In 2017, the Texas Supreme Court went out of its way to write about an unlikely topic—the form in which computer records must be produced in civil discovery. In re State Farm Lloyds, 520 S.W.3d 595 (Tex. 2017). But “form fights” are not mere technical matters of interest only to the pocket-protector crowd. Beneath the surface, battles over form of production are usually battles over content and proportionality. Sure enough, the Texas Supreme Court’s opinion wasn’t just—or even mostly—about form of electronically stored information (ESI) productions. It was about how relevance and proportionality converge to define the scope of discovery, and the trial court’s duty to keep discovery costs reasonable.
The plaintiffs in the case alleged that their insurer, State Farm Lloyds (SFL), underpaid hail damage claims, and they sought claims-processing documents in discovery. Ordinarily, one might expect an insurer to pull claims files from scattered offices to meet that request. But years before, State Farm had centralized all claims processing into its Enterprise Claims System (ECS). SFL agreed to produce the requested documents and proposed to produce them from the ECS repository.
Not so fast, said the plaintiffs. They had requested that SFL produce the ESI in native form—i.e., the form in which the ESI had been created. But the ECS respository stored everything as searchable read-only image files (e.g., PDF, TIFF, or JPEG). Many of the documents in the ECS had been created in a nonimage application like Microsoft Word and converted to an image file when entered into the ECS. Native versions of those documents might still exist, but SFL would have to look outside the ECS repository to find them.
For some of the data that had been converted to static image form when entered into the ECS repository, the plaintiffs stood on their request that SFL find and produce their native precursors. SFL refused, saying that the image files in the ECS repository were sufficient. The parties were at a stalemate.
What’s Really at Stake
At this point, you might be shaking your head. A party refused to accept documents because they were in the “wrong” form? Isn’t that formality taken to madness? Not necessarily. Image forms may or may not be adequate substitutes for their native precursors. Understanding why is the key to understanding what’s really at stake in this dispute.
The biggest issue lurking behind form fights is loss of content. Image files don’t contain the same content as their native precursors. Chiefly, what gets left out is metadata—information that the computer application automatically records about the creation or modifications of the document. For example, Microsoft Word tracks who worked on a document and when. That type of information is contained in a native Word file but usually not a PDF because the PDF captures only what appears on the screen. Similarly, tracked changes are often available in a native Word file but missing from a PDF version.
So, if content is missing from image files, does that mean image files are insufficient? Not necessarily. The critical question is whether the missing information is relevant to any of the disputed issues. For example, information about who worked on a document and when might be relevant in some cases but wholly irrelevant in others. If it is relevant, howimportant is it? And how costly would it be to get it? Those questions, of course, go to proportionality.
Sure enough, much of the form fight in In re State Farm Lloyds was really about content. For example, the plaintiffs argued that comments were sometimes written on claims-scene photos, and these comments were missing from the versions of the photos stored in the ECS. That’s clearly relevant content (and not even metadata). But SFL countered that it had made that information available from other sources and provided a log listing the omitted comments. There were also concerns about missing metadata. The plaintiffs complained, for example, that they wouldn’t be able to see the formulas behind Excel spreadsheets, that they wouldn’t get threading information in emails, and that some Word documents might be missing tracked changes.
The other issue lurking behind form fights is usability. In the past, the biggest problem was that early-generation image files could not be term searched. Today, that issue has receded as image-file technology has evolved to support term searching. But other usability issues can arise. For example, the plaintiffs argued they could sort data, analyze the relationship of data, and develop email threads and timelines more readily with native data than with the ECS documents. They also said that the conversion to PDF caused higher vendor and storage fees because it resulted in larger file sizes. SFL’s expert countered that static image files are easier to Bates stamp, redact, and use as exhibits, and that using static image files avoids alteration or integrity disputes.
The Ruling(s) in In re State Farm Lloyds
Now that we know what the parties were reallyfighting about, we can return to the dispute and how it was resolved. The trial court sided with the plaintiffs. SFL then sought mandamus relief, and the Texas Supreme Court (while technically denying the petition) used the case as a platform to weigh in and provide guidance for the future.
One question that the court quickly disposed of was which side gets to choose the form of production. Neither, the court said. Either side can propose a form; and if they agree, that will control. But neither side gets to dictate the form of production; and if they can’t agree, then the court must choose.
That left the bigger question: When the parties can’t work it out and turn to the judge to select the form of an ESI production, how is the judge supposed to decide? In this context, the court said, judges should do what they usually do when resolving discovery disputes: apply the doctrine of proportionality. Proportionality, the court said, is the “polestar” guiding principle for all discovery. In re State Farm Lloyds, 520 S.W.3d at 615.
