calendar entries and 473 text message documents.
Now, we’re before the Court because the plaintiffs want not just what was produced, but they want everything from those text messages and calendar entries that were preserved that are relevant. So, essentially, they want to say, we want you to ignore that we agreed to search terms here. And so the Court is looking at whether there is a basis to do so.
Now, the documents that those three defendants preserved but were not produced encompass a much larger swath of text messages and calendar entries. For example, Huntsman has approximately 62,000 unproduced calendar entries and approximately 86,000 text messages. WCA has 39,000 unproduced calendar entries and 97,000 unproduced mobile transcripts. BASF has approximately 269,000 unproduced calendar entries and more than 6300 unproduced text messages.
So those are all documents that have been preserved but were not produced. Following the application of search terms and review of the search term hits for responsive documents, the plaintiffs, as I mentioned, and these three defendants, agreed in writing to an ESI Protocol to apply negotiated search terms to these categories of data.
Following the production, then the plaintiff said, we want more information than those search term hits provided. Now we’re asking for everything that’s responsive out of the total. So, as I mentioned, those numbers of the three defendants, we’re talking about a substantial amount of data that needs to be reviewed. And review of text messages is difficult. And review of calendar entries is difficult, frankly, because you need to understand the context of the people involved, lots of different parameters. It’s not something that’s very easily assigned as a standard contract linear review.
The Court begins with a discussion of the proportionality considerations of Rule 26(b)(1), and we’ve covered those factors ad nauseam on the #caseoftheweek. But we’re going to find that the Court does a similar analysis here. And it’s very important. This dispute is not over the relevance of the calendar entries or the text messages, but whether “wholesale production” of those two sources of ESI without search terms is required here.
The plaintiffs argue that “courts routinely order the full production of calendars without first applying search terms, and because the search term review improperly risks excluding relevant evidence.” The Court completely disagrees, looks at the cases that the plaintiff cited, and distinguishes them as not being factually similar — and frankly, they really weren’t at all supportive of the plaintiff’s argument for blanket production of such a large number of entries for such a large number of custodians.
The Court also notes that
[W]hile it is true that the use of search terms is imperfect and may exclude some relevant documents, not using search terms when reviewing a voluminous universe of documents also risks expending time and resources to cull through and produce large amounts of irrelevant documents. Consequently, search term methodologies are routinely used, particularly in large and complex commercial and antitrust cases, as an acceptable, if not preferred, approach. Only a reasonable search for responsive information pursuant to a reasonably comprehensive search strategy is necessary. And in an era where vast amounts of electronic information is available for review, courts cannot and do not expect that any party can meet a standard of perfection.
That’s very important language because, essentially, the Court is validating the use of search terms as a methodology for limiting a preserved set of documents for what should be reviewed and produced under the Federal Rules of Civil Procedure. What’s key next to the Court is the ESI Protocol that the parties entered into that requires the parties to meet and confer to reach agreement on search methodologies to limit the scope of document collection and review for production.
The Protocol also provides for a process by which the parties are to develop and apply search terms. In the Protocol, the parties agreed to use search terms for review and production of text messages and calendar entries, and the plaintiffs here really provided no basis to change those agreements that are set forth in the ESI Protocol. However, the Court did note that after a review of correspondence between the parties that followed after the protocol, the parties were not wholly in agreement that the use of search terms would end the inquiry and that plaintiffs would have follow ups after production of messages based on the results. That makes sense, right? Because when we choose search terms, we never know about the efficacy of those search terms unless and until we see documents.
The Court then says that while, “plaintiffs have not demonstrated on the current record that certain defendant’s search terms or search methodologies pertaining to calendar entries or text messages are unreasonable or inadequate,” the parties did agree that there would be follow up requests. Now, rather than request a full-scale production of all preserved and responsive materials, the court encouraged the plaintiffs to issue targeted follow up requests — an approach would be more proportional to the needs of the case than a full production or a full linear review and following production of the remaining preserved materials. In making that decision, the Court looked at the proportionality factors of Rule 26(b)(1) and found that the most important factor was the burden and expense of re reviewing the entire universe of calendar entries and text messages saying search terms when balanced against the likely benefit. That’s the Court’s language here.
I kind of took issue with that a little bit because they wouldn’t have to review the entirety of the universe of calendar entries and text messages, just those that weren’t responsive to the search terms. So likely the Court understands the difference here, but the language of the decision doesn’t reflect that. Based on that analysis, the Court denied the motion to compel, but did advise the parties that the plaintiffs could issue follow up requests to seek relevant evidence proportional to the needs of the case, and that any such follow ups should be non-cumulative, specifically targeted, and informed by the documents produced to date. All of which are themes that we talk about regularly on the case of the Week.