The court dismissed the mandamus petition without prejudice, effectively remanding for the trial court to apply its new guidance on proportionality. The trial court would need to make more specific findings about the potential benefits of native documents. What specific content—metadata or otherwise—would be missing from the ECS image files? Is it relevant? How? Is native production the only way of getting it? How important is it? The trial court would also need to balance any costs against the benefits. SFL argued that the costs of collecting native precursors would be “extraordinary and burdensome,” but it made no effort to quantify those costs. This time, SFL would need to be more specific. And what about the parties’ competing views on the advantages and disadvantages of image files versus native files? The trial court would need to weigh all of these competing benefits and costs and exercise its judgment about what would be reasonable under the circumstances.
What happened in In re State Farm Lloyds holds important lessons for lawyers practicing in any jurisdiction that views discovery disputes through the lens of proportionality—and that clearly includes federal practice after the 2015 amendments to the Federal Rules of Civil Procedure. Here are our top seven takeaways:
#7: Resolve “form fights” up front. Federal Rule of Civil Procedure 34(b)(1)(C) invites requesters to specify the form they’d prefer, and Rule 34(b)(2)(D) requires producers to specify the form they plan to use, to smoke out potential form fights early in the discovery process. Fed. R. Civ. P. 34 advisory committee’s note (2006). If you produce in the form requested, or as the court orders, you’re probably done. Fed. R. Civ. P. 34(b)(2)(E)(iii) (“A party need not produce the same electronically stored information in more than one form.”). But if you don’t lodge a timely objection and instead just produce in your preferred form, you run the risk of the court ordering a costly do-over or worse. See Morgan Hill Concerned Parents Ass’n v. Cal. Dep’t of Educ., 2017 WL 445722, at *6 (E.D. Cal. Feb. 2, 2017) (untimely objection to the form requested not valid under Rule 343(b)(2)(D)); Fed. R. Civ. P. 34 advisory committee’s note (2006).
#6: There are no entitlements and no presumptions, just reasonableness. Neither party gets to dictate the form in which the ESI will be produced. When they cannot agree, the judge must decide. The rules don’t presume that the ESI is to be produced in any particular form. Questions about form of production turn on the same relevance and proportionality considerations that govern discovery generally. Each situation must be decided on its own facts.
#5: Cost matters, but content is king. Counting the times that the Texas Supreme Court opinion referred to cost or expense (by our count, 55), one might conclude that cost is the major factor in resolving discovery disputes. While cost is important, it’s just one factor. Discovery is not unreasonable just because it is expensive. The real question is not what it costs but whether it’s worth it. That analysis starts by figuring out what the value of the content is and then deciding what burdens are fairly borne to get it.
#4: Be specific and offer support. The first and most important question in form fights is whether image-file production will omit relevant information. Arguments that it will must be informed and specific. “Hypothetical needs, surmise, and suspicion should be afforded no weight.” In re State Farm Lloyds, 520 S.W.3d at 609. If relevant information is shown to be at issue, then the analysis turns to the importance of the information and the cost of getting it. Be specific. Boilerplate proportionality objections, without providing any information concerning burden or expense, are insufficient. See Liguria Foods, Inc. v. Griffith Labs., Inc., 320 F.R.D. 168, 171 (N.D. Iowa 2017); Carr v. State Farm Mut. Auto. Ins. Co., 312 F.R.D. 459, 468 (N.D. Tex. 2015).
#3: It’s not all or nothing. Neither the process nor the outcome needs to be all or nothing. Some probing into the native files may be needed to make an informed decision about what information is at stake and its value. Courts increasingly are taking this type of exploratory approach to achieve proportionality when the costs and benefits are initially unclear. And the ultimate answer may be that native production is justified for some things but not others.
#2: There is no free pass for deliberate degradation. One message that we hope readers don’t take from the SFL opinion is that they should immediately scan their records into image files as a means of short-circuiting native production. State Farm’s ECS was a preexisting businessdatabase, not a litigation repository. Parties cannot deliberately degrade their ESI for litigation purposes and expect to be insulated from having to produce the more usable native files. Fed. R. Civ. P. 34 advisory committee’s note (2006); see White v. Graceland Coll. Ctr. for Prof’l Dev. & Lifelong Learning, Inc., 586 F. Supp. 2d 1250, 1264 (D. Kan. 2008) (finding emails converted to PDFs and produced in paper format not “reasonably usable” and requiring reproduction in native form). “Self-imposed burdens and expenses related to forms generated for the purpose of litigation, rather than in the ordinary course of business, do not factor into the analysis.” In re State Farm Lloyds, 520 S.W.3d at 607 n.41.
#1: Courts must impose reasonable limits. The Texas Supreme Court had one point it wanted to stress above all else—the duty of trial courts to impose reasonable discovery limits. The basis of that process is determining not just what is relevant but what is proportional to the needs of the case. For those decisions to be made, both sides will need to marshal arguments and evidence—not just surmise or guess as to what is relevant and proportional. And trial judges will need to roll up their sleeves and engage.