I’m going to start with one takeaway that we came up with last week from the University of Florida Levin College of Law E-Discovery Conference, and that is there are no takebacks. As we noted during our session last week, Judge McCoy stated that where the parties have agreed and that agreement is entered by the Court, whether it’s an ESI protocol, an ESI order, some other order that the parties have stipulated to, and then the court is asked to sign off on — unless there is a good rationale supported by documented facts, the parties will be held to that agreement. There are no takebacks.
Now, here, the parties had exchanged written correspondence that follow ups would be necessary. But at least as the facts are laid out here, that’s not what the plaintiffs did. According to the decision, plaintiffs moved for a wholesale production contrary to the language of the protocol. If you’re going to negotiate an ESI protocol, you need to think it all the way through. And by thinking it all the way through, I mean you need to think about how the language you are agreeing to will apply to each different source of ESI, and you need to be confident that the search terms that you are negotiating will apply across the board to the various sources of ESI. We’ve discussed multiple times on #caseoftheweek that the language that custodians use varies depending on the platform they are creating and sending ESI on.
For instance, what I say in an email may be very different than how I code a calendar entry, or what I say in a text message, or what I say in a slack message. We all communicate differently depending on the type of platform on which we are communicating. That needs to be factored into your analysis on an ESI protocol now.
Next takeaway. Calendar entries are really key when you’re setting up a timeline and showing interactions between people. And they are particularly relevant on antitrust claims where you’re trying to show that individuals from various parties met and conferred to engage in the alleged conspiratorial behavior.
But I’m telling you that there are technologies available, and we covered some of them at the UF Law eDiscovery Conference — that can help you take those calendar invites as well as GPS location data and emails and create a timeline for you that allows you to pinpoint locations of individuals and meeting dates. And those are relevant not just in the antitrust context, but in many, many cases that can show all kinds of different issues on both civil and criminal litigation. So, keep in mind that calendar entries are discoverable. They’re going to be collected large scale in email. But having to produce calendar entries is something that you absolutely need to think about when that is a source of ESI, that’s going to be important for purposes of discovery.
Next takeaway. We’ve covered this one a number of times on the #caseoftheweek. Search terms are fraught. Unless the party with the data is identifying the correct search terms to be used to identify responsive data, you will not get all of the responsive data. It is impossible to guess what language and terms individual custodians use to create, store and send data without looking at the data. This case is just one of hundreds of examples of this very problem.
I encourage you, if you are plaintiffs, even in asymmetrical litigation, that you go to the Court and require defendants to be able to produce reasonable search terms based on sources of ESI. That means that in this particular instance, as an example, there would have been specific search terms that would garner calendar entries. Not just search terms across the board for all sources of ESI, but search terms that are specific to sources of ESI.
Next takeaway. This one is about form of production. As I looked at the information provided by the defendants on the number of text messages and calendar entries that were provided, counsel really needed to be able to determine which documents in their platform fit that criteria — meaning which documents were text messages and which were calendar entries. And to do that you have to look at metadata. I want you to think, as you’re loading documents, think about how you want to code those documents going in or whether you can create a metadata field for document type that allows you to sort very effectively. So essentially, if you’re in a review platform and you’re ordered by the Court to provide an affidavit on a motion to compel like the one before us, where you have to provide the numbers that I cited to you earlier (about volume of text messages and calendar entries), that you can easily sort on document type and then be able to have those numbers at your fingertips. Otherwise, you’ve got a lot of manual review and entry to do on document numbers. And you want to make sure that the document numbers you submit to the Court are accurate.
Using that metadata field and loading that information when you load the documents is really crucial, and I encourage you to do it both for when you’re handling your own client’s data, for purposes of producing information, but also for purposes of dealing with productions once you get them in the door.
There was a specific session at the University of Florida Levin College of Law E-Discovery Conference last week with Scott Milner (Morgan Lewis), Stephanie Clerkin (Korein Tillery), and Rebekah Bailey (Nicholas Kaster) that talk specifically about tips for dealing with incoming productions. I recommend you take a look at that session if you are registered for the conference. Those are our takeaways.
That’s our Case of the Week for this week. Thanks so much for joining me.
We’ll be back again next week with another decision from our eDiscovery Assistant database. If you’re interested in doing a free trial of our Case Law and Resource database, you can sign up for a free trial.
We will be at Legal Week in New York City in March, and if you’re interested in sitting down with us and seeing what is new in eDiscovery Assistant or how we can help with training and education in your organization, drop us an email at firstname.lastname@example.org and we’ll be in touch to set something up.
Thanks so much and have a great week